Full text of "Armature winding and motor repair; practical information and data covering winding and reconnectig procedure for direct and alternating current machines, compiled for electrical men responsible for the operation and repair of motors and generators in industrial plants and for repairmen and armature winders in electrical repair shops"

ENGINEERING LIBRARY

ARMATURE WINDING

AND

MOTOR REPAIR

ARMATURE WINDING

AND

MOTOR REPAIR

Practical Information and Data Covering Winding and Reconnecting

Procedure for Direct and Alternating Current Machines, Compiled

for Electrical Men Responsible for the Operation and Repair

of Motors and Generators in Industrial Plants and for

Repairmen and Armature Winders in

Electrical Repair Shops

BY
DANIEL H. BRAYMER, A. B., E. E.

AMERICAN INSTITUTE OF ELECTRICAL ENGINEERS MANAGING EDITOR
OF ELECTRICAL WORLD FORMERLY EDITOR OF ELECTRICAL
ENGINEERING AND OF ELECTRICAL RECORD

FIRST EDITION
TWENTY-SEVENTH IMPRESSION

McGRAW-HILL BOOK COMPANY, INC.

NEW YORK AND LONDON

HfiiJl

UttW

ENGINEERING LIBRARY

PRINTED IN THE UNITED STATES OP AMERICA

THE MAPLE PRESS - VORK PA

PREFACE

In this book no attempt has been made to discuss the sub-
ject of armature winding from theoretical or design stand-
points. On the contrary, it is a compilation of practical
methods tha't are used by repairmen and armature winders.
In selecting the material a special effort has been made to
include as far as possible details of those methods which have
been found by actual experience to represent best practice in
a repair shop of average size. In this work the writer has
drawn from his own experience in repair work, from the ex-
periences of repairmen and armature winders in large and small
repair shops and manufacturing plants which have been
visited, from descriptions of practical methods and the pro-
cedure followed in the solution of special problems as presented
by practical men in technical journals.

The title of repairman as used throughout this book is one
that a good engineer can bear with pride when he measures
up to all its qualifications. Such an engineer is one who in
in the majority of cases not only knows what to do in the case
of an electrical trouble but just how to proceed to do that
particular thing and who seldom guesses without a good per-
centage of the probabilities of being right in his favor. The
main difference between the designer and the repairman is
that the former must know what to do while the latter must
know how to do it. A capable repairman combines both
qualifications through years of experience.

When called upon to locate troubles in motors and genera-
tors, electricians and repairmen whose experience in this kind
of work has been limited often find themselves wondering just
what to do first. It is from just this viewpoint that the infor-
mation on winding procedure and the hunting and correcting
of troubles has been presented. That is, instead of discussing
the fundamentals involved in any method of working out a
repair problem, the actual problem or job as the case may be
is discussed from the " how-to-do-it" standpoint. Then for
each individual operation or procedure the applications of

M184967

vi PREFACE

fundamental laws and rules are worked out. Considerable
repetition of some details of similar methods will therefore be
discovered in connection with information covering such pro-
cedure as the rewinding of machines of the same class but
of different types. This has been considered advisable since
a repairman should not be required to study a complete
volume when details and information are desired at any one
time on the procedure for a particular type of winding for a
particular design of machine.

Liberal use has been made of practical data and practices
in repair shops so as to combine the good features of a book of
methods with handbook information covering these methods.
If this book shall help young repairmen to absorb information
that can be secured otherwise only through years of experience
in handling one job after another, and if the older and more
experienced repairmen find it a handy source of reference as
a supplement to their own stock of information, then the aim
of the author will be accomplished.

When material has been taken from the experiences of
engineers and their recommendations on repair methods as
published in the technical journals, it has been the aim to
give credit to both the author and the journal in the para-
graph, or section where the material is used. Special acknowl-
edgment is made to A. H. Mclntire, editor of the Electric
Journal, for permission to make liberal use of information con-
tained in several articles compiled at his suggestion and pub-
lished in the Journal. This material has been incorporated
in Chapters 3, 8, 9 and 11. To A. M. Dudley, engineer of the
industrial division, Westinghouse Electric & Manufacturing
Company, the author is also especially grateful for sugges-
tions and for permission to use details of methods which he
has developed for reconnecting and testing induction motor
windings. This information appears in Chapters 9 and 11.
The diagrams at the end of Chapter 11 have been selected
from a series of eighty-one devised by Mr. Dudley and
shortly to be published in a valuable treatise on "Connecting
Induction Motors."

The author also desires to acknowledge the assistance ren-
dered by H. S. Rich and Alex R. Knapp in the form of data

PREFACE Vll

:''' !"''.-

and information compiled from their own experiences in solving
a variety of motor troubles met with in industrial plants and
in making repairs. To Henry Scheril, formerly a member of
the engineering department of the Crocker- Wheeler Company,
acknowledgment is made for helpful suggestions in arranging
the material and for assistance in checking and reading the
proof. Credit is also due and is hereby accorded to the elec-
trical manufacturers who furnished the photographs from
which many of the halftone illustrations were made.

DANIEL H. BRAYMER.
NEW YORK CITY,
December, 1919.

INTRODUCTION

Through the courtesy of the author of this book the writer
has had the privilege of reading the proofs. I have found,
with great delight, that the treatment of the subjects discussed
is not only clear and easily understood but always from the
practical man's standpoint. While the book will appeal
strongly to practical men engaged in repair shop work, power
station work and the maintenance of motors in industrial
plants, it will also appeal, in the opinion of the writer, to
students of electricity. Since the material presented in this
book, which I dare say is unique in its field, has been ob-
tained from actual practical experiences and outlines the
practical remedies that have been applied by repairmen in the
solving of puzzling problems, it will be of decided assistance
to men who are in need of such practical help.

It is an " electrical book of knowledge," for in its pages
readers will find answers to practically all armature winding
questions and solutions of many of the repair problems that
they will meet in practical work. The diagrams are clear and
easily followed by the shop man and run in synchronism with
the text. Theory with mathematical considerations have been
resorted to only in a very few cases so that the reader of the
book can make use of the information and understand the
discussions of all phases of armature winding even though he
may have only a limited knowledge of mathematics.

A book of this kind, in spite of the errors that are bound to
creep in, is a very valuable asset to any practical man who de-
sires to enlarge his own stock of knowledge by learning how
other men in similar positions have solved the many electrical
problems that come to the repairman.

HENRY SCHEBIL.

NEW YORK CITY,

December, 1919.

CONTENTS

PAGE

PREFACE ........ f . ,. v

INTRODUCTION. . \ .... ..... ix

CHAPTER I

DIRECT-CURRENT WINDINGS

Action of a Commutator .'..'. . 1

Types of D-C Armature Windings i . . ^ . . . 2

Winding Parts and Terms 1 ....'.. 2

Armature Conductor or Inductor '...?. . 3

Winding Element or Section . .-. V.. :; .'H'. , V . . ". ... 3

Armature Coils. . ^i~. '..?-. , ; i . . N . . ; ' . "". - . '. , . . . 4

Winding Pitch or Coil Pitch . 6

Front and Back Pitch. . ; vV v . . , : . . . . . . . . . ; '%.

Full Pitch and Fractional Pitch Coils . . . -. . . \ . . '. . . '^ - ' t

Symbols Used in Winding Formulas . . . . .":..,/ . . / 8

Numbering Coil Sides in Armature Slots . .VV.Y . ... 8

Lap Multiple or Parallel Windings . . . . ......... 9

Formulas for Lap Windings Multiple; Single, Double and
Triple Windings Meaning of the Term Reentrant Multi-
plex Lap Windings.

Wave Series or Two-circuit Windings 16

Formulas for Wave Windings Multiplex Wave or Series-
Parallel Windings Formulas for Series- Parallel Windings
Symmetrical Windings.
Possible Symmetrical Windings for D-C Machines of Different

Numbers of Poles 21

Equipotential Connectors T . .' ; . . : . . . . 21

Best D-C Windings for a Repair Shop to Use . .,;;..'.. . 23

Number of Armature Slots .....';. . 23

Voltage between Commutator Segments . . .... ...... 24

Number of Commutator Bars 24

Usual Speeds and Poles of Different Sizes of Generators. ... 25

Safe Armature Speeds . . . ... r .;.v v .......';. . 25

CHAPTER II

ALTERNATING-CURRENT WINDINGS

Types of A-C Windings , i . 27

Distributed Windings . . 27

xii CONTENTS

PAGE

Concentrated Windings . .7. .... ..'..' 27.

Spiral or Chain Windings . . - . . . . 27

Lap and Wave Windings .'. : . . ; ... 28

Whole-coiled and Half-coiled Windings . . * v V^' v *' ^L'V . . 30
Single-phase and Polyphase Windings .......,;.... 30

Coil Pitch ../I' 32

Phase Spread of Winding 32

Two-phase from Four-phase Windings . .../..... . . . 33

Three-phase from Six-phase Windings . . . ... ....... 33

Wire, Strap and Bar Wound Coils 34

Methods of Laying out and Connecting A-C Windings .... 35
Group Windings Full and Fractional Pitch Windings
Simple Winding Diagram Reconnecting a Winding Simple
Method for Indicating Polarity of Coil Groups Changing Star
to Delta Connection A-C Wave Windings -Progressive and
Retrogressive A-C Wave Windings Connections for Coils
of Polyphase Windings Double-layer Winding, Lap Con-
nected Connecting a Chain Winding Other Common
Windings.

Easily Remembered Rules for Arrangement of Coils in an Induc-
tion Motor . . ; , , 51

Simple Rules for Checking Proper Phase Relationship in Two-
or Three-phase Windings . . ,. -., .... 52

CHAPTER III

REPAIR SHOP METHODS FOR REWINDING D-C ARMATURES

Dismantling a D-C Armature 56

Winding Data Needed for a Dismantled Armature 57

Removing Old Coils 57

Winding D-C Armatures Having Partially Closed Slots .... 60
Winding a Threaded-in Coil Insulating Lining for Slots
Inserting Coils in the Slots Insulating Overlapping End Con-
nections of Coils Connecting Finish Ends of Coils to Com-
mutator Loop Windings for Small Motors.

Winding D-C Armatures Having Open Slots 66

Winding and Insulating Coils Insulating Open Slots In-
serting Coils in Open Slots Shaping End Connections-
Truing up the Heads of the Winding Insulation Between
Commutator End Connections.

Winding Large D-C Armatures 77

Coils for Large D-C Armatures Lap and Wave Windings for
Large Armatures Insulating the Core Inserting the Coils
Banding Wire Balancing Large Armatures Rotary Con-
verters Three-wire Generators.

CONTENTS xiii

PAGE

Winding Railway, Mill and Crane Types of Armatures .... 90
Railway Type Armature Coils Coil Insulation Insulating
the Core of Railway Armatures Inserting the Coils Con-
nections with Dead Coils Hooding and Banding.

CHAPTER IV

MAKING CONNECTIONS TO THE COMMUTATOR

Locating First Connection to Commutator 101

Testing Out Coil Terminals 102

Commutator Connections for a Lap Winding. . . . . . . ' . . 102

Requirements of a Lap Winding 103

Commutator Connections for a Wave Winding 104

Locating First Connection to Commutator for a \Vave Winding . 105

Requirements for a Wave Winding 107

Progressive and Retrogressive Wave Windings . . . .'". . . . 107

Wave Winding with Dead Coils . V. : 108

Cutting out Coils of a Retrogressive and Progressive Wave

Winding 109

Tables for Placing Coils and Connecting Them in a D-C Winding 110

Wave vs. Lap Windings '....,' l! . ..'... . . 112

Lap Windings for D-C Armatures . . . \ ''.*/." '". . . '.' . . . 115

Lap Windings for A-C Machines '.,-. . . . 116

Wave Windings for D-C Armatures ,; V/'V . . . 117

Wave Windings for A-C Machines ;,v. .- . . . .118

Single vs. a Number of Independent Windings . ; . . . . . .118

Lap Windings vs. Multiple Wave Windings 119

Use of Equalizer Rings .-. T ..../:...'.;:' ^ .... 120

CHAPTER V

TESTING DIRECT-CURRENT ARMATURE WINDINGS

Causes of Short Circuits in an Armature. . . ... ... .. ,. . . . 122

Test for Short Circuits in an Armature . . . ...... . .' . v . . 123

Testing for Short Circuits and Open Circuits with a Small Trans-
former ...-;. / v ... : ,....,. . . 125

Causes of Open Circuits in an Armature 127

Tests for an Open Circuit in an Armature 128

Cutting Out Injured Coils 129

Causes of Grounds in an Armature 130

Tests for Grounds in an Armature 131

Use of a Bar Magnet and Millivoltrneter to Locate a Reversed

Armature Coil

Use of a Compass to Locate a Reversed Armature Coil . . . .132

Locating Low Resistance or Dead Grounds 133

xiv CONTENTS

PAGE
Use of a Telephone Receiver in Testing for Short Circuits, Open

Circuits and Grounds 135

Testing for Reversed and Dead Field Coils 136

The Commutator : ,.. . s . . 136

Testing Equipment for a Repair Shop . .,,.-. ... . . . . . 138

CHAPTER VI

OPERATIONS BEFORE AND AFTER WINDING D-C ARMATURES

Stripping Off an Old Winding 139

Cleaning and Filing Slots 140

Testing Commutator -, , y- .... 140

Making New Coils 140

Forms for Winding Coils f ..... 141

Insulation of Core and Slots 144

Testing Out the Winding V:* j y ;v.^ A . . . 144

Soldering Coil Leads to the Commutator . , . .,-.,,. . . . . . 144

Hoods for Armatures . . ...... ... . . .,.-.. . . . . 145

Banding Armatures ,. . ..... . 146

Seasoning and Grinding a Commutator . . ,- ..... ... . . 148

Undercutting Mica of a Commutator . .... . '*. 150

Balancing an Armature ., . ... '...-.-. .. ^ . ... * - , 150

Painting the Winding . . . . N . ..;.* ... .,.<.. ... 151

Relining Split Bearings . . . ... , . :. . , '. . , . .... . 151

CHAPTER VII

INSULATING COILS AND SLOTS FOR D-C AND A-C WINDINGS

Insulation for Armature Coils and Slots . . . . . 153

For Mechanical Protection and Electrical Insulation For
High Temperatures and Electrical Insulation For Electrical
Insulation Only.

Descriptions and Uses of Insulating Materials ..... . . . 156

Treated Cloths Pressboards, Fibres and Papers Coil and
Slot Insulation Used in One Large Repair Shop Micarta-
folium.

Thickness of Insulation Required in Slots 163

Insulation of Formed Coils -. ". . . . . 163

Insulation for Coils Used in 240-Volt and 500-volt D-C Machines. 165

Coil Insulation for Induction Motor Windings . . . / ; . . . 167

Coil and Slot Insulation Employed by a Large Manufacturer. . 168

Insulation of End Connections of Coils .... . ' / :r. .' . . 172

Phase Insulation When Reconnecting from 2-phase to 3-phase . 173

Mica Insulation for Armature Coils .!.;-.. . . 173

Repairing Coils Damaged in Winding Process . -. . .'"'. . . . 174

Voltage to Use When Testing Coil and Commutator Insulation. 175

CONTENTS xv

PAGE

Field Coil Insulation . 179

Varnishes and Impregnating Compounds for Coils 176

Characteristics of Insulating Varnishes 178

Solvent Chart for Insulating Varnishes 178

Method for Making Tape from Cotton Cloth . 181

Drying Out Insulation of D-C Generators 181

Drying Out Insulation of Synchronous Motors 182

Drying Out Induction Motors 183

Measuring Insulation Resistance >;;';>-. . 185

CHAPTER VIII

REPAIR SHOP METHODS FOR REWINDING A-C MACHINES

Winding Small Single-phase Motors. . ...'.. . i- ;''';".'".''.. . 186
Insulating Lining for Slots Winding the Skein Coil Insert-
ing Skein Coil in Slots Winding for a Repulsion-start Motor
Winding Small Motors by Hand Windings for Odd Fre-
quencies Connections for Main and Starting Windings
Testing Small Induction Motor Windings Windings for
Small Polyphase Induction Motors.

Winding Small Induction Motors with Formed Coils in Partially

Closed Slots ; *'..;.. . . 194

Insulation for Slots Basket Coils Winding a 3-phase
Stator with Basket Coils Threaded Diamond Coil Winding
a 3-phase Stator with Diamond Coils.

Winding Induction Motors Having Open Slots. . . . ; . . . 201
Winding a 2-phase Stator Having Open Slots Testing the
Windings Inserting a New Coil in a Winding Connecting
the Coils Points to Consider when Connecting Coils Cleats
and Terminals Painting Winding.

Induction Motor Secondaries . 207

Squirrel-cage Secondaries Phase- wound Secondaries.

Winding Large Alternating Current Stators . . . ' . v . . . . . 210
Coils for Partially Closed Slots Coils for Open Slots Lap
and Wave Connections Insulation of Coils Inserting
Shoved Through Concentric Coils Bar and Connector Wind-
ing Diamond Coils Double Windings Testing Windings
of Large Machines Connecting the Coils Bracing Needed
for Heavy Windings.

Winding the Stator of Alternating Current Turbo-generators . . 223
Coils for A-C Turbo-generators Forming the Coils Insula-
tion for Turbo-generator Coils Testing Turbo-generator
Windings Inserting the Coils in a Turbo-generator Bracing
for Windings Connecting the Winding Break-down Test.

xvi CONTENTS

CHAPTER IX

PAGE

TESTING INDUCTION MOTOR WINDINGS FOR MISTAKES AND FAULTS

Testing for Grounds and Short-circuits . . . . .^ .,'.. . 231

Reversal of One or More Coils or Groups . . . . .> , . . . . 233

Open Circuits v . ..> . . . 234

Placing Wrong Number of Coils .,:.,.*.. . N . . 234

Using an Improper Group Connection . . . 234

Order in Which Tests Should be Made f; - . . . 235

Connecting for the Wrong Number of Poles 235

Applying Direct Current and Exploring with a Compass. . . . 236

CHAPTER X

ADAPTING DIRECT-CURRENT MOTORS TO CHANGED OPERATING CON-
DITIONS
Changes in Speed. . . .^ ...... ..... r ,.'...-.-.. . . 237

Changes in Operating Voltage . . . fe , .>,:.: . . , . ." . 238
Operating a Motor on One-half or Double Voltage . . . , . . 239
Size of 'Wire for D-C Armature Coils . > . .. / . ^ . . . 240
Operating a Generator as a Motor and Vice Versa . ..... . . 241

Motor Speed when Reconnecting a D-C Motor Winding Wave

to Lap 242

Adjusting the Air Gap on Direct-Current Machines. ;. N . v '.,. . . 242
jChange in Brushes when Reconnecting D-C Motor from a Higher

to a Lower Voltage 243

Rewinding and Reconnecting D-C Armature Windings for a

Change in Voltage 243

Reconnecting a Lap Winding Reconnecting a Wave Winding

Reconnecting Duplex Windings.

CHAPTER XI

PRACTICAL WAYS FOR RECONNECTING INDUCTION MOTORS

Points to Consider before Making Reconnections ....... 262

Diagrams for Different Changes of Connections . .... 262

Diagrams for Three-phase Motors 265

Use of Table of Connections V?^> ; . . 267

Two-phase Diagrams . .. ^ V , >-...- v .-. . 267

Meaning of the Term Chord Factor. . .. . ... v .-, \ : * . . . . 269

Phase Insulation V. , I '-.- . . . 271

Reconnecting Motors to Meet New Conditions. / 272

Procedure when Considering a Reconnection of Windings
Practical Example for Reconnection Changes in Voltage
only with all Other Conditions Remaining the Same
Changes of Phase Only Changes in Frequency Changes in
Number of Poles Testing a Reconnected Winding.

CONTENTS xvii

PAGE

Effects of High and Low Voltage on Motor Operation 283

Operating Standard A-C Motors on Different Voltages and

Frequencies. ,,-.... 284

Factors which Limit a Change in Number of Poles of an Induction

Motor . . ". ., 286

Single-circuit Delta and Double-circuit Star Connections . . . 287

Cutting Out Coils of an Induction Motor 288

Procedure When Connecting Coils of an Induction Motor

Winding 288

Connecting Pole-phase-groups of a Winding General
Theory on Which Connection Diagrams are Constructed
Determining Number of Poles from Slot Throw of Coils
Typical Circle Diagrams for Connecting Induction Motors.

CHAPTER XII
COMMUTATOR REPAIRS

Causes of Commutator Troubles ...'..... . . 301

Troubles Resulting from High Mica 301

Remedy for High and Low Bars . . '" . 302

Burn-out Between Bars . . . . . . . . . . ..... . , . 303

Plugging a Commutator. .V. ; V': :. ..,...; . . . . . 304

Removing Bars and Mica Segments for Repairs . . . . . . 304

Repairing a Burned Commutator Bar . . ..'..: ... . . 305

Replacing a Repaired Commutator Bar . . . . -, : . . >. . . . 306

Tightening up a Repaired Commutator . . ... .... ... 306

Baking Commutator with Electric Heat . .. . ,. ; ; ;-> 308

Removing and Repairing Grounds in a Commutator 308

Turning Down a Commutator without Removing Armature . . 309

Temporary Cover for Use When Turning Down a Commutator. 310

Refilling a Commutator 311

Boring Out the End of a Commutator , \ ^ . . . 313

Mica Used in Commutators ... > .- -315

Shaping Mica End Rings . . 316

Templet for Making Mica End Rings '. . . . /:./,. 316

Micanite as a Commutator Insulation . . . ; . '. . 317

Precautions when Tightening a Commutator. ... , . ... '. . . 317

Making Micanite End Rings ' . . .

Causes of Excessive Commutator Wear . . . . . ' . . . .

Copper Used for Making Commutator Bars ..... . . 319

Test for Oil-saturated Mica in a Commutator .
Blackening of a Commutator at Equally Spaced Points .

Undercutting Mica of Commutators 320

Tools for Undercutting Mica Size of Circular Saw Required
Finishing Slots and Commutator Surface after Undercut-
ting Brushes for Use on Undercut . Commutators Hand
Tools for Undercutting Mica of Commutators.

xviii CONTENTS

CHAPTER XIII

PAGE

ADJUSTING BRUSHES AND CORRECTING BRUSH TROUBLES

Fitting or Grinding-in Brushes . . . v 326

Adjustment of Brushholders 329

Causes of Rapid Brush Wear , . % , . .331

Methods for Locating the Electrical Neutral in Setting Brushes . 332

Angle at Which Brush is Set. . . v 333

Checking Brush Setting 334

Brush Pressure /. ', .,!.= .'. . . 334

Common Brush Terms . v vs. : , . . 334

Procedure for Locating Causes of Brush Troubles. . . ... . 337

Too Low Brush Pressure Incorrect Spacing of Brushes
Brushes Not Operating on Electrical Neutral Incorrect
Thickness of Brushes Using Brushes of Wrong Character-
istics.
Causes and Remedies for Sparking at Brushes . .... . ... . . 341

Causes and Remedies for Flat Spots on Commutator ..... 342

Causes and Remedies for Blackening of Commutator . . . . . 343

Causes of Heating in a Motor or Generator . . .- \ . ... . 343

Causes and Remedies for Honey-combing of Brush Faces . . . 343

Causes and Remedies for Brushes Picking up Copper 343

Causes and Remedies for Brushes Chattering .... N ^ 344
Causes and Remedies for Loosening of Brush Shunts .... . 344

CHAPTER XIV

INSPECTION AND REPAIR OP MOTOR STARTERS, MOTORS AND GEN-
ERATORS.

Cost of Repairs for Polyphase Motors 346

Points to Consider When Estimating Cost of Motor Repairs . . 347
Inspection and Overhauling of

Direct-Current Motor Starters. . >. . X 349

Auto-starters for A-C Motors . .... 351

Drum Type Controllers . ... . . % . . . 355

Large Compound D-C Motor 358

50-Hp. Induction Motor . ... . . . , 363

25-Hp. Slip-ring Motor ,. . ... . . . . . ;.*'. 366

Single-phase Commutator Motor . . . . . :.'. . . '-. . . . 368
Direct-Current Engine Type Generator. . . . < . .... . 371

CHAPTER XV

DIAGNOSIS OF MOTOR AND GENERATOR TROUBLES

Lack of Proper Cleaning >..* 376

Operation in Damp Places . . . . 7 . *^,. . ;..,.*.:> 377

Exposure to Acid Fumes and Gases 377

Lack of Frequent Inspection and Replacement of Worn Parts . . 378

CONTENTS xix

PAGE

Operating Temperatures too High 378

Electrical Defects 379

Causes and Remedies for Troubles in A-C Machines 386

Induction Motor Troubles Locating Troubles in Windings
of Induction Motors Mechanical Adjustments Troubles
Due to Electrical Faults Troubles with Synchronous Motors
Causes of A-C Motor Fuses Blowing Inspection of
Motor Starting Devices Testing Motor for Grounds Hot
Stator Coils Tension of Belts Troubles in Rotor Wind-
ings Examination of Stator Winding Sparking at Slip
Rings.

CHAPTER XVI

METHODS USED BY ELECTRICAL REPAIRMEN TO SOLVE SPECIAL
TROUBLES

Sparking at Commutator Caused by Poor Belt Joints 394

Plugging a Commutator Uv;.'. , . v v . .' * . . 395

Knock in an Armature Due to Band Wires Being too High . . 395
Heating of Armature Traced to Poor Soldering of Commutator

Connections 396

How a Commutator was Repaired under Difficulties 397

Holder for Sandpapering Commutator ; . . . 400

Use of Portable Electric Drill to Undercut Mica 400

Jerky Operation of New Commutator Traced to Burred Commu-
tator Bars . . . . .'-..' ..... 402

Why Brush Studs Heated on an 8-pole Machine . . . ..- , . . . 403

An Accident Due to Incorrectly Set Brushes 404

Wrong Setting of Brushes for Direction of Rotation Caused 1 lotor

to Flash .V. .-,'. 405

Proper Adjustment of a Reaction-type Brush-holder 406

Heating of Brush-holders Traced to Defective Contact Springs . 407

Simple Scheme for Banding Armatures 408

Use of a Crane to Band an Armature . . . '.' . . . ". . . . . 409
Method Used to Band a 2000-Hp. Rotor . '., . V : . .... . 410

Improvised ft ethod Used to Turn a Commutator 411

Cause of a Motor Reversing its Direction of Rotation on High

Speed _.;':

Checking Connections of an Interpole Motor

Heating of Field Coils Traced to Wrong Type of Starting Box . .414

Safe Operating Temperature of Portable Desk Fans 415

An Adjustable Shunt for Series Fields of Exciters. . .416

A Peculiar High-speed Motor Trouble. . . ...'.... . 417

Ways that End Play Variations Show Up .418

Connections for Two 220-volt Motors When Operated on 440
Volts. .\. 419

XX CONTENTS

PAGE

Cleaning Motors with Compressed Air 420

Testing out Phase Rotation 420

An Induction Motor Trouble Due to Wrong Stator Connections . 421
Stalling of Wound Rotor Induction Motor Explained . .... 422
Loose Bearing Caused Induction Motor to Fail to Start .... 423

Three-Phase Motors Used on Single-phase Lines 425

An Apparent Overload Trouble That was Traced to a Defective

Fuse Block 427

Cause of Noise in Three-phase Motor Driving Exhaust Fan. . 428
Cause of Burned out Starting Winding in a Single-phase Motor. 429
Cause of One Motor Failing to Start While Another was Running

on Same Circuit 430

Cause of Synchronous Motor Failing to Start . .... . . . 432

Effect of Decreased Frequency on Induction Motor- Generator Set 433
Simple Rules for Reconnecting A-C Motors . . . . . ... . 434

Changing 440-volt Motor for Operation on 220 Volts . ; -' 435
Multiple Connection Diagram for A-C Motor Windings .... 436

Brush and Slip-ring Sparking Traced to Absence of Rotor Balanc-
ing Weights 437

Overheating of an Induction Motor Traced to Variation in Fre-
quency. . . . . V . . * , ". . . . . . . . V , .. . 438

Relief for a Hot Bearing. . . : . . . . . . ',. ,, ... . . . , . . 440

Static Sparks from Belts . . . . _. -. ... o ... . V. ,- ... . 440

Ratings of A-C Generators. . . : . . ...*,.*.*. . . 440

Alternating Current Motor Phase Rotation . . '.',, ;, ,- .... 440

End Bells or Heads . ....;..... ,'..,.,... . 441

Brushes and Brush Holders . ',. ; .;.. . . . . .,;.,..... . 441

The Rotor ,= ,>.!..:. -.'...-. . ... > s A ..\ . 441

The Stator , . >.,-,,;.;. . 441

Sizes of Fuses for A-C Motors ...... .";..-.. . . . 442

CHAPTER XVII

MACHINE EQUIPMENT AND TOOLS NEEDED IN A REPAIR SHOP

Armature Winder's Tools 445

Device for Shaping Insulating Cells of Armature Slots 447

Tool for Cutting Cell Lining at Top of Slot . : . . . . ...,.'. . 447

Special Winding Tools 449

Repair Tools that can be Made from Old Hack-saw Blades . . 451

Special Coil- winding Device . . . . . . . ,, rt^-> . . 452

Steadying Brace for Repairing Small Motors. . , . ... . . 454

Tension Block for Use when Banding Armatures ... . . 454

Armature Sling . . , . . . . . V . . 455

Pinion Puller ...'... .456

Coil Winding Machines ! . . .. . ... . . . . 456

Qoil- taping Machines . ....,..,.,,,;,*.>'< 4^7

CONTENTS xxi

r ( k

PAGE
Commutator-slotting and Grinding Machines . ; .!. '-. . . . . 459

Armature Banding Machine . V . . . ' . .-. . v . . . . . . 462

Combination Machines ...... .; '.-.', . . 462

APPENDIX

DATA AND REFERENCE TABLES

How to Remember the Wire Table . . . - . - - \'', . . ,-. . . . 465

Copper for Various Systems of Distribution . . . . V .... . 465

Classification of Wire Gauges > 1, . .... 466

General Wiring Formulas for A-C and D-C Circuits . . . . . ' 468

Data for Connecting Motors to Supply Circuits .... . . . 473

Voltage, Horsepower and Speeds of Motors . . > ^- , . . 478

Transformer Rating for A-C Motors ... ........ 478

Sizes of Fuses. Switches and Lead Wires for Motors of Different

Sizes on Different Voltages ...-.. . . . 484

Circuit Breakers for Overload Protection of Motors 489

Belting .''..... . . .490

Rules for Pulley Sizes . v . . . . . 492

Speed of Pulleys ... , . . v . . . ... 492

Chain Drives _ . . 493

Points to Consider when Calculating Size of Chain 493

Horsepower Transmitted by Steel Shafting . . A . v 495

Horsepower Transmitted by Single Ropes . -. . ; . . . . . . 495

Gear Table .'..-. 496

Some Handy Rules . . . , , .% . /.'..;.... 498

INDEX . . 500

ARMATURE WINDING S

AND

MOTOR REPAIR

CHAPTER I
DIRECT-CURRENT WINDINGS

The essential physical differences between a complete direct-
current and a complete alternating-current armature winding
is that the former is wound on the rotating member of the
machine while the latter is wound on the stationary member
and that the direct-current winding requires a commutator
while the alternating-current winding does not. However,
since the practical make-up and construction of windings
will be discussed later for particular types of direct and
alternating-current machines, the general theory of arma-
ture windings will likewise be taken up first for direct-current
and then for alternating-current machines (see Chapter II).

Action of a Commutator. The emf and current produced
in each armature conductor of a direct-current generator is
alternating in character. It is the function of the commutator
to deliver from the armature winding an electromotive force
and current that is unidirectional, that is, such that one termi-
nal will be always of positive polarity and the other of nega-
tive polarity. The commutator and its brushes accomplish
this by being connected in series between the generator leads
and the armature windings so as to reverse (in effect) the con-
n3ctions of the armature coils (connected to the commutator
bars) with respect to the machine leads every time the emf
and current induced in these coils reverse upon moving out of
the influence of one pole into the field of the next adjacent

2 ARMATURE WINDING AND MOTOR REPAIR

pole. The alternating emf and current generated in the
armature winding is thus rectified or commutated into a uni-
directional emf and current.

In the case of an alternating-current generator no such
rectifying of induced emf and current is necessary so that the
coils or elements making up the armature winding can be
connected directly together with the resulting terminals of
the winding becoming the terminals of the machine.

Types of D.-C. Armature Windings. In general, armature
windings are either of the open circuit or the closed circuit
type. The latter is used in all modern direct-current ma-
chines, while alternating-current machines may have either
open or closed windings. In the closed circuit winding of
the direct-current machine the end joins up with the begin-
ning or re-enters itself with the commutator tapped to the wind-
ing at equally distant points. In the case of open circuit
winding of an alternating-current generator wound on the
revolving member, the ends terminate in collector rings and
the winding is thus open until closed by the brushes of the
external circuit. When the winding is on the stator of an
alternating-current machine the ends are joined through the
load circuit. The following classification of closed circuit or
direct-current windings may be made:

Direct-current windings (closed circuit).

1. Lap multiple or parallel.

(a) Single lap.
(6) Multiplex lap.

2. Wave series or two-circuit.

(a) Single wave.

(6) Multiplex wave or series-parallel windings.

Winding Parts and Terms. In formulas for armature
windings and in laying out a repair job, certain terms are used
which refer to parts of the armature winding, the armature
core and details of the arrangement of the former in the slots
on the surface of the latter. In what follows these terms are
explained. In most cases they are used alike both in windings
for direct-current and for alternating-current machines.

DIRECT-CURRENT WINDINGS 3

Armature Conductor or Inductor. That part of a wire
which lies in an armature slot and cuts the magnetic lines of
force or field flux as the armature rotates, is called an armature
conductor or an inductor.

FIG. 1. Different types of coils used in armature windings.
(A) One-piece series diamond strap coil. Leads at end of straight part. (B) One-piece
series diamond coil. Leads at end of straight part. (C) Two-piece series diamond coil.
(JD) One-piece multiple diamond coil. Leads at point of diamond. (E) Two-piece
multiple diamond coil. (F) Concentric coil bent down at both ends. (G) Concentric
coil, straight. (HI One-piece wire wound involute coil. Leads at point of involute.
/) Two-piece involute coil. Leads at point of involute. (J) Threaded-in type diamond
coil. Leads at point of diamond before and after pulling. (K) Basket coil. (LI Same
as B of threaded-in type. (M) Same as D of threaded-in type. (AO Bar and involute
end connector. (0) Group of concentric end connectors. (P) Concentric shoved
through type coil bent down on one end.

Winding Element or Section. That part of an armature
winding which is connected between two commutator bars is

4 ARMATURE WINDING AND MOTOR REPAIR

called a winding element. In its simplest form a winding
element consists of a coil of one turn of wire or two conduc-
tors. An element therefore must have at least two conduc-
tors but may consist of more than one turn of wire or even
number of conductors.

Armature Coil. When a winding element consists of more
than one turn of wire or two conductors, it is usually known
as a coil and the winding is a coil winding as distinguished
from a bar winding, where the conductors in the armature
slots are copper bars.

From a mechanical standpoint an armature winding con-
sists of a number of coils connected to a commutator in the
case of a direct-current machine or connected together in the
case of an alternating-current machine to form a series or
group. Each coil may be made up of one turn of wire with
each side forming one armature conductor or inductor, or a
coil may be made up of several turns of wire or of copper
strips. A classification of the different types and uses of
armature coils which has been made by R. A. Smart
(Electric Journal, Vol. VII, No. 6) is given in the accom-
panying table.

The so-called form-wound, diamond coils mentioned in the
table are formed and completely insulated before being as-
sembled on the armature. They can only be used in open
slots Concerning the advantages of diamond coils, involute
coils and concentric coils, Mr. Smart has the following to
say: "The great advantage of diamond coils is the easy and
simple manner in which they can be manufactured, especially
in large quantities, which makes them well adapted for stand-
ard machines. Since all the coils used on one machine are
of the same size and shape, only one winding mould over
which to form them is necessary. Moreover, the number of
spare parts which must be kept on hand for repairing is re-
duced and repairs can be made easily and quickly. From
the electrical point of view, the diamond type of winding
possesses the advantage of being absolutely symmetrical.
Hence there is no tendency for unbalancing of voltages due
to differences of self-induction; and in closed windings there
is no tendency to produce internal circulating currents."

DIRECT-CURRENT WINDINGS

CLASSIFICATION OF ARMATURE COILS ACCORDING TO SLOTS IN WHICH

THEY ARE USED, THEIR FORM AND TRADE NAMES AS EMPLOYED

IN BOTH DIRECT- AND ALTERNATING-CURRENT-WINDINGS

( Leads at ends of straight part

1 Leads at point of diamond

Involute

j Leads, at ends of straight part

Open slots

\ Leads at point of involute '

Short type

involute

Leads at point of involute

Concentric

Straight

Bent at both ends

Mould wound
coil of insulated

Shoved through con-

Straight
Bent at one end

wire or ribbon

centric

Bent at both ends

Partially

,'
Threaded

. ' , J Leads at end of straight part
.Diamond < .
|^ Leads at point of diamond

closed slots

Shuttle

Leads at ends of straight part

Basket

Leads at point of diamond

Hand-wound, pulled j S trai S ht

,, , i Bent at one end
through concentric ._ . . .
1 Bent at both ends

Diamond <

Open slots <

Involute

Form wound

coil of bare

Concentric

wire or strap

Shoved through con-

Partially

centric.

closed slots

Threaded, diamond <

Leads at end of straight part

Leads at point of diamond

Leads at end of straight part

Leads at point of involute.

Straight

Bent at both ends.

Straight

Bent at one end

Leads at ends of straight part

Leads at point of diamond

Bars and con-
nectors of
bare strap

Partially
closed slots

Involute end connectors
Concentric end connectors

Involute Coils. " Involute coils share the advantages of
the diamond coils in that all are of a standard size and shape.
They also require less space for end connection than any other
form of coil. They are, however, difficult to insulate properly
on account of the larger number of bends and are difficult to
assemble in position in the armature. For this reason their
use is restricted. The bar type of coil with involute end
connectors is easy to insulate and assemble and can be readily
repaired. Their principal use is for direct-current and indus-
trial motors where end space must be reduced to a minimum. "

Concentric Coils. "Concentric coils can be used on any
kind of slot. They can be hand-wound, machine-wound, or
'shoved through' (a combination of the other two methods),
as best suited. The shape of coils is simple, hence they are

6 ARMATURE WINDING AND MOTOR REPAIR

easy to wind either on a mould or by hand. They can be ade-
quately insulated, and can be securely braced with simple and
reliable coil support. However, the coils belonging to the
same group are of different size, and the coils in different
groups, except on single-phase machines, are bent in at least
two and often in three different shapes. This is a disadvantage
from the electrical point of view, since there will always be a
tendency toward unbalancing due to differences in self-
induction, and toward the production of circulating currents
in closed windings. And it is also a disadvantage from the
mechanical point of view, since for one machine, a large
number of different moulds will be required and coils cannot
be interchanged. Hence, the number of spare parts necessary
for repairs is greatly increased, and both the manufacturing
and repairing of the winding will require more time and be
more expensive."

Winding Pitch or Coil Pitch. The distance between the
beginning of one winding element or coil side to the beginning

Commutator Bars

CommutatprJJars

FIGS. 2 and 3. Comparison of wave and lap windings.

At the left a wave winding showing front pitch of coils as yi; back pitch as j/a; total
pitch as y and commutator pitch as yk. At right, a lap winding with the same notations
for the different pitches of coils.

of the next element or coil side connected to the first one is
called the total winding pitch. This is shown as y for both lap
and wave windings in Figs. 2 and 3. Winding pitch is meas-
ured in number of slots or by number of elements or coil
sides spanned by a single coil or by the number of commutator
bars between the connections to the commutator of the two
sides of a coil. In the third instance the measurement is
known as commutator pitch and indicated as yk in Fig. 2

DIRECT-CURRENT WINDINGS

and Fig. 3. In this chapter unless otherwise stated coil
pitch will be designated by the number of winding spaces
or coil sides spanned. There are two winding spaces per
slot in a double layer winding.

Front Pitch. The distance between the two coil sides con-
nected to the same commutator bar, measured in coil sides
at the front or commutator end of the armature, is called the
front pitch. It is indicated as y\ in Figs. 2 and 3.

Back Pitch. The distance between two sides of a coil,
measured in coil sides, at the back end of the armature is
known as the back pitch. It is shown as y 2 in Figs. 2 and
3. The total winding pitch y is equal to the algebraic sum

I I

Fro. 4. Lap and wave windings showing connections of formed coils to the
commutator in double layer windings.

of the front and back pitches. That is, since in a lap winding
the front and back pitches are of opposite sign, being laid
off in opposite directions on the armature, the total pitch y
will be the difference between them. In a wave winding where
the front and back pitches are laid off in the same direction, the
total pitch y will be their sum.

Full Pitch and Fractional Pitch Coils. When a coil spans
exactly the distance between the centers of adjacent field poles
(known as pole-pitch) it is said to be a full-pitch coil. In
cases where a coil is less than full pitch, it is said to have a
fractional pitch or to be a short-pitch coil. Such a winding is
often referred to as a short- chord winding.

Fractional pitch windings are much used because of the
following advantages: (1) They make possible shorter end-
connections and therefore call for less copper. (2) Armature

ARMATURE WINDING AND MOTOR REPAIR

reaction is reduced since the currents in the neutral zone flow
partly in opposite directions and neutralize each other. (3)
In alternating current generators a smoothing out of the wave
of induced emf is produced so as to more nearly approach a
sine form.

Symbols Used in Winding Formulas. The symbols which
are much used in winding formulas for the parts of windings
defined in the preceding paragraphs are as follows:

Z = Total number of armature conductors, (2gC) or (2gK).

C Number of winding elements or coils usually equal to K.

g = Number of turns per winding element or coil.

2g = Number of conductors per winding element or coil.

N Number of coil sides of winding.

y = Winding pitch or coil pitch.

2/i = Front pitch of winding element or coil.

2/2 = Back pitch of winding element or coil.

2/fc = Commutator pitch.

K = Number of commutator bars.

p = Pairs of field magnet poles.

2p = Number of field magnet poles.

2a = Number of sections of armature winding in parallel.

s = Number of slots on the armature.

Numbering of Coil Sides in Armature Slots. Form wound
coils are usually arranged in two layers with one side of each

5|D
6| D

FIG. 5. Winding with two
coil sides per slot.

FIG. 6. Winding with four coil sides
per slot. The coil pitch in this case is
7 winding spaces or coil sides.

coil placed in the bottom of a slot and the other side in the
top. For convenience in laying out windings, the coil sides
forming the top layers in the slot are given odd numbers and
those forming the bottom layers even numbers. This scheme
of numbering is shown in Figs. 5 and 6. In those cases
where the winding element consists of a coil of a number of
turns of wire instead of two conductors or bars, numbers are

DIRECT-CURRENT WINDINGS 9

assigned to the consecutive bundle of conductors or wires of
which the coil is made instead of assigning numbers -to the
individual conductors or inductors as they are sometimes
called. When such coils are used, the front and back pitches
are determined by counting consecutive half coils between the
the two coil sides of any coil beginning from one coil side and
counting forward as the coil is laid off and considering the
first side of the coil as one. That is, in case of a double layer
lap winding, with one coil side in the bottom of slot 3, and
the other side in the top of slot 7, the back pitch would be 7
counted as follows, there being two coil sides in each slot:
First coil side 6, next coil sides spanned are 7, 8, 9, 10, 11 and
12. The number of slots spanned by the coil is called the
slot pitch or throw of the coil. The throw of the coil, more
often called coil throw, in the case mentioned, where one side
of a coil is in slot 3 and the other in slot 7, is 4 slots. In the
case of a four-pole machine and an armature having 16 slots,
this coil throw is a full pole pitch, that is, from the center of
one field pole to the center of the next, and the winding is a
full pitch winding.

LAP MULTIPLE OR PARALLEL WINDINGS

-i
->

A lap winding is so called from the lapping back of the coils
or winding elements as they are wound on the armature core.
This is shown in Figs. 7 and 8. A better name is a multi-

FIG. 7. Lap winding showing the relative position of one winding element
or coil of two turns on the armature of a 4-pole machine.

pie or parallel winding because of the fact that such a winding
consists of as many circuits as there are field poles and these
circuits are connected in parallel between the brushes. The
number of brushes required therefore equals the number of

ARMATURE WINDING AND MOTOR REPAIR

armature sections in parallel and also equals the number of
poles. That is a = p. A lap winding can be distinguished

--fh

FIG. 8. Winding diagram for the lap winding of Fig. 7 showing how
the finish end of one winding element or coil of two turns is joined to
the commutator and to the start end of the next coil under the same pair of
poles.

FIG. 9. (a) Lap winding showing both end connections of each coil side
bent toward center of the coil. (6) Wave winding showing end connections
of each coil side bent in opposite directions.

from a so-called wave winding from the appearance of the
end connections of coils. In the wave winding the front and

DIRECT-CURRENT WINDINGS 11

rear ends of the coils lead in opposite directions while in the
lap winding they continue in the same direction around the
armature, as shown in Fig. 9.

Formulas for Lap Winding. The following rules and restric-
tions govern the assembling and use of a lap winding:

1. The front and back pitches (in winding spaces or coil sides)
must both be odd numbers and differ by two or some multiple
thereof.

2. The front and back pitches are of opposite sign (one positive
and the other negative) since they are laid off in opposite direc-
tions on the armature.

3. The winding pitch is equal to the algebraic sum of the front and
back pitches. That is, in case the front pitch is + 9 and the
back pitch 7, the algebraic sum is (9 7) or 2.

4. The total number of armature conductors or inductors must be
a multiple of the number of slots on the armature.

5. The number of armature slots may be odd or even.

6. The number of current collecting points or brushes on the com-
mutator must equal the number of poles.

7. The maximum emf between two consecutive coil sides (top
and bottom) in the same slot, is equal to or a little less than
the terminal voltage.

8. The end of one coil is joined to the commutator and to the
start of another coil (usually the next) under the same pair of
poles.

The following formulas apply in lap windings :

N b
2p

N b

Front pitch = y\ = ^ - 2

Back pitch = y k

Winding pitch = y = algebraic sum of yi and y% = 2
Commutator pitch = y k = 1

In the formulas for front and back pitches, N (as given on
page 8) is the number of coil sides in the winding; p the
number of pairs of poles and b any even number which will
make both t/i and y 2 odd whole numbers and about equal to
the pole pitch. The value of b should be taken as large as
allowable but not too large for then there is liability of the
two sides of a coil coming under the influence of poles of the
same polarity so that the induced emfs would oppose each

ARMATURE WINDING AND MOTOR REPAIR

other. For 6 = 0, the back pitch y 2 becomes equal to the pole
pitch. When b is positive t/ 2 becomes greater than the
pole pitch, while with b negative, 1/2 becomes less than
the pole pitch. In most cases b is taken negative so that 2/ 2
is equal to or less than the pole pitch.

FIG. 10. Lap winding for a 4-pole machine showing the direction of current
flow in one of the four current paths for the brush positions as illustrated.

The current (i) which flows in each conductor of a lap
winding is,

Where 7 is the total armature current or that flowing in the
external circuit. The value of (i) should be considered when
selecting the proper size of wire in making up or ordering coils.

The only practical lap winding from the standpoint of
symmetry is the one in which the number of pairs of circuits
(current paths) in parallel equals the number of pairs of poles
or a = p.

A lap winding is symmetrical when s -f- a and s -r- p is a

DIRECT-CURRENT WINDINGS 13

whole number. Here s equals the number of slots, a the
pairs of circuits in parallel and p, the pairs of poles. With
four-pole motors it is not always found that s -f- a and s -f- p
are whole numbers, since manufacturers sometimes make the
number of commutator bars and slots odd so as to make the
armature suitable for a two-circuit wave winding as well as a
lap winding.

In lap windings where the commutator pitch (y k ) equals
one, when a = p, the common factor of the number of commu-
tator bars, and yk will always be unity and there will be a single
winding.

In a double layer winding with two coil sides per slot, the
number of slots equals the number .of coils in the winding.
When there are more than two coil sides in a slot, the number
of coils, C = sN 8 -T- 2 with no idle coils; where s equals the
number of slots and N 8 equals the number of coil sides per slot.
More than two coil sides per slot greatly reduces the number
of slots required. Such a winding is symmetrical only when
s -r- a is a whole number.

For a lap winding the potential pitch, y p = K -~ p, where K
equals the number of commutator bars and p is the number
of pairs of poles.

When equalized rings are needed with lap winding, it is
good practice to use one ring for every 6 or 12 commutator
bars.

Multiplex, Single, Double and Triple Windings. In cases
where it is necessary that the armature shall carry a very
heavy current more than one winding may be used on the
armature with an equal number of commutator bars for each.
Both lap and wave windings may be so made up with one,
two or three entirely separate windings. In the case of a
double lap winding, the sides of a winding element or coil
of one winding are sandwiched between the coil sides of the
other and likewise the commutator bars of ona winding are
sandwiched between those of the other. Each brush must
be thick enough to always touch two commutator bars so that
both windings will always be connected to the brushes and both
deliver or receive current evenly. Three windings so sand-
wiched make up a triple winding.

14 ARMATURE WINDING AND MOTOR REPAIR

A single lap winding always has the same number of current
paths between brush sets as there are field poles while a double
lap winding has twice the number of current paths as there are
field poles. The triple lap winding has three times the number
of current paths as there are field poles. A single wave wind-
ing always has two current paths between brush sets while the
double wave winding has four current paths between brush
sets and the triple wave winding has six such paths between
brush sets.

An armature has one winding (single) when the number of
commutator bars K and commutator pitch y k have no common
factor; it is double when their common factor is two, and it
is triple when their common factor is three.

Meaning of the Term Reentrant. A winding is often said
to be single or double reentrant. In the case of a single
winding, this means that the winding closes on itself or re-
turns to the beginning point after being traced through all the
coils upon passing once around the armature core. A wind-
ing is doubly reentrant if it only re-enters itself after making
two passages around the coils of the armature. A single
winding may be either single reentrant or doubly reentrant.

In the case of double and triple windings, the term reentrant
is sometimes used in an improper sense. For this reason it is
advisable to specify types of armature windings by the num-
ber of separate windings, used. Thus a single winding as a
single-closed; two windings as double-closed and three wind-
ings as triple-closed etc. r A winding made up of two single
windings, each of which re-enters itself, will therefore be a
double-closed winding not a "double-reentrant" one.

Multiplex Lap Windings. As explained under the heading
of "Multiplex Windings," a lap winding may be made up of
one, two or three separate windings in order to handle heavy
armature currents. In the case of a double lap winding, two
similar windings insulated from each other are placed in the
armature slots with the even numbered commutator bars
connected to one winding and the odd numbered bars con-
nected to the other winding. In the same way for a triple
lap winding one-third of the commutator bars provided would
be connected to each winding.

DIRECT-CURRENT WINDINGS 15

The formulas which apply in multiplex lap windings are as
follows :

Front pitch = y l = ^= 2m
zp

Back pitch = y 2 =

Winding pitch = y = y 2 y = 2m
Commutator pitch = y k = - 2 = m

In these formulas w is 2 for a double winding and 3 for a
triple winding, etc. When the number of commutator bars
is exactly divisible by m, the windings will be entirely separate
from each other.

A double lap winding will have 2 X (2p) current paths be-
tween brushes, where 2p is the number of poles. That is, each
winding for a 4-pole machine will have four current paths so
that a double lap winding on a 4-pole machine will have eight
current paths between brushes.

In order that the two sides of a winding element or a coil
may move simultaneously under field poles of opposite
polarity, the total number of coil sides that make up the com-
plete winding divided by the number of the field poles, that is
N -T- 2p, is the approximate value for both the front and back
pitches as in the case of the single lap winding. Under such
conditions the electromotive force induced in the two sides of
a winding element add up. The smallest front or back pitch
to satisfy this condition, is the distance across a single pole
face and the largest front or back pitch is the distance from
one pole tip to the nearest pole tip of the same polarity.
If the front and back pitches are much less than N -5- 2p,
then a chorded winding results.

Since y\ and ?/ 2 must be odd numbers and approximately
equal to N -f- 2p, where 2p is the number of field poles, this
value will help in determining the value of b to be used in the
formula for front pitch (7/1) and back pitch (t/ 2 ).

The multiplex lap winding is largely confined to windings
for small and medium sized machines carrying large currents
and using coils made up of wire rather than copper strips or

16 ' ARMATURE WINDING AND MOTOR REPAIR

bars. This is for the reason that in such a winding for a
large multipolar machine the bars would be too thin and a
mechanical construction result which would make another
type of winding advisable, probably a series-parallel design.

WAVE SERIES OR TWO-CIRCUIT WINDING

The wave winding is so called from the zig zag or wave path
that the winding takes through the slots of the armature, as
shown in Figs. 11 and 12. This type of winding is more

FIG. 11. Wave winding showing relative position of one winding element or
coil of two turns on the armature of a 4-pole machine.

definitely described as a series or two-circuit winding because
of the fact that half of the armature coils or sections are con-
nected in series and the two halves are connected in parallel.
This winding therefore has only two current paths in parallel

FIG. 12. Winding diagram, for wave winding of Fig. 11 showing how the
finish end of one coil is joined to the commutator and connected to the start
of the next coil under the next pair of poles. &' is a continuation of 6.

between brushes regardless of the number of poles. Only two
sets of brushes are required for a machine of any number of
poles but improved commutation is brought about when the
number of brushes equals the number of field poles. The wave
winding is used in small and medium sized machines where it
is desired to keep the number of coils as small as possible.

DIRECT-CURRENT WINDINGS 17

Formulas for Wave Winding. The following rules and
restrictions govern the assembling and use of a wave winding:

1 . The front and back pitches (in winding spaces or coil sides) must
both be odd.

2. The front and back pitches may be equal or may differ by two
or some multiple thereof. The former condition is usually the
case.

3. The front and back pitches are of the same sign since they are
laid off in the same direction.

4. The winding pitch is equal to the sum of the front and back
pitches.

5. The commutator pitch and number of commutator bars must
not have a common factor.

6. The number of current collecting points or brushes on the
commutator is always two for any number of poles but the
number of sets of brushes may equal the number of poles.

7. The maximum emf between two consecutive coil sides (top
and bottom) in the same slot, is equal or nearly equal to the
terminal voltage of the machine.

8. The finish of one coil is joined to a commutator bar and to the
start of another coil which lies under the next pair of poles.

The following formulas apply in wave winding:

N 2
Winding pitch = y = yi + y 2 =

Front pitch = y\
Back pitch = y z
Number pairs of poles = p

^ u 2/i + 2/2 Kl

Commutator pitch = y k = ~

4 P

N
Number commutator bars = K = -^=pXyk^-'

N 2
The number of coil sides (N) must be such that

will be an even number with y\ and y% both odd -numbers.
yi and y 2 are usually taken equal. Under these conditions
the back pitch (y 2 ) is nearly equal to the pole pitch and the
sum of the front and back pitches is nearly equal to double
the pole pitch. If y 2 is reduced then yi must be increased so
that 7/1 + 2/2 will be constant. This is what happens in a
chorded or fractional pitch winding.

ARMATURE WINDING AND MOTOR REPAIR

The current (i) which flows in each conductor of a wave
winding is always one-half the total armature current, i =
I -f- 2. This value of (i) should be considered when select-
ing the proper size of conductors for coils of wave windings.
The closed winding formula for a wave winding is yk =
(K a) -f- p. Where yk is the commutator pitch, K the num-

FIG. 13. Wave winding for a 4-pole machine showing the direction of
current flow in one of the two current paths for the brush position as illus-
trated.

ber of commutator bars, a the pairs of parallel circuits in the
winding and p the number of pairs of poles. When a = 1, it
is only possible to make a single winding. The highest com-
mon factor of yk and K gives the type of winding, whether
single, double, etc. The number of coil sides in a wave wind-
ing equals twice the number of commutator bars.

The number of slots without idle coils must satisfy the
formula, s = (2K) -i- N 8 . Where K is the number commu-
tator bars and N a the number of coil sides per slot.

A wave winding is symmetrical when K -f- a, s -f- a and
p -5- a, are whole numbers. Where K is the number of com-

DIRECT-CURRENT WINDINGS 19

mutator bars, a, the number of pairs of parallel circuits in
the winding, s the number of slots, and p the number of pairs
of poles.

When equalizer rings arc needed with a wave winding, the
use of one ring to every 15 to 20 commutator bars is good
practice.

Multiplex Wave or Series-Parallel Winding. In a single
wave or series winding the number of armature circuits in
parallel is always equal to two, while in a lap or multiple wind-
ing the number of circuits in parallel is equal to the number of
field poles. The series-parallel winding is a design of wave
winding in which the number of armature circuits in parallel
may be larger than two and yet smaller than the number of
field poles. It is especially suitable for large multipolar arma-
tures which use winding elements made up of copper bars.

Formulas for Series-Parallel Winding. The following
rules govern the assembling and use of this winding:

1. The front and back pitches (in winding spaces or coil sides)
must be odd numbers. They may be equal to or different from
each other although they are usually equal.

2. The winding should be symmetrical as shown by the number
of field poles divided by the number of armature circuits in
parallel being a whole number. If this is not the case the final
winding should be made up of a sufficient number of independent
windings to make the number of field poles divided by the number
of armature circuits in parallel a whole number for each of the
independent windings.

3. It is necessary that the number of commutator bars divided by
one-half the number of armature circuits in parallel shall be
a whole number. If the conditions of (2) are fulfilled, the
condition named here is likewise satisfied at the same time.

4. If the number of commutator bars and the commutator pitch
have no common factor, the winding is a single one; if the com-
mon factor is 2, the winding is double or if it is equal to m, the
winding is multiplex of m separate windings. The common
factor must, however, be smaller than half the number of circuits.
An illustration will bring out these conditions. In the case of
an 8-pole machine with 27 slots, 54 commutator bars, 4 Con-
ductors per slot, and a commutator pitch of 13, the two series
windings are singly closed forming one winding, since 54 and 1 3

20 ARMATURE WINDING AND MOTOR REPAIR

have no common factor. For a machine of 4-poles, with 20 slots,
40 commutator bars, 4 conductors per slot, commutator pitch of
18, the winding may be made up of two series windings, the
values 40, 18 and 4 have a common factor of 2. This makes the
winding doubly closed and consists of two series windings with
each in turn consisting of two other series windings, each of the
latter forming one continuous winding.

5. For large machines using series-parallel windings, it has been
found advisable in most cases to use equipotential connect-
ions and connect about every fourth or eighth commutator
section. (See heading " Equipotential Connectors" on page
21.)

6. The number of current collecting points or brushes on the com-
mutator is equal to the number of armature sections in parallel.

The following formulas apply in the case of a series-parallel
winding :

Front pitch = yi
Back pitch = y 2

Winding pitch = y = y l + y 2 =

Commutator pitch = y k = -

N
Number commutator bars = K = -^

Where p is number of pairs of poles and N is the number
of coil sides in the winding. N must be such a value that

N 2a.

- is a whole number.
P

Symmetrical Windings. All armature windings should
be made symmetrical if possible. A winding is symmetrical
when tLe total number of slots divided by one-half the number
of armature sections in parallel is a whole number. If this
is not the case the emfs induced in different circuits will pro-
duce circulating currents. A winding is also symmetrical
if the number of field poles divided by the number of armature
sections in parallel is a whole number or the number of com-
mutator bars is divisible by one-half the number of armature
sections in parallel. To be sure that any winding is symmetri-
cal all three of these conditions should be fulfilled.

DIRECT-CURRENT WINDINGS 21

Possible Symmetrical Windings for D.-C. Machines of
Different Number of Poles. In a discussion on armature
windings, Stanley Parker Smith (London Electrician, June 16,
1916) has selected the following as the most suitable windings
for machines of different number of poles:

1. For a two-pole machine either a two-circuit lap or wave winding
can be used. The former is usually preferred.

2. For a four-pole machine either a two-circuit wave or a four-
circuit lap winding can be used.

3. For a six-pole machine we are limited to the two-circuit wave
and six-circuit lap windings, since the four-circuit wave wind-
ing (a = 2, p = 3) is unsymmetrical. This is probably the
most objectionable restriction of all, because so many cases
arise where four circuits are desirable in a six-pole machine.

4. For an eight-pole machine wave windings with two or four
circuits and an eight-circuit lap winding can be used. This
is the first instance of a symmetrical wave winding with
a > 1< p, and can be used with great advantage in many cases.

5. For a 10-pole machine there are again only two symmetrical
windings the two-circuit wave and 10-circuit lap windings.

6. For a 12-pole machine the possibilities are much greater, for
wave windings with two, four or six circuits and a lap winding
with 12 circuits can be used.

7. The list can be continued further, if desired, but it is plainly
seen that certain very important advantages are open to the
designer by making use of wave windings with more than two
circuits in 8, 12 and 16-pole machines. It is necessary,
however, to observe the other conditions of symmetry, namely
s/a must be a whole number and idle coils must be avoided.

Equipotential Connectors (Equalizing Rings and Phase
Tappings). In a symmetrical winding, that is, a winding
with identical K/a phase systems, there are always a coils
at the same potential, and these can be joined together if
desired. If no dissymmetry whatever were present, however,
there would be little object in making such connections, unless
they are needed as phase tappings to obtain an alternating
pressure. Actually there are many cases of dissymmetry in
a machine, apart from those due to the arrangement of the
winding already mentioned. Thus, the magnetic material
may not be uniform, the pole-shoes may not be properly

22 ARMATURE WINDING AND MOTOR REPAIR

spaced, the gap may not be uniform, and so on. In general,
perfect symmetry must be regarded as an unattainable ideal
in practice according to Mr. Smith, who calls attention to the
following conditions:

Owing to these dissymmetries, the pressure induced in the
several armature circuits varies, and causes equalizing currents
to flow through the brushes. These equalizing currents, if of
sufficient strength, produce sparking and in any case, load the
brushes in an undesirable manner. It is interesting to note
that, even when no equalizing rings are present, the positive
and negative collecting rings act as equalizers, and the tend-
ency of these equalizing currents is to neutralize the inequali-
ties in the magnetic field to which they are due. When
equalizing rings are used, however, large equalizing currents
will flow along these and strongly damp out any inequalities
in the field, and so reduce the difference of potential between
corresponding points in the winding. Consequently the
brushes are relieved, and are so much better able to perform
their proper function of collecting the current.

Equalizing rings must not be regarded as in any way es-
sential, and many machines work quite well without them.
Nevertheless, they add a certain factor of safety which the
manufacturer is often glad to purchase at so small a cost, for
he is not only surer of his machine passing the test satis-
factorily but also knows that after-effects, like wear of the
bearings, cannot give rise to such serious trouble as when equal-
izers are absent. Consequently, equalizers are seen on many
large machines with lap-wound armatures, or with wave
windings with more than two circuits. Equalizing rings
should not have an extremely low resistance. Such practice
requires not only an excessive amount of copper, but leads to
considerable loss and heating in the winding. All that is
really necessary is to provide an alternative path of negligible
resistance compared with that of the brushes, and for this
purpose it is usually sufficient to make the section of the rings
about half that of the conductors.

Regarding the number of equalizing rings, much depends
on the opinion of the designer. With lap windings, one ring
for every 6 to 12 segments is common practice, but this is

DIRECT-CURRENT WINDINGS 23

scarcely feasible when p > a, for here the potential pitch, 1
y p = K/a, may be fairly large and the number of rings be-
comes prohibitive. In such cases (wave windings with more
than two circuits), one ring for every 15 to 20 segments may
suffice. There is no need to make the pitch between all the
rings the same, but designers generally prefer to split up y p
into a whole number of parts. In this case the tappings form
a symmetrical polyphase system of pressures.

Best D.-C. Windings for a Repair Shop to Use. To secure
good commutation and eliminate heating troubles, all windings
should be made symmetrical. This among other things
means that there must not be more armature circuits in paral-
lel than there are field poles. Where the number of armature
circuits equals the number of poles either the lap or the wave
winding may be used. The lap winding is however mostly
used in such a case. The wave winding then need only be
considered in those cases where the number of armature circuits
is to be less than the number of field poles. A single winding
is more simple and should be used in preference to a compli-
cated one where conditions permit either a single winding
or two or more separate windings except in those cases where
a very heavy current is to be carried in the armature. In
general then, where changes in armature windings are neces-
sary that call for a choice of types of windings, it is advisable
to employ either a lap winding with as many circuits as there
are poles or a wave winding with two circuits. Usually a lap
winding is possible if the number of coils is a common multiple
of the number of poles and the number of slots. The number
of coils is equal to the number of commutator bars.

Number of Armature Slots. The open slot with parallel
sides is most used for large armatures so that form wound
coils can be easily assembled. These coils are held in place
by wire bands or by wood or fiber wedges driven into grooves
at the tips of the teeth as shown in Chapter III. The number
of slots per pole is usually not less than 10. For multipolar
armatures there are at least from three to four slots in the space

1 The potential pitch is the number of commutator bars between suc-
cessive equipotential points in the winding which can be joined together
when equalizing rings are needed.

24 ARMATURE WINDING AND MOTOR REPAIR

between pole tips. In the case of high speed machines having
large pole pitch there may be from 14 to 18 slots per pole.
For machines above say 5 hp. the area of the slot will approx-
imate one square inch. A rough rule for the capacity of
a slot of this area is about 1,000 amp. turns for machines
under 500 volts.

Voltage between Commutator Segments. For direct-cur-
rent machines without interpoles, about 15 volts between
segments of the commutator is the maximum allowed while
double this value can be considered as the maximum for ma-
chines employing commutating interpoles. This value of
voltage between segments can be obtained for an existing
machine by measuring this voltage between plus and minus
brushes and dividing by the number of commutator segments
between these brushes. The voltage between a positive and
nagative brush will depend upon the number of inductors (wires
making up one side of a coil) connected in series between the
brushes. For calculating windings for machines, Prof. Still
gives the following relationships between machine voltage
and volts between commutator segments for good operation:

VOLTAGE BETWEEN SEGMENTS OF COMMUTATORS

Machine voltage

Volts between commutator segments

110

Ito6

220

2.5 to 10

600

5 to 18

1200

9 to 25

Number of Commutator Bars.^-The proper number of
commutator bars for a particular winding depends on the volt-
age between commutator bars. The number of bars may be v
a multiple of the number of slots. For low voltage machine
there may be one, two or three bars per slot while in higher
voltage slow speed machines, there may be as many as four
or five bars per slot. While improved commutation results
from a large number of commutator bars the difficulty in
repair and assembly from a mechanical standpoint offset
their advantage over fewer bars for the same conditions of
winding and machine operation.

DIRECT-CURRENT WINDINGS

Usual Speeds and Poles for Different Sizes of Generators.

For direct current generators Prof. Still 1 gives the following
relationship between number of poles and speed for different
ratings of machines. This table represents usual practice
by designers.

NUMBER OF POLES AND USUAL SPEED LIMITS OF D.-C. GENERATORS

Output in kw.

Number of poles

Speed in rpm,
maximum and minimum

Oto 10

2400 to 600

10 to 50

1300 to 350

50 to 100

4 or 6

1100 to 230

100 to 300

6 or 8

700 to 160

300 to 600

6 or 10

500 to 120

600 to 1000

8 or 12

400 to 100

1000 to 3000

10 or 20

200 to 70

Safe Armature Speeds. The safe speed of an armature
varies with the armature construction. Direct-current ma-
chines are built with a peripheral speed of 2,500 to 3,500 feet
per minute. It is not advisable to exceed an armature periph-
eral speed of 6,000 feet per minute in machines which are
not designed to take care of the mechanical stresses incident
to the higher speeds.

1 Principles of Electrical Design, page 81.

CHAPTER II
ALTERNATING -CURRENT WINDINGS

In the main, the windings for alternating-current motors
and generators are alike. For this reason details of the wind-
ings that can be used for either machines will be given. It
should be noted at this point that direct-current and alter-
nating-current windings differ essentially by the former being
of the closed-circuit type while most alternating-current wind-
ings are of the open-circuit type. Either open- or closed-
circuit types of windings may however be employed in alter-
nating-current machines, but the open type is in most common
use. By a closed winding is meant one which has a continuous
path through the armature and re-enters itself to form a
closed circuit. Such a direct-current winding always has at
least two current paths between brushes. In the case of an
open-circuit alternating -circuit winding, there is a continuous
path through the conductors of the coils of each phase of the
winding, with the ends of this path forming two free ends.
Such a winding does not close on itself.

In the closed-circuit windings of direct-current machines,
the bars of the commutator are simply connected at equally
distant points around the winding. In the open-circuit wind-
ing of an alternating-current generator with revolving arma-
ture, the terminals of the completed winding are connected to
collector rings and the winding is open-circuited until closed
by the connections between the brushes. Closed-circuit wind-
ings are only used for special alternating-current machines.
The conditions which usually call for a closed-circuit wind-
ing are the following:

1. An alternating-current machine which must deliver a large
current at low voltage. In such a case the winding usually
consists of several similar paths connected in parallel to the

ALTERNATING-CURRENT WINDINGS 27

terminals of the armature, thus forming one or more closed cir-
cuits within the armature winding.

2. In designs of machines that must handle direct and alternating
current, as in double current generators and in rotary converters.

In general it may be said that a direct -current winding may
be changed so that the machine can be used as an alternator,
but an alternating-current winding cannot be used for a direct
current generator since it is not reentrant.

Types of A.-C. Windings. With reference to the arrange-
ments of coils used in an alternating-current armature, wind-
ings may be divided into two general classes as follows:

I. Distributed Windings.

1. Spiral or chain.

2. Lap.

3. Wave.

II. Concentrated Windings.

1. Lap.

2. Wave.

Distributed Windings. An armature winding which has
its inductors of any one phase undigr a single pole placed in
several slots, is said to be distributed. When these inductors
are Dunched together in one slot per pole, per phase, the
winding is called concentrated. It is usual in a distributed
winding to distribute the series inductors in any phase of the
winding among two or more slots under each pole. This
tends to diminish armature reactance and gives a better emf .
wave, besides offering a better distribution of the heating due
to armature copper loss than in concentrated windings.

Concentrated Windings. The uni-slot or concentrated
winding gives the largest possible emf from a given number of
inductors in the winding. That is for a definite fixed speed
and field strength in an alternator, the concentrated winding
requires a less number of inductors than a distributed winding,
but increases the number of turns per coil.

Spiral or Chain Winding. In this winding as shown in
Fig. 14 there is only one coil side in a slot. An odd or even
number of inductors per slot may be used but several shapes
of coils are required since the coils enclose each other and must

ARMATURE WINDING AND MOTOR REPAIR

have special end shapes to clear each other. This arrange-
ment, however, makes possible good insulation of end con-
nections through adequate separating air spaces in high vol-
tage machines. The number of coils required in this winding

D -

f j

y,,

(

1 -

FIG. 14. A spiral or chain winding for a 2-phase, 4-pole machine with
the coils of one phase in place and connected. The coils of the other phase
go in the slots shown by the full lines inside the coils.

is also small compared with other windings. This type of
winding is mainly used in alternating-current generators.

Lap and Wave Windings.- Both distributed and concen-
trated windings make use of lap and wave connections. These

FIG. 15. Single-phase lap winding for the same conditions as the wave
winding in Fig. 16.

FTG. 16. Single-phase distributed wave winding with two slots per pole per
phase and one coil side per slot.

arrangements are in principle the same as used in direct-current
windings. (See Chapter I, pages 9 and 16.) The diagrams of
Figs. 15 and 16 show single-phase distributed lap and wave

ALTERNATING-CURRENT WINDINGS

windings for a four-pole armature, having two slots per pole
with one inductor per slot.

In a double-layer lap or wave winding for an alternating-
current armature, the coils are usually of the same general
shape as used in direct-current windings. Instead of con-
necting the terminals of the coils to a commutator, they are
connected together in a definite order (see Chapter XI, page
288) for each phase. The phase windings are then connected
together in star or delta as shown in Fig. 22. In that diagram
a three-phase wave winding is shown for a four-pole machine.
In the double-layer winding, the number of conductors per

FIG. 17. At the left is shown the appearance of one formed coil in a 2-
layer winding. One side of the coil lies in the top of the slot and the other
side in the bottom of the slot. The top side bears an odd number and the
bottom side an even number. The coil pitch in this case is 6 slots or 13
winding spaces. At the right the method of connecting coils for a 2-layer
winding using form coils is shown. The finish (F) of one coil is joined
to the start (S) of the next.

slot must be a multiple of two. It lends itself to a variety
of connections, particularly to a fractional pitch lap winding
where the two sides of a coil are not similarly placed in respect
to the center lines of the poles. The double-layer winding
is not, however, as well suited for high voltages as the single-
layer winding. For this reason many water-wheel types of
generators have been built with a single-layer winding, which
in its most common form, is known as the spiral or chain
winding.

When laying out or changing a double-layer winding, it
is usual to assign odd numbers to the sides of coils in the top
of the slots and even numbers to the sides in the bottom of
the slots. This is important when the pitch of the armature.

30 ARMATURE WINDING AND MOTOR REPAIR

coils is expressed in terms of coil sides (winding spaces) in-
stead of slots.

Whole-coiled and Half-coiled Windings. When the coils
of an alternating current winding are connected so that there
are as many coils per phase as there are poles, the winding is
called "whole-coiled." When the coils are connected so that
there is only one coil per phase per pair of poles, the winding is
called " half-coiled." The main difference between these two
connections is in the method of making the end connections

A Whole-coiled winding B Half-coiled winding

FIG. 18. A 6-pole stator with whole-coiled and half-coiled windings.
The whole-coiled winding, A, has as many coils per phase as there are poles.
The half-coiled winding B has only one coil per phase per pair of poles.

for the coils. In the " whole-coiled " winding each slot con-
tains two coil sides. It is not, however, strictly a double-
layer winding, as the coil sides are placed side by side and
not one above the other. In the " half -coiled" winding, how-
ever, each coil may have twice the number of turns of a " whole-
coiled" winding or the two coils under a north or south pole
of the latter type may be connected in series and taped to-
gether to form one coil in case of a change in connections.

The " half -coiled " winding has the advantage that, when
used with large generators the armature frame may be split
into two sections for shipment or repair, without disturbing
many of the end connections.

Single-phase and Polyphase Windings. The winding of
a single-phase motor or generator has only one group of induc-
tors per pole, placed in one slot or several slots depending
upon whether or not the winding is concentrated or distrib-

ALTERNATING-CURRENT WINDINGS

FIG. 19. A simple single-phase winding.

FIG. 20. Simple 2-phase winding.

FIG. 21. Simple 3-phase winding.

FIG. 22. A 3-phase winding showing how it may be connected in delta

or star.

32 ARMATURE WINDING AND MOTOR REPAIR

uted. Such a single-phase concentrated wave winding for
a four-pole armature is shown in Fig. 19.

Two-phase and three-phase windings may be considered
as made up of single-phase windings properly placed on the
same armature. For the two-phase windings two separate
single-phase windings are used spaced 90 electrical degrees
apart. This is shown in Fig. 20. For the three-phase wind-
ing, three single-phase windings are used, spaced 120 degrees
apart, as illustrated in Fig. 21. Although the single-phase
windings are independent of each other, their terminals are
connected in star or delta as shown in Fig. 22.

Coil Pitch. In the case of a two-phase winding, the total
number of slots should be just divisible by two so that each
phase will have the same number of winding elements or
coils per pole. In the same way, for a three-phase winding
the total number of coil sides or the total number of slots
should be just divisible by three (the number of phases)
and sometimes by the number of poles. This will result in
a full pitch winding, that is, a winding in which a coil spans
exactly the distance between the centers of adjacent poles.
If the coil spans less than this distance, so that its two sides
are not exactly under the centers of adjacent poles at the same
time, it is said to have a fractional-pitch. When a fractional-
pitch is used in alternators on account of the electrical factors
of the design, such as to secure as nearly as possible a sine
wave shape of emf, the total number of slots per phase must
be a whole number. A fractional-pitch is also widely used
in induction motors.

Coil pitch is expressed as a fraction of the pole pitch, in
slots, in electrical degrees or in winding spaces (coil sides).
In the case of a six-pole machine having 72 stator slots, and
a double-layer winding, the pole pitch would be 12 slots. If
the coil pitch were given as %, this would be 120 degrees
or eight slots or 13 winding spaces (coil sides). A full coil
pitch for this winding would be 180 degrees, 12 slots or 21
winding spaces.

Phase Spread of Windings. The spread or space occupied
by each single-phase winding is known as the phase spread
of the winding. For a two-phase winding the phase spread

ALTERNATING-CURRENT WINDINGS 33

is (180 -r- 2) or 90 degrees. For a three-phase winding, it is
(180 -T- 3) or 60 degrees. In a single-phase winding, the phase
spread is theoretically 180 degrees. Prof. Alfred Still points
out, however, in his book on " Principles of Electrical Design, "
that nothing is gained by winding all the slots on the armature
surface of a single-phase machine. After a certain width of
winding has been reached the filling of additional slots merely
increases the resistance and inductance of the winding with-
out any appreciable gain in the developed voltage. In prac-
tice only about 75 per cent, of the available slot space is
utilized making the phase spread for a single-phase winding
about 135 electrical degrees.

The fact that, in polyphase machines, the whole of the arma-
ture surface is available for the winding, while only a portion
is utilized in a single-phase alternator, accounts for the fact
that the output of the latter is less than that of the polyphase
machine using the same size of frame. In a three-phase
machine it is only necessary to omit one of the phase windings
entirely and connect the two remaining phases in series to
obtain a single-phase generator. Such a modified generator
will give about two-thirds of the output of the polyphase
connection. A three-phase star connected induction motor
can also be used as a single-phase motor by properly con-
necting two phases of it (see Chapter XI).

Two-phase from Four -phase Windings. In many cases the
two-phase induction motor is designed as a four-phase machine
with the connections between conductors of the winding
arranged so as to permit operation on a two-phase supply
circuit. As shown in Fig. 23, a two-phase winding maybe
secured from a four-phase grouping of coils by connecting the
first and third groups in series and the second and fourth
groups in series.

Three-phase from Six-phase Windings. Few strictly
three-phase induction motors are built. The design may be
more properly called a six-phase winding with the three phases
' spaced 120 electrical degrees and the connections of coils
such as to permit the motor to be operated on a three-phase
circuit. As shown in Fig. 24 in a six-phase winding the coils
of the six-phases are spaced 60 electrical degrees. For three-

ARMATURE WINDING AND MOTOR REPAIR

phase operation the coils of phases one and four, three and
six, and five and two are connected in series. The terminals
of phases two, four and six are connected to a common point.

FIG. 23. Connections for four windings spaced 90 electrical degrees apart
to secure a 2-phase motor. There are really four phases shown, the first and
third and the second and fourth are connected in series. The full lines are
front connections of coils and dotted lines the back connections. $ and F
indicate the start and finish of the different groups of coils.

FIG. 24. Connections for a 6-phase winding design so that a 3-phase
winding is secured. The six windings are spaced 60 electrical degrees
apart. In this diagram phases one and four, three and six, and five and two
are connected in series and the terminals F 2 , F4 and Fe jointed to a common
point. The 3-phase leads are Si, S 3 , and SB. The full lines indicate
front connections and dotted lines back connections. S and F indicate the
start and finish of the different groups of coils.

Wire, Strap and Bar Wound Coils. For the coils used in
small motors round insulated wire is most employed. These

ALTERNATING-CURRENT WINDINGS 35

coils are either wound in the slots by hand or assembled by
use of specially formed coils wound in forms and insulated
before being placed in the slots. Such formed coils are usu-
ally used except in cases where the slots are closed or nearly
closed. For further information on formed coils see page 4,
Chapter I, and page 141, Chapter VI.

For large motors and generators where the amperes to be
carried in each armature circuit is a large value, copper straps
are frequently employed for making up the armature coils.
In very large machines a copper bar is used instead of the
copper straps. In such a case one bar serves as the inductor
of a coil having one turn per slot. A two-layer bar winding
made up of two bars per coil and four bars per slot is also
used. The copper bars are connected to the end connections
of the coils by brazing, welding or bolting. In all cases,
whatever the construction of the coil used, the slots must be
properly insulated with fullerboard, mica, fish paper or other
suitable insulating material. For data on slot insulation see
Chapter VII.

The current density in alternating-current windings is
about 2500 amp. per square inch of armature conductor in
small machines, 2000 amp. in medium sizes and 1500 amp. in
high-voltage designs. Except in high-speed machines it is
not safe to use the maximum limit owing to damage to wind-
ings from over heating.

METHODS FOR LAYING OUT AND CONNECTING ALTERNATING-
CURRENT WINDINGS

In rewinding an alternating-current machine, the number of
slots on the stator, the operating voltage, speed, phase and
frequency of the supply circuit are points that must be con-
sidered in laying out a new winding or reconnecting an existing
one. The fundamental requirements of windings and the ways
in which they can be fulfilled as outlined by M. W. Bartmess
(Electric Journal, Vol. VIII, No. 5) are given in what follows.

Group Windings. Group winding may be defined as that
class wherein the total winding is divided into separate parts,
composed of adjacent coils or conductors. The grouping is,
in the case of lap and wave windings, an arbitrary one, the

ARMATURE WINDING AND MOTOR REPAIR

coils being all similar and divided into groups solely by their
connections. The number of coils per group may equal the
number per pair of poles divided by the number of phases, or
the number per pole divided by the number phases. The latter
method of grouping is generally used on modern machines.
In the case of a six-pole, three-phase winding of 36 slots, the
number of coils per group is 36 -j- (6 X 3) or 2. When the
number of slots is not evenly divisible by the product of poles

FIG. 25. Methods of connecting pole-phase groups shown in (a), (6) and (c).
(a) Four-pole winding with alternately positive and negative pole-phase groups,
(fc) Four-pole winding of the consequent pole type, (c) Two-pole winding obtained
from (b) by reconnecting the pole-phase groups alternately positive and negative.

and phases, dissimilar groups must be employed. In such
cases it is advisable to arrange the grouping so that all the
phases have an equal number of coils, and if possible the group-
ing should be arranged symmetrically with respect to the core
itself. To prevent local currents, which may prove injurious,
all circuits which are in parallel must have an equal number of
coils and should be symmetrically arranged with respect to
each other and to the other phases.

Although one turn coils only are shown in Figs. 27 to 29,
the same connections are applicable to windings having

ALTERNATING-CURRENT WINDINGS 37

any number of conductors per coil. These conductors may
all be in series, in which case there is one lead at each end of the
coil or the conductors may be divided into any number of equal
parallels, in which case there are as many leads at the ends of
the coils as there are parallel circuits. The leads at the
beginning and end of the coils are connected in the same man-
ner as indicated for the one turn per coil winding. For the
sake of simplicity the number of coils per group and hence the
total number of coils in the diagrams has been kept lower than
is generally found in commercial machines.

Full and Fractional Pitch Windings. The number of
slots in the core, divided by the number of poles gives a value
of the pole arc expressed in terms of the slots. A full pitch
winding is one in which the effective span of the coils is equal
to the pole arc, and a fractional pitch winding is one in which
the effective span of the coils is not equal to the pole arc. For
a two coil per slot, lap or wave winding, the effective span of the
coil is equal to the actual span of the coil. In this case the
full pitch winding is one where the coil throw is equal to

/total number of slots , A ^

I r e~ r~ ~ plus 1 ) . For a one coil per slot lap

\ number of poles /

winding the effective span of the coil may be greater or less than
its actual span. In Fig. 26 (a) and (6) show two different coils,
in each of which the effective span is the full pitch of 12 slots
while the actual span in (a) is only 11 slots and that in (6)
is 13 slots. Needless to say, (a) is more generally used on
account of the saving in copper and space for end connections.
A coil with a span either less or greater than that shown
would result in a fractional pitch, as in (c) and (d) .

Representative cases of concentric group windings are
shown in Fig. 26, (e) and (f), (e) representing a three-bank
winding, in which the number of coils per group equals the
total number of coils per phase divided by the number of poles,
while (f) represents a two-bank winding of the consequent
pole type in which the number of coils per group equals the
total number of coils per phase divided by the number of pairs
of poles. Neither of these types can be conveniently wound
with a fractional pitch, especially with formed coils. Where
dissimilar groups are employed, that is where the number of

ARMATURE WINDING AND MOTOR REPAIR

slots is not evenly divisible by the product of phases and poles,
the full pitch is frequently not a unit and hence a fractional
pitch is necessary. In the case of a three-phase, four pole
winding with 30 slots, 30 coils, two coils per slot, full pitch
covers a span of 7.5 slots and the nearest lower even pitch gives
a throw of 1-8.

In general, fractional pitch affects the performance of the
apparatus similarly to a reduced number of turns in the wind-

+ + + +O OOO 4 -1-4.4.

(X) (/)

FIG. 26. Possible pitch for one coil per slot windings.

(a) Full pitch, effective space 12, actual space 11, throw 1 to 12. (6) Full pitch,
effective space 12, actual space 13, throw 1 to 14. (c) Fractional pitch, effective
space 10, actual space 9, throw 1 to 10. (d) Fractional pitch, effective space 10,
actual space 15, throw 1 to 16, (e) Concentric group, full pitch, effective space 12,
actual space 9 and 11, throw 1 to 11. (/) Concentric group, full pitch, consequent
poles, effective space 12, actual space 9, 11, 13, and 15, throw 1 to 13.

ing, but not in the same proportion. In a generator this re-
duces the voltage of the machine. In an induction motor, the
maximum available torque is increased but the densities in the
magnetic circuit are also increased with a resulting reduction of
power-factor. For either motor or generator, considerable
copper may thus be saved in the coil ends and a standard
frame may frequently be used for special purposes.

Simple Winding Diagram. It is evident that for all com-
binations the number of diagrams necessary would be unlimited.
A simplified diagram may be employed which will not only
reduce the required number of such diagrams but will also

ALTERNATING-CURRENT WINDINGS

minimize the labor in tracing out the connections. Thus it will
be seen that the diagram in Fig. 27, will satisfy many require-
ments for connections of groups. In addition to this it will
apply for any similarly connected three-phase, four-pole, series
star lap-winding, irrespective of the number of coils per
group (provided the groups are regular) or of the throw of the
coils, that is, whether the winding is full or fractional pitch.
This information for the throw of the coils and the number of
coils per group may be carried
on the same specification with
the remaining winding con-
stants. The groups are formed
by connecting the required num-
ber of coils together, the end of c
the first coil to the beginning of
the second, etc., the beginning
of the first coil and the end of
the last coil in the group forming
the beginning and end of the
group. Such diagrams may be
made for any number of phases,
poles or possible parallel circuits,
and for any desired method of
connection of the groups. In

case the coils per group are irregular or unbalanced, it is advisa-
ble to have a special diagram giving the number of coils in each
group, their location and any other information necessary.

Reconnecting a Winding. It is obvious that if a winding
gives satisfactory operation on a certain voltage, a similar
winding of one-half the number of series conductors, but of
double the current carrying capacity, will give satisfactory
operation on one-half the voltage. This latter condition may
be obtained by paralleling the groups, as in Figs. 28 and 29, or
where this is impossible by paralleling the series conductors in
the slots. For example, if the full voltage connection of a
14-pole motor corresponds to the parallel connection the only
method to change to half voltage would be to change the wind-
ing itself since 14 poles does not permit of a four-parallel
connection. Again an irregularity of coils per group will at

FIG. 27. General winding dia-
gram for a 3-phase, 4-pole motor
with pole-phase groups connected
in series star.

ARMATURE WINDING AND MOTOR REPAIR

times prevent doubling the parallel circuits where otherwise
this might be possible.

When it is desired to use one winding for either full or half
voltage, the winding, if possible, is laid out for equally satis-
factory operation on either connection, and for the minimum
amount of labor required to connect from one to the other.
This is exemplified by Figs. 28 and 29. It is evident that any
eccentricity of the rotor with respect to the stator will affect
equally the circuits which are in parallel.

Simple Method for Indicating Polarity of Coil Groups.
A simple method for obtaining the proper polarity of the groups
is indicated in Figs. 27 and 28 for three-phase and Fig. 29

FIG. 28. Winding diagram for a
3-phase, 4-pole motor with pole-
phase groups connected in 2-parallel
star. This is the winding of Fig. 27
connected in parallel.

FIG. 29. Winding diagram for a
2-phase, 14-pole motor with pole-
phase groups connected in 2 parallels.

for two-phase winding. In a three-phase star winding by
traveling from each of the three leads to the star points, the
direction of travel is reversed in adjacent groups. In a two-
phase diagram the only necessary precaution in determining
the proper polarity is to remember that adjacent groups of the
same phase are reversed. By marking the groups A, B, C,
etc., and indicating the direction of travel, it is a simple matter
to connect them in the proper direction. Additional index
marks may be given on the diagram to aid in connecting the
winding, by marking those group ends which are joined by the

ALTERNATING-CURRENT WINDINGS 41

same connector, with the same numeral. For further details
in using this method see Chapter XI.

Changing Star to Delta Connection. Any star diagram can
be readily changed into a corresponding delta diagram by
opening up the star points and connecting the inner end of
phase A to the outer end of phase B or C, the inner end
of phase B to the outer end of phase C or A, and the inner end
of phase C to the outer end of phase A or B. If the star
diagram is not symmetrical with respect to the three phases
it is never advisable to change over to delta.

A.-C. Wave Windings. In a wave winding, correspond-
ingly placed conductors under adjacent poles are connected
in series, the circuit proceeding from pole to pole one or more
times around the core, and not forward and back upon itself
as in a lap winding. The circuits are then interconnected in
such a manner as to give the requisite phase relations. The
total number of these circuits must be a multiple of the number
of phases and is ordinarily twice the number of phases. Due to
certain limitations, this type of winding is not used to as great
an extent as the lap or the concentric windings. Its use on
small motors is limited to phase-wound secondaries. Since a
two-phase secondary would require four collector rings, or if
connected for a three-wire system, would overload one of the
rings, while a three-phase winding requires but three rings, the
latter only is general for such applications.

Progressive and Retrogressive A.-C. Wave Windings. The
number of slots for a wave winding (plus or minus one) is
so chosen as to be divisible by the number of pairs of poles
or preferably by the number of poles. If plus one, it is said
to be a progressive winding, since after traveling once around
the circuit it returns to the starting slot plus one. If minus
one, it is said to be retrogressive since the circuit returns the
winding to the starting slot minus one. With this arrange-
ment it is impossible to balance the phases exactly, but the
effective unbalancing is small with a large number of slots.
Since an unbalanced three-phase winding is less objectionable
than a two-phase winding, the scheme is used chiefly for the
former. Again, since an unbalanced winding is less detri-
mental in a secondary circuit than in a primary, the principal

ARMATURE WINDING AND MOTOR REPAIR

application of this type of winding has been in secondary
circuits.

The diagram of Fig. 30 represents a two conductor per slot
winding, such as a bar and end conductor type but is equally
applicable to strap or wire wound coils of two or more series
turns per coil, in which case the connector on the rear end
of the coil takes care of itself and the front end is connected

in a manner similar to the
sketch. There are several
methods of connecting up the
three-phases depending on the
desired voltage. The princi-
pal connection is indicated in
Fig. 30. By connecting the
end of each series circuit to
the beginning of the next and
taking off leads to the point
of connection of series 12,

FIG. 30. Wave winding for a 3 ~4, 5-6, instead of the COn-

3-phase, 4-poie motor, having 19 slots, nections between the series

19 coils, two coils per slot, throw 1 to , -n- on

6, with unbalanced phases connected shown in Fig. 30, a COnnec-

in series star. tion is obtained for one-half

First series begins in bottom slot 6, ends vo l tage . An 86 per cent.
in top slot 12. Second series begins bot-
tom slot 17, ends top slot 4. Third series Voltage tap in terms of the

connections shown in Fig. 30
is secured by connecting the

begins bottom slot 9, ends top slot 15,

Fourth series begins bottom slot 1, ends

top slot 7. Fifth series begins bottom

slot 12 ends top slot 18. Sixth series end Q f geries 2 to the begin-

begins bottom slot 4, ends top slot 1. p v'

ning of series 3, the end cf

series 4 to the beginning of series 5, the end of series 6 to
the beginning of series 1. To connect in star join the ends
of series 1, series 3, and series 5, and take off leads at the
beginning series 6, series 2, and series 4. Since this connec-
tion reduces the voltage without increasing the cross section
of the copper the winding will be less efficient on account of
higher copper loss for the same output. This is also true of
the 50 per cent, voltage connection, since for the output the
current density in the windings is 15 per cent, greater than
for the full voltage connection.

It is possible, by choosing a number of slots which is divisible

ALTERNATING-CURRENT WINDINGS

by the product of the number of phases by the number of
poles, to lay out a winding which is balanced. For such a
winding the circuit after passing once around the armature,
returns to the starting slot. It is then only necessary to
supply a special connector to join it to the conductor in the
starting slot plus or minus one. This winding thus embodies
the best features of both types as, for example, an electrical
balance, and a minimum number of special connections, which
means a very compact and easily assembled winding. The
number of special connections is comparatively small with
respect to the number of coils, and this feature is more pro-
nounced as the number of poles is increased. Hence for a
winding for a large number of poles, 12 to 40, the number of
special connections becomes insignificant.

Connections for Coils of Polyphase Windings. 1 A two-
phase winding for a four-pole alternator is shown in the sim-
plest possible form, as a
radial diagram, in Fig. 31.
It is a concentrated wave
winding, having one slot per
pole per phase, full pitch.
Each phase has one element
of winding, or one slot, under
each pole, and the elements
of phase B are distant from
the corresponding elements
of phase A by exactly 90
electrical degrees. The

FIG. 31. Radial diagram of a 2-phase
winding for a 4-pole machine.

whole winding is thus di-
vided into two exactly sim-
ilar halves, electrically dis-
tinct from each other, just like two single phases, their positions
being relatively fixed so that S a passes under middle Ni just
one-quarter period after (or before, depending on direction
of rotation) Sb passes the same point. They are thus tied

1 The following descriptions and illustrations of windings, pages 43 to
51, have been taken by permission from Alternating-Current Electricity
by W. H. Timble and H. H. Higbie (John Wiley & Sons, Inc., New York,
N. Y.).

ARMATURE WINDING AND MOTOR REPAIR

together in phase relation through the magnetic field and
the mechanical distribution of the winding. The slip rings
of phase A would be connected to S a and F a , and the rings
of phase B to Sb and F b .

A three-phase winding for a four-pole machine is shown in

Figs. 32 to 34. The winding
is the simplest practicable one
to make, occupying only one
slot per pole per phase. This
winding is laid out as follows:
First, make sure that the total
number of slots is divisible by
3 (phases) and 4 (poles). Then,
holding the rotor stationary,
mark One element (or slot)
under the middle of each pole

FIG. 32. Radial diagram of a as belonging to phase A, and

connect them properly to-
gether in series so that their
emfs add together when they are in the position where the
maximum instantaneous emf for the whole group is induced
(as shown in Fig. 32). Mark one end of this series S a as in

3-phase winding for a 4-pole ma-
chine, star-connected.

FIG. 33. Developed diagram of the 3-phase winding of Fig. 32, delta-
connected.

Fig. 32, and the other end F a . With the rotor still fixed in
the same position and starting from S a , proceed to count the
slots or coils in one direction around the armature until two-
thirds of those which lie between the middle of adjacent

ALTERNATING-CURRENT WINDINGS 45

poles have been passed. Since the distance between the
middle of adjacent poles is 180 electrical degrees, the dis-
tance passed over is 120 electrical degrees. Label this slot, ele-
ment or coil S b and locate the other three B elements or coils with
respect to each other, exactly as the A elements are related
to each other. Connect the B group in additive series exactly
as the A group was connected, and mark F b on the finishing
end of the B series. Now from $5 continue to count slots or
coils around in the same direction until you have passed as
far ahead of Sb as Sb is ahead of S a . This will be 120 electri-
cal degrees from S& or 240 electrical degrees from S a . Mark
this slot or coil S c? and locate the other C slots or coils in simi-
lar positions with respect to all other poles. Connect the C
slots together in additive series just as the A slots were con-
nected, and put the F c label on the finishing end of the series.
If the emf . in phase A is at its maximum value from S a to F a
at the position shown in Figs. 32 and 33, it is apparent that
the emf S a to F a must pass through one-third cycle or 120
electrical degrees before the emf. from Sb to Fb reaches its
maximum value (or is brought into the same position in the
magnetic field by a counter-clockwise rotation of the rotor).
Also, the emf from S a to F a must pass through % cycle or
240 electrical degrees before the emf from S., to F c reaches its
maximum value. Since these three emfs reach their maxi-
mum value in a direction away from S just 120 electrical de-
grees apart, consecutively, the S ends must be connected
together to neutral, to get a star-connection. The terminals of
the three-phase armature thus connected in star are F a , Fb,
F c , as shown in Fig. 32.'

A developed view of this same winding (three-phase, four-
pole, one slot per pole per phase) is shown in Fig. 33, connected
in delta or mesh. Notice that if only that end of each phase
is marked which is separated by 120 electrical degrees from
the similar end of the preceding phase, the connection be-
comes very simple, because we merely connect the finishing
end of one phase to the starting end of the next phase 120 de-
grees ahead, and so on. Thus connect F a to S b , F b to S c and
F c to S a - These junction points are then the terminals of
the delta winding. The equal emfs (S a to F a ), (S b to Fb),

ARMATURE WINDING AND MOTOR REPAIR

FIG. 34. Three-phase winding
having two slots per pole per phase.
Two phases only are shown.

(S c to F c ) are 120 electrical degrees apart successively in

the same direction through the closed mesh or series.

This relation gives a resultant emf of zero volts around

the mesh.

If the winding is to occupy two slots per pole per phase,

requiring 3 X 4 X 2 = 24 slots altogether, the connections

for a two-layer lap winding are
shown in Fig. 34. There are as
many coils as slots, and all
coils are exactly alike. Each
slot contains two coil-sides (Fig.
35). The one at the bottom is
the right-hand side of a coil
lying to the left of the slot, and
the one at the top is the left-
hand side of a coil lying to the

ri ht of the slot ' The elements
at the bottom of the slot are

shown in dotted lines.
Such an arrangement of coils is typical 'of all lap-wound
direct-current machines, and synchronous converters; but
it may be used for any sort of an alternating-current winding,
closed or open. If the similar end of every coil at its starting
end is labeled S, and the other end F (as it is wound up on a
form, for instance), then a closed winding is obtained by simply
soldering the F of one coil to the S of the one lying in the next
slot, and so on all around the armature until the last F is
soldered to the first S. This closed winding could be tapped
at equidistant points, depending on the number of poles.
Or this closed winding could be opened up at two or more
points and the parts connected in series as an open winding for
any number of phases. The winding of Fig. 35 may be
recognized for a four-pole stator, because each coil spans one-
quarter of the circumference. To get the greater emf in a
coil, its opposite sides should both come as nearly as possible
simultaneously under the middle of adjacent poles. There are
24 coils altogether, for three phases and four poles, which allows
two coils per pole per phase. These two will, of course, be adja-
cent coils, in order that their emfs shall be as nearly as possible

ALTERNATING-CURRENT WINDINGS

in phase with each other so as to get the greatest possible
resultant emf . from the series.

Double-layer Winding, Lap Connected. In connecting a
winding such as shown in Fig. 35, choose a coil located exactly
under the middle of NI and Si, and label it AI. Mark its
starting end S a . Connect the finish end of coil AI to the
starting end of coil A 2 which is adjacent to coil AI. Now

FIG. 35. TWO layer winding for a 3-phase, 4-pole machine.

F B S * F c

L_^

S 8

FIG. 36. Winding of Fig. 35, delta-connected.

F B S A F c S a S c

999 ? Neutral 9

FIG. 37. Winding of Fig. 35, star-connected.

locate coils A z and A 4> also belonging properly to phase A,
because they are located with relation to pole Si and N 2 jus.
exactly as coils A\ and A 2 are located with relation to pole
NI and Si. Similarly, locate A 5 and A$ under poles N 2 and
2 , and A 7 and A 8 under poles S 2 and NI. Then group A 3
and A 4 in additive series by soldering the finish of As to the
start of A 4 . Similarly, connect A 5 and A G together, and A^
and AS together. Now, since the emf is clockwise around
coils AI and A 2 , counter-clockwise around coils A 3 and A 4,

48 ARMATURE WINDING AND MOTOR REPAIR

clockwise around A 5 and A 6 , and counter-clockwise around
A 7 and A 8 , it will be seen, that in order to get these groups of
coils together into additive series AI, A 2 , A 5 and A 6 must be
connected together similarly but A 3 , A 4 , A 7 and A 8 oppositely.
If the series of coils composing phase A are carefully traced
starting at S a and going right through to F a , it will be seen
that the instantaneous emfs are all in the same direction at
about the same time when the emfs induced in the coils of
phase A are greatest (which is about the position shown in the
diagram).

Phase B has been started at one end, Sb (similar to the end
/S ), of a coil located 120 electrical degrees from coil AI, and
from this point through to F b the connections and arrangement
of coils are an exact duplicate of phase A, except as to actual
position in the magnetic field. Likewise phase C is a duplicate
of phase AI, but S c is located 120 degrees further along in the
same direction from Sb, or 240 degrees from S a - This gives
the six terminals of the three phases all properly labeled. In
Fig. 36 are shown the proper connections between these six
terminals to give a three-phase delta. In Fig. 37 are shown
the connections between the same six terminals to give a
three-phase star.

Connecting a Chain Winding. The chain winding has
been sometimes used by makers for alternating-current genera-
tors, in addition to the two-layer windings. Fig. 38 repre-
sents a three-phase four-pole chain winding, using two slots
per pole per phase on the same 24-slot armature which we have
been using throughout for illustration. In order not to confuse
the diagram, phases A, B and C have been drawn out sepa-
rately, in Figs. 38, 39 and 40. Notice that they are exactly
alike, except as to relative position on the armature. Phase
B is 120 electrical degrees from phase A, and phase C is 120
electrical degrees further in the same direction, from phase B,
or 240 electrical degrees from phase A, the positive direction
of emf in each phase being from the S end to the F end.
When the three phases are assembled altogether as in Fig. 41,
it is seen that there must be a different shape or length of coil
for each phase in order that the ends of the coils shall not
interfere with each other. This is expressed usually by saying

ALTERNATING-CURRENT WINDINGS

that the end-bends of the coils are in three ranges. This is due
to the fact that the winding is a single- layer winding (the

22 23

19 20 21 22

FIG. 38. Phase A of a 3-phase chain winding. Two slots per pole per

FIG. 39. Phase B of the 3-phase chain winding of Fig. 38.

21 22

FIG. 40. Phase C of the 3-phase chain winding of Fig. 38.

FIG. 41. -Three-phase chain winding for which the separate phases are
shown in Figs. 38, 39 and 40. End connections of phase A only are shown.
Three forms of coils are necessary.

coils of a two-layer winding are all exactly alike), and also
because all coils in each phase have been made the same shape.

ARMATURE WINDING AND MOTOR REPAIR

In Fig. 41 the end-connections between coils are shown only
for phase A, to avoid confusion. The system of connections
would be exactly the same as in other figures which are
complete

Other Common Windings. A few other typical forms of
windings are illustrated in Figs. 42 to 46. Fig. 42 is a

FIG. 42. Three-phase bar winding (wave) using one slot per pole per phase,

star-connected.

FIG. 43. Skew winding with all coils alike, each having one side shorter

than the other.

FIG. 44. Short-coil winding with each coil only two-thirds of the pole
pitch. The emf's of the two sides do not add to such good advantage as
in other types.

three-phase bar winding (wave) using one slot per pole per
phase. It may be extended as shown for any number of
pairs of poles and is drawn star-connected. Fig. 43 is known
as a " skew-coil" winding; although there is only one coil-
side in each slot, all coils are of the same shape for all phases.
Conflicts of the coil-ends are avoided by making one side of

ALTERNATING-CURRENT WINDINGS

each coil longer than the other side. Fig. 44 illustrates
what is called a " short-coil winding" for a three-phase ma-
chine using two slots per pole per phase. By making the
breadth of each coil only % of the pole pitch, overlapping of
coils is altogether avoided, and all coils in the entire winding
are exactly alike. The series emfs composing each phase are
not added to as good advantage as in other types of winding
using coils nearer full-pitch, and, therefore, more copper would

FIQ. 45. Creeping winding in which coils have a fractional pitch. Three
coils cover four poles. The small dash lines represent slots left vacant
for clearness.

Mil

1 Mil

HIM

^<^^r^L*

Mill

' ^ < K

L ^-^- _*

INN

M iml

Cl^2S^->

Mill
<^~^^

JIIM

FIG. 46.-

-Single-phase, whole-coiled winding for 8 poles using 3 slots
per pole. Armature has 64 slots.

be needed for the same capacity. The wave-form is also
likely to be more peaked. Fig. 45 shows a " creeping wind-
ing" in which the coils are of fractional pitch and the series of
coils in each phase are arranged so as to gain or lose one or more
poles around the armature. In Fig. 45 three adjacent coils
each spanning 240 electrical degrees, together cover 720 degrees
or four poles. Fig. 46 shows a single-phase whole-coiled
winding for eight poles, using three slots per pole, for an
armature having altogether 64 slots.

Easily Remembered Rules for Arrangement of Coils in an
Induction Motor. The following are rules that can be
easily remembered and cover conditions frequently encoun-
tered in laying out alternating-current windings.

1. Number of coils per pole-phase-group = No. slots -f- (No,
poles X No. phases).

52 ARMATURE WINDING AND. MOTOR REPAIR

2. When the number of slots is not evenly divisible by the number
of poles times the number of phases, dissimilar groups must be
employed. These groups should be arranged so that all the phases
have an equal number of coils. The grouping should also be sym-
metrical, with respect to the core.

3. A full pitch winding is one in which the span of a coil equals
the number of slots divided by the number of poles. That is in a
36 slot, six-pole, three-phase winding, 36 -f- 6 = 6 slots or the arc of the
stator covered by one pole. A coil with a span of 6 slots makes the
winding a full pitch winding. When the coil span is less than this,
the winding is known as a fractional pitch winding.

4. If a particular winding has given satisfaction on any particular
voltage, a similar winding of one-half the number of series conductors
but of double the current-carrying capacity will also give satisfaction
on one-half the voltage. This latter condition can usually be ob-
tained by paralleling the groups of coils. Or when a machine has
two windings connected in series, say for 440 volts, they can be con-
nected in parallel for 220 volts.

5. In a wave winding, correspondingly placed conductors under
adjacent poles are connected in series, and the circuit proceeds from
pole to pole one or more times around the core. The circuits are
then interconnected to give the requisite phase relations. The
total number of these circuits must be a multiple of the number of
phases and is ordinarily twice the number of phases. The number
of slots for this winding (plus or minus one) should be divisible
by the number of pairs of poles or preferably by the number of poles.

"If plus one, the winding is progressive, since after traveling once
around the circuit it return to the starting slot plus one. If minus
one, it is said to be retrogressive since the circuit returns the winding
to the starting slot minus one. To have a balanced wave winding,
the number of slots must be divisible by the product of the number
of poles and phases. This winding returns to the starting slot after
going once around the core arid special connectors must be used to
connect the finish end to the conductor in the starting slot plus or
minus one.

Simple Rule for Checking Proper Phase Relationship in
a Two- or Three-phase Winding. A fundamental considera-
tion when checking the instantaneous flow of current in a three-
phase circuit, is to imagine that when the current flows in
the same direction in two legs of the circuit, it flows in the
opposite direction in the third leg. This principle can be

ALTERNATING-CURRENT WINDINGS

applied to both motors and generators. When this scheme
is applied to an alternating-current winding in checking the
connections of the coils a great deal of experience is required

FIG. 47. Simple scheme of alternately reversing arrows of pole-phase
groups to check correct phase polarity of a 3-phase winding. It is supposed
in this case that current flows in the three leads toward the star points of the
winding which are indicated thus ( *) .

FIG. 48. Developed winding diagram for a 4-pole induction motor showing

series-star connected end connections.

This winding shows the groups of phase coils marked A, B, C with arrows pointing
in opposite directions on adjacent pole-phase groups. It is another way of showing the
connections illustrated in Fig. 47.

in order to make sure that the leads of the three-phases of the
motor are brought out at an electrical angle of 120 degrees
apart. This is usually a separation of two poles.

ARMATURE WINDING AND MOTOR REPAIR

FIG. 50. A 3-phase, 6-pole wind-
ing having three coils per pole-phase
group on a stator of 54 slots. This
winding can be connected for 2-phase
operation with an odd grouping of
coils as shown in Fig. 51.

FIG. 51. A 2-phase, 6-pole wind-
ing on a stator of 54 slots. The
3-phase, 6-pole winding is shown in
Fig. 50.

FIG. 52. A 3-phase, 6-pole wind-
ing showing the arrangement of odd
groups of coils. The groups of two
coils are located diagonally opposite
each other. The small arcs num-
bered I to VI each span one pole
containing 8 coils. The stator in
this case has 48 slots.

ALTERNATING-CURRENT WINDINGS 55

A simple and more reliable method for the repairman to
use is shown in Fig. 47. In this scheme it must be supposed
that current flows in all three leads of the star connection to-
ward the point of the star connection. And that in the case
of a delta connection the current flows around the three sides
of the delta in the same direction. Then in either case for
a three-phase winding, the polarity of each of the pole-phase-
groups will alternate regularly around the winding and can
be indicated by arrows as in Fig. 47. As shown, for a three-
phase winding there will be three times as many pole-phase-
groups as there are poles. By the use of this scheme there is
no chance for a reversal of a phase to be passed by not noticed
when checking the winding.

In adopting this scheme for a two-phase winding it need
only be remembered that both groups of coils in each phase
must reverse alternately, that is, they should be so indicated
on the diagram by arrows, as shown in Fig. 29, page 40.

By marking each of the pole-phase groups on a diagram,
A, B, C, to indicate a complete phase-group, and placing the
arrows on each single group as shown in Fig. 47, the armature
winder will have little trouble in understanding the diagram
and making the proper connections. To make it still easier,
the group ends that are to be joined by the same connector
can be marked on the diagram with the same number.

The use of this method for regrouping coils of a three-
phase winding for operation on a two-phase circuit and for
any odd grouping of coils is shown in Figs. 50 to 52.

CHAPTER III

REPAIR SHOP METHODS FOR REWINDING
DIRECT-CURRENT ARMATURES

Dismantling a D.-C. Armature. When an armature is to
be entirely rewound, it must be stripped and reinsulated
throughout. First remove the wedges in the slots when
banding wires are not used. If there are banding wires,
remove them by filing in two parts. When using a hammer
and chisel in cutting banding wires be careful not to mash the
armature teeth out of shape. Now look over the connec-
tions to the commutator and the end connections of the coils
to determine whether a lap or a wave winding was used. In
a lap winding with formed coils, the start and finish terminals
are connected to adjacent commutator bars while in the wave
winding the terminals of any coil are a considerable distance
apart around the commutator. This distance is approxi-
mately equal to the number of commutator bars divided by
one-half the number of poles. Also in a lap winding the end
connections of the coils at both ends of the armature bend
toward the center of the coil or in the same direction, while
in a wave winding they bend in opposite directions.

The wave winding usually has either two or four sets of
brushes. Two sets of brushes only are needed regardless of
the number of poles in the machine since there are only two
current paths in parallel through the armature winding
(see Fig. 13). In the lap winding there are as many paths
for the current as there are poles in the machine (see Fig. 10)
and there are usually as many sets of brushes as there are
poles. Simply because there are only two brush holders on a
multi-polar machine, however, does not always mean that a
wave winding is used. The commutators of some few lap-
wound machines have internal cross-connections so that the
commutator bars of the same potential are joined together.

REWINDING DIRECT-CURRENT ARMATURES 57

This permits the use of only two sets of brushes. In such a
case an extra long commutator is used which will serve as an
indication of such connections.

Winding Data Needed for the Dismantled Armature.
After having made an inspection of the winding, enter the
following data in a note book:

1. Number of armature slots.

2. Number of coil sides per slot.

3. Number of commutator bars when there are two coil
sides per slot, the number of bars will equal the number
of armature slots in both lap and wave windings.

4. Coil throw in slots.

5. Commutator pitch.

6. Number of turns per coil and size of wire used.

A convenient chart and diagram for recording these data are
shown in Figs. 53 and 54 as used by a large manufacturer. In
case new coils must be ordered, this is the information that the
manufacturer needs. It is a good plan to have a note l^ook
with duplicate printed pages so that the record made on the
first page can be transferred to the next by apiece of typewriter
carbon paper. The first sheet can then be given to the
armature winder and the copy kept in the book as an office
record for use in case the winder destroys his copy, for use in
making up a bill for the job and for reference in case further
repairs should be later needed on the same machine. Such a
record of repairs often helps in the location of new trouble in a
repaired machine.

Removing the Old Coils. The next operation is to unsolder
the leads from the commutator, and proceed to remove the
coils. This can be done usually by raising the top sides of
the coils for a distance of the coil throw, when the bottom side
of a coil can be reached and the others taken out one after the
other. In removing the coils, try and preserve one in its
original shape, to use as a guide for winding new ones. Enter
in the note book the number of turns per coil, and with a wire
gauge determine the size of wire. Also note whether the wire
is single or double cotton covered. If the commutator contains
twice as many bars as there are slots on the armature, then

ARMATURE WINDING AND MOTOR REPAIR

each armature coil will contain two coils taped together.
Some armatures, especially those designed for low speed, may
contain three or even four times as many bars as slots. By

ARMATURE WINDING AND CONNECTING DIAGRAMS

Eig.A

K. W.
H. P.~~

Serial No.

Volts

Xo. Armature Slots

No. Commutator Bars

Refer to Figure

Throw of Coils, A-B

Throw
of Leads

Left, E-F.

Total, C-F.
Right, C-D

Remarks.

FIG. 53. Convenient form for use in recording winding data when new coils
must be ordered from the manufacturer.

this construction, the inductance and the current per coil is
kept low, and the voltage between bars reduced.

When the armature has been stripped, all old insulation
should be removed from the slots by scraping and burning with

REWINDING DIRECT-CURRENT ARMATURES

Electrical Specification
For D.C.Xachinei

Armature- CoU same asLNo

She of strand Material Arrangement
D.UCXRlbbon Wide Deep
D.C.U.Ku.\\ire

Method of Coil Layout

B.C. C. Ribbon

D.C.C.Kd.Wlre
D.C.C.8q.Wlre X-

Bare Strap
__i Turns per single coil

_J Single coils per cofl

12 Total Strands per Coil"!

_!___ Wide X ___L__deiJ
-Silots-Two Ooih per Slot
-Com. Bars

i ouptd Ends-Strap Commutation Coil Drilled Ends
Frame side JramejiJe_

Numbering of Slots and liars to bo frju, right to left

Coil In bottom of dot
Ko.l

Connect this single
Coil to bar No.1

M. Btry,,. j.

tfr "H

Wave winding
progressive
retrogressive
with dead coil
rlthout dead coll

bottom of slot No.l

Connect firet act he
single coil on rurht
of slot No 1 to bar

- 1 Center punch marks on \
ends of com bars >'

bottom of slot No.1

Bar No. I Bar No.
Bar No. i. | . Bar No. Bar No

Bar No.

.....Number of special coils with right hand leads stretched one bar
batween half idle bmr No and bar No

Fia. 54. Diagrams for recording coil throw and commutator throw when

stripping an armature or when ordering new coils.

These diagrams are useful for checking up a wave winding that must duplicate the
original one. In this case center punch marks are made on the commutator bars and
slots to indicate the coil throw and commutator connections of one of the orignal
coils as indicated in section (E) of the illustration. Here two punch marks indicate
terminal for coil side in bottom of slot and one punch mark for the terminal of the
coil side in the top of the slot. Two crosses (XX) indicates the slot in which the coil
side is at the bottom and one cross (X) the slot in which it is at the top. When the
center line of the coil is marked on the drawing to show whether it falls on a commutator
bar or a mica, the commutator can be removed and replaced to line up with the original
markings. For other details of determining the commutator throw, especially with
dead coils, see "Practical Method for Locating First Connection to Commutator for a
Wave Winding" on page 105. Section (D) of the above illustration shows the use of the
diagram for marking the coil throw of a lap winding. Section (A) is filled out to in-
dicate its use in specifying new coils for a wave winding.

60 ARMATURE WINDING AND MOTOR REPAIR

a torch and each slot filed to remove any burrs or rough places.
Then clean the core thoroughly with a blast of compressed air.
The core is now ready for its insulation and a new winding.
Details are given in the following paragraphs for insulating
and winding the different types of armatures according to the
types of slots used and the requirements of different sizes and
types of machines.

I. WINDING D.-C. ARMATURES HAVING PARTIALLY CLOSED

SLOTS

Direct-current motors of the industrial type in sizes of
from one to five horsepower are often built with partially
closed slots. When winding the armature of such a motor, a
wire-wound, threaded-in coil is used. Practical details for
rewinding an armature of this kind with recommendations
for the insulation of slots and testing as outlined by G. I.
Stadeker in the Electric Journal, Vol. VII, No. 7, are given in
what follows:

Winding a Threaded-in Coil. Each complete coil is usually
wound with from one to four separate wires double cotton
covered, on a special form or mould. When winding the coil
the two or more wires are held together in the hand and wound
as one wire. Such a coil is then made up of two or more
separate coils, and there will be as many beginning ends or
terminals and finish ends to the coil as there are wires-in-
hand when winding. Each of these ends or terminals should
be provided with an extra insulation in the form of woven
cotton sleeves. As an aid when connecting the terminals to
the commutator, sleeves of different colors should be used.
For a coil wound with three wires, black, white and red sleeves
can be used. The same color of sleeve must be used on the
start and finish ends of each of the wires used in winding the
coil. To make sure of this, two sleeves of the same color
should be slipped over each wire before beginning to wind the
coil. When starting to wind the coil, one set of the cotton
sleeves should be slipped down to the end of each wire and
adjusted to reach about three-quarters of an inch along the
body or slot side of the coil. The other set of cotton sleeves
can be slipped back along the wires as the coil is being wound.

REWINDING DIRECT-CURRENT ARMATURES

Then fasten the ends of the wires on the form or mould. As
the spindle on which the winding form is mounted is revolved
slowly, the winder can guide the wires into the mould, placing
strips of tape under them at the corners when the first turn is
applied. The wires should be kept under some tension
to make them conform closely to the shape of the mould or
form. On the last turn another sleeve should be slipped down
each wire and allowed to extend along the body of the coil
a sufficient distance to be bound in place with the tape which
was previously inserted. The leads can then be cut off to the
proper length.

In case a coil of a large number of turns is to be wound and
the cotton sleeves used do not slip easily over the wires, it may
l>e quicker to use a test lamp to find the correct finish end on
which to use the proper color of sleeve. In such a case one of
the start ends of the coil can be placed on one terminal of the
test lamp and each of the other finish ends tried until the lamp
will light. On the two ends thus located the same color of
sleeve should be used. The second pair of ends can be located
in the same way.

Insulating Lining for Slots. The slots should be insulated
with an outer protective layer of fish paper (for thickness see
pages 163 to 172, on ''Insulation for
Slots") about three-quarters of an inch
longer than the slot and two inner cells
of treated cloth. One of these cells
should enclose the lower and the other
the upper coil as shown in Fig. 55.
For machines having a terminal vol-

r/\f\ ij. "j. j

tage of 500 volts or over it is good
practice to use a third cell of treated
cloth placed next to the fish paper cell and enclosing both of
the coils. This provides a better insulation of the winding
from the core. These cells should be cut so that they will
project about an inch beyond the slot opening to serve as a
guide when the strands of the coil are being inserted in the slot.
Inserting Coils in the Slots. Since the slots are partially
closed, all the bottom sides of the coils can be inserted in the
slots before the top sides are inserted. In each case the indi-

FIG. 55. Slot insulation

for a double layer winding

in partially cl J ed slots .

62 ARMATURE WINDING AND MOTOR REPAIR

vidual strands may be forced into the slot with a flat fiber
drift. As each coil is put in place the lower leads should be
inserted into the slits of the proper commutator bars, care
being taken that the different colored leads are connected
always in the same order. (For details of connections for
lap and wave windings, see Chapter " Making Connections
to the Commutator," page 101).

After all the coils are in position in the bottoms of the slots,
the protecting edges of the inner cell enclosing the lower
coil side should be drawn up as far as the coil side will allow,
then cut off close to the slot and folded in. With a fiber drift
and mallet force the wires and cell into proper position in
the lower half of the slot in order to make room for the
upper coil side. The unprotected wires of the coil which cross
the end of the core should now be taped up with cotton tape
for about two-thirds of their length, starting up close to the
core and enclosing the ends of the lower cells, which project
from the slot.

Steel driving
1 Slot wedge slide

FIG. 56. Device for driving wood and fiber wedges in slots.

When ready to insert the top sides of the coils, the upper
cells of treated cloth should be inserted in the slots preferably
one at a time. Now face the commutator and count the
throw already calculated for the winding in a counter-clockwise
direction from the first slot prepared with its insulating cell,
in order to determine which coil should go into this slot. The
proper coil should then be bent into shape, its strands waxed
and inserted into the slot. The edges of the projecting cell
can now be clipped and folded in and hammered tightly into
place with the mallet and drift. Then drive in a fiber wedge
over the coil by using a wedge driver such as shown in Fig. 56.
This consists of a hollow rectangular piece of steel fitted with a
sliding steel strip about the same size as the wedge. The wedge

REWINDING DIRECT-CURRENT ARMATURES 63

is inserted into the driver, its end beveled and forced into the
slot by tapping on the steel strip of the driver with a mallet.

Insulating Overlapping End Connections of Coils. After
the wedge has been driven in, take a piece of cotton tape and
wrap firmly around the coil and the projecting tip of the wedge
close up against the core and continue to tape up the end con-
nection of the coil to the point where it was insulated from
the bottom upward when the other side was placed in the
bottom of the slot. Glue the ends of the tape where they
meet.

When two or three of the top sides of coils have been placed
in position, insert one or more strips of treated duck
between the end connections of the upper and lower coils
where they cross each other at each end of the armature.
As the remaining upper sides of the coils are placed in the
slots, these strips should be wound around the armature
so that they finally form a complete band of insulation between
the upper and lower coils. Each coil as it is finally placed
should be shaped at its ends with a mallet and fiber drift so
that one coil fits .snugly against the other and there is a
rigid construction when the armature is completed.

Connecting Finish Ends of Coils to Commutator. The
ends of the upper sides of the coils can now be connected to
the commutator using a test lamp. Connect the colored leads
in proper order as explained on page 101 under " Making
Connections to the Commutator." Before soldering the
coil ends, the entire winding must be tested for grounds
with proper voltage and for short circuits and open circuits.
(See page 122 under the heading of " Testing Armature Wind-
ings.") If no faults are discovered the ends of the coils can be
soldered to the commutator and the latter turned and finished
with sandpaper. The armature is now ready for banding.

Loop Windings for Small Motors. Small direct-current
armatures for fan motors and general utility use, are often
wound by hand using what is sometimes called a loop winding.
In the case of a 2-pole armature having 15 slots and 15 com-
mutator bars, such a winding might be made up as follows:
Wind slots Nos. 1 and 8 half full, using double cotton-covered
wire, leaving the ends a little longer then needed to connect to

64 ARMATURE WINDING AND MOTOR REPAIR

FIG. 57. Winding the first coil in the slots of a small direct-current motor.

FIG. 58. The operator is here shown completing the last turn of the winding.

FIG. 59. In this illustration the operator iri driving down the coils before
inserting fiber wedges in the slots of the completely wound armature.
In Figs. 57 to 59 three steps are shown in \vinding a small direct-current armature.
Before starting to wind the coils, the cores are insulated on the ends as shown in the
lower right-hand corner of Fig. 57. The armature illustrated is for a 32-volt motor and
has 19 slots with a coil pitch of 1 to 8 slots. Ten turns of No. 11 wire per coil are used
with two coils per slot. There are a total of 38 coils and 38 commutator bars (Robbing &
Myers Company).

REWINDING DIRECT-CURRENT ARMATURES

the commutator and without cutting the wire wind a similar
coil in slots 2 and 9 and bring out a second loop. Proceed
in this way throughout all the pairs of nearly opposite slots
until on the second round the slots are completely filled when
the beginning and the end of the wire can be twisted together
for the 15th loop. Cotton sleeves can now be inserted over
the loops and connections made to the commutator. Figs.
57 to 59 show an armature being wound by this method.
In the illustrations the necessary insulation on the shaft close
to the core and on the core at the ends of the slots is shown.
The slots should be insulated in this case with heavy fish
paper about 10 mils thick.

Banding a Small D.-C. Armature. When applying the
banding wire, the armature should be inserted in a lathe.

FIG. 60. Dissembled view of a small direct-current motor showing banded
armature (Fidelity Electric Company).

The first operation is to hammer down the ends of the coils
until their diameter at the armature ends is no greater than
that of the core. Extreme care must be taken not to injure
the insulation with the mallet. As a base for the banding
wires, two bands of cotton tape separated by a band of var-
nished paper should be wound over the end connections near
the core and the whole tied down with several layers of twine.
Short strips of tinned copper about 0.02 by 0.25 inch in cross-

66 ARMATURE WINDING AND MOTOR REPAIR

section should be slipped under the temporary banding twine
at intervals of two or three inches with two extra ones used
where the banding wire is started and ended. For the banding
wire of small armatures No. 14 to No. 17 B. & S. gauge tinned
steel wire can be used. The start should be fastened to a peg
slipped into an air duct or fastened to the end of the banding
twine and the first layer guided so that it crosses itself to
relieve the strain on the end fastening. After two or three
revolutions of the band wire the temporary banding twine
can be cut off and the banding wire wound on tightly across
the protecting tape. After the required width has been
wound, the copper strips at the start end should be turned
up and clipped off so that about one-quarter inch projects
from under the banding wire. Then bend the strips over to
hold the banding wire in position.

Without cutting the banding wire, it should be guided across
the core to the opposite end of the armature and wound in the
same manner as before to within about a quarter of an inch
of the edges of the end connections. The clips at beginning
and end can now be bent over and soldered. The banding
wires on both ends of the armature should now be driven up
close together, all the clips bent over and the wires soldered
together. In the soldering of band wires no acid should be
used. A solution of rosin in alcohol is recommended. After
the soldering operation is completed the surplus turns and
crossovers of the banding wire can be cut off. The armature
is now completed except for the balancing. (See page 150
under the heading of "Balancing an Armature. 7 ')

II. WINDING D.-C. ARMATURES HAVING OPEN SLOTS

Motors in sizes above five horsepower that nave armature
cores with open slots may be wound with coils made up of wire
or of copper strap. The coils may be of the form wound or
pull in types but are fully insulated before being inserted into
the slots. In general two types of open slots are found, one
with wedge grooves in the top of the slot in which wedges can
be driven to hold the coils in position; while the other is a slot
with smooth sides requiring banding wire to hold the coils in

REWINDING DIRECT-CURRENT ARMATURES

place. Grooves are usually provided on the armature for
bands in addition to those over the end ' connections. Very
often the same armature core with open slots is used for differ-
ent types and sizes of machines. The coils may therefore be
smaller in some cases than the original ones and do not fill the
slots. In these cases fillers must be used to fill out the sides
and below the coils in the slot so that the wedges or band wire

FIG. 61. Direct-current armature showing duck or canvas pad held by

cordage on ring which supports coils

This insulation is used to prevent coils from rubbing on the frame and gives a larger
margin of safety against grounds. Slot insulation made of treated pressbpard is shown
in place. Separators between coils in each slot are made of the same material but a little
thicker (Roth Brothers & Company).

will press firmly on the top of the coils and prevent any pos-
sible motion of the coils in the slots that will result in chafing
and damage to the insulation. Fillers most used consist of
strips of treated fullerboard or treated wood. If a coil is
only slightly loose it is better to add insulation to it.

Winding and Insulating Coils. In case coils must be wound
for a particular job proceed as outlined under the heading

68 ARMATURE WINDING AND MOTOR REPAIR

of " Forms for Winding Coils," on page 141. In rewinding
a bar wound armatAire, it will seldom be necessary to form new
coils, because each coil consists of a single turn of heavy copper
strip, which is not as easily damaged as are wire wound coils.
In this case, all that will generally be necessary is to reinsulate
the coils and the armature core. When all the coils have been
wound, they should be covered with cotton tape about three-

FIQ. 62. Armature shown in Fig. 61 with a partial lap winding in place.

The insulation on ends of coils and method of interlacing coils so that they occupy
a minimum amount of space are shown. Coil ends are later to be fastened to supporting
spider ring with treated canvas or duck and cordage. Band wires are wound on core
with stripes or pads of thin canvas underneath, after winding is completed (Roth Brothers
& Company).

quarter inch in width. The end of the tape winding can be
made fast with a piece of thread, care being taken to have the
taped end on a part of the coil that does not go into the slots,
as the thickness of the thread on each side may be sufficient
to cause the coil to stick while being placed in position on the
core. The finished coils should go into the insulated slots
without undue driving. The coils should now be given a coat

REWINDING DIRECT-CURRENT ARMATURES 69

of moisture repelling varnish and allowed to dry. When thor-
oughly dry the armature can be rewound.

Insulating Open Slots. When the armature is not too large
it can be wound in a lathe or bench stand. For large armatures
a suitable floor stand can be easily made. Place the armature
in the lathe or on the stand with the commutator at the
winder's right. Before inserting the insulation for the slots,
they should be thoroughly cleaned and all burrs and sharp edges
removed with a file. The same arrangement of slot insulation
described for partially closed slots on page 61 can be employed.
However, for low-voltage machines (not over 250 volts) to
be used in dry places a satisfactory slot insulation consists of
two layers of fish paper each 0.005 inch thick between which is
placed a layer of empire cloth 0.010 inch thick. This insula-
tion should be cut so that it will extend about J^j inch past the
end of the slot and project about one inch above the entrance
bo the slots. This will form a mechanical protection for the
coils and serve as guides through which the coils can be slid
into place. The corners of the projecting edges should
be clipped to keep them from interfering with the insertion of
the coils. If end-ring insulation is to be used, hold it in place
by winding thread over the ends of the core through the slots.

Inserting Coils in Open Slots. After the slots have been
insulated, place one side of a coil in a slot and force it to the
bottom with a fiber drift a little narrower than the width of
the slot. The other half of the coil (or top side) should be left
out of the armature for the present. Insert the bottom half
of coil No. 2 in the next slot. turning the armature in a counter-
clockwise direction (looking toward the commutator end).
Proceed in this manner until the top side of one coil is to be
placed into the slot containing the bottom side of the first
coil placed on the armature. This will be the first coil in
which the top side can be placed in a slot over the bottom side
of coil No. 1. Before doing this, reference should be made to
the note book data in case an armature is being rewound
exactly as before, in order to make sure that the proper number
of slots have been spanned.

To make this method of winding clear, refer to Fig. 63.
Here the coils span five slots, that is, the top side of coil No. 1

70 ARMATURE WINDING AND MOTOR REPAIR

is removed five slots from the bottom side of the same coil.
The top half of coils Nos. 1, 2, 3, 4, and 5 (called the throw
coils because they cover a part of the armature equal to the
throw of a coil) are left out of the slots as shown, as the bottom
half of other coils will have to be placed in these slots before
these top halves can be put in. When the bottom half of coil
No. 6 is placed in slot 11, its top side may be placed in slot 6,
because the bottom side of coil 1 is located in this slot. Con-
tinue to place the coils on the core, traveling in the direction
indicated by the arrow. The bottom half of coil No. 7 is

FIG. 63. Method of placing throw coils on the armature of a lap winding.

placed in slot 12, and the top half in slot 7. Before inserting
the coils in the slots it is a good plan to rub the sides with
paraffin wax. This helps in inserting the coils and prevents
damaging the insulation.

As the armature is being wound, a strip of heavy pressboard
(about 0.050 inch thick) should be placed in the slots between
coils to thoroughly insulate them from each other. When all
the coils have been inserted, the top sides of coils Nos. 1, 2,
3, 4, and 5 can be placed in the slots. As each coil is put in
position, the end connections should be carefully shaped to
the core.

Shaping End Connections. The end connections are shaped
by means of a winding drift. This consists of a steel bar about
12 inches long, one inch wide and tapered in thickness from
one-half to one-eighth inch, having all the corners rounded

REWINDING DIRECT-CURRENT ARMATURES

and smooth. The tip of this drift should be placed against the
inner side of the end connection of the coil and tapped with a
mallet, forcing the upper part of the end connection out from
the armature and away from the lower half. Each coil as it

FIG. 64. Rear and commutator ends of a wave-wound armature using
strap coils. Note the insulation of slots and the insulated support of rear
end connections (General Electric Company).

is put in place can be similarly shaped so that when the arma-
ture is completely wound a circular air chamber is formed
between the upper and lower halves of the end connections at
both front and rear. This process should be continued until
the first slot is again reached.

72 ARMATURE WINDING AND MOTOR REPAIR

Six Steps in Winding a Small Direct-current Armature

No. 1

FIG. 65. This illustration and those of Figs. 66 to 71 show successive
steps in winding a small direct-current armature (Crocker-Wheeler Company).
The slot and shaft insulation and first two coils are shown in place here.

No. 2

FIG. 66. Appearance of the winding before the last three coils are inserted in
bottoms of the slots.

REWINDING DIRECT-CURRENT ARMATURES 73

No. 3

iG. 67. All bottom sides of coils in place and strips of insulation inserted
between bottom and top coil sides ready for the top layer of the winding.

No. 4

FIG. 68. Inserting the last few top sides of coils. Note the shaping of the
end connections of this winding alternately in and out to give the compact
appearance shown in Fig. 71.

74 ARMATURE WINDING AND MOTOR REPAIR

No. 5

Fiu. G9. Connecting the first two leads of bottom coil sides to the commu-
tator.

No. 6

FIG. 70. Connecting the leads of top coil sides to the commutator,
the tape insulation between coil terminals.

Note

REWINDING DIRECT-CURRENT ARMATURES

FIG. 71. Completed armature banded and treated with insulating com-
pound. Note winding insulation over coil terminals at commutator end.

FIG. 72. Portable floor stand for winding small direct-current armatures-
The operator is shown shaping the end connections with a fiber drift.

ARMATURE WINDING AND MOTOR REPAIR

Truing Up the Heads of the Winding. When the winding
is completed, bend the leads back over the surface of the core
so as to expose the head of the winding. This must be trued
by revolving the armature and marking the high places with
chalk. These high spots can be driven down with a rawhide
mallet, and the low places raised even with the others. The
back head should also be carefully trued up so as to present a
good appearance when running. Trim off the projecting
slot insulation even with the surface of the armature if banding
wires are to be used. In case the coils are to be held in place
by . wood or fiber retaining wedges, lap this insulation down
over the coils and drive in the wedges. If the armature leads
are not covered with tape, they must be protected in some
manner and for this cotton sleeving can be used.

FIG. 73. Insulation of end connections showing friction cloth blanket i
place at the commutator end as a protection for the leads.

Insulation Between Commutator End Connections. A

friction cloth blanket should be placed over the end connec-
tions on the commutator end, as shown in Fig. 73, as a
protection for the terminals of the coils. It is good practice
to shape the ends or terminals of the side of the coils in the bot-
toms of the slots as shown in this illustration and to bend
back over the core the terminals of the top sides of the coils.
This helps in arranging the connections from the coils in the
bottoms of the slots, which can now be made.

REWINDING DIRECT-CURRENT ARMATURES 77

Now wrap several thicknesses of treated cloth in a belt
over the bottom leads and bind with thread. The top lead,
located by a lamp tester of magneto can now be connected
to the commutator. In the same manner connect the remain-
ing leads. Cut off the ends of the leads that project past
the neck of the commutator, and if there is space after the lead
have been driven down, these pieces can be used as " dummies"
to fill such spaces. When this stage of the winding process
has been reached, a short-circuit and open-circuit test must
be made. (See page 122 under heading of " Testing D.-C.
Armature Windings.") If the armature tests clear, the leads
can be soldered in.

When the wedges are not used to hold the coils in place some
means must be provided to prevent them from becoming
loosened from the slots, during the subsequent operations
before banding. To accomplish this a single band of wire
can be tightly fastened around the coils at each end of the
armature.

After these operations the commutator can be turned and
polished and the armature banded if this is required. When
balanced and painted with insulating varnish, the armature
is ready for use.

m. WINDING LARGE D.-C. ARMATURES

In contrast with the winding of small armatures, the winding
of large direct-current armatures is not a particularly compli-
cated operation. Although the insulation throughout must be
moisture and oil proof, no such elaborate precautions as in
the case of industrial motors is necessary in the larger sizes
of machines for they are in most cases installed in dry clean
places. The centrifugal strains, however, may be higher in
the larger machines and the windings must also be braced
against the magnetic strains produced by heavy short circuits
which may occur on account of the very low resistance of
circuits using wire of large cross-section. The following
recommendations are given for the winding of large armatures
by a writer in the Electric Journal, Vol. VII, No. 11.

Coils for Large D.-C. Armatures. For the large sizes
of direct-current machines the armature coils are usually

78 ARMATURE WINDING AND MOTOR REPAIR

formed of bare copper strap. They are usually of the one
piece or two piece, one turn, diamond type with the number
of coils equal to the number of commutator bars. To secure
best possible space factor, and for other mechanical reasons,
single coils are often bound together into a larger coil, each
single coil being insulated from the other and electrically
separated. In such a case the number of slots is only a
fraction of the number of commutator bars.

FIG. 74. Lap-wound armature using two-part strap coils. Note the
four layers of insulation between end connections at commutator end (General
Electric Company) .

The method of insulating this type of coil depends upon
the size, voltage and operating conditions of the machine and
on the number of single coils composing a complete coil.
When there are less than four single coils per complete coil,
the ends of each single coil are taped with one layer of cotton
tape, half overlapped. This taping extends a sufficient dis-
tance along the straight part to assure that the joint between
it and the rest of the insulation will be well protected. The
straight parts are then wrapped with a fish paper and mica
wrapper, interwoven between the straps in such a manner as to
furnish insulation between the single coils, and then wrapped
several times around the complete coil, the exact number of

REWINDING DIRECT-CURRENT ARMATURES 79

turns depending on the size, voltage and operating conditions
of the machine. (See "Coil Insulation," pages 163 to 172.)

When there are four or more single coils per complete
coil, alternate single coils should be wrapped with one turn
of fish paper and mica, held in place by a non-overlapping layer
of cotton tape. The single coils are then assembled and a
cell of fish paper and mica wrapped over the whole. The coil
should then be taped with a layer of cotton tape, non-over-
lapping over the wrapper and half lapped over the ends.
Then brush with or dip in a black finishing varnish and air
dry. After this dip twice in insulating varnish and dry in
an oven for twelve hours after each immersion. Before
the coil is used in the armature, the leads should be cleaned
of insulation and varnished and thoroughly tinned.

Lap and Wave Windings for Large Armatures. In the
main the same conditions outlined on page 112 under the
heading "Wave Lap vs. Direct-current Windings" apply in
the case of largo armatures. The wave winding has the
decided advantage, however, that no cross-connections are
required. It has, therefore, a wide use in machines where the
size of coil required does not become excessive nor the voltage
between commutator segments too great to permit of good
commutation. Under ordinary conditions this consideration
limits the wave winding to four- or six-pole machines. Where
the number of poles is greater the lap winding seems to suit
the conditions of good operation best. With this winding
the voltage between segments is kept down and high conduc-
tivity through the armature is made possible without using
coils of large cross-sections. For both lap and wave windings,
one-piece coils can be used, but a two piece or half coil has
some advantages in that less skill is needed to wind the
armature and the coils are easily repaired if the damage is only
to the top half. The half coil however calls for a soldered
joint at the rear of the armature which must be made with
great care.

Insulating the Core. Before starting the winding operation
the core should be thoroughly cleaned with an air blast, thus
removing any iron filings or other foreign matter from the
slots. The commutator necks must then be carefully ex-

80 ARMATURE WINDING AND MOTOR REPAIR

amined to see that all are straight and that the openings at
the top are wide enough to admit the coil leads easily. Test
the commutator for breakdown to ground, and for short-
circuit between segments, with the standard test voltage for
the machine. (See Chapter V, also page 175.) All parts of
the spider which come in contact with the coils, such as coil
supports, etc., should be carefully insulated with either tape,

FIG. 75. Medium-sized heavy duty armature showing use of strap coils
for a wave winding (Westinghouse Electric & Mfg. Company).

fullerboard channels, or two or three thicknesses of fish paper
strips. When tape is used it is wrapped in overlapping layers
over the entire support. At the point where the spurs which
hold the coil support in position prevent winding on the tape,
the iron should be covered with insulating cloth, and held in
place by the tape on each side. Each layer of the tape should
be shellaced as it is wound. When fullerboard strips are
used, the first layer is frequently screwed to the iron to pre-
vent lateral motion. Other layers are shellaced over this,

REWINDING DIRECT-CURRENT ARMATURES

and the whole is usually bound with twine. Special care
should be taken to stagger all joints.

Inserting the Coils. The assembly of the different types of
coils is essentially similar. Mark two slots with chalk to
receive the first coil, and count off and mark the commutator
necks into which its leads will be connected. Fish paper cells
can then be inserted into the slots, and the coils driven into
position one after another with a mallet and a fiber drift. If
a two-piece coil is used, the lower half coils should be inserted
first all the way around the armature and then the upper half

FIG. 76. Wave-wound armature partly completed showing insulation used
between end connections of coils.

coils. If one piece coils are used the coils should be inserted
in regular succession, the bottom half of the coil being driven
into the bottom of its slot first. The other half is driven into
close contact with the coil which is already in the bottom of
the slot. If there is no coil in the bottom of the slot, as hap-
pens with the throw coils, this top half is inserted only tem-
porarily until the winding has been carried entirely around the
armature. Then the throw coils must be removed so that
the coil sides can be placed in the bottom of the slots.

When a one-piece coil is used in a wave winding, the throw
coils span so large a part of the armature that it is not usually
advisable to insert the upper part of the coils and then remove
them but allow them to hang free as shown in Fig. 76 until

82 ARMATURE WINDING AND MOTOR REPAIR

all the coils have one side in place in the slots. The upper
sides can then be driven into place in regular order. As in
the smaller windings, the upper and lower coil ends or termi-
nals should be separated with bands of oiled duck or drilling.
With one piece coils, this should be threaded into place as the
coils are inserted in the upper part of the slots. With two-
piece coils it is simply wound over the lower set of coils before
the others are placed in the slots.

FIG. 77. Core of a large armature showing construction for good ventilation
(General Electric Company}.

The coils must be a close fit in the slots, in order to prevent
any possibility of chafing. If necessary, strips of fullerboard
or treated wood should be inserted at the sides or bottom of the
slot, to make the coils a tight fit: As each top coil is put in
place it should be driven into the slot, the protecting cells then
cut off, and folded over it, and fiber wedges driven into the
wedge grooves. The slots on a large-sized machine are too
long to allow one wedge to be used, so that one or more must
be driven in from each side of the slot to furnish complete
protection for the face of the coil. The armature should then
be tested for grounds, before the connections are soldered.

After the winding is completed, the armature may be banded
temporarily at both ends. Then drive wooden wedges loosely

REWINDING DIRECT-CURRENT ARMATURES

in between the commutator necks, all around the armature to
insure even spacing. After this drive them in tightly, to
force the necks and coil ends into tight contact and hold them
rigidly in place. With two-piece coils, connecting clips are
placed over the leads at the rear end, and wedges driven in
between these in the same manner. The connections to the
necks, and the rear end connections, if any are used, can be
soldered. This soldering should be done on the side of the

FIG. 78. Large direct-current engine type armature showing method of
ventilation (Westinghouse Electric & Mfg. Company).

'machine instead of the top, as a better joint can be made in
this manner, and there is less liability of the melted solder
running along the necks and short-circuiting the commutator
segments.

Before removing the wedges the armature should be mounted
in a lathe, or if no suitable lathe is available, in its bearings
with the field frame removed, and the soldered connections
turned down. If the armature is mounted in its bearings, a
suitable tool holder 'must be fastened to the frame, or some
rigid support. The commutator may then be turned down and
given its final polishing at the same time. Now knock out the

ARMATURE WINDING AND MOTOR REPAIR

wedges from between the leads, and round off the sharp cor-
ners with a file. When the bare copper is not covered with
insulation, insulating material can be inserted between the
adjacent leads to prevent accidental contact. At the rear
end, this usually takes the form of asbestos tape which can be
interwoven between the leads or of canvas hoods, sewed in
place. At the front end, the necks may be separated at
the tops by strips of heavy fish paper, bent over the tops of the
leads so as to be held in place by the band wires. Where the

FIQ. 79. Large direct-current engine type armature partly wound with*
strap coils (Westinghouse Electric & Mfg. Company).

necks are quite long, additional separators may be inserted
half way up from the commutator. These may be in the form
of fiber buttons or may merely consist of heavy twine, inter-
woven between the long commutator necks.

Banding Wire. Bands of steel wire are ordinarily placed
over both ends of the armature, and frequently another over
the connections to the commutator necks. No bands need be
used over the surface of the core, as the wedges are sufficient to
retain the body of the coils in place. The coils can be pro-

REWINDING DIRECT-CURRENT ARMATURES

tected from the mechanical pressure of the banding by layers of
surgical tape separated by strips of cement paper, over which
the bands are wound. The wire should be wound on under
heavy tension, secured by clamping it between blocks of wood.
These blocks can be held from moving by heavy straps of wire,
fastened to some rigid object usually the machine frame. If
desired, a spring balance may be inserted, which will give the

FIG. 80. Armature of a 350 kw., 250-volt, engine type generator showing
the core construction for good ventilation and the use of copper strap coils
(Fairbanks-Morse & Company).

exact tension that is being applied. This should run between
300 and 400 Ib. The bands should be firmly soldered in place.
When it is desired to secure extra mechanical strength,
the wire is sometimes wound on two or three layers
deep, each layer being soldered separately. The proper
banding of a large armature by this method is often a serious
problem on a repair job. A sectional band wire such as shown
in Fig. 82, is therefore convenient. When using this type of
banding, the two ends of any section are keyed together into an

ARMATURE WINDING AND MOTOR REPAIR

FIG. 81. Completed armature of a 350 kw., 250-volt, direct-current engine
type generator, lapwound for 8 poles with equipotential connections. Four
coils per slot are used (Fairbanks-Morse & Company).

FIG. 82. Tool for use in applying section bands on large armatures.

REWINDING DIRECT-CURRENT ARMATURES 87

open loop and then applied to the armature. In making the
final connection the special clamp shown in Fig. 82 is needed.
The two jaws of this clamp grip the ends of the band and by
means of the handle whose lower end is formed into a cam, the
jaws can be forced together so as to interweave the loops of the
band wire and permit the steel key B to be inserted. In this
operation the clamp can be held in any position by inserting

FIG. 83. Large direct-current armature partly wound with strap coils.
Note the insulating strips between end connections and double insulation in
slots (Crocker-Wheeler Company).

the pin A through the movable jaw and beam. For other
details of banding see Chapter VI, page 146.

Balancing Large Armatures. After the armature is banded
it is ready for balancing. Suitable balancing ways may con-
sist of heavy steel beams with polished steel plates mounted
on their upper edges. The surface of the polished plates
must be accurately leveled. The shaft should rest on the
inner surface of a polished steel ring, which in turn rests on
the polished plates. In this way, an almost frictionless

ARMATURE WINDING AND MOTOR REPAIR

bearing surface is obtained, and the armature tends to roll
until the heaviest part is at the bottom. Melted lead can be
poured into recesses in the spider, or cast-iron weights bolted
to the spider arms to correct any unbalanced condition, until
the armature will lie with any part uppermost. The armature
should then be thoroughly cleaned with an air blast and

FIG. 84. Method of cross connecting a large direct-current armature having
a lap winding ( Westinghouse Electric & Mfg. Company).

sprayed inside and out with black finishing varnish. Special
care should be taken to reach all exposed parts of the core,
to prevent rusting.

Rotary Converters. There is no essential difference in the
winding operations as described between rotary converters,
and other direct-current machines. At the rear of the arma-
ture, however, taps are brought out from the coils at regular

REWINDING DIRECT-CURRENT ARMATURES

intervals. On a three-phase machine, this will be two-thirds
of the pole pitch; on a two-phase machine, one-half the pole
pitch, and on a six-phase machine one-third the pole pitch.
These taps are connected to the collector rings. A two-phase
or a six-phase rotary converter cannot be wound with a wave
winding on account of the necessity of having an equal num-
ber of coils between taps on the armature.

Three -wire Generators. Practically any standard genera-
tor can be adapted for use as a three-wire machine by the

FIG. 85. Completed armature for a large 3-wire direct-current generator
(Crocker- Wheeler Company).

addition of suitable collector rings and balancing coils. These
coils are entirely self-contained and may be installed apart
from the generator. The collector slip rings are usually much
smaller than those of a rotary converter, as each one carries
only a fraction of the unbalanced current. The current
which they carry is largely unidirectional, only enough alter-
nating current flowing to excite the core of the balancing coils.
They are accordingly made of iron to avoid the blackening
from electrolysis which takes place when the direct current
flows from copper to carbon. They may be placed at either
end of the armature, but are usually placed at the end of the
commutator, for greater convenience (see Fig. 85) .

ARMATURE WINDING AND MOTOR REPAIR

IV. WINDING RAILWAY, MILL AND CRANE TYPES OF
ARMATURES

The methods of winding and insulating armatures for rail-
way, mill, mining and crane types of armatures are practi-
cally the same, since the service is somewhat similar and the
windings must stand up against much hard usage and abuse
of the motor. Thorough insulation and protection against
vibration and chafing of the insulation are essential points
in the winding of these armatures and must be given more
attention than in any other.

Railway Type Armature Coils. This type of armature coil
is completely formed and impregnated with insulating com-
pounds. It needs no shaping when placed on the armature.

Fullerboard Protecting Strip
.-'Coil in Top of Slot Only

Plaetlo Varnish

Cotton Tape

Fuller Board Separators

ullerboard Protecting Strip"
'Coilln Top of Slot Only
,D.C.C. Wire

Treated Cloth

Fuller Board Separators

Black Plastic Varnish

Fro. 86. Wire wound coils show-
ing single coils arranged in vertical
layers.

| -^ Cotton Tape

FIG. 87. Wire wound coils
showing single coils arranged in
horizontal layers.

It is usually advisable for the repairman to purchase these
coils from the manufacturer of the motor which is to be re-
paired. Usually the coils are made up of two or more single
coils wound with round wire or copper strap. These single coils
are separately wound but grouped together for mechanical
reasons in handling and insulation. The windings of these
motors are of the wave or two-circuit type, except for the
largest sizes of mill or locomotive motors. The insulation
for coils and core and the details of the winding of railway
armatures as given here, are the recommendations outlined by a
writer in the Electric Journal, Vol. VII, No. 10.

Wire Coils. The coils for smaller machines are made of
double cotton-covered wire. The single coils which form a

REWINDING DIRECT-CURRENT ARMATURES

complete coil are insulated from each other by fish paper or
fullerboard separators and may be arranged radially or cir-
cumferentially in the slot as shown in Figs. 86 and 87. The
leads from wire wound coils are secured along the diamond end
and leave the coil one after the other, each being firmly tied
and taped in position so that there is no possibility for them
to chafe against one another. They are reinforced with
cotton sleeves.

Strap Coils. Most of the larger railway motors have the
armature coils made from rectangular conductors instead
of round wire, as with this form of conductor a greater pro-
portion of the slot space may be filled with copper, without

Fish Paper Protecting Strip
'"Coil in Top of Slot Only

'ullei board Protecting Strip
Coil in Top of Slot Only

FIG. 88. Section of slot showing
strap wound coil of two turns. -

FIG. 89. Section of slot showing
strap wound coil of one turn.

sacrificing the insulation requirements. The pressure on
the insulating surfaces is also more evenly distributed, as
with the round wire the pressure bears on a line, while with
the rectangular conductors it is distributed over a flat surface
and is much less liable to injure the insulation. Figs. 88
and 89 show a cross-section through two-turn and one-turn
coils, respectively.

The one turn coil readily lends itself to bringing out the
leads in position to enter the top and bottom of the com-
mutator necks respectively, by the use of the standard form
of diamond end. A two turn coil requires a special turn at
the rear end of the coil, in order to bring the leads out in
the proper position. By the use of this form of coil, all the

92 ARMATURE WINDING AND MOTOR REPAIR

advantages of the strip windi ig can be secured for the smaller
machines on which more than one turn per slot is usually
required.

In some large motors, coils of the rectangular types are
made in two pieces, and are known as two-piece coils. Their
advantage lies in the fact that if a coil becomes damaged, only
one-half of the complete coil need be removed to overcome
the defect. As damage to the coils nearly always occurs on
the outside of the armature, this type of coil is peculiarly
adapted to railway type armatures. It requires, however,
a soldered connection at the back, in addition to the usual
soldered connection at the commutator, and hence is used
only on the large motors where the saving of copper for repair
parts would be great.

Coil Insulation. The insulation used on motors of the type
under consideration depends largely on the type of coil used.
Where double-cotton-covered coils are used there is little
advantage in using materials for the remainder of the slot
insulation which have higher heat resisting ability than the
cotton strands which are in immediate contact with the con-
ductors, since it is necessary under the circumstances to limit
the temperature to values consistent with the cotton insulation.
For this reason on certain types of mill motors, where the
temperature conditions are exceptionally severe, asbestos
covering is used instead of the cotton, with mica insulation
around the complete coil, cotton being used only in the protec-
tive taping over the outside.

With strap wound coils it is possible to use built up mica
in immediate contact with the conductors and cotton only
on the outside protective coverings, where it is in contact with
the air or the relatively cooler iron. Hence the copper can
be safely worked to a higher value and continued overloads
and abuse will not be so liable to cause breakdown. The coils
should be vacuum impregnated before insertion in the arma-
ture. This process renders them thoroughly moisture and
oil proof and prolongs the life of the coils over that of un-
impregnated ones, especially where they are subject to mois-
ture, acid fumes and deleterious gases. It is applied to all
railway type coils except those for mine locomotives, whose

REWINDING DIRECT-CURRENT ARMATURES

armatures are usually impregnated as a whole after the wind-
ing is completed.

Insulating the Core of Railway Armatures. In this type of
armature it is advisable to supply extra protection at every
point of special electrical or mechanical stress such as where
the coils leave the slots, where the leads leave the coils and
where the leads cross one another or cross the ends of the coils.
Before applying the core insulation, the core should be
thoroughly cleaned with an air blast and burrs taken off
with a file. When wire wound coils are used the coil supports
have curved surfaces and should be insulated with treated
cloth in strips or with layers of rope paper and treated cloth.

Comtnuta'tor

Coil Support

FIG. 90. Insulation of support for front coils of railway armature.

Slits should be made in the strips where necessary to make
them lie smooth as over the end bell, care being taken that
the slits in the successive layers are staggered. These strips
should be bound together with shellac and ironed smoothly
into place. They should be built up to a thickness of about one-
eighth inch (six layers of rope paper and five layers of treated
cloth are sometimes used) over the entire support and to the
level of the bottom of the slots and the commutator necks
at each edge. Where this would require an excessive amount
of insulating material, as occurs in the rear of the commutator
on certain types, a bed of rope can be built up, as shown in
Fig. 90, and bound in place with an insulating cement. A final
layer of friction tape can be applied over all the insulation,
great care being observed to make the layers lie smooth, and
to build up a firm support for the coils where they leave the
slots.

94 ARMATURE WINDING AND MOTOR REPAIR

On the cores for strap wound coils, the coil supports are
usually straight and insulated with built-up mica bushings or
with heavy bands of treated cement paper. No tape is used
in this case, but the bushings are arranged to come up level
with the bottom of the slots.

On both wire and strap wound armatures, the slots for
about an inch at the ends may be slightly wider than the coils.
In such cases narrow strips of heavy fish paper, projecting
slightly from the slots, should be inserted for additional pro-
tection to the coils at this point. The slots should be further
insulated with regular fish-paper cells for the mechanical
protection of the coils.

Inserting the Coils. Before starting to wind the armature,
the commutator should be tested for breakdown with 5000
volts to ground and 200 volts between segments. Mark two
slots, separated by the proper throw, with chalk for the first
coil. Then count off from the bar opposite the center of the
first slot, the commutator bars into which the leads from these
slots must fit. In a lap winding these bars should lie adjacent.
In a wave winding, the number of bars between them must be
determined. (See page 105, Chapter IV.) The first coil is
then placed in these two slots, the bottom half being driven
into the lower half of the slot, and the top half being merely
caught in its proper slot, as it will have to be removed later,
to allow a coil to be inserted beneath it.

Wire Coils. In wire wound armatures the lower leads
should be taped with friction tape when necessary to make
the insulation continuous from the coil to the commutator.
These leads should then be laid along the coil supports of the
armature core in smoothly fitting rows and the bare ends
driven into the proper commutator necks. Heavy insulation
is needed between the coil ends and the upper and lower
leads and between the upper and lower coil ends. This may
be of different form in different types of armatures. In one
type treated canvas strips may be inserted so as to furnish
extra insulation between the ends of adjacent coils and be-
tween the coil ends and the lower leads In addition a friction
cloth strip, doubled oVer a piece of rope, may be inserted at
each end between the upper and lower coil as the coils are

REWINDING DIRECT-CURRENT ARMATURES

inserted, the rope fitting in the point of the diamond. The
coils should be shaped with a fibre drift and rawhide mallet
so as to fit snugly against one another at both ends. It is

FIG. 91. Inserting the top sides of coils in an armature. The leads of
the bottom coil sides are shown connected to the commutator. Note the
band of insulation between end connections.

essential that they be made to fit closely together when first
inserted, otherwise the armature will bulge at the ends. Any
attempt to shape the coils in a completed armature is liable
to injure the insulation.

96 ARMATURE WINDING AND MOTOR REPAIR

After all the coils have been inserted and the top parts of
the throw coils have been replaced, the ends of the canvas
strips which project up between the coil ends should be trim-
med off level with the top of the coils at both ends of the arma-
ture. Those strips which project out from beneath the
coils, should be turned up over the coil ends and bound in
place with friction tape. This tape when wound completely
across the upper surface of the coil ends, serves as a protecting
and insulating layer between the coils and the upper leads.

Another method of providing extra insulation on some
armatures is to slit two strips of treated canvas and insert
them between the lower leads and the coil ends with the slits
staggered. No strips are inserted at the rear end and no
insulation is required between the ends of adjacent coils,
beyond that on the coils themselves. Strips of friction cloth
and rope may be inserted between the upper and the lower
coil ends, and the canvas strips folded over the ends of the
coils and covered with friction tape as just described. In
this case, however, strips of fish paper should be slipped over
each coil and wedged between the coil ends close to the core,
for further protection to the upper leads.

Before connecting the upper leads to the armature they
should all be tied together with bare copper wire and the coils
subjected to a break-down test of 3600 volts. Any defective
coil must be replaced. The armature can then be trued up.
This can be done by tapping down with a mallet all the high
coils as located by holding a piece of chalk so that it will rub
against the high parts when the armature is revolved. A some-
what better method is to squeeze the coils into place by means
of a flexible metallic strap and turn-buckle. The latter
method is less liable to damage the insulation and all the coils
receive uniform treatment so that a better balance of the
armature results. Both ends must form a compact mass.
Where end room is especially short, as in mine motor?, a special
form can be used on the rear end of the armature so as to
press all coils against this when being inserted to give uni-
formity to the arrangement.

Where the top of the coil is above the top of the commutator,
there is sometimes difficulty in keeping the leads properly

REWINDING DIRECT-CURRENT ARMATURES 97

separated in bringing them down to the commutator. In
such cases a canvas strip may be interwoven over every other
one, making it possible to have two layers of leads on the
vertical part. The leads are then inserted into the slits in the
proper commutator necks and copper wire dummies driven
over them, to prevent any possibility of a portion of the leads
being removed when the necks are turned down. Both leads
and dummies should be tinned and make a driving fit in the
necks.

Strap Coils. The leads of strap wound armatures are
formed to shape and, therefore, require little bending during
their installation. The coil supports of the two turn strap
coils are shaped and insulated in a manner very similar to
that for wire wound coils. In addition a bed of the insulating
cement is made over the insulation at the rear of the commuta-
tor into which the lower leads are forced as they are inserted
into the commutator necks. They are thus held rigidly in
place after the paste hardens. Strips of treated canvas or
of friction cloth folded over fish papsr and mica should be
threaded between the upper and lower coil ends, at each side
of the machine, as the coils are inserted. A length of rope
should also be threaded through the diamond point at each
end. Between the coil ends and the upper and lower leads,
strips of treated canvas with slit edges should be inserted
so that the openings will be staggered. The edge toward
the core must be shaped to fit up between the coils and furnish
added protection to the leads. After all the coils have been
inserted, the upper leads* may be bent up slightly, and the
edges of the various insulation strips cut off even with the
commutator. Friction tape should then be wound smoothly
over the treated canvas, holding it in place, and forming a bed
for the upper leads. These leads can then be bent down and
inserted into the proper commutator necks.

Coil supports for one-turn strap coils, whether two piece or
one piece, are straight, and can be insulated with built up mica
forms, or strips of fish paper or treated cement paper, shellaced
and tied in place. Insulating cement should be plastered
over the insulation back of the commutator, to hold the
leads from the individual coils in place. Separate the ends

98 ARMATURE WINDING AND MOTOR REPAIR

of the coils by two thicknesses of treated canvas, threaded in
place as the coils are inserted.

The coil supports for two-piece coils may be insulated in the
same manner as the one-piece coils. The end bells for this
type of armature, however, are usually separate from the core
and are not put in place until the winding is complete, so that
the winder has plenty of room to work on the rear end of the
coils. The winding operation can be greatly facilitated by
the use of a steel winding jig, consisting of a slotted disk with
a hub bored to fit the armature shaft. The number of slots
should be equal to the number of single coils in the armature,
and the thickness of the disk equal to the width of the connect-
ing clips. As each coil is placed in the armature slot, the
straps composing it should be placed in the proper slots in the
jig and in the commutator necks until all the lower half of the
winding is in place. The leads on the upper and the lower
half should be separated by a couple of thicknesses of treated
cement paper, or by a layer of fish paper and mica. After all
the coils are in place, the straps which are to be connected
together at the rear end lie one above the other in the slots
of the jig. These can be cut off even with the surface of the
jig, a temporary band wrapped around the coil ends, and the
jig removed. Copper connector sleeves should then be slipped
over the coil ends and wooden wedges driven in between them.
The connectors may then be soldered and the coil ends turned
down at the top and side. Next knock out the wedges and
interweave asbestos braid between the connectors to prevent
accidental contact. The end bell, properly insulated with
moulded mica or moulded paper, can then be bolted into
position.

Connections with Dead Coils. If in a four-pole motor with
a two-circuit winding there is an even number of single coils
in a complete coil, one single coil in the armature must be cut
out in order that the winding may be made continuous. This
coil is called a dead coil, because it is not connected to the
circuit. It is necessary first to determine the number of single
coils in a complete coil, by dividing the number of complete
coils or slots by the number of commutator bars. If there are
more leads from each side of the coil than there are single

REWINDING DIRECT-CURRENT ARMATURES 99

coils, each single coil is composed of two or more wires in
parallel and these must be treated as a single lead. In strap
coils each strap corresponds -to a single coil. The coil is cut
out by cutting off the leads on both sides of a single coil about
an inch from where they separate from the coil. Then care-
fully tape them up. The body of the coil of course, must be
left in the slot for uniformity of the winding.

Hooding and Banding. As a final protection to the arma-
ture coils, heavy hoods may be put on over the ends of the
coils, covering the armature from the commutator to the core
and from the core to the end bell. At the commutator end
the hood may be of woven asbestos sewed to a conical shape,
and impregnated in a moisture and oil repelling compound.
It should be put in place while wet. The small end should be
drawn up over the commutator, turned inside out, and firmly
tied over the leads and commutator necks with heavy twine.
The body of the hood must then be turned back over the arma-
ture. If the commutator necks are lower than the level of the
core, another layer of twine should be wound over the hood
near the commutator and a band of canvas sewed over the
whole. Then stretch the hood tightly back over the armature
and tie with twine.

Around the rear end of the armature, a band of canvas may
be wrapped so that the greater part of the strip extends out
over the shaft, only enough being wound over the armature to
permit a secure fastening. This may be bound in place with
a band of twine wound tightly in the groove between the coil
ends and the end bell. The canvas should then be turned
back over the armature and bound smoothly in place.

The number and size of bands depends upon the size and
speed of the armature. All armatures should have bands on
each end, placed as far out as possible, so as to cover the
greater part of the coil ends. When such a band would be
quite wide, two separate narrower bands may be wound on
each end. These bands should be insulated from the coils
by three turns of canvas tape separated by treated paper
which extends at least one-eighth inch beyond the band on
each side. The bands around the body of the core in the band
grooves should be insulated from the core and coils by strips

100 ARMATURE WINDING AND MOTOR REPAIR

of fish paper. They should also be put on with a tension
of about 350 pounds sufficient to make a good firm band and to
bring the coils down so that they will not project above the
surface of the core at any point. The individual turns of the
banding should be well soldered together and held at several
places by thin copper clips- placed before the banding was
started. For other details of banding see Chapter V, page
146.

After the banding is completed, the mica insulation between
the commutator bars should be undercut to a depth of about
one-sixteenth inch, with a special milling cutter. The entire
armature except the commutator, can now be sprayed with an
air-drying finishing varnish.

CHAPTER IV
MAKING CONNECTIONS TO THE COMMUTATOR

Before starting to insert coils in the slots of the armature,
the commutator should be tested for grounds. This can be
done by touching one lead of a high-voltage transformer
(1200 to 2000 volts) to the shaft and moving the other over the
surface of the commutator and at the edges. If there is no
arcing, the commutator is properly insulated. The winder
can now insert the coils. As each coil is put in place the
sleeving on the ends of the lower leads should be fastened to the
wire by a few turns of friction tape and these leads inserted
into the slits of the proper commutator bars. In case the
coil has two or more start and finish ends, care must be taken
that the different-colored sleeves are connected to the com-
mutator always in the same order.

Locating First Connection to Commutator. Before con-
necting the first coil to the commutator, the winder must
examine the setting of the brushes, to see whether they are
centered between pole tips or opposite the center of the pole.
When the brushes are centered between the pole tips, the
start end of each coil must be connected straight out to the
bar opposite the slot in which the beginning of the coil is
located. When the brushes are opposite the center of the poles,
the start end of the coil must be swung a certain number of
bars (equal to 90 electrical degrees) to the right or left of the
bar opposite the slot in which the beginning or bottom side
of the coil is located. The number of bars right or left (equal
to 90 electrical degrees) can be determined by the following
formula: Total number of commutator bars -r- (Number of
poles X 2). If this number is mixed such as 6.5, use the next
higher whole number as 7. The reason for these connections
of leads is that the coils must be commutated or short -cir-

101

102 ARMATURE WINDING AND MOTOR REPAIR

cuited by the brushes while the coil sides are in a neutral
position or outside the pole flux.

The spacing of the brushes from the heel of one brush to the
heel of the next for a lap winding will be equal to the num-
ber of commutator bars divided by the number of poles.

Testing Out Coil Terminals. After the commutator pitch
has been determined, take the bottom lead of one coil and
connect it to a commutator riser, using the same throw as
employed on the old winding in case it is being duplicated.
After the position of the first lead has been determined, the
remainder of the bottom leads can be connected in rotation.
When all the bottom leads are in place, a lighting out test
should be made to see if the leads are connected to the proper
bars. A short circuit and ground test must also be made at
this time.

For a lighting-out test, place one terminal of the lamp tester
on a commutator bar, and with the other touch the top leads
of several coils until the lamp lights. This will locate the top
side of the coil corresponding to the bottom side connected
to the test lamp. If the lamp lights on more than one lead,
it indicates a short circuit between coils. In an armature
containing twice the number of commutators bars as there
are armature slots, the same procedure is followed in locating
the two leads of the same coil.

Commutator Connections for a Lap Winding. After all
the beginning ends of the coils or the ends for the coil sides in
the bottom of the slot have been connected to the commutator
and all the coils tested out for open circuit, short circuit
and grounds, the finish ends or the ends of the coil sides in
the top of the slot, can be connected to the commutator In
the case of a coil wound with one wire or two wires in parallel,
the finish end of coil No. 1 is connected to the commutator
bar next to that to which the start end is connected. Thus,
if for coil No. 1 the beginning end is connected to bar No. 1
the finish end will be connected to bar No. 2. The beginning
end of coil No. 2 will also be connected to bar No. 2
and its finish end to bar No. 3 and so on until for the last
coil the finish end will connect to bar No. 1 and close the wind-
ing (see Fig. 97).

MAKING CONNECTIONS TO THE COMMUTATOR 103

On the armature of some large direct-current machines
a commutator is provided which has twice or three times as
many commutator bars as slots or winding coils. The object
of this is to improve commutation and prevent sparking by
commutating only a part of the current at a time. In a case
where the commutator has twice as many bars as slots the
coils are made up with two wires-in-hand when winding.
For the case where there are three times as many bars as slots,
coils made up with three wires-in-hand when winding are used.
The coils in either case are placed in the slots the same as
those wound with one wire but connected to the commutator
differently.

In the case where the number of bars is double the number of
slots and coils, the start and finish ends of the coils will have
two wires. For coil No. 1 connect the start ends as follows:
One wire to bar No. 1 and the other end to bar No. 2; for
coil No. 2, one end to bar No. 3, and the other to bar No. 4;
and so on. When connecting the finish ends proceed as
follows: For coil No. 1, connect one end to bar No. 3 and the
other to bar No. 4; for coil No. 2, one end to bar No. 5 and the
other end to bar No. 6 and so on until for the last coil one of
the finish ends will connect to bar No. 1 and the other to bar
No. 2. In this case, each brush will cover two bars.

When there are three start and three finish ends to each
coil and three times as many commutator bars as slots,
proceed in the same way by connecting the start ends to
adjacent bars and the end of one coil to the start of the next.
In this case each brush will cover three bars.

Requirements of a Lap Winding. A lap winding can be
wound on an armature having any number of slots provided
each slot will accommodate two coil sides for each commutator
bar and the number of bars is an even multiple of the number
of slots. When the total number of coils is not exactly di-
visible by the number of pairs of poles the winding pitch
can not be equal to a pole pitch. This, however, is not an
obstacle except where it is desirable to use equipotential
connectors. When the winding pitch can not be equal to a
pole pitch it is made so as nearly as possible. When the front
and back pitches are specified to be odd for a lap winding,

104

ARMATURE WINDING AND MOTOR REPAIR

the coil pitch in winding spaces is meant and not the slot
pitch. It will be found that with the usual type of form
wound coils, the pitches for back and front are necessarily
odd, because one side of the coil is in the bottom of a slot
and the other side in the top of the slot. The terminals of the
coil side in the bottom of the slot are usually considered the
start ends and the terminals of the side in the top of the slot
the finish ends.

Commutator Connections for a Wave -Winding. For a
wave winding, the start and finish ends of coils are connected
to the commutator differently than for a lap winding. In
the case of a coil wound with one wire, having one start and
one finish end, these are not connected to adj acent commutator
bars as in a lap winding, but a number of bars apart. When
the coils are inserted in the slots the start ends of the coil

sides in the bottom of the
slots should be connected
to the commutator as in
the lap winding when each
coil is placed. When ready
to connect the ends, the
commutator pitch must be
determined. The formula
for commutation pitch in
numbers of bars is (y k =
2/i + 2/2) -5- 2.

Where 2/1 is the front

FIG. 92. A 4-pole, wave winding for pitch and 2/2 is the back

pitch of the coil, each
counted in winding spaces.
With a double layer winding or two coils per slot, there will
be two winding spaces per slot.

In the case of a 4-pole, double layer with 13 slots and 13
commutator bars, 2/2 equals 7 and 2/1 equals 5. Then y k =
(5 + 7) -T- 2 or six bars. With a back pitch of 7 winding
spaces and a front pitch of 5 winding spaces coil No. 1 would
lie in slot No. 1 and slot No. 4. Its start end would be con-
nected to commutator bar No. 1 and its finish end to bar
No. 7. Coil No. 2 would he in slots No. 2 and 5. Its start

an armature having 13 slots and 13 com-
mutator bars.

MAKING CONNECTIONS TO THE COMMUTATOR 105

end would be connected to bar No. 2 and its finish end to bar
No. 8 and so on until coil No. 13 in slots NGS. 13 and 3 would
have its start end connected to bar No. 13 and its finish end
to bar No. 6.

It will thus be seen that in a 4-pole machine the finish of
one coil is joined to the commutator and to the start of another
which lies under another pair of poles. (See Fig. 92.) The
finish end of the latter coil is connected to the commutator
bar adjacent to the one to which the start of the former coil is
connected. In Fig. 92 it can be seen that the coils referred
to are the ones connected to commutator bars, No. 1, 7 and 13.
In other words the two coils are connected in series with
their start ends under different north poles and their finish
ends under different south poles. This explains why the
wave winding is called a series or two circuit winding.

Top.

25 Slot.
^123 Bars
"Ooil Slots 1 and 7

Lead Bart 1 and W

FIG. 93. Wave winding with an odd coil throw and an odd throw of coil

leads.

Practical Method for Locating First Connection to Com-
mutator for a Wave Winding. When there are an odd num-
ber of leads per coil, or in case of a dead coil an odd number of
coils remain, the following procedure will locate the first

106

ARMATURE WINDING AND MOTOR REPAIR

connection to the commutator. Locate the center line be-
tween the slots of the coil throw. If the throw of the leads of
the coils is an odd number of bars, this center line will fall on
a commutator bar. If it is an even number, the center line
will fall on the mica between bars. In either case the bar or
mica so located is the starting point for laying off the connec-
tions to the commutator.

If there are an odd number of bars in the throw of the leads,
take one less than the number of bars and count off half of this

FIG. 94. Wave winding with an even coil throw and an even throw of coil

leads.

number in each direction from the starting bar, and this will
give the first and last bar of the commutator throw. If
there is an even number of bars in the throw, count off half
the number in each direction from the starting mica. A
check is to count from the first to the last bar, and see if it
agrees with the information given. As the first coil put down
will have an odd number of leads, the center one of the top and
bottom leads should be placed in the first and last bar of the
throw as determined.

MAKING CONNECTIONS TO THE COMMUTATOR 107

When there are an odd number of coils per slot and an even
number of slots, a somewhat different method is required.
This, however, seldom occurs. In such a case, if the lead
throw is an odd number of bars, the center, as indicated by the
coil throw, will line up on the mica and, if an even number of
bars, it will line up on a bar. If there is an odd number of
bars in the throw, take one less than the number of bars and
count off half this number to the left and one more than half
to the right, and this will give first and last bar of the commuta-
tor throw. If there is an even number of bars in the throw
count off half the throw to the right and one less than half to
the left. If there are two leads in the first coil, No. 1 lead
should lie in No. 1 bar, and if there are four leads in the first
coil, No. 2 lead should lie in No. 1 bar.

Winding conditions with odd number
of commutator bars

Center line of coil lines up

with

Coil throw and lead throw even

Mica and tooth

Coil throw odd, lead throw odd

Bar and slot

Coil throw even, lead throw odd ....

Bar and tooth

Coil throw odd, lead throw even

Mica and slot

Requirements for a Wave Winding. For the wave or two-
circuit winding there should always be an odd number of
commutator bars. With an odd number of slots on the arma-
ture core, and an odd number of coil sides per slot, a balanced
wave winding with no dead coils is possible. The actual
number of coil sides per slot can be determined by counting
the terminals of each taped up coil in the slot. There may be
only two taped-up coil sides in the slot but each coil may be
wound with three wires in-hand while winding. In such a
case there will be six coil sides per slot for a two-layer winding.

Progressive and Retrogressive Wave Windings. When a
wave winding passes once around the complete armature and
has its start and finish ends connected to the commutator as
shown in Fig. 95 (at left), it is said to be a progressive winding.
When after passing once around the armature the start and
finish ends are connected as shown in Fig. 95 (at right) it is
said to be retrogressive.

108 ARMATURE WINDING AND MOTOR REPAI"R

The conditions under which a wave or series winding can be

kp
used are shown by the formula: K = -^ 1.

Where K is the number of commutator bars, p the number of
poles and k a whole number. When the minus sign is used
the winding will be progressive and when the plus sign is used
it will be retrogressive in a case where the number of com-
mutator bars is odd.

FIG. 95. The illustration at the left shows one turn of a progressive wave
winding; the one at the right a retrogressive wave winding.

Wave Winding with Dead Coils. For a wave winding, as
already mentioned, the number of commutator bars must
not be a multiple of the number of poles, otherwise the wind-
ing would close after it has passed once around the armature,
instead of advancing or falling behind one commutator bar
as required by the wave winding. In other words, there must
be as many coils in series between adjacent commutator bars
as there are pairs of poles. Also with two or more coil sides
per slot, the total winding pitch of the end connections (front
pitch + back pitch) should be a whole number. It frequently
happens that is found to be a fraction. By dropping the
fraction when the number is odd, a wave winding can usually
be connected with a dead coil. A rule which is often used by
engineers to determine when a wave winding is possible with-
out a dead coil is that given in another paragraph, namely:

Number commutator bars = -^ 1 ; where p is the num-
ber of poles and k must be a whole number. When an even
number of coil sides per slot are used and an odd number of

MAKING CONNECTIONS TO THE COMMUTATOR 109

commutator bars, there will always be a dead coil. Also
when there are an even number of slots on the armature
core, an odd number of commutator bars, and an odd
number of coil sides per slot are used, there will always be a
dead coil.

Another rule for a possible wave winding is that the total
number of slots times the number of coil sides per slot may
be any number divisible by the number of poles.

When it is found that a dead coil must be used, one coil,
(any one) on the armature should be cut back a short distance
from the commutator and taped up with friction tape. The
other coils can then be connected as in any other waye winding.
That is, connect all the leads of the bottom sides of the coils
to the commutator first, then connect the leads of the top
sides of the coils the proper commutator pitch away from the
bottom leads.

When there are twice as many commutator bars as coils in a
wave winding, each coil is wound with two wires and there are
two start ends and two finish ends. In making connections
to the commutator the two start ends should be connected to
adjacent bars and the two finish ends to adjacent bars, spaced
the proper commutator pitch apart.

Cutting Out Coils of a Retrogressive and Progressive Wave
Winding. In Fig. 96a is shown a portion of the winding diagram
of the armature, for the case of a retrogressive winding with a
connecting pitch of 1 to 49. If the burned out coil is between
commutator bars 47 and 95, disconnect it from the commutator
at a and b and cut it at c if it. has more than one turn. Then
connect commutator bars 46 to 47, and 94 to 95 and disconnect
the coil between bars 95 and 46 at a' and b'. Or connect bar

95 to 96 and 47 to 48 and disconnect the coil between bars

96 and 47 at a" and b" '.

For the case of a progressive winding with a connecting pitch
of 1 to 50, the winding diagram is shown in Fig. 966. If the
burned out coil is between commutator bars 3 and 52, dis-
connect it from the commutator at a and 6 and cut it at c,
if it has more than one turn. Then connect bars 3 to 4 and
52 to 53 and disconnect the coil between commutator bars
52 and 4 at a' and b'. Or connect bars 2 to 3 and 51 to 52

110

ARMATURE WINDING AND MOTOR REPAIR

and disconnect the coil between bars 51 and 3 at a" and 6".
For definition of retrogressive and progressive wave winding,
see page 107, Chapter IV.

94 95 9C 97

45 4G 47 48 49

93 94 95 9C 97

FIG. 96a. Method of cutting out a damaged coil in a retrogressive wave

winding.

t 2

FIG. 966. Method of cutting out a damaged coil in a progressive wave

winding.

345

FIG. 97. Double layer winding for a 4-pole armature having 24 slots and
24 commutator bars. The connections for this diagram are given in tabu-
lated form on page 111.

Tables for Placing Coils and Connecting Them in a D.-C.
Armature Winding. To avoid mistakes in reading a complete
winding diagram and to avoid the necessity of making such a
diagram, Fig. 97, it is advisable to furnish an inexperienced
armature winder with a table for laying the coils in the slots
and also a table for connecting the proper leads to the com-
mutator. These tables can be made up as follows:

MAKING CONNECTIONS TO THE COMMUTATOR 111
TABLE FOR PLACING COILS IN ARMATURE SLOTS LAP WINDING

Coil number

Place sides of coil in

Winding space numbers

In slot numbers

1 and 12

1 and 6

3 and 14

2 and 7

5 and 16

3 and 8

7 and 18

4 and 9

9 and 20

5 and 10

11 and 22

6 and 11

13 and 24

7 and 12

15 and 26

8 and 13

17 and 28

9 and 14

19 and 30

10 and 15

etc.

47 and 10

24 and 5

This table is for a four-pole, double layer, lap winding on an
armature having 24 slots and 24 commutator bars. By giving
the throw of the coils both in winding spaces and in number of
slots the winder will not become confused when either is used
in referring to the winding. A table of connections to the
commutator for this winding is made up as follows:

TABLE FOR CONNECTING COIL LEADS TO THE COMMUTATOR LAP

WINDING

Connect terminals of coil

tJoil number

Start end to bar number

Finish end to bar number

10.

etc.

112 ARMATURE WINDING AND MOTOR REPAIR

Such tables are particularly useful in the case of a double
layer wave winding where each coil is wound with two or more
wires in-hand. The following tables for a four-pole, double
layer, wave winding with two terminals per coil, 15 slots and
30 commutator bars, will illustrate their usefulness in such a
winding:

TABLE FOR PLACING COILS IN A.RMATURE SLOTS DOUBLE WAVE

WINDING

Place sides of coil in

0011 numoer

Winding spaces numbers

In slots numbers

1 and 8

1 and 4

3 and 10

2 and 5

5 and 12

3 and 6

7 and 14

4 and 7

9 and 16

5 and 8

etc.

29 and 6

15 and 3

TABLE FOR CONNECTING COIL LEADS TO THE COMMUTATOR DOUBLE
WAVE WINDING

Connect terminals of coil

Start ends to bars numbers

Finish ends to bars numbers

1 and 2

15 and 16

3 and 4

17 and 18

5 and 6

19 and 20

7 and 8

21 and 22

9 and 10

23 and 24

etc.

29 and 30

13 and 14

Wave vs. Lap Windings. The wave winding is mostly
used on the smaller types of direct current machines of multi-
polar construction. This winding has but two paths through
the armature, regardless of the number of poles, and half the
coils in the armature are connected in series in each path,

MAKING CONNECTIONS TO THE COMMUTATOR 113

whereas the lap winding has as many paths as there are poles,
and a correspondingly smaller number of coils in series. For
instance, in the case of a four-pole armature, wound with
coils of the same number of turns and same size conductors;
if all are connected in a wave winding, the armature will be
suitable for double the voltage at half the current that would
be used if the same armature were connected as a lap winding.
Thus on small machines, where the number of turns is neces-
sarily limited, the wave winding is usually much cheaper and
easier to use. The wave winding is widely used in armatures
for railway, hoist and crane motors.

The limitation in the use of the wave or series winding is in
the main the amount of current which can be carried by one
armature circuit. Since the wave winding consists of two
circuits, the current per circuit is equal to one-half the total
armature circuit of the machine. Where the value of this
current exceeds that which has been found to be consistent
with good practice, (up to 250 amperes per circuit in non-
interpole machines and up to 550 amperes in interpole ma-
chines) then it becomes necessary to arrange more circuits on
the armature, each circuit carrying a part of the total current.
In this case the lap or multiple winding is usually employed,
which as explained has a number of circuits equal to the number
of poles of the machine. However, the choice of wave or lap
windings is usually not only determined from the amount of
current to be handled but is greatly influenced by the require-
ments for good commutation and by the size and design of the
armature to be wound.

It is customary with some motor manufacturers to use wave
windings wherever the current to be handled permits the use
of a conductor of a sufficient size to form it into a wave shaped
coil and where the number of coils in the slot does not become
too large. The latter condition is objectionable on account of
the space required for insulating the coils. Other manufac-
turers employ the lap winding wherever possible, that is, in
cases where the number of turns per slot does not become
too small. As a general rule, it can be said that the wave
winding often works out to advantage in cases of compara-
tively large machines for low voltage whereas the lap winding

114 ARMATURE WINDING AND MOTOR REPAIR

is preferable in cases of large or small machines for high vol-
tage as well as for small machines for low voltage.

Except under the special conditions named above, such as
heavy current to be handled by the armature, machines of
high voltage, and small machines of low voltage where the lap
winding has advantages over the wave winding, there is little
choice between the wave or lap winding for those machines
where the number of poles may equal the number of circuits
of which the winding may be made up. It may be said that
in this case, American practice favors the lap winding and
European practice the wave winding.

It might be pointed out further that in a wave or two-circuit
winding all of the coils between two diametrically opposite
points on the winding are in series with each other while in a
lap winding only the coils between adjacent pole centers are
in series. Therefore, under similar conditions, the voltage
in a wave winding will equal the voltage of a lap winding
multiplied by the number of pairs of poles. The voltage
between the commutator segments will also vary in the same
ratio. In order to obtain the same voltage of the machine
the relative number of coils in the two types must vary in-
versely with the number of poles and the size of the coils
in the wave winding must increase in proportion. Where the
size of armature coil does not become excessive nor the voltage
between segments too great to permit good commutation,
the wave winding can be used. These conditions, however,
limit this type of winding under ordinary conditions to four
or six pole machines. In cases where the number of poles
is larger the lap winding has a wider application for direct-
current armatures.

In regard to the selection of the proper type of winding for
direct-current and alternating-current machines, Henry Scheril,
formerly a member of the engineering department of the
Crocker- Wheeler Company, has made the following comment
(Electrical Record, January, 1919) :

The selection of the proper kind of winding depends upon
the capacity of the machine, the voltage and speed. Wave
windings are used on machines of small capacity. They are
also being used for low speed and medium ratings. On small

MAKING CONNECTIONS TO THE COMMUTATOR 115

high voltage machines, like the railway motor, the wave wind-
ing is being used to advantage because it has only two sets of
brushes and it can be inspected very easily. Wave windings
are also used for the rotors employed in induction motors. In
all other cases, lap windings are to be preferred.

Lap Windings for Direct -current Armatures. In general,
the ends of a coil of lap winding in a direct-current machine
are connected to the adjacent commutator bars. The sides
of the coil arc spread over and placed in the slots of the arma-
ture corresponding to the distance equivalent to 180 electrical

FIG. 98. Lap or parallel winding for a 6-pole machine.

degrees. That is to say, if one side of the coil is under the
center of a north pole, then the other side of the coil will be
placed in a slot having a similar position under the adjacent
south pole. This brings one to a consideration of the wind-
ing pitch. The pitch is the number of slots spread over the
periphery of the armature corresponding to the arc equal to the
distance between two similar points of two consecutive poles.

116 ARMATURE WINDING AND MOTOR REPAIR

For instance, if a machine has 72 slots and eight poles, the
full pitch will be 72 -5- 8 = 9, and the coil will lie in slots of
one and ten.

In a direct-current machine the lap or multiple winding has a
number of paths for the current to travel through the winding
from positive to negative, equal to the number of poles. It
becomes necessary, sometimes, to wind a machine with what
is known as the "short chord" or fractional pitch winding.
In this kind of winding, using a fractional pitch, the coil is
spread over the distance which takes in less than the number
of slots between two similar points on two consecutive poles.
The use of this kind of winding on a direct-current machine
is made when there is a desire to improve the commutation.
That is, because with a fractional pitch winding the emf.,
due to inductance of the armature, is lessened at the time of
commutation. In direct-current machines it is not advisable
to use a shorter pitch than about 90 per cent, of the full pitch.
If this figure is exceeded then the advantages which this wind-
ing would give are practically eliminated and liable to bring
about worse results.

All interpole machines are built with lap windings and
equalizer bars are used. The setting of the brushes on inter-
pole machines must be done very accurately because their
position is fixed once and for all, but due to irregularities in
manufacture, in order to eliminate any possibility of poor com-
mutation equalizer connectors are being used.

Lap Windings for Alternating-current Machines. Practi-
cally all alternating-current machines use windings having a
short pitch. The reasons for using the short pitch winding,
especially on alternating-current machinery, are many. To
begin with, most manufacturers standardize their frames,
standardize their shields, standardize the clearances between
the rotary elements and the stationary elements. In some
machines, the clearances between the windings and the shields,
especially in machines carrying heavy currents, and also in
high voltage machines, become very small due to the form of
end connections which are being used in lap windings. It
has been found that by using the short pitch winding, the
end connections can be shortened considerably and thus ob-

MAKING CONNECTIONS TO THE COMMUTATOR 117

tain more clearance. Another reason for using the short
pitch winding in alternating-current machinery is because if
the machine is to be designed with existing punchings, there
is a possibility that in designing the winding, one may get
too many turns by using the full pitch due to a greater number
of slots than required. By reducing the pitch from full to
fractional, the same results are brought about as when the
number of slots is reduced. In alternating-current machinery,
short pitches are used as low as % of the full pitch. Practic-
ally all two-speed induction motors have % pitch.

Wave Windings for Direct -current Armatures. The
series or wave winding consists of coils spread over the per-

P N 1

FIG. 99. Wave or two circuit winding for a 6-pole machine.

iphery of the armature in the same way as the multiple winding
but with the ends of the coil in the direct-current machines
connected to the commutator bars whose relative positions
correspond to about double pitch. From the nature of this
winding, there are only two paths for the current to flow from

118 ARMATURE WINDING AND MOTOR REPAIR

positive to negative brush independent of the number of poles.
Now since the winding goes around the armature several
times, depending upon the number of slots, in order that the
winding may not close upon itself, the number of coils in
series must be one more than, or one less than the number of
poles. Since this winding has only two circuits, regardless of
the number of poles, only two sets of brushes are needed.
However, when this winding is being used on machines of
large capacity, there will usually be as many sets of brushes
as there are poles. The purpose of this is twofold. One is
to reduce the size of the commutator and the other is to reduce
the brush density.

Wave Windings for Alternating -current Machines. In
alternating-current machines wave windings are mostly used
in rotors of slip ring induction motors. They are either
single or double circuit and their use for that purpose is two-
fold. One reason is because the end connections are much
easier to make than in the case of the lap winding and the
second reason is because of certain definite ratios to be ob-
tained. Double or triple wave windings are also used on
alternating-current machines of very low voltage and high
currents where lap windings are found to be not practical
because of the form of end connections. All wave windings
should be symmetrical.

For a wave winding in an alternating-current motor the
number of slots (plus or minus one) is chosen so as to be divis-
ible by the number of poles. If the number of slots is plus
one, the winding is called progressive, since after traveling once
around the stator it returns to the starting slot plus one. If
it is minus one, it is called retrogressive since the circuit
returns the winding to the starting slot minus one. In this
winding the correspondingly placed conductors under adja-
cent poles are connected in series with the circuit proceeding
from pole to pole several times around the stator. The circuits
are then interconnected to give the required phase relations.

Single vs a Number of Independent Windings. Most
machines are built with a single winding with the exception of
some three-wire generators, having double windings, the two
windings being independent of each other. All alternating-

MAKING CONNECTIONS TO THE COMMUTATOR 119

current machines are built with windings so that they can be
easily changed from multiple circuit to single circuit. That is
to say, the design of a winding for an alternating-current ma-
chine is made such that if it is built in two circuits, for say
220 volts, that same machine can have the windings recon-
nected for 440 volts.

Double or triple windings, either lap or wave, have been
found to give much trouble in direct-current machines, espe-
cially in commutation. Sparking at the commutator be-
comes very pronounced and the commutator wears down very
rapidly. For this reason this kind of winding is not widely
used on direct-current machines.

Lap Windings vs Multiple Wave Windings. Although they
have not been much used in the past, multiple wave windings,
sometimes called series-parallel windings since they are wave
windings with more than two circuits, have some advantages.
The points in favor of their use are outlined as follows by
Albert A. Nims of the engineering department of the Crocker-
Wheeler Company (Electrical Record, February, 1919) :

There seems to be no reason for avoiding wave windings
with more than two circuits on ratings where they would be
desirable, provided they be made perfectly symmetrical.
This condition would bar out a four circuit winding on a six-
pole machine. Arnold, who is generally regarded as the origi-
nator of this type of winding, did not always insist on complete
symmetry, but the writer believes that the unsuccessful
experiences with series-parallel windings in the past has been
largely due to an incomplete understanding or observance of
the laws of symmetry of armature windings.

In windings with circuits equal in number to the poles mul-
tiple or lap windings are ordinarily used. They possess the
advantage of having each coil commutated at only one brush,
and also the disadvantage of having each circuit under a dif-
ferent pole, or pair of poles. If the field strengths under the
various poles are not identical, the induced emf s in the various
circuits are not the same. Undesirable currents, which may
be of considerable magnitude because of the low resistance of
the winding, then circulate between the different circuits,
overloading some brushes and causing them to spark, and

120 ARMATURE WINDING AND MOTOR REPAIR

uselessly heating the armature. The brushes may be pro-
tected by diverting these currents through internal equi-
potential connections, but the useless heating of the armature
still remains. This, however, is usually not excessive com-
pared with the heating of other losses, so that this winding may
be called standard today.

Series-parallel wave windings with circuits equal in number
to the poles possess the disadvantage, common to all series
wave windings, that the terminals of each coil lie under two
different brushes, which are of the same polarity and, therefore,
connected by a low resistance conductor. That is, a coil is
commutated by two brushes and the connector between them
and there is a tendency for selective commutation or unequal
division of the current among the brushes to occur, especially
in large non-interpole machines. This generally results in
sparking at some of the brushes. They are eaten away so
that the effective position of their contact surfaces are changed.
Other brushes then get more current than they should and
eventually all the brushes are damaged. Series-parallel
wave windings do have the advantage that each circuit is
influenced by all the poles, so that the induced emfs in all
circuits are equal and there is little tendency for circulating
currents to occur due to unbalanced induced emfs. To
eliminate the disadvantages and combine the advantages of
these two classes of windings, multiple wave windings have
been used, particularly on four-pole machines. Although they
are " special" and, therefore, cost more than standard windings,
they accomplish their purpose admirably. There seems to be
no reason why they would not be equally successful on ma-
chines with six or more poles.

Use of Equalizer Rings. Equalizer rings are being used on
lap windings only on large multipolar machines, where un-
balanced conditions in the magnetic circuit are liable to cause
circulating currents between the several paths through the
armature. This condition will occur for instance in the
course of operation where the air gap will not be uniform
so that the emf induced in a coil opposite a small gap will
be larger than that induced in the coil which is opposite the
large gap and this difference in the emfs will bring about

MAKING CONNECTIONS TO THE COMMUTATOR 121

a circulating current which will flow between the windings and
thus interfere with the performance of the machine. This
condition will also occur if the machine has been in operation
any length of time, the inequality of the air gap being brought
about by the wear in the bearings. The same condition will
occur if the brushes are not spaced properly. In a wave
winding, each circuit has its conductors pass under all poles
and, therefore, there is no necessity of using equalizer rings.

CHAPTER V
TESTING D.-C. ARMATURE WINDINGS

The common causes of trouble in armatures are practically
the same as in any electrical circuit, namely, a short circuit
in or between coils, an open circuit, reversed coils and grounds.
Of all the faults inherent to armatures, probably the most
dangerous is the short circuit between coils. If it is not
detected and remedied as speedily as possible, the result in
most cases is the burning out of the coils affected, and possibly
the whole armature.

Causes of Short Circuit in an Armature. There are numer-
ous ways in which a short circuit of coils may occur. In the
case of wire-wound coils, it sometimes happens that one of the
turns forming the coil becomes twisted during the process
of inserting the coil in the slot, and in order to force the wind-
ing down to an even depth, the turn of wire was driven down
upon other turns, cutting through the insulation and causing
a short circuit between turns of the same coil. When this
occurs, the resistance of the coil is reduced, allowing more
current to flow, increasing the temperature of the coil and
eventually causing a deterioration of insulation on other wires
at that point. In the majority of cases this results in short-
circuiting the entire coil upon itself.

Also a frequent cause of trouble is the short circuit between
coils This is often caused on the back end of the armature
by oil soaking into the coils by leakage from the out-board
bearing, which, together with the dust that will invariably
work in between the windings, break down the insulation and
cause an electrical leak between coils. A short circuit may
also result from the top and bottom armature leads coming
in contact with each other. A short circuit between commuta-
tor bars is often the cause of burned out coils. In soldering
leads to the commutator, great care must be exercised not to

122

TESTING D.-C. ARMATURE WINDINGS 123

allow any of the molten solder to run down behind the bars,
as this very frequently short circuits the bars. A good way
to avoid this is to raise the back end of the armature a trifle
so as to allow the solder to run to the front where it can be
easily removed.

Tests for a Short Circuit in an Armature. Probably the
best way of detecting the presence of a short circuit in the
armature of a motor in operation, is to carefully watch it
when starting as soon as faulty operation is noticed. Some-
times the armature will not start upon the first few points
of the rheostat, and will then take an excessive current. This
will cause it to run with a slow and unsteady motion (especially
at low speeds), due to the fact that every time the short-cir-
cuited coil comes under the influence of a pole, it will have a
tendency to retard the motion of the armature. By running
the motor for a short time, the bad coil will heat much more
than the others, and its location can usually be detected by
passing the hand over the end windings.

When a short circuit is suspected in an armature, the ma-
chine should be shut down at once and the necessary repairs
made. One sure method of locating a short-circuited coil
is to disconnect all the leads from the commutator and test
out the coils with a test lamp. A test lamp consists of two
wires about 10 feet long, connected to a 110-volt circuit with
an incandescent lamp connected in series in the circuit. This
method requires a great deal of work in unsoldering all the
leads, which may not be necessary since in the majority of
cases, the seat of the trouble will be found in the commutator
itself, due to short circuits between bars.

A rapid test often used to locate the trouble without dis-
connecting any of the wires on the commutator during the
test, is the bar to bar test. This test can be applied to arma-
tures with any style of winding connections, for there will be
exactly the same drop of potential between any two adjacent
commutator segments no matter which scheme of connection
is used. Fig. 100 shows the connections necessary for a test of
this kind. A steady current, taken from a 110-volt circuit
should be sent through the armature at opposite sides of the
commutator. The brushes B should only be wide enough to

124

ARMATURE WINDING AND MOTOR REPAIR

cover one bar. C is a fiber block holding the copper contact
points, so spaced as to rest on adjoining segments as shown.
Adjust the lamp bank until the voltmeter gives a readable
deflection when C is in contact with what are supposed to be
good coils- The deflection of the voltmeter will depend upon
the difference of potential between the bars. If everything
is all right, practically the same deflection will be obtained
all around the commutator regardless of what pair of bars C

Mail

FIG. 100. Connections for testing out armature coils with a millivoltmeter.

may rest upon. Pass the contact points over each pair of
bars and note the deflection on the voltmeter. When the
short-circuited coil, to which the bars are connected, comes
under the contacts there will be very little if any movement of
the needle, because there will be little or no drop through the
coil A more satisfactory test for use on a removed armature
is the transformer test discribed on page 125.

When the coil at fault has been found in this manner, its
leads should be disconnected from the commutator, together
with the leads adjoining it on either side. The commutator
should now be tested by use of the test lamp to determine if
the bars to which the coil was connected are short-circuited.

TESTING D.-C. ARMATURE WINDINGS 125

The banding wires should be removed from the armature
next, and the defective coil taken out by raising the top sides
of other coils clear of the armature as far around as the bottom
side of the damaged coil, when it can be lifted out. In many
cases the insulation on the wire of the coil has reached such a
stage of deterioration that a new coil will be necessary, in
which event a new one should be formed with exactly the
same number of turns and size of wire as the old one. Care
should be taken not to wrap a thicker layer of tape on the
new coil than was on the old one, for if this is done trouble
will be experienced in forcing the coil back into the slot. The
finished coil should be given a coat of insulating varnish. It is
a good plan to re-insulate the armature slots also, before re-
turning the coil. If the commutator is free from short-cir-
cuits, the coil may be replaced, the raised coils returned to
their slots, and the leads soldered to the commutator again.

Testing for Short Circuits and Open Circuits with a Small
Transformer. While the method of testing between adjacent
commutator bars with a millivoltmeter will indicate short-cir-
cuited or poorly soldered leads by a low reading and open-
circuited or poorly soldered leads by a high one, it often occurs
that an armature is rewound and reinstalled with considerable
time and labor and found to be defective after all. The
millivoltmeter or drop of voltage method merely measures
the resistance of each coil but when an armature is subjected
to magnetic induction, an emf is induced in its windings
which will cause current to flow in the turns that are short-
circuited. When one turn in a coil has been forced so hard
against another that the insulation is broken and the turns
become short circuited, the millivoltmeter test may not serve
to detect the short circuit and as a consequence the turns, and
probably the coil, would be destroyed by the immense current
which would flow in the short-circuited turns when running
in the magnetic field of the machine.

A simple way of detecting such defects is by the use of a
small transformer often called a "mill" or "bug" and con-
structed as shown in Fig. 101. When this transformer is
applied to the armature core as shown, the alternating flux
produced by it flows through the core and produces an emf. in

126 ARMATURE WINDING AND MOTOR REPAIR

the coils. If the winding is correct no current will flow as
the voltages will balance each other. If, however, a coil is
short-circuited, a current will flow in the turns that are short-
circuited. To locate the defective coils take a sharp piece of
steel or a knife blade and pass it around the commutator so as
to short circuit in succession, the coils which have one side
under the transformer. A decided sparking, indicating a
potential difference between the bars, shows that the coil is
in good condition. Absence of sparking indicates either

FIG. 101. Method of testing an armatuie for short circuits and open circuits
in coils by use of the special transformer shown.

an open circuit or a short circuit. The latter can be readily
determined by running a light piece of sheet iron over the
surface of the armature core so as to bridge the slots in suc-
cession. If there is a short circuit in one of the coils which
has one side under the transformer, a local current will flow
through this coil generating a magnetic flux which will attract
the piece of sheet iron. If it is held away slightly it will be
made to vibrate very rapidly. The coil will also heat very
rapidly and if the transformer is large enough for the armature
being tested, the coil will be burned out completely.

In case there is no sparking at the commutator when the
coils are short-circuited as described above and there is no

TESTING D.-C. ARMATURE WINDINGS 127

local magnetic flux when moving the piece of sheet iron over
the slots of the core, an open circuit is indicated.

A transformer cf the dimensions shown in the accompanying
illustration can be operated on a 110- volt, 60-cycle circuit and
will serve for testing many sizes of armatures. The one
illustrated was made up by E. W. Copeland (Electrical World)
using 60 turns of No. 6 magnet wire. When using such a
transformer it should be fastened under the armature so that
there is just enough clearance between the transformer and
armature core that neither will touch. Current should always
be off while placing the transformer and also when it is not
in use.

When testing an armature with a small exploring trans-
former, the number of poles in the machine must be taken
into consideration. In a two-pole armature, a single short
circuit may heat up two coils, in a four-pole armature, four
coils and so on. A good way to locate the defective coil is
suggested by Maurice S. Clement (Electrical Record, Novem-
ber, 1918) by applying the telephone receiver test. In apply-
ing this test the terminals of a telephone receiver are placed
on adjacent commutator bars which are connected to one of
the affected coils, and the volume of sound transmitted to the
receiver noted. The same should be done to all the other
affected coils. A short circuit will have a greater volume of
sound than a perfect coil.

Causes of Open Circuits in an Armature. An open circuit
may result from a number of causes. In the first place, when
the armature was wound, the coil may have been driven into
position in such a manner that one of the wires was strained
or partly cut in two. The momentum of the armature, and
constant vibration of the machine will finally break the wire,
and in this way form an open circuit. Sometimes an open
circuit of this kind will only show up when the armature is up
to speed, the centrifugal force causing the wires to separate,
thus opening the circuit. When the motor is at rest, the wires
will come together again, and a test will reveal nothing. This
condition is known as a " flying" open circuit, and occurs
quite frequently. The same state of affairs may result with
a short circuit between overlapping coils. An open circuit

128 ARMATURE WINDING AND MOTOR REPAIR

may also be caused by the armature leads being drawn too
tightly when they are soldered to the commutator. This will
cause a break due to expansion and contraction of the wire
from the constant heating and cooling.

Another common cause of open circuits is poor workman-
ship when the leads are soldered to the commutator. If the
lugs or risers are not perfectly tinned before attempting to
solder the leads into them, the solder will not take hold over
the entire area, and a lead may be held in place only by a
thin film of solder on the outside surface. When the current
through the armature is heavy, the contact area between the
riser and the leads may not be sufficient to carry the necessary
current without excessive heating. This will melt out what
little solder there is, and an open circuit will result. Sometimes
a commutator will become so hot from excessive brush friction,
resistance drop, overloads, or the like, that it may throw
solder, and cause an open circuit.

Tests for an Open Circuit in an Armature. The symptoms
of an open circuit are often very prominent. A vicious green-
ish-purple spark will usually appear at each brush as the
open-circuited coil passes from one pole to the next. This
spark has a tendency to leap out from the brush and follow
around the commutator for quite a distance. The bars to
which this coil is connected will be found to be burned and
roughened, and the mica insulation between eaten out to a
considerable depth.

In a lap wound armature the position of an open-circuited
coil is easily located, because each end of the coil is connected
to adjoining bars. In a wave winding this is not the case.
Each end of a coil is connected to a bar removed a certain
distance around the commutator from the other, depending of
course, upon the number of poles and the winding pitch
employed.

If an open circuit exists in an armature for any length of
time, the burned condition of the commutator bars will usually
indicate where the trouble is located. Both lap and wave
wound armatures may be tested for open circuit by a testing
transformer, by use of the ordinary test lamp, and a bar to bar
test, or by ringing out between adjacent bars with a magneto.

TESTING D.-C. ARMATURE WINDINGS

129

If the test lamp is used, test the commutator from bar to
bar and note the brightness of the lamp on each pair of bars.
When the bars are reached to which the open-circuited coil is
connected, the light will dim considerably, and may go out,
depending upon the resistance of the winding. When more
than one coil is open-circuited, the winding will be divided into
two or more sections, and the test lamp will only light when
the test leads are in connection with the bars in each section.

The testing set described for locating short circuits (page
124) may also be used for open circuits. Proceed in the same
manner as when testing for short circuits. When the con-
tact points C (Fig. 100) are connected to the open-circuited
coil (indicated at D), there will be a violent throw of the needle,
because the voltmeter will then be connected to brushes B
through the intervening coils. When C is moved to the next
segments, there will again be no deflection, thus locating the
break definitely.

If an open circuit results from a lead breaking off at the
commutator, it is an easy matter to solder it back again.
When the break occurs within the coil itself, a new one must be
substituted, as described for the short-circuit test.

y. y

'I''

i i ' ' '

' i i

i i M

1 234 S

B B'

FIG. 102. The illustration at the left shows method of bridging a coil of a
wave winding. That at the right for cutting out a coil of a lap winding.

Cutting out Injured Coils. In case of emergency, the bad
coil can be cut out of circuit and the commutator bars con-
nected together by a wire large enough to safely carry the
current. This wire should be well insulated from the other
leads, as any connection with them would constitute a short-
circuit. In Fig 102 the method of bridging out a coil in a

130 ARMATURE WINDING AND MOTOR REPAIR

wave winding is shown. When one of the coils is short-cir-
cuited as shown at A, the top side of the coil is disconnected
from bar B, and the bottom side from bar B' Jumpers
should be soldered in as shown by the dotted line. The ends
of the coil leads can then be cut off close to the armature core
and taped. The coil should be cut completely in two at X
and X', and the ends taped. This will prevent self-induced
currents from being generated within the coil, which might
cause heating and injure the insulation on other good coils. In
Fig. 102 (at right) the method of cutting out a coil in a lap
winding is shown. The coil is open-circuited at A. The top
side of this coil should be disconnected from bar 3, and the
bottom side from bar 4. In this case the jumper must be run
from bar 3 to bar 4. The dead coil can be taped up the same
way as the series or wave coil mentioned above.

One coil cut out of an armature will not perceptibly affect
the running of a motor, and several of them can usually be
cut out with safety, providing they are not bunched together.
It is not wise to cut out too many coils, as this increases the
heating and speed of the armature and lowers the efficiency of
the machine.

Causes of Grounds in an Armature. A ground occurs when
current leaks from the current carrying parts of the armature
into those parts that are not intended to carry current. A
single ground will have little effect on the operation of a motor,
but it should be removed as soon as possible, as there is always
danger of a second ground coming on at some other point,
which would produce the same effect as a short circuit. When
a ground occurs, a small hole will be found burned through the
insulation and into the iron parts of the armature. Across
this carbonized insulation, current will pass. Grounds occur
very frequently on the ends of the armature core at the points
where the coils leave the core. If the bend has been too
sharply made, or has been hammered too hard, the sharp
edge of the core will cut through the insulation. To avoid
this, the slot insulation should extend at least one-quarter
to one-half inch past the end of the core on each end.

Grounds also frequently occur in the commutator, caused
by oil creeping up on the mica ring. Combined with the

TESTING D.-C. ARMATURE WINDINGS 131

copper and carbon dust from the commutator, this forms a
good path for leakage of current.

Tests for Grounds in an Armature. A ground can usually
be located by using the test lamp. Disconnect the leads from
four of the commutator bars on one-quarter of the circumfer-
ence. This will determine the section of the winding in
which the ground is located. Raise the leads of the defective
section out of the commutator. Place one wire of the test
lamp on the shaft, and with the other, test each coil separately
to locate the ground. Sometimes the trouble may not b& in
the armature, but may be caused by a grounded commutator.
If the coil is at fault, it should be removed and reinsulated.

Reversed Coils. A reversed coil, that is, one with the leads
to the commutator reversed, frequently occurs. A practical
way of locating a reversed coil is to pass a current through
the armature at opposite points. The lamp bank and con-
nections of Fig. 100 can be used for this test. Then with a
compass or small bar of magnetized steel explore around the
armature to determine the direction of magnetism from slot
to slot. When the compass is over the reversed coil, the needle
will reverse, giving a very definite indication of the coil which
is connected wrong. The leads of this coil should simply be
reversed.

Use of a Bar Magnet and Millivoltmeter to Locate a
Reversed Armature Coil. In armatures where both leads of a
coil are taped together, and led out from an identical point,
there is considerable danger of getting the leads crossed while
connecting, thus reversing the direction of flow of the current
in that coil. Such a reversal will in all cases "light out"
as though perfect when tested with a lamp, but will cause
bucking when the machine is run. A reversed coil of this
sort is unusually difficult to locate and in many cases a whole
machine has been stripped because ordinary methods of testing
failed to locate the trouble.

An efficient test which has been used by J. G. Yoerns
(Electrical Record, September, 1918) is illustrated in Fig. 103.
It will be noticed that a millivoltmeter is used rather than a
voltmeter because of the greater sensitiveness of the former.
Both terminals of the meter are connected to adjacent

132

ARMATURE WINDING AND MOTOR REPAIR

commutator bars. Next, take a piece of metal which has
been magnetized and move it in a direction corresponding
to the revolving of the armature, directly above the coil
to which the meter is connected. It is well to keep in mind
the fact that if a clockwise motion is used on the first coil, the
same direction must be retained on all remaining coils,
otherwise the meter reading will be backward. On the
downward stroke of the magnetized bar and as it approaches
the coil to be tested, magnetic lines of force will travel from
the 'bar to the coil, thence to the meter, causing the needle
to fluctuate slightly. If the needle jumps ahead on the
downward stroke, it will jump backward on the upward
stroke, and vice versa. The reversed coil will read opposite
from the others. To change the direction of the fluctuations
of the needle, either reverse the meter terminals or reverse
the motion of the magnet.

iwrfvorfrnefer Compass

FIG. 103. Method of using a
magnetized bar and millivoltmeter
to locate a reversed armature coil.

FIG. 104. A convenient method
for using a compass when testing
for a reversed coil.

Use of a Compass to Locate a Reversed Armature Coil.

In connecting up small armatures with lap windings wound on
by hand with four or more leads coming out of each slot,
the leads may be easily confused as already mentioned so that
some of the individual armature coils will be reversed. E. C.
Parham (Electrical Record, August, 1918) has, therefore,
suggested the use of the testing device shown in Fig. 104.
While several reversed coils distributed around the armature

TESTING D.-C. ARMATURE WINDINGS 133

may not cause sufficient effect to excite suspicion, they are
likely to give trouble in time. It is conceivable that if alter-
nate coils were reversed, a highly improbable condition, the
armature would be inoperable because then there would
be an equal number of coils tending to turn the armature
in opposite directions.

The accompanying diagram (Fig. 104) shows a simple cheap
method of readily locating any reversed coils that may exist.
The armature rests in a support A that permit of rotating the
armature as the test progresses. A strip of copper B is bent
over the armature, as indicated, and a compass placed upon it.
Current from an incandescent lamp test circuit is then applied
to adjacent commutator bars that are connected to the coils
that lie in the slot that is immediately under the compass.
Suppose that the compass needle is deflected to the right on
touching bars 1 and 2 and bars 2 and 3, there being two coils
per slot. Rotate the armature until the next slot comes under
the compass and touch the test points to bars 3 and 4 and then
to bars 4 and 5, and so on all round the commutator.

The compass deflections obtained should be always in the
same direction. Any pair of adjacent bars touched by the
test points causing a reversed deflection, includes a coil the
leads of which have been brought down to the commutator
in reversed order. In order to test the effectiveness of the
method, it is necessary to only apply the test points to adjacent
commutator bars in reversed order and observe that the com-
pass deflection is thereby reversed.

Locating Low Resistance or Dead Grounds. It is often
difficult to locate a low resistance or "dead ground" in a low
resistance armature owing to the very low resistance of the
windings themselves. In such cases the following method
can be used:

First, short circuit all commutator bars by winding several
turns of bare copper wire around them; then apply a source
of energy, direct current being preferable, to the commutator
and shaft. The voltage to be used depends upon the resistance
of the "ground. " This produces a circuit from the commuta-
tor through the grounded coil to the ground and out through
the shaft, thus setting up a field around the conductors in

134

ARMATURE WINDING AND MOTOR REPAIR

this coil. By applying a small piece of iron to the surface of
the armature core and gradually moving it around, the
grounded coil can be located by means of its field, which will
attract the iron.

The same method can also be applied to alternating current
windings although not quite so readily. For example, in
the case of a three-phase, single-circuit, F-connected armature,
first disconnect the F, splitting the winding up into three
separate circuits. Then test out each circuit with a magneto

i i*

i -.

y j/

I
1

O 1

W Ground to Armature Shaft
FIG. 105. Method of locating grounds in an armature winding.

or test lamp. Next apply a current to one end of the grounded
circuit and to the shaft. Assume that there are 12 coils as
shown in Fig. 105, coil No. 7 being the grounded coil while
coil No. 1 is connected to the line as shown. There will then
be a circuit through coils Nos. 1, 2, 3, 4, 5, 6, and 7 which can
be readily detected with a piece of iron as already explained,
while coils Nos. 8, 9, 10, 11, and 12 are dead. It is then, of
course, obvious that if coils Nos. 1, 2, 3, 4, 5, 6, and 7 carry a
current while coils Nos. 8, 9, 10, 11, and 12 carry no current,
the ground must be in some section of coil No. 7, the circuit
being completed at that point.

TESTING D.-C. ARMATURE WINDINGS

135

Use of a Telephone Receiver in Testing for Short Circuits,
Open Circuits and Grounds in an Armature. The telephone
receiver on account of being very sensitive to sound, makes
a convenient testing instrument in a repair shop. The ways
it can be employed to discover a short circuit, open circuit,
or ground are given here as used by Maurice S. Clement
(Electrical Record, December, 1918).

This testing device can be used either with alternating cur-
rent or direct current. If alternating current is to be used, a
test lamp and a pair of
leads from a 110- volt cir-
cuit are connected to the Source of Energy

commutator as in Fig. 1C6,
from one-fourth to one-half
of the circumference apart.
Next, take the receiver
which has about two feet
of two-wire telephone cord

attached and hold it to the FlG i 06 .-Connections for using a
With the Other hand telephone receiver for locating short-
i i circuits and open circuits in an armature

ear.
press

the

recever

winding.

firmly to the commutator,

taking care to touch adjacent segments. Move from one lead
of the test lamp to the other segment by segment and repeat
the operation until the commutator has been circled.

If the wire in the coils is of too low a resistance to make a
buzz in the receiver, put a rosette fuse in place of the test
lamp and cut a rheostat in series with the fuse. This will
bring the resistance up sufficiently to make a buzz in the
receiver. A low buzz indicates a good flow of current; if
no sound whatever can be heard from two commutator bars,
there is a dead short circuit. An open circuit is indicated by a
very loud buzz. A "cross connection will produce a defective
sound on three segments. These three leads should be taken
off and reconnected immediately and the receiver test once
more applied.

For a dead ground which persists in remaining invisible,
place one side of the receiver line to the shaft, as in Fig. 107,
and the other side of the test line to the commutator, then,

136

ARMATURE WINDING AND MOTOR REPAIR

Test Lamp
Source of Energy j '^

with the receiver to the ear, buzz each segment. The

grounded coil will buzz louder than the rest.

If direct current is to be used, the source of energy should be

a battery. A buzzer connected in series on one side of the

battery completes arrangements.

Sometimes, when testing an armature with a transformer,

a single short circuit will heat up coils in two, four, six or eight

places, according to the
polarity of the winding.
By applying the receiver
leads to the bars con-
nected to each coil thus
affected, a short circuit
can be located quickly.
Testing for Reversed

Ground on -^ and Dead Field Coils.

Machine Frame In multipolar machines

FIG. 107 Connections for using a tele- tne po l a rity of field coils

phone receiver for locating grounded coils J .

in an armature winding. can be tested with a

carbon filament lamp.

By placing the lamp while lighted between the pole tips, the
loops of the filaments will draw close together or separate far
apart depending upon the direction of the magnetic flux. A
dead pole can be quickly found in this way.

Another very simple and positive method for testing the
polarity of field coils is by means of two ordinary iron nails.
To do this, pass a current through the field coil winding and
place the nails on two adjacent poles. If the polarity is right,
they will attract each other and if wrong they will repel each
other. This method can be used also for detecting a dead coil.

The Commutator. Trouble in a commutator may be traced
to either a short circuit or a ground. In a commutator a
short circuit may be caused by any of the following: Particles
of metal touching two adjacent bars, or mica burned away
allowing current to arc across. In an undercut commutator
be sure the slots between bars are thoroughly free from all
dirt before testing. A grounded commutator must invariably
be removed from the shaft, and the damaged part well cleaned
and reinsulated. In a case where a motor has been through a

TESTING D.-C. ARMATURE WINDINGS

137

138 ARMATURE WINDING AND MOTOR REPAIR

fire, the segments will in nearly all cases be undamaged, but
the mica will be burned beyond future use, and in such a case
refilling will be necessary.

Testing Equipment for a Repair Shop. In addition to the
facilities necessary for testing out armatures and machines
for short circuits, open circuits, grounds and the like, every
repair shop must be equipped to test and load motors of differ-
ent sizes requiring voltages from 110 to 2300. A voltage of
from 1200 to 1500 volts is needed for testing an armature for
grounds. A transformer is therefore required with taps and
connections so that the voltage can be varied from about 500
volts up to 2500 volts.

For testing for short circuits and open circuits a small
transformer with specially shaped laminations about which a
coil of wire is wound should be available (see Fig. 101 on page
126). When applied to an armature with a short-circuited
coil, the defective coil will show up by getting hot.

For testing out circuits for opens, and to match up ends of
coils when inserting them in the armature core, a test lamp and
terminals is convenient. A magneto is also used for this
purpose. In the case of testing for open circuits, two leads
connected to a 110- volt circuit with a lamp connected in series
in one of the leads is sufficient. For selecting coil terminals
or " lighting-out " an armature winding while in the machine
to discover trouble, a 6- or 12- volt automobile storage battery
provided with two leads having snap testing clips and a lamp
in series with one lead, is a convenient outfit that can be taken
about the shop and to any job.

At least one portable ammeter and one voltmeter are needed
and one ringing out magneto and a pair of head telephone
receivers.

For testing motors a small switchboard wired so that con-
nections can be made on the front by plugs and the voltage
increased by using jumpers, is a convenient outfit that saves
much time in making a test. This switchboard should have
mounted on it, a voltmeter, ammeter, frequency meter and
the necessary switches to connect and operate both motors and
generators on either direct or alternating current.

CHAPTER VI

OPERATIONS BEFORE AND AFTER WINDING
D.-C. ARMATURES

Before an armature goes to an experienced armature winder
and after it leaves his hands, there are certain steps in the
process which can be properly termed "before" and "after"
operations. In a case where a damaged armature must be
Completely rewound, these operations may be outlined as
follows :

Operations before winding

Operations after winding

Stripping off old winding.

Cleaning slots and ends of core.

Filing burrs off slots.

Testing commutator.

Repairing commutator.

Making new coils.

Insulating ends of core and slots.

Testing out winding.
Soldering leads to commutator.
Hooding armature.
Banding armature.
Turning commutator.
Undercutting mica of commutator.
Balancing armature.
Painting armature.
Relining bearings.

Reference has already been made to a number of these
operations in connection with the procedure in rewinding
machines as outlined in Chapters III and VIII. The details
that are given here refer to the requirements in all cases, with
references to other Chapters where the operation has been
taken up as a special subject.

Stripping Off an Old Winding. In this operation when
band wires must be removed, they should be cut or filed in
several places. When a chisel and hammer are used care must
be exercised not to mash down the armature teeth. The next
step is to unsolder the leads to the commutator and clean out
the slits hi the commutator necks carefully. By pulling out

139

140 ARMATURE WINDING AND MOTOR REPAIR

the top sides of the proper number of coils according to the
throw, the bottom sides can be reached that will allow the
coils to be removed easily. (See Chapter III.)

Cleaning and Filing Slots. After the coils have been re-
moved, all old insulation must be thoroughly taken off the
core and the slots. A solution composed of 25 per cent, alco-
hol and 75 per cent, benzole is good for loosening the varnish
of old insulation so that it can be scraped from the slots in
the armature. This will produce no bad effects on the lami-
nations or on the winding when the armature is rewound.
Alkali solutions such as caustic soda will also loosen the insula-
tion without injury to the laminations, but will creep be-
tween the laminations and after the armature is rewound the
alkali fumes are liable to damage the insulation. A tool
made from bar steel one by one-sixteenth inch, of suitable
length and drawn down to a long thin point like a chisel
answers very well to remove old insulation after it has been
softened, and will also give very satisfactory results with-
out the use of any chemicals whatever. After removing
the insulation in this way a file drawn through the slots will

remove small pieces of insulation
and smooth off all roughness. The
edges of the slots should be filed to
X siit 8 for remove sharp edges and burrs that
thread to tie would injure the new coils while

coiltogether ^.^ ^^ ^ ^ ^^ ^

entire core should then be thor-
with a blast of

-KJ

oughlv cleaned

FIG. 109. Shuttle form for
winding a coil that can be pulled compressed air.
into a diamond shape in a spe-
cial pulling machine (Fig. 291).

Testing Commutator. Details
of this operation are given in
Chapter XII as well as the steps in repairing a commutator,
testing it out after reassembly, baking the insulation and
tightening end rings.

Making New Coils. This operation calls for an inspection
of the old coils and winding data as outlined on page 57,
Chapter III. In case of changes in speed or voltage it requires
certain calculations which involve the size of wire to use. These
calculations are given in Chapter X, page 240. When the

OPERATIONS BEFORE AND AFTER WINDING 141

size of wire is known the winding of the coils can be done on

dne of the forms shown in Figs. 109 to 114.

In making coils in large repair shops three methods are
used, namely winding on a mould, on a former or on a shuttle.
Mould coils are usually those made -on a form rotated in a
lathe with all necessary shaping done with very little
pounding on the conductors. By formed coils is usually
meant, those coils made over a stationary form with the bends

FIG. 110. Three steps in the construction of shuttle wound coils (Fairbanks-
Morse & Company).

(a) The coil at the bottom is shown as wound on a shuttle form using square copper
wire. (6) At the left it is shown after being pulled into final shape, dipped in insulating
varnish, thoroughly baked, and then taped, (c) The finished coil is shown at the right
after four to six alternate dipping and baking treatments.

made by the use of levers and mallets to force the coil to the
proper shape. Shuttle or "pulled" coils are first wound on
a simple shuttle such as shown in Fig. 109 which is fastened
to a lathe and then pulled on a coil puller to the shape
required for the particular throw of the coil as shown in
Fig. 110.

Forms for Winding Coils Like Those Previously Used.
A simple form that can be made from a pine board for dupli-

142 ARMATURE WINDING AND MOTOR REPAIR

eating the shape of coils previously used in an armature that
is being rewound is shown in Fig. 111. The dimensions and
general shape can be secured from a sample coil preserved
from the old winding. In the illustration A is a small hole
for mounting on a spindle. Small pins B are driven into the
form to hold the coil in place while winding. C and C 1 are
slots cut into the sides so that the several turns of the coil
may be tied together with thread. This keeps the wires
together after the coil is removed from the form. At the ends
D and D f two pins are used in order to make the turn shown
on the end section at the right in Fig. 111.

FIG. 111. Form for shaping diamond armature coils to match old ones used.

Now mount the form on a spindle so that it can be revolved
and wind on the required number of turns, using the pins as a
guide. If the completed coil is to consist of two coils, take
two wires and wind them together. When the proper number
of turns have been made, cut off the wire and tie the coil with
thread at C and C'. Remove the coil from the form and com-
pare it with a sample if such a sample was preserved from the
old windings. It is also well to try this first coil in the arma-
ture, and see that it has the proper span, and that the leads to
the armature are long enough.

The following method for winding coils for small motors
has been found convenient by Maurice S. Clement (Electrical
Record, November, 1918) in those cases where a repair must

OPERATIONS BEFORE AND AFTER WINDING 143

be made on the job or coils made up where a coil winding
machine is not available. The form is made up as follows:
First take a flat piece of wood about three-fourths of an inch
thick and plot out the shape of the coil. Then place right
angle screw hooks, at all angle points. The screw hooks
should point away from the center of the form. Mark out the
center and run a breast drill through as far as it will go. Place
the breast drill in a vise in a horizontal position with the coil
form toward the left, taking care to place it so as to allow free

Bolts /
extending
through
form

FIG. 112. FIG. 113.

FIG. 112. A six gang form for winding coils in series.

Six coils can be wound in series and removed in a group so they can be inserted in the
slots of an alternating current motor stator without the necessity of makihg the series
connections. This form can also be used to wind a single or double coil. It is built up in
sections which are held together by the bolts at A. The slots in the divisions between
sections enable the winding and removal of the coils in series.

FIG. 113. Continuous jointless phase-group coils used on rotors of
Fairbanks- Morse phase wound motors. These coils are wound on a form
similar to that shown in Fig. 112.

movement of the handle. Turn the wire of which the coil is
to be made, once or twice around a nail on the back of coil
form, and lead the wire over the edge to the face of the form
and turn the handle. The form can be revolved at any de-
sired rate of speed and the wire run over all the screw hooks.
Keep sufficient tension on the wire to permit each turn to lie
snugly beside its predecessor.

This is a convenient way to wind coils, for the reason that
unnecessary crossing of wires can be prevented. When the
specified number of turns have been wound, twist a short
piece of wire around each end of coil to hold it in proper shape.

144

ARMATURE WINDING AND MOTOR REPAIR

To finish the operation, turn all screw hooks toward the center
of the form and slip the coil off.

Another hand-made form also recommended by Mr. Cle-
ment for use in winding larger coils than the one described
above is shown in Fig. 114.

Insulation of Core and Slots. The insulation needed with
coils used in partially closed slots and in open slots of alternat -
ing current machines is given in Chapter VIII and in connection
with the winding of direct-current machines in Chapter III.

Slit through whicli
First Lead of Coil
Passes

Slits through which
Strings are Laid
to Bind Coil

-Handle

-Clamp

Plate

FIG. 114. Construction of a convenient bench winder for forming armature

coils.

Testing Out the Winding. Details of this operation both
before the leads to the commutator are connected and when
trying to locate troubles in the winding are given in Chapter V.

Soldering Coil Leads to the Commutator. After the wind-
ing of an armature has been thoroughly tested out, the coil
terminals can be soldered to the commutator. In the solder-
ing operation great care should be taken that pieces of solder
do not fall or run down back of the commutator to later pro-
duce a short-circuit and cause the armature to be returned
for further repairs. To prevent damage to windings, acid
fluxes should never be used in soldering a commutator. A
solution of rosin in alcohol is recommended instead. A tin
and lead solder is considered best in soldering leads to the
commutator but a pure tin solder is used in making all other

OPERATIONS BEFORE AND AFTER WINDING 145

joints on the coils as the insulation is less liable to be damaged
with this solder on account of its lower melting point. When
tin is used the best results are obtained by working upon
the side of the armature, so that the joint is level. After
soldering, the armature should be mounted in a lathe and the
rough solder on the necks of the commutator turned down,
the commutator polished and the wiper rings turned to give
the exact distance between bearings. In some armatures
wedges are inserted in the slots above the coils. These extend
above the surface of the banding grooves and should be turned
down while in the lathe if they are used, so the banding grooves
will present a smooth bed for the band wires.

Hoods for Armatures. In order to protect armature end
connections of railway motors and of mill motors that must
be used in places where dirt and dust is liable to accumulate
on the armature, a heavy hood is often put over the ends of
the coils. For the commutator end a hood of woven asbestos
is suitable. This hood is usually sewed in a conical shape and
impregnated with a moisture and oil-repelling compound
and fastened in place while wet. The small end should be
drawn up over the commutator and turned inside out and
firmly tied over the coil leads and commutator necks with
h?avy twine. The body of the hood should then be turned
back over the armature. If the commutator necks are lower
than the level of the core, another layer of twine should be
wound over the hood near the commutator and a band of
canvas sewed over the whole. The hood should then be
stretched tightly back over the armature and tied with twine.

Around the rear end of the armature a band of canvas should
be wrapped so that the greater part of the strip extends out
over the shaft only enough being wound over the armature to
permit a secure fastening. This should be bound in place
with a band of twine wound tightly in the groove between the
coil ends and the end bell. The canvas should then be turned
back over the armature and bound smoothly in place.

In case the armature is to be banded with steel wire, two
strips of cotton tape separated by a band of varnished paper
should be wound over the hood and end connections near the

core of the armature, as a base for the banding wires.
10

146 ARMATURE WINDING AND MOTOR REPAIR

Banding Armatures. In repair shops where banding
machines are not available, the banding of armatures is done
in a more or less indifferent way with the result that the
banding may only perform a part of its function, namely,
to prevent the coils from being thrown off the armature core.
Unless the banding is placed on the armature with great care,
there may be sufficient movement of the coils in the slots to
wear the insulation and cause grounds. Such movement is
also liable to cause breaks in the copper leads where they are
soldered rigidly into the commutator. The details for banding
an armature given in what follows, are taken from an article
in the Electric Journal and represent good shop practice that
can be followed in both large and small repair shops.

Shrinking Coil Insulation. Since the coil insulation shrinks
upon being heated, it is necessary to shrink it as much as
possible before the final banding wire is applied. This is
done by heating the whole armature to about 75C. (167F.),
when the insulation becomes pliable and can be pressed down
into permanent shape.

Temporary Bands. The hot armature should then be
mounted in a lathe and a protecting strip of cloth placed over
the end windings. Wind a temporary banding wire over the
coils with enough tension to draw them down into place, and
fasten the ends by soldering tin clips over the wire. The
armature should then be allowed to cool. After the tem-
porary wires are removed the armature is ready for the
permanent banding.

Banding Machine. When a banding machine is used, the
tension in the wire is regulated by passing it over a train of
friction pulleys, mounted on the carriage. The friction of the
pulleys can be adjusted to any desired value by the regulating
screws. In the absence of such a device, fair results can be
obtained by passing the wire two or three times around a
round wooden banding stick approximately two inches in
diameter and adjusting the tension by hand.

Core Bands. When core bands are used, the grooves should
be fitted with thin strips of tin, which protect the coils from
the cutting action of the bands. In starting the permanent
banding, wind a few turns at one end fco secure the necessary

OPERATIONS BEFORE AND AFTER WINDING 147

Then wind all the banding groups continuously, to
eliminate the necessity of fastening the ends of each group as
they are wound. The bands should be held together and the
ends fastened by means of narrow tin strips (about 0.012 to
0.02 inch thick and 0.25 inch wide) placed under the wires
and bent back over the top and held by pure tin solder. These
strips should be inserted while the wire is being fed on and
located about every three inches around the armature, with
closer spacing at the beginning and end of each band. For
the core bands, these strips should be placed in the slots and,
being wider than the groove, they prevent any tendency for
the bands to slide around the armature. The ends of the
groups should then be cut and secured by being bent back
outside one clip and inside the next one. Pure tin solder
should be applied to the whole surface of the bands to form
a solid web. ;

End Bands. The end windings should be secured by
groups of wire wound on insulating hoods to protect the coils.
On the commutator end, strips of thin mica with overlapping
ends are usually placed on the commutator neck and held in
place with a few turns of twine. If a hood or head of canvas
is to be used it should be wrapped around the neck, extending
about an inch from the edge and turned inside out. After
this end is secured with twine, the free end of the hood should
be pulled back over the windings, bringing the outside of the
hood at the surface and making a neat folded-under edge.
This hood can be held temporarily with twine until the wire
is applied. The other end of the armature should be similarly
covered with a hood and banded.

Tension to be Applied to Band Wire. The proper tension
for banding wire when being applied varies with the size of
wire and the construction of the end windings. When
the end coils have no rigid support and extend out a con-
siderable distance from the core, the tension should be gradu-
ally reduced, as shown in the accompanying table.

Wire. The best material is a high-grade steel piano wire,
having a final breaking strength of 200,000 Ib. per sq. in. For
temporary bands a cheaper grade can be used. The band
wire should be tinned.

148 ARMATURE WINDING AND MOTOR REPAIR
POUNDS TENSION FOR BANDING WIRES

End bands

Diameter of wire,
in.

Core bands

At core

At end of wdg.

0.045

200

175

160

. 0641

300

250

225

0.0803

400

300

260

Solder. Pure tin should be used, as this gives a band that
will hold together for a longer time than half-and-half solder.

Flux. About 1.5 lb. of powdered rosin, dissolved in one
quart of denatured or wood alcohol makes a good flux.

Tin Clips and Strips should be of commercial sheet tin
about 0.012 to 0.02 inch thick.

Precautions. Use a band wire that is strong enough to
prevent movement due to high speed and vibration.

Secure the ends of all band wires under the clips.

See that the bands are below the surface of the core to
keep them from rubbing on the poles.

Before applying core bands, see that the tops of the coils
are about ^ 2 inch above the band groove, so that the bands
pull the coils down even with the core. If the coils are too
high and the bands do not rest on the cores, loose bands will
result when the insulation dries out.

Wind all core bands in one operation.

In soldering, use a 4-lb. clean iron, well tinned.

Bands to be effective should be kept tight. If they are
allowed to become loose, grounded armatures and broken leads
may result. It is considered good practice to reband new or
newly rewound railway or hoist armatures after about 12 to
18 months as a safety-first measure; and to make renewals on
old ones whenever the bands start to loosen. The service
conditions temperature and speed largely determine the
length of time that bands will hold tight. One large operating
company rebands all armatures every two years.

Seasoning and Grinding a Commutator. A new or reas-
sembled commutator becomes "seasoned" that is the
insulation baked out and all parts in their final set position,

OPERATIONS BEFORE AND AFTER WINDING 149

only after being in operation for a time with the necessary
tightening and grinding. This is particularly true of a large
commutator. Need of attention to a commutator will be
indicated by roughness, high or low bars, flat sections result-
ing in poor commutation. If the commutator is in very bad
condition it may be necessary to turn it down, but for ordinary
cases a grinding: tool is preferable and recommended such as
shown in Fig. 115 (Instruction Book, Westinghouse Electric &
Mfg. Co.).

Commutators should always be ground at from 100 to 120
per cent, normal speed. Turning requires a much lower
speed; it should not be higher than 150 feet per minute.

FIG. 115. Grinding device for truing commutators when they do not require

turning down.

Before grinding a commutator, the machine should have
been in service a sufficient length of time to bring the tem-
peratures up to a constant value. Before grinding, the
brushes should be lifted off the commutator as the copper and
stone dust will rapidly wear them off. The dust will also
become imbedded in the brush contact surface and later
damage the commutator or cause poor commutation. The
armature winding should also be thoroughly protected during
this operation to prevent an accumulation of dirt and metal
chips which may result in an insulation failure when the
machine is again put in service. This protection can usually
be obtained by using a circular shield of fullerboard, or similar
material, around the commutator at the end next to the
armature. This shield can be easily supported from the

150 ARMATURE WINDING AND MOTOR REPAIR

brush-holder arms and should extend from the commutator
surface to an inch or two above the surface of the armature.
It may also be desirable to put a temporary canvas hood
over the armature winding. This protection can be best
provided by carrying the copper dust away by means of a
vacuum system. Even when this is done the armature should
be protected as described. After grinding, the complete
machine should be thoroughly cleaned by the methods already
described. It may be nec.essary to repeat the heating, tight-
ening and grinding one or more times before the commutator
is in first-class condition. Emery cloth or paper should never
be used for this purpose on account of the continued abrasive
action of the emery which becomes embedded in the copper
bars and brushes. Even when sandpaper is used the brushes
should be raised and the commutator wiped clean with a piece
of canvas lubricated with a very small quantity of vaseline or
oil. Cotton waste should never be used and an excess of
lubricant should be avoided.

The grinding device shown in Fig. 115 can be mounted in one
of the brush-holder arms or brackets of a large machine.
The grinding stones should be adjusted against the rotating
commutator until a clean cutting effect is secured, but should
be carefully shaped to the commutator surface before being
placed in the grinding device. The stones can be moved
across the surface of the commutator while it is running, by
means of the handle shown at the right in the illustration.

Undercutting Mica of Commutator. Several devices and
hand tools are used for this operation. Details of their use
and descriptions of the devices are given in Chapter XII,
page 320.

Balancing an Armature. After an armature has been wound,
banded and its commutator trued it must be balanced. In
the case of small medium speed armatures this operation can
be successfully done on a pair of steel knife edges mounted
parallel to each other and perfectly level. The ends of the
armature shaft are placed on these knife edges so that the
armature is free to roll. It is then given a slight roll with
the hand and when it comes to rest, the bottom marked with a
piece of chalk. This rolling and marking should be repeated

OPERATIONS BEFORE AND AFTER WINDING 151

several times. If the marks fall well distributed around the
armature core the armature is in sufficient balance for service
and can be placed in the motor. In case the marks fall close
together or all on one side of the armature, it must be balanced
by adding weight to the opposite side or removing weight
from the heavy side. This is done in different ways depend-
ing upon how much the armature is out of balance. Before
the weight is permanently fixed it must be determined exactly.
This can be done by the use of slugs of lead properly attached
and shaved down with a knife until the proper balance is
secured. Frequently a nut or bolt head can be filed on the
heavy side when the amount to be removed is small. Other-
wise some method of attaching a piece of metal or solder must
be devised equal in weight to the lead slug determined by
the test.

For high-speed armatures and those machines where a
perfect balance must be secured, the armature should be
tested in a special balancing machine.

Painting the Winding. The kinds of insulating varnish
and impregnating compounds that should be used on an arma-
ture winding are given on page 176 of Chapter VII. Only
good grades of insulation paints should be used on windings.
When the winding is subjected to acid fumes, or the machine
located in damp places the manu-
facturer should be consulted on the
treatment that should be given the
winding.

Relining Split Bearings. The
following suggestions (Lnstruction
Book, Westinghouse Electric &
Mfg. Co.) can be followed in re-
newing split bearings of the oil ring
type. Melt the old bearing out
of its shell and prepare an iron
mandrel such as shown in Fig.
116 having the same diameter / as the shaft and dimensions
c, d and e taken from the bearing to be renewed. Iron pieces
BB and C should be attached by screws to the mandrel to
form the oil ring slots and the horizontal inspection opening

FIQ. 116. Mandrel for
use when relining split bear-
ings.

152 ARMATURE WINDING AND MOTOR REPAIR

in the top half of the bearing. The pieces BB should be
tapered so that they will withdraw easily from the cast metal.
A shpulder D, so placed as to fit against the end of the bear-
ing shell serves as a guide.

For the Lower Half. Warm the madrel and bearing shell
and while both are still warm so as not to cool the metal too
rapidly when pouring, place the mandrel in the lower half of
the shell with the shoulder D tight against the end of the shell
and the straight bottom portions of the pieces BB resting on
the split plane of the bearing. Close the joints x with putty
and fill the openings between the shell and the housing lead-
ing to the oil well with wet waste. Pour the molten metal,
heated just enough to flow readily, into the space between the
shell and the mandrel until the metal is flush with the split
surface of the housing. The metal will harden very quickly
and the mandrel can then be removed.

For the Upper Half. Fill the openings in the upper half
shell with putty and lay the mandrel in the shell with the
split side up. Block the openings leading to the oil well with
waste, and pour as described for the lower half. Remove the
mandrel, smooth all rough edges in the bearing by chipping
or filing, and chip the oil grooves in the lining of the upper
half. The bearing surface of both halves should be eased off
by scraping and the edges along the split surface should be
filed flush with the shell.

CHAPTER VII

INSULATING COILS AND SLOTS FOR DIRECT -CUR-
RENT AND ALTERNATING -CURRENT WINDINGS

On account of the fact that the voltage between commutator
segments in a 1 10- volt direct-current machine is not often more
than six volts, in a 600- volt machine not over 18 volts and in
a 1200- volt machine a maximum of about 25 volts, the voltage
between the conductors of an armature coil is relatively low.
These conductors are usually provided with either a single,
double or triple cotton covering, the triple covering being used
when the voltage between the conductors is near the upper
limit of 25 volts. Where the winding space is small a silk
covering is sometimes used in small machines. After the coils
have been formed and bound together to hold the strands in
place, the entire coil receives a special insulation as a protection
against breakdown between the copper of the coil and the iron
of the slot in which it is laid. The insulation for small low
voltage machines may consist of a wrapper of treated cloth
or mica held in place by a layer of cotton tape wound so that
it does not overlap. The end connections of the coil are then
protected by an overlapping layer of cotton tape with cotton
sleeves used on the leads as further protection. The entire
coil is then dipped in an insulating varnish and baked.

Insulation for Armature Coils and Slots. In addition to the
insulation provided on the coils, it has been found necessary
to pay particular attention to the insulation of the winding as a
whole from the armature core. The insulation used performs
two functions, namely, to provide mechanical protection and
to serve as electrical insulation. The following classification
can be made of the materials that are much used as slot and
coil insulation:

For Mechanical Protection. Pressboard, presspahn, vul-
canized fiber, hard fiber, fish paper, rope paper, Japanese
and Manila paper.

153

154 ARMATURE WINDING AND MOTOR REPAIR

For High Temperatures and Electrical Insulation. Mica,
micanite, mica paper and mica cloth.

For Electrical Insulation Only. Cotton tape and oiled or
treated cloth which includes cotton or linen muslin, varnished
cambric, varnished muslin, and empire cloth.

FIG. 117. Type of wire wound coils used in stator of Fairbanks-Morse
alternators (Fig. 155). These coils are thoroughly insulated and require
only a non-abrasive material in the slot to give mechanical protection to the
coils.

For Mechanical Protection and Electrical Insulation.

Pressboard or fullerboard is a material resembling cardboard
and made from cotton rags and paper clippings. It varies
from seven to 125 mils in thickness and when properly treated
and varnished has a dielectric strength of about 500 volts per
mil in thickness up to 25 mils and then reducing to about 200
volts per mil in the thicker sheets.

Presspahn is the name given to the pressboard which is made
in Germany.

Vulcanized fiber also known as hard fiber is a dense hard
material with a dielectic strength of about 200 volts per mil
at thicknesses of from 50 to 150 mils. It is used wherever
an insulating material of exceptional mechanical strength is
needed such as wedges in armature slots and coil braces.

Horn fiber is a material made in different colors that has
a high tensile strength and also a good dielectric strength
(250 volts per mil when 10 mils thick) which can be increased
by impregnation with oil or varnish.

Fish paper is made from rag stock and through a treating
process becomes a hard fiber-like paper which is very strong.
This material is not affected by heat and on account of this

INSULATING COILS AND SLOTS

155

and its mechanical strength is much used as cell lining in
armature slots.

Manila paper is made from linen or Manila fiber producing a
tough strong paper which when dry has a dielectric strength
varying from 100 to 230 volts per mil in thicknesses of from
1.8 to 28 mils.

For High Temperatures and Electrical Insulation. Mica
is one of the very few materials which maintains a high dielec-
tric strength at high temperatures. It is not, however, me-
chanically strong. Flexible sheets of mica are made up by
sticking thin splittings of mica on one side of a sheet of paper
or cloth with a suitable varnish in such a way that the joints

FIG. 118. Different types of coils used in rewinding motors.

(1), (2), and (3) are wire wound direct-current coils. Two sections are shown in (l)
before assembling the complete coil in (2). The coil in (3) is ready to be spread into a
diamond shape in a pulling machine. (4) is a so-called mush or basket coil for partially
closed slots. The coil as shown is ready for use in A. C. motors from about H to 15 hp.
(5) is a completed strap coil ready for final treatment or dipping. (6) is a shuttle
wound coil, three wires wide by eight wires deep in series for A. C. motors with open
slots. This coil is spread into shape in a pulling machine. It is insulated with
varnished cambric at the end where wires cross in winding. The coil at (7) is the same
as (6) after being spread and insulated with tape.

are staggered. Such built-up sheets are known under different
names, such as " Japanese Paper and Mica," "Fish Paper and
Mica" and " Treated Cloth and Mica." These are also
referred to as Mica paper and Mica cloth.

Micanite is a form of reconstructed mica made both in plate
and flexible forms. The latter is used for armature slots and
the former for commutator segment insulation. Its dielectric
strength is very high ranging in the flexible form from about
600 volts per mil for five mil thickness to 500 volts per mil in

156 ARMATURE WINDING AND MOTOR REPAIR

125-mil thicknesses. Micanite paper and Micanite cloth is
also made with Japanese paper and with muslin.

For Electrical Insulation Only. Cotton is used in the form
of tape or cloth. When used primarily for its dielectric
strength, it is treated with an insulating compound. After
such treatment cotton cloth will withstand about 1000 volts
per mil thickness. Because this insulation is flexible and
tough it is 'much used as an insulation in the insulating of coils
and other parts of electrical machinery. Since ifc is quite
susceptible to damage and abrasion it is mostly used with a
protective covering such as untreated cotton tape or friction
tape or a tough paper such as fish paper. Different weights
and thicknesses of cloth are used for insulating purposes such
as cambric five mils thick; muslin eight mils thick; heavy cotton
11 mils thick; drilling about 17 mils thick; and duck about 30
mils thick. When duck is treated with linseed oil or varnish
it is frequently used as a protective covering over coil supports,
between coil ends and over the ends of armatures. Cotton
insulating material when treated with an insulating compound,
varnish or linseed oil, is known under several names such as var-
nished cambric, varnished muslin, Empire cloth, Kabak cloth, etc.

Any of the cotton materials can be cut into narrow widths
for use as tape when insulation is more important than me-
chanical strength. Tape cut on the bias is used to tape-up
coils of irregular shapes.

Descriptions and Uses of Insulating Materials. In the
following paragraphs the composition of the insulating ma-
terials already mentioned together with many others that are
available are given with their various uses. This information
is taken from a comprehensive classification of treated cloths,
pressboards, fibers and papers by Hugh E. Weightman,
Chief Engineer, Engineering Service Company, Chicago, 111.
(Electrical Record, July, 1919).

Treated Cloths. The different kinds of treated cloths
which are available are as follows:

Black varnished cambric Yellow varnished cambric

Japanned muslin Oiled muslin

Japanned duck Yellow oiled canvas

Varnished silk Yellow oiled cotton drill

INSULATING COILS AND SLOTS 157

Uses and Properties of Treated Cloths. Each of the materials
in the foregoing list has somewhat different insulating pro-
perties. Black varnish cambric is a varnish coated cloth
used in the form of straight-cut tape for wrapping wire and
cables and as bias cut tape for armature coils. It is usually
supplied in 0.010-in. and 0.012-in. thicknesses in rolls about
36 in. wide. It is also supplied in ready cut tapes. The
material is obtainable in two or more grades, the cheaper
grades being used for phase insulation.

Japan muslin is unbleached muslin cloth treated with black
japan and baked to produce a waterproof material of good
insulating properties. It is used for wrapping where a coarse
cloth is permissible and is usually supplied by the manufacturer
in 0.017-in. thick by 30-in. wide rolls.

Japan duck is a high grade 8-oz. duck approximately
0.025 in. thick and of close weave. It is treated with japan
and oven cured. It is chiefly used under the binding bands
of railway motors as a protecting and moisture excluding
fabric. This material is usually supplied in rolls 36 in.
wide.

Varnished silk is made of Japanese silk treated with a
high grade insulating varnish and oven cured. This makes
a thin, tough insulating material of high dielectric strength
which is used where light weight and a minimum of thickness
is required. Varnished silk is also employed on meter coils
and as insulation in airplane apparatus. The material in
addition to being light does not become brittle in extreme cold
nor gummy in heat. It is quite expensive and for that
reason is not generally applicable. The usual thicknesses
are 0.003 in. and 0.005 in. Sheets are 27 in. wide.

Yellow varnished cambric is a strong, closely woven cotton
cloth having an especially soft finish and treated with high
grade varnish. The varnish is baked in place, producing
a material having a very high dielectric strength and a hard
smooth surface. Its insulation resistance is usually not as
high as black varnished cambric, especially at high tempera-
tures. It is more easily handled than the black varnished
cambric. It is usually supplied in rolls and sheets 36 in. wide
and in tapes straight-cut or bias-cut, 0.010 in. and 0.012 in,

158 ARMATURE WINDING AND MOTOR REPAIR

thick. This material is used for much the same purposes as
black varnished cambric.

Oiled muslin is a linen finished cloth coated with oil and
oven cured to set the film to a hard smooth surface. The
product is a flexible cloth having high insulating properties
and good resistance to deterioration through vibration or
aging. It is used for a large variety of purposes, especially
for wrapping armature coils and, in tape form, for taping coils
and leads. It is supplied 0.007 in. thick in 36-in. rolls or in
standard width tapes.

Yellow oiled canvas is a high grade duck treated with
oil to produce a flexible waterproof material. It is used
for pads under railway motor field coils and in similar places.
This material is commonly obtained in one thickness of 0.045
in. in rolls 36 in. wide. It is also used for tarpaulins for genera-
tors, switches and such apparatus. Yellow oiled cotton drill
is a light unbleached cotton drill treated with oil and oven
cured to set the film to a firm surface. It is made 0.017 in.
thick and in rolls 30 in. wide. It is often used for coil separa-
tors in transformers and regulators.

Pressboards, Fibres and Papers. Fibers, leatheroids and
such materials, are especially selected for their high insulating
properties, and they are treated to render them more pliable
and easily worked. The treatment does not appreciably
affect the thickness except in the case of the shellacked
materials. This treatment adds approximately 0.005 in.
in thickness to the size of each sheet. When ordering treated
material from the manufacturer the thickness always refers
to the untreated material. The following kinds of this class
of treated materials are available:

Pressboard Varnished rawhide fiber

Japanned pressboard Leatheroid

Oiled pressboard Express parchment paper

Shellacked pressboard Shellacked express paper

Varnished pressboard Varnished express paper

Horn fiber Red rope paper

Japanned horn fiber Oiled red rope paper

Oiled horn fiber Shellacked red rope paper

Shellacked horn fiber Varnished red rope paper

Varnished horn fiber Shellacked bond paper

INSULATING COILS AND SLOTS 159

Rawhide fiber Asbestos paper

Japanned rawhide fiber Oiled asbestos paper

Oiled rawhide fiber Varnished asbestos paper
Shellacked rawhide fiber

Pressboards. Pressboard is a specially prepared paper which
is dense, flexible and easily worked. It readily absorbs oils
and varnishes which render it less hygroscopic than in its
untreated state. It is used very extensively as layer insulation
and spacers in various types of generator, motor, transformer
and regulator coils. Collars and shields on high-voltage
transformers are also made of this form of insulation. One
manufacturer carries this material in stock in the following
sizes :

Thickness in inches Size of sheets in inches

0.009 30 X 40

0.020 33 X64
0.030 " v 34X40

0.030 36X84

0.60 34X40

0.60 36X84

^2 24 X 60

H 40 X 60

Japanned pressboard is used for separators and fillers
in armature and field coils. It is made in the same sizes
listed in the preceding paragraph. Oiled pressboard is used
in the same way as japanned pressboard, and it is sometimes
preferred since it is more flexible. It is supplied in the same
sizes as given for pressboard. Shellacked pressboard finds
its principal application as separators in bonding railway
armature coils. The shellac upon the application of heat,
melts, forming a close union between the separator and the coil.
The principle thickness used is 10 mils or 0.010 in. although
other sizes are available. This material is often made up
locally by manufacturers. Any sizes previously listed are
available.

Pressboard varnished and oven cured is used for fillers and
separators. The usual sizes are available as listed above.

Horn Fibers. Horn fiber is a tough flexible insulating
material of high mechanical and dielectric strength. It is
used for separators, slot channels, angles and joint insulation.

160 ARMATURE WINDING AND MOTOR REPAIR

and wherever a high degree of flexibility is essential. This
material can be bought in both sheets and rolls. All horn
fibers are obtainable in the same thicknesses as listed for
pressboard. Japanning greatly increases the flexibility and
dielectric strength of horn fiber. One special application of
japanned horn fiber is as fillers for bracket insulation and
around rocker arms.

Treated with oil and oven cured horn fiber makes a superior
dielectric material for general work. Shellacked horn fiber
finds its greatest application as slot armors, it being customary
to form the armors before shellacking. Varnished horn fiber
is used for fillers and separators in preference to pressboard
on better machines and especially on high speed units because
it is more flexible and does not disintegrate as rapidly when
subjected to vibration.

Rawhide Fibers. Rawhide fibre is a harder pressed material
than horn fiber and consequently lacks some of the flexibility
of the latter. It is very tough and can be rendered more
flexible by treatment with japan, oil shellac or varnish.
Untreated it is used where toughness is essential and where
great flexibility is not required. The thicknesses of the stock
sizes of the different forms of rawhide fiber usually found on the
market are: 0.005 in., 0.010 in., 0.015 in. and 0.020 in. Sheets
or rolls are 40 in. to 48 in. wide. Sheets are usually 72 in.
to 120 in. long. Japanned rawhide fiber has increased diel-
ectric strength and is more flexible than the untreated fiber.
It is used for the more simple shapes of slot armor and for
fillers and separators. Rawhide fiber treated with a high
grade oil and oven cured to produce a firm surface is used for
the same purposes.

Shellacked rawhide fiber is used for the same purposes as
shellacked horn fiber, where great flexibility is not required.
Varnished rawhide fiber is the most flexible of the rawhide
fibers and is used for fillers and separators.

Leatheroids. In appearance leatheroid is quite similar to
rawhide fiber. It is used for the same general purposes as
the rawhide and horn fiber. It is, however, more resistant
to heat, is more dense and molds more readily than either of
the others. The usual thicknesses available in sheets or rolls

INSULATING COILS AND SLOTS 161

42 in. wide are 0.010 in., 0.015 in., 0.020 in., 0.030 in., 0.060 in.,
and % m - The sizes apply to all forms of leatheroids. The
japanned leatheroid is used similarly to japanned horn or
rawhide fiber where these materials are unable to stand the
heat. Oiled leatheroid is used in place of oiled horn or raw-
hide fiber where a greater resistance to heat is desirable.

Paper Insulation. Express parchment paper is a strong
high-grade wood fiber paper selected to insure freedom from
pin holes and metallic particles. It is used for layer insulation
and in making up pasted mica sheets. The commercial
sizes are as follows:

Rolls 0.003 in. thick, 30 in. wide

Rolls . 005 in. thick, 32 in. wide

Rolls 0. 009 in. thick, 32 in. wide

Sheets 0.005 in. thick, 30 in. X 35 in. wide

Sheets 0.009 in. thick, 30 in. X 35 in. wide.

When it is shellacked, express parchment paper is used
for armature and field coil bonding. The 5- and 9-mil
sizes are most frequently used for this purpose. Varnished
express parchment paper is used chiefly for slot insulation,
its best recommendation being its glossy surface to which coils
will not readily adhere when putting them in place. The
same commercial sizes are available.

Red rope paper of a good hemp rope stock, selected to
avoid pin holes and metallic particles is used in conjunction
with parchment paper and other insulations in building
up layer insulations. Treated with hot oil red rope paper is
excellent for troughs, slot insulation, etc. Where heavy bar
wound armature coils are used which are subjected to steam
mold bonding shellacked red rope paper is usually used. The
sizes listed for express parchment apply equally to these red
rope papers. The 5- and 9-mil sizes of shellacked red rope
paper are the most used. Red rope paper varnished and
baked is used principally for slot insulation. As a rule it is
made only in 5- and 9-mil thicknesses in sheets 30 in. by 55 in.

Shellacked bond paper is bond paper treated with a medium

bodied shellac solution. It is used for separators in steam

molds and is usually obtainable in rolls of any desired width.

The thickness is 0.010 in. or 10 mils. Asbestos paper is used

162 ARMATURE WINDING AND MOTOR REPAIR

in making layer insulations where heat resistance is needed.
This material must be purchased under specifications having a
test or inspection clause in order to guard against the presence
of conducting particles of iron oxide. It is furnished usually
in rolls 36 in. wide and in the following thicknesses: 6 mil,
15 mil and 20 mil, J 2 m -> KG in. and ^ in. Asbestos paper
treated with hot oil and oven cured is used in railway motor,
field and armature coils and in machinery operating at high
temperatures.

When the asbestos paper is treated with a special black
plastic varnish and oven cured it becomes a great moisture
proof insulation and is used in forms for cores or spools. It
is used as a moisture-proof, heat-resisting insulation for motors
intended for naval, mine or similar services.

Coil and Slot Insulation Used in One Large Repair Shop.
The following grades and thicknesses of coil and slot insulation
is carried in stock and used by one large repair shop where all
types, makes and sizes of motors are rewound.

Fish Paper: In sheets 4, 7, 10, 15 and 23 mils thick.

Fish Paper with Mica Splittings: In sheets 13 mils thick.

Fullerboard: In sheets 7, 10, and 15 mils thick.

Shellacked Fullerboard: In sheets 7, 10 and 15 mils thick.

Treated Cement Paper and Mica Splittings: In sheets 14 mils
thick.

Barren Paper: In sheets 3 and 5 mils thick.

Empire Cloth: In sheets 9 mils thick.

Cotton Duck: In sheets 12 mils thick.

Cotton Tape: In widths % and 1}^ in., 7 mils thick.

Treated Cloth Tape: In width % in., 8 mils thick.

Under the different headings of Chapter III dealing with the
windings of particular machines, details are given concerning
the uses of the different insulating materials described here.

Micartafolium. A special insulating material has been
developed by the Westinghouse Electric & Manufacturing
Company which is known as micarta. In the form of
micartafolium it can be used as a wrapping. This insulation
is used on the larger alternating-current armature windings,
particularly those of alternating-current turbo-generators.
It is also used for direct-current armature windings. For

INSULATING COILS AND SLOTS 163

commutators it is used between the bottom of the segments
and the sleeve on the driving spindle. It is also used in
direct-current machines for flash guards between the com-
mutators and the armature winding of large units. For
insulating studs and the capping of nuts and bolt heads it
has also been found satisfactory. This type of insulation
is made in sheets, blocks and tubes.

Thickness of Insulation Required in Slots. Prof. Alfred Still
in his book on Electrical Design, under the heading of slot
lining points out that the insulation may be placed around
the individual coils or in the slot before the coils are inserted.
Part of the insulation may also be placed around the coils
and the remainder in the form of slot lining. The essential
thing is to have sufficient thickness of insulation between the
cotton-covered conductors and the sides of the slot. In this
connection the following figures are given by Prof. Still for
the thickness of slot lining required for machines of different
voltages. The thickness in inches is for one side only of the
slot, and is as well the thickness that should be provided be-
tween the upper and lower coil sides in the slot.

Oper i ting voltage Slot insulation one side

for direct-current machines

AI B

Up to 250 0.035 inch

Up to 500 . 045 inch . 035 inch

Up to 1000 0.060 inch 0.055 inch

Up to 1500 0.075 inch

For alternating-current machines

500 to 1,500, same as for direct-current machines
Up to 2,000 0.080 inch 0.090 inch

Up to 4,000 . 120 inch . 130 inch

Up to 8,000 0.190 inch 0.185 inch

Up to 12,000 . 270 inch . 220 inch

Insulation of Formed Coils. The practice of most motor
manufacturing companies is to insulate coils for two or three
ranges of voltage. In making changes in the connections of

1 Prof. Alfred Still Principles of Electrical Design.

2 H. M. Hobart Design of Polyphase Generators and Motors.

164 ARMATURE WINDING AND MOTOR REPAIR

induction motors, therefore, where the changes involves an
increase in voltage, it is important to know that the changes
do not subject the coils to a voltage outside the range for
which they were insulated. This precaution does not, of
course, apply when reconnecting for a lower voltage.

FIG. 119. Diamond shaped coil insulated and impregnated with compound
ready for use (Westinghouse Electric & Mfg. Company).

The practice of one large manufacturer is to insulate coils
for three ranges of voltage as follows:

/. Direct-current and Alternating-current Group Coils for Vol-
tages up to 220.

For plain coils other than phase coils :

1. Wound with double cotton-covered wire.

2. Binding paper over slot side, 0.003 inch thick.

3. 1^ turns treated cloth on slot side, 0.008 inch thick.

4. Cotton tape, one-half lapped all around, 0.004 to 0.007 inch
thick.

5. Entire coil dipped in insulating compound and baked.

For phase coils the same insulation is provided and in addition a
layer of empire cloth and tape over each end of the coil up to the
straight sides. The diamon'd points are painted red to dis-
tinguish these coils from plain coils. Phase coils are used at
beginning and end of each phase group.

II. Direct-current and Alternating-current Group Coils for Vol-
tages up to 500.

For plain coils other than phase coils:

1. Wound with double cotton-covered wire.

2. Binding paper over slot side, 0.003 inch thick.

3. 1^ turns mica and fish paper on slot side, 0.012 inch thick.
This is made up of fish paper, 0.004 inch; three layers mica
splittings, each 0.001 to 0.003 inch thick ; Japanese paper, 0.001
inch thick. All shellacked together.

INSULATING COILS AND SLOTS 165

4. Cotton tape, one-half lapped all around, 0.007 inch thick.

5. Entire coil dipped in insulating compound and baked.

For phase coils the insulation of ends is the same as for 220 volts of
the first group.

Ill . Alternating-current Group Coils for Voltages from 500 to
2300.

For plain coils other than phase coils :

These coils are insulated the same as group II except 3^ turns
instead of 1^ turns of fish paper and mica are used. A layer
of treated cloth on the ends of the coils is also used.

For phase coils two layers of treated cloth and a layer of cotton tape
are added.

Insulation for Coils used in 240-Volt and 500-Volt Direct-
current Machines. The following recommendations are given
by Prof. Alexander Gray (Electrical Machine Design, Chapter
IV on Insulation) for the windings of 240- and 500- volt direct-
current machines.

For a 240-volt, double-layer winding with two turns per coil
and four coil sides (eight conductors) per slot, the conductors
being of strip copper wound on edge:

(a) After the copper has been bent to shape, tape it all over with one
layer of half-lap cotton tape 6 mils thick. This forms the insulation
between the adjacent conductors in the same slot.

(6) Tape together the two coil sides when they form one group (as in
this case) with one layer of half-lap cotton tape 6 mils thick all around
the coils. This forms the end-connection insulation and also part of the
slot insulation.

(c) Bake the coil in a vacuum tank at 100C. (212F.) so as to expel
all moisture, then dip it into a tank of impregnating compound at 120C.
(248F.) and leave it there long enough to become saturated with the
compound.

(d) Put one turn of empire cloth 10 mils thick on the slot part of the
coil and lap the ends. This insulation should extend % inch past the
ends of the slot.

(e) Put one turn of insulating paper 10 mils thick on the slot part of
the coil and lap it over at the ends. This paper should be long enough
to also extend past the edge of the slot % inch on each end. It is not
put on for insulating purposes but to protect the other insulation which
is liable to become damaged when the coils are being placed in the slots.

(/) Heat the coil to 100C. (212F.) and then press the slot part to
shape while hot. The heat softens the compound and the pressing forces
out all excess of compound. The coil can then be allowed to cool while

166

ARMATURE WINDING AND MOTOR REPAIR

INSULATING COILS AND SLOTS 167

under pressure and will come out of the press with such a shape and size
that it will slip easily into the slot.

(gr) Dip the ends of the coil in elastic insulating varnish.

For a 500-volt, double-layer winding having five turns per
coil, and thirty conductors per slot, with the conductors covered
with double-cotton coverings:

(a) Put one turn of insulating paper 5 mils thick around the slot
part of the two conductors that form the individual coils. This paper
should extend % inch beyond the edge of the slot at each end. It forms
part of the insulation between the individual coils in the same slot and
also part of the insulation from winding to core.

(6) Put one turn of empire cloth 10 mils thick around the three coils
that form one group and lap it over at the ends. Allow this insulation
to extend as before % inch at each end of the slot.

(c) Put one turn of paper 5 mils thick on the slot part of the coil and
lap it on the top. As in the other cases allow the paper to extend over
the slot length % inch.

(d) Tape the ends of the groups of three coils with one layer of half-
lap cotton tape 6 mils thick and carry this tape on to the paper for ^
inch to seal the coil.

(e) Wind the machine with these coil groups, putting a lining of paper
10 mils thick in the slot and then hold the coils down with band wires.

(/) Place the armature in a vacuum tank and bake it at 100C. (212F.)
to expel moisture, then force impregnating compound into the tank at a
pressure of 60 Ib. per square inch and maintain this pressure for several
hours until the winding has been thoroughly impregnated.

(g) Rotate the armature while it is still hot at a high speed so as to
get rid of the excess of compound which will otherwise come out when the
machine is carrying a heavy load.

(h) Paint the end connections with elastic finishing varnish, taking
care to get into all the corners.

Coil Insulation for Induction Motor Windings. The fol-
lowing recommendations are given by Prof. Alexander Gray
(Electrical Machine Design, Chapter XX on A.-C. Insulation)
for the coil and slot insulation of induction motors :

For 440-volt, wire-wound coils of double-layer windings:

(a) Double cotton covering on the conductors.

(6) A layer of empire cloth six mils thick between horizontal layers of
conductors.

(d) One layer of half-lapped empire cloth tape six mils thick all around
the coil.

168 ARMATURE WINDING AND MOTOR REPAIR

(e) One turn of paper 10 mils thick on the slot part of the coil to protect
the empire cloth.

(/) One layer of half-lapped cotton tape six mils thick on the end con-
nections to protect the empire cloth.

(g) The coil should be baked and impregnated before the paper and
cotton tape are put on and dipped in finishing varnish after they are all
put on to make it water and oil proof.

For a 2200-volt induction motor with strip copper coils and
a double-layer winding:

(a) One layer of half -lapped cotton tape six mils thick on each conduc-
tor to form the insulation between the conductors.

(6) One layer of half-lapped cotton tape six mils thick all around the
coil to bind the conductors together.

(d) Two layers of half -lapped empire cloth six mils thick all around the
coil.

(e) One turn of paper 10 mils thick on the slot part of the coil to protect
the empire cloth.

(/) One layer of half -lapped cotton tape six mils thick on the end con-
nections to protect the empire cloth.

(g) Bake and impregnate the coil before the paper and the last taping
of cotton tape are put on. After that, the slot part of the coil should be
hot pressed and allowed to cool under pressure. Then the coil can be
dipped in finishing varnish to make it water and oil proof.

Coil and Slot Insulation Employed by a Large Manu-
facturer. The following insulation for coils and slots repre-
sents the practice of a large manufacturer specializing in the
construction of motors and generators in a wide range of
sbes and for all commercial voltages.

I. For Small Direct-current Machines Using Wire-wound Coils:

1. Coils are wound with double cotton-covered wire.

2. Dipped in varnish and baked to about 100C. (212F.) until coils
are free from stickiness.

3. Taped with 6-mil linen tape.

4. Coils are then placed in armature and soldered to commutator
risers. After cleaning, the armature is dipped or sprayed with
varnish and baked.

5. The final operation is to spray with shellac.

6. For slot insulation 10-mil pressboard is used.

II. For Engine Type Direct-current Armatures Using Strip Copper Coils:

1. Copper strips are taped with 6-mil linen tape, % lapped. Width
of linen tape is %, 1 or 1 % inch, depending upon size of the coil.

2. Coil is then dipped in good insulating varnish and baked until
free from stickiness.

INSULATING COILS AND SLOTS

169

3. When coils are used in groups, the groups are taped with 6-mil
linen tape % lapped.

4. Operation two is then repeated.

5. Coil is now ready to be inserted in armature slots. When arma-
ture is completed, and bands are on but before assembling in
frame, the windings are saturated with air-drying varnish or it
is baked until free from moisture. As a rule, machine is well
baked by running under full load. It can then be sprayed with
air drying varnish.

6. For slot cells, 10-mil pressboard is used not so much for insulation
as for mechanical protection.

FIG. 121. At the left, armature coil insulated for a 250-volt direct-current
motor. At the right, stator coil insulated for a 220-volt induction motor
(Crocker- Wheeler Company}.

Variation from above practice depends upon the voltage. For high
voltages, say 550 volts, the coil will perhaps be taped with 9-mil oil
muslin half lapped and an additional varnish and baking treatment.

For mill motors the coils are usually insulated with flexible mica. The
coil is heated in a form and connected in the secondary of a transformer.
While the current passes through the coil, it is tightened so as to give
the proper thickness, alter which it is insulated with asbestos tape.
Asbestos tape is used because of the severe duty to which these machines
are subjected and because they are completely enclosed.

III. General Method for Insulating Direct-current Field Coils:

1. Coils are wound with double cotton-covered wire.

2. The coil is then taped with 6-mil linen tape about % lapped on
the outside of coil. Width of tape one inch.

3. Coils are next placed in vacuum tank until free from moisture.

170 ARMATURE WINDING AND MOTOR REPAIR

4. Dipped in good insulating varnish and baked. Or if the varnish
has air drying qualities, the coils are dried in air until tree from
stickiness.

IV. Insulation for Stator Coils of Alternating-current Machines for 600
Volts and Under Using Wire-wound Coils:

1. Coils are wound with double cotton-covered wire.

2. Heated and dipped in insulating varnish and baked.

3. Operation two repeated.

4. Taped with 6-mil linen tape half lapped.

5. Coil is then shaped.

6. Dipped in varnish and baked.

7. Operation six repeated.

8. Operation four repeated.

9. Dipped in good moisture-proof varnish and baked until all
crevices are well filled.

10. For slot insulation 10-mil pressboard is used.

V. Insulation for Stator Coils of Alternating-current Machines, Made up
of Strip Copper:

1. One-turn coils are taped with 6-mil linen tape one inch wide,
half lapped. Two turn coils are taped with one turn as above.
Three turn coils have middle turn taped. Four turn coils have
first and third turns taped.

2. Coil is then taped with 6-mil linen tape half lapped.

3. Dipped in insulating varnish and baked.

4. Operations seven, eight and nine as for wire coils are then applied.

5. For slot insulation 10-mil pressboard is used.

VI. Insulation for Coils Used in 2200-volt Machines:

1. Coil is wound with double cotton-covered wire.

2. Wrapped with 6-mil linen tape half lapped.

3. Heated and dipped in insulating varnish and baked.

4. Operation three repeated.

5. Operation two repeated.

6. Coil is then formed.

7. Last tape is now removed from coil, which was put on in order
to protect the first linen tape while the coil was being shaped.

8. Wrapped with 9-mil oil muslin half lapped and brushed with best
quality insulating varnish.

9. Operation eight repeated.

10. Wrapped with 6-mil linen tape half lapped.

11. Coil is then dipped in insulating varnish and baked.

12. Operation ten repeated.

13. Operation eleven repeated.

14. For slot insulation, 10 to 20-mil pressboard is used and for insu-
lation between layers 10-mil pressboard.

In making up coils for any kind of electrical machinery, the main
thing to bear in mind is to apply such processes as to eliminate moisture
from the coil. Even for 2200-volt machines, except in special cases, the

INSULATING COILS AND SLOTS

171

S .S ,
CH _^ C

u S 8

S -o 8?

I M

'S T3

c as .a

* a.|

^ M

111

ill

-2

H >

? ||

Jj if

s !

o 5 I

172 ARMATURE WINDING AND MOTOR REPAIR

practice of the company employing the insulating methods described is
to use oil muslin and depend to a great extent upon the proper varnish
treatment. After much experimenting with varnishes it has been found
possible to wind coils that will withstand high voltages equally as well as
insulated coils wound for the same voltages.

Insulation of End Connections of Coils. In a double-layer
winding, the voltage between the end connections of two
coils where they cross each other at the ends of the slot may
be about equal to the terminal voltage. Suitable insulation
should be used at this point. A belt of cotton duct is usually
used between the end connections for this protection, and also
to protect the coils from mechanical injury by rubbing against
each other.

Also the slot insulation should be allowed to extend out each
end of the slot a certain distance depending upon the machine
voltage. Prof. Alexander Gray gives the following values as
typical of good practice.

Terminal voltage of machine Length of insulation out of slot

Not over 800 . 75 inch

800 to 2500 1 . 25 inches

2500 to 5000 2.00 inches

5000 to 7500 : 3 . 00 inches

7500 to 11,000 4 . 50 inches

Coils for alternating-current windings must be wound to
stand the line voltage to ground. In a Y-connecJied machine
the total insulation from copper to copper between two coils
should stand about 1.7 times this value. The first coil of
each phase in a Y-connected machine has a voltage equal to
the line voltage to ground (volts between lines -s- \/3) against
which to insulate and should be given a protection equal to
the line voltage. The first and last coils of each phase group
must also have extra insulation for protection at the points
where they cross. These are called the phase coils. One
layer of six-mil empire cloth on the end connections covered
with six-mil cotton tape half lapped in addition to the coil
insulation will give sufficient insulation to the phase coils
except in cases of voltages past 2300. (See page 165 for
phase coil insulation for 2300 volts.)

INSULATING COILS AND\8JbQTS 173

Phase Insulation when Reconnecting fro.nl Two-phase to
Three-phase and Vice Versa. In reconnecting the winding
of an induction motor from two-phase to - three-phase or
vice versa and in reconnecting a winding for a different number
of poles to change the speed, it is necessary to. rearrange the
phase insulation because the spread of the coils per-phase,
per-pole is being changed.

Mica Insulation for Armature Coils. Mica has been found
to be a first class insulating material. Its insulation re-
sistance increases with temperature, a valuable characteristic
for machines operating at high temperatures and in direct
contrast with the properties of treated tapes, in which the
insulation resistance and loss increases rapidly at temperatures
above 100C. or 212F. It is unaffected by temperatures
far in excess of those encountered in the modern, well-ven-
tilated alternator. It is also impervious to the static dis-
charges present in all high voltage machines. Furthermore,
it is resilient and retains its resiliency indefinitely thus
helping to hold the coil tight in its slot.

Mica is a mineral obtained in the form of large crystals.
These split readily into thin, parallel-sided laminae, or flakes.
The flakes can be pasted uniformly on cloth or paper to
facilitate handling and to provide a mechanical support dur-
ing application. In the form of a " wrapper/' that is, pasted
on large sheets of specially treated paper, mica is mostly used
on the straight sides of each armature coil, to provide in-
sulation between conductor and iron, the operating voltage
of the machine determining the number of turns, or the thick-
ness of this ineulation wall.

All known insulating materials are relatively poor heat
conductors. This is equally true of mica and treated tapes.
Therefore, the tighter and the thinner the wall, the better the
heat radiating characteristics of the coil. For the lower
voltage machines the mica wrapper is applied as tightly as is
possible by hand. For the higher voltage windings, 6600
and above, where the insulation wall must be relatively thick,
special, patented machines are used which apply the wrapper
under heat and pressure, and finish it to a solid, compact wall.

In general, all of the larger capacity generators have

174

ARMATURE WINDING AND MOTOR REPAIR

relatively wide cores. Internal "hot spot" temperatures,
considerably higher than those measurable by thermometer,
exist. On all such machines, each conductor of the coil is also
insulated with mica tape.

Repairing Coils Damaged in Winding Process. It often
occurs that the insulation of a coil becomes damaged from
charing or from carelessness in the use of tools when laying
the coils in the .armature slots. Such damage should be re-
paired at once to prevent possible trouble later which will
make the repair more difficult and expensive. To repair
a damaged coil where the injury to it is only slight, the coil
should be removed from the slot and all the insulation removed
from the injured section. Then apply an overlapping wrapper

FIG. 123. At top, one whole and two damaged Eickemeyer coils taken
from an elevator motor. At bottom, fiber drifts, slot insulation, hammer,
parallel jaw pliers and coil lifter.

of treated cloth and around this a protecting covering of cot-
ton tape. Glue down the ends of the tape securely and apply
a good heavy coat of shellac. The shellac can be dried quickly
by touching a lighted match to it when the alcohol in which
it is dissolved will burn with a blue flame. Care must be
taken that the tape is not burned in doing this. In such a
case the flame will turn yellow and should be smothered at
once. The heat of the burning alcohol is not usually sufficient
to burn the tape. After this the coil is dry and can be put
back in the slot. While this method of repair can be safely
used in small armatures, it is not good enough for large arma-
tures. In the latter case a new coil should be used or the old
one stripped and completely reinsulated, dipped and baked.

INSULATING COILS AND SLOTS

175

On account of the stiffness of the insulation on the terminals
of formed coils, before connections are made to the commutator
it is a good plan to soften the leads with ^armalac" or a similar
armature compound at the points where the leads leave the
bottom coils in the slots. This will prevent breaks while
handling the leads.

Voltage to use when Testing Coil and Commutator Insula-
tion. Under certain conditions the difference of potential
between coils and the iron core of a machine may be equal to
the terminal voltage. Under abnormal operating conditions
it may even be greater. It is important, therefore, that the
insulation of coils and commutator shall be sufficient to stand
a voltage considerable larger than the terminal voltage of the
machine without developing grounds. For this reason a
high-voltage test is made on armature windings of both direct-
current and alternating-current machines and the commutator
of the former before the winding is finally completed.

The following values of test voltages that should be applied
are given by Prof. Alexander Gray (Electrical Machine Design,
page 34) based upon the standardization rules of the American
Institute of Electrical Engineers.

Rated terminal voltage
of machine

Rated output of machine

Testing voltage

Up to 400

Under 10 kw.

1000

Up to 400

10 kw. and over

1500

400 to 800

Under 10 kw.

1500

400 to 800

10 kw. and over

2000

800 to 1200

Any

3500

1200 to 2500

Any

5000

2500 and over.

Any

Double-rated voltage.

Some insulating materials will withstand very high voltages
before used on coils. For instance, a good quality of oiled
muslin will withstand as high a test voltage as 1500 to 2000
volts per mil for 9-mil thickness, however, when applied to
coils, its insulating properties will diminish because of handling.
This is, of course, the reason why so much insulation is used
on coils, to protect them from becoming grounded. For

176 ARMATURE WINDING AND MOTOR REPAIR

coils that are to be used for machines below 2200 volts, no
steps are usually taken to test coils before placing them in
the machine. However, for voltages over 2200, the coils are
tested before they are placed in the slots. For instance, coils
that are to be used on machines of 6600 volts, are tested for
ground at 20,000 volts by wrapping them with tin foil before
placing them in the slots. Then they are tested with 15,000
to 16,000 volts after being placed in the slots. Finally a test
of at least twice normal voltage is made for one minute.

Field Coil Insulation. For the insulation of field coils,
10 mil paper, J^g inch cardboard and 6 mil tape can be used
where the field coil has a cardboard spool. The insulation
can be applied in the order named using 2 layers of tape and
paper with the paper next to the wires on the spool. The
coils should then be baked in a vacuum tank and impregnated.

Varnishes and Impregnating Compounds for Coils. The
varnish used over the outside of insulated coils should be
water, oil and acid proof and dry quickly in air forming a hard
smooth surface. Such varnishes are made by a number of
manufacturers.

Compounds for use in impregnating coils are also available
in the open market and are usually an asphaltum or a paraffin
base dissolved in a suitable thinning solution. It is important
that the material will not attack copper, iron or insulating
materials used and form a solid at all temperatures below
212F. without contraction when changing from the fluid to
the solid state. This material should not be applied at
temperatures above the breakdown of cotton materials,
that is, above a temperature of 248F.

Because the desirable characteristics for a perfect varnish
cannot be combined into one compound, a number of varnishes
have been developed, each having its own characteristic.
The purposes of some insulating varnishes of the Sherwin-
Williams company are indicated in the accompanying table.
Where more than one varnish can be successfully used the
different types are indicated as first, second and third choices.

All of the varnishes mentioned in the table are of the
baking type. However, insulating varnishes in general
may be divided into several general classes, such as clear var-

INSULATING COILS AND SLOTS

177

nishes, black varnishes, baking varnishes and air-drying
varnishes. The most marked difference between the clear
and the black varnishes is the color, but owing to fundamental
differences in the characteristics of the ingredients entering
into their composition, there are also some differences in the
physical properties of the varnishes themselves. As a general
rule clear varnishes possess greater mechanical strength and
resist oil better than black varnishes. An exception to this is
the black elastic baking varnish shown in the table. Where
extreme mechanical strength is required as on small high-speed
armatures clear varnishes are almost always used.

TABLE SHOWING SUITABILITY OF INSULATING VARNISHES

Characteristics of clear and black
baking insulating varnishes and
uses for which they are
recommended *

Clear varnishes

Black varnishes

Clear
quick
baking

Clear
quick
elastic
baking

Clear
elastic
baking

Black
quick
baking

High
Heat
resist-
ing
baking

Black

plastic
baking

Black
elastic
baking

Dielectric strength
Mechanical strength

3
3
3

1
2

1
2
3

1
1
1
2
2
1
1
1

1
1
1

1
2

1
2
1

1
1
2

2
2
2
1

2
2
2
2
2

3
3
3

2
2
4

2
3
3
3
3

1
3

2
4
3

3
2

3
3

3
2
2

1
2

2
5
2
1
4
1
1

2
3

3
2
1

1
1

1
2

1
2
1

2
2
2
1
1

2
2
2
2
2

Flexibility

Plasticity

Oil resistance

2
2

2
3
3
3
3

Water resistance

Treating cloth, paper and thin fib-
rous materials
Treating fullerboard and heavy
fibrous materials

Small high-speed armatures
Intermediate-speed armatures
Large low-speed armatures
Field and stator coils

Automobile-starting motors
Vacuum-cleaner motors

Washing-machine motors

Street railway and electric locomo-
tive motors

Fan motors

Magnetos and induction coils
High-potential apparatus

Transformers

Average repair shop conditions. . .

* Numbers indicate order of suitability.

Black varnishes are not quite so strong mechanically
as clear varnishes but are sufficiently strong for most purposes.

178 ARMATURE WINDING AND MOTOR REPAIR

On stationary windings, as on alternating-current stator
windings, the varnish is not subjected to centrifugal stresses
and there is no advantage in using a clear varnish. Certain
black varnishes are made from plastic materials and have the
ability to withstand long-continued heating without hardening.
Black varnishes as a rule are cheaper than clear varnishes and
are more commonly used for that reason.

The chief difference between baking varnishes and air-
drying varnishes is in the proportion of oxidizing ingredients
contained. The baking varnishes are tougher, more elastic,
more resistant to oil and water, and have longer life under
heat. Speed in drying is always accomplished at the expense
of these characteristics, and the air-drying varnishes are less
durable and elastic than the baking varnishes. The air-
drying varnishes find their principal field of usefulness on
apparatus where severe conditions of usage are not encountered
and for quick repair work.

Characteristics of Insulating Varnishes. An important
factor concerning insulating varnishes is that the dielectric
resistance increases directly with the length of baking or dry-
ing period, the slow varnishes imparting the highest degree of
insulation and flexibility and producing a tough, flexible film
which may also be depended upon for mechanical strength and
extreme durability. Black varnishes are claimed to be
better for work where a transparent coating is not absolutely
essential because they produce a more flexible and highly
insulating film than clear varnishes of the same class. Var-
nishes should not crystallize under prolonged vibration, and
their mechanical structure should be elastic and homogeneous.
The accompanying table (page 179) gives the characteristics,
uses, drying time and solvents of several varnishes.

Solvent Chart for Insulating Varnishes. Since varnish is
usually sold in concentrated form care must be taken in
dissolving to avoid wrinkles or stringy drip forming on coils
when varnish is too heavy. Benzine is preferable to gasoline
as a solvent but gasoline can be used. The solvent and varnish
should be approximately the same temperature and neither
should be under 60F. (15.5 C C.). The chart on page 180
shows the percentage of 58 degrees benzine to be added to every

INSULATING COILS AND SLOTS

170

Solvent

0-
G

|
1

"ol

1
.a

a ^

3d*

MU3

bio

III

< N

<J N

2cS

^, ^

Jj rH

^~-s

^^^->

CO "t^

resisting

durable .

1
"I

resisting

I| :

1
I

t? i

05 _

S-

c .2
ll

"5 ft !

Prevents ab
ure, resists
and acid. .

Oilproof, to

Water, oil a
Heat cond

Waterproof

ft :

ft ;

2 g

- "05

G
IS

"*^ <D

S ro

(-1

a 03

JS o

*o3

S fl"

03 Jj

^.2

bfi

a3 '^3
03

03 aS

a ,zf

1 1

S
6

g |

tt >FH

j .

CH ^

o 'S

* 1

bJO

> <a

o |

.22
G

;

bfi

a
la

a 3
1

'3
ya

180

ARMATURE WINDING AND MOTOR REPAIR

JS o
2 a cs o>
n '43 a 9 *s

- 05 g

ivf* -j 02

f- CX ^

i;i 11

1 ! ". " 2

DOMPA

w$3

II? ^ ""

g M pq

FACTU

05
2>S

So^ M OiOOJ'O'-HOO
3 00 i-< (N Cl

o^ IY

p
fc

10 ,
00 CO

tigs

SS| o^^.e^
.- 1 ^ 1

|^| X
J| 13 '* oc 22^co^
-^ ** >. S

I| S 6 b * 00 5 5" w ?

J g 60 ^-<<N<NCO^iO

t* o

^ V<N

*g^-sai|:

00 N

\IN \IN \N

OcOOO-iiOOiOOCOCOOOOI^
^i-i(N(NCOCOrJ(iOiOtD

^-; 2 f^^g 3 K

^v -^\ -\

-A,

SSS5SSSSSSS8SSSJ5S8S

1 1

OJOOOOGOOOOOOOOOOOoroOOOGOOOGOGO

dddddddddddddddd

^TABJO ogioadg pajisaQ

INSULATING COILS AND SLOTS 181

100 gallons of insulating varnish to bring it to the correct
specific gravity. The most suitable gravity should be estab-
lished by trial and the specific gravity kept constant at this
value. With 63 degrees gasoline 15 per cent, less should be used
than that shown and with 54 degrees benzine 10 per cent, more
should be added.

Method for Making Tape from Cotton Cloth. Maurice
S. Clement has described the following method (Electrical
Record, October, 1918) for making tape as used by a middle
western electrical repair shop owing to a shipment of cotton
tape being held up on account of war conditions. The
repair job was urgent and the repairman was forced to
make up an amount of cotton tape of different widths rang-
ing from J-^ inch to 2 inches, sufficient to keep its main-
tenance men going until the order arrived. After considerable
thought, the following was decided upon as a temporary
remedy.

A bolt of white cotton cloth was purchased in a nearby dry
goods store. The cloth was rolled tightly on a half -inch dowel
and then marked off to the various widths on the outside. A
band-saw was used to good advantage cutting the cotton cloth
into tape; as this tape had no salvage edge, steps had to be
taken to prevent it from unravelling. Both edges of each
roll of tape were given a heavy coat of shellac and before it
dried the alcohol was burned off. This has a sort of semi-
baking effect which tends to strengthen the tape. It also
prevents the tape from curling.

Drying Out Insulation of Direct-current Generators (In-
struction Book, Westinghouse Electric & Mfg. Co.). Drive
the generator by a motor connected by a belt and short-
circuit the armature beyond the ammeter using a very weak
field excitation. If the generator is shunt wound, low voltage
separate excitation must be employed; if compound wound
the armature may be short-circuited through the series field
coifs. Direct - current generators are very sensitive when
operated as series machine and there is danger of generating
an excessive current. Consequently this method should be
undertaken only by experienced operators.

The field coils may be dried by applying from some separate

182 ARMATURE WINDING AND MOTOR REPAIR

source of excitation approximately two-thirds of the normal
direct-current voltage.

There is always danger of serious injury to the windings
when drying out with current since the heat generated in the
inner parts is not readily dissipated; furthermore, coils con-
taining moisture are much more susceptible to injury from
overheating than when thoroughly dry. The temperature
of all accessible parts should be carefully observed during
the drying out process and never allowed to exceed 80C.
(176F.), total temperature. Several hours or even days
may be required for thoroughly drying out large machines.
During the drying out process the temperature should
not be allowed to drop below that of the surrounding
air as moisture then condenses on the coil surfaces and the
effect of the previous drying would be largely lost. At
regular intervals during the drying out run, readings of the
insulation resistance (see page 185) may be taken at regular
intervals and plotted as a curve, using time for the horizontal
scale and resistance for the vertical scale. The drying should
continue until the resistance has begun to increase. If the
insulation contains appreciable moisture the resistance will
decrease during the first part of the drying out process.

Heating windings by current is more effective than any
process of heating from the outside, such as enclosing the
machine and heating the air by resistance or fires, because in
the former method the inside of the coils becomes hotter than
the outside and .moisture is driven outward. With external
heating the reverse is true.

Drying Out Synchronous Motors and Generators. Syn-
chronous motors and generators can be dried out by rotat-
ing the motor or generator at any convenient speed and short-
circuiting the armature beyond the ammeters. The field
should be excited so that the desired heating current will flow
in the armature winding. For windings of 2400 volts or lower,
the temperature, as measured by thermometers properly
applied to the hottest accessible part of the winding should not
be higher than 80C. (176F.). For 6600-volt windings the
temperature should not be higher than 75C. (167F.) and for
11,000 to 13,200-volt windings not higher than 65C. (149F.).

INSULATING COILS AND SLOTS 183

The reason for specifying the lower temperature for the higher
voltage windings is the greater difference in temperature be-
tween the inside of the coil and the outside (when the tem-
perature is measured with a thermometer) in the coils having
the thicker insulation.

If a low voltage (5 to 15 per cent, of normal) can be obtained
from the taps on a transformer for example, the armature
winding can be dried out by applying this low voltage to the
armature terminals, the rotor remaining stationary. The
field winding should be short-circuited and the temperature
of the cage winding on the rotor should be watched. Less
than normal current will be necessary on account of the
absence of ventilation.

For medium sized alternators and synchronous motors a
satisfactory way of drying out both field and stator windings
is to connect the machine to an alternating-current circuit
and run as a motor with fields overexcited at zero power factor.
This method is both cheap and effective.

Drying Out Induction Motors. Small motors can be baked
in ovens. The temperature should be raised gradually taking
several hours to bring it to the maximum value which should
not be more than 80C. (176F.) at the hottest point. The
temperature should be maintained constant for from one day
to a week depending on the size and voltage of the machine
and the history of its exposure to moisture. Induction motors
can also be dried by operation at no load on low voltage (the
primary current and heating increases as the voltage is re-
duced) or by a still lower voltage that will circulate a suffi-
ciently heavy current with the rotor blocked. If a sufficiently
low alternating-current voltage is not available, direct current
may be used. There is always more or less danger of over-
heating the windings of a machine when drying them with
current as the inner parts which cannot quickly dissipate
the heat generated in them and which cannot be examined,
may get dangerously hot while the exposed and more easily
cooled portions are still at a comparatively moderate tempera-
ture. The temperature of the hottest part accessible should
be measured during the drying out process and not allowed
to exceed 80C. (176F.).

184

ARMATURE WINDING AND MOTOR REPAIR

Insulation Test. During the drying out of a machine,
insulation resistance tests should be made at regular intervals
and plotted in the form of a curve using time on the horizontal
scale and values of insulation resistance on the vertical scale.
The drying out should be continued until the resistance
reaches it proper value. The insulation resistance is at best
only a rough guide in determining the condition of the ma-
chine as to moisture and relative values in the same winding
during a drying out run are of more value than the relative
values of windings in other machines.

500 to 600 Volt D.C. Circuit

[

1 2

3 4

Double throw,
double pole switch

To resistance to
be measured

Voltmeter

FIG. 124. Connections of double-pole, double throw switch and 500-volt
v voltmeter for measuring insulation resistance.

A megger is also sometimes used for testing the condition
of the windings during the drying out prcoess.

The insulation resistance of a machine in good condition
and at its operating temperature will usually not be less than
the value given by the following formula:

T , . . . Machine voltage

Insulation resistance in megohms = ^ . , T ^

Rated Kva. + 1000

For example a 1000-Kva., 11,000-volt motor should have an
insulation resistance, if clean and dry, of 5.5 megohms.
The insulation resistance of field windings will, in general, be
much higher in proportion to the operating voltage than that
of the armature. Since large armatures have much greater
areas of insulation, their insulation resistance will be propor-
tionately lower than that for small machines. The insulation

INSULATING COILS AND SLOTS 185

resistance of any machine will also be much lower when hot
than when cold, especially when the machine is heated rapidly.
Measuring Insulation Resistance. Insulation resistance
may be measured with a megger or by the use of a 500-volt
direct-current voltmeter and a 500 volt-direct-current circuit.
Connect the voltmeter as shown in Fig. 124 and read first, the
voltage of the line; then connect the resistance to be measured
by throwing the double throw switch and read the voltmeter a
second time. The insulation resistance is then calculated by
the following formula:

Insulation resistance (R) = , ^ MOfrOOO ' ';

Where V is voltage of the line; v the voltage reading with
insulation in series with the voltmeter; r the resistance of
the voltmeter in ohms which is generally marked inside the
instrument cover, and R the resistance in megohms. A meg-
ohm is equal to one million ohms.

If a grounded circuit is used in making the measurement,
care*must be taken to connect the grounded side of the line to
the frame of the machine to be measured and the voltmeter
between the windings and the other side of the circuit.

CHAPTER VIII

REPAIR SHOP METHODS FOR REWINDING ALTER-
NATING-CURRENT MACHINES

I. WINDING SMALL SINGLE-PHASE MOTORS

The complete stator winding of most small single-phase
induction motors is made up of two windings; the main wind-
ing of many turns of heavy wire and what is known as the
"teaser" or starting winding. The latter is necessary because
a single-phase motor is not self -starting and, therefore, requires
some means of producing a rotating field to overcome this
deficiency. This the starting winding does. It is of smaller
wire than the main winding and of high resistance.

The method of winding single-phase motors differs from^that
used for other alternating-current motors in that a skein
winding is often used. That is, the winding coil is in the
form of a skein of wire which is looped many times through
several slots to form a pole of the winding. The details of
this method of winding as given in what follows, are based
on articles that have appeared in the Electric Journal by
G. I. Stadeker and C. A. M. Weber.

Insulating Lining for Slots. The slots should be lined with a
protecting cell of fish paper cut to fit the slot. Inside this a
cell of treated cloth should be placed cut so that its edges will
project about % inch beyond the entrance to the slot. In
those slots which will contain both the main and the starting
winding, an extra treated cloth cell should be inserted over
the main winding to enclose the starting coils. End plates of
fullerboard or fiber are used to insulate the core from the
windings. The end connections of the main and starting
windings should be separated by friction cloth.

Winding the Skein Coil. In repairing a motor, the number
of times the skein is to be looped through the slots and .the
length of the skein can best be obtained from a skein taken

186

REWINDING ALTERNATING-CURRENT MACHINES 187

from the burned-out machine. In removing the old winding
care should be taken to preserve one entire skein if possible.
If this is impossible a satisfactory scheme for the repairman
is by trial with a single wire. This wire should be laid in the
slots exactly as the skeins of wire will be laid, proper allowance
being made for the building up of the skein ends from slot to
slot. The wire should then be removed and measured. Make
up a trial skein of this length and wind it in the slots. Cor-
rections if necessary can be made on the next skeins made up.
Inserting the Skein Coil in the Slots. After the skein length
and distribution have been obtained from the old motor, the

abode f g

FIG. 125. Successive steps in applying a skein coil for main winding of a
split-phase, GO-cycle, 4-pole, 24-slot induction motor.

exact procedure in winding is as shown in Fig. 125 (a) to (g).
The distribution of a 24-slot primary winding, indicating the
number of times the main winding skein is wound into each slot,
is shown in Figs. 126 to 129. The distribution in Fig. 126 is
one commonly used. Other distributions* may be used, but in
all skein windings the wires in any slot must be a multiple of
the wires in the skein.

A developed view of the primary or stator winding is shown
in Fig. (125a), looking at the teeth with the first operation of
putting the skein winding in slots 3 and 5 completed. The
end of the coil thus formed should be firmly pressed against
the core, using a rawhide or fiber mallet or a piece of smooth
wood. A half twist is next made in the skein, as shown in

188

ARMATURE WINDING AND MOTOR REPAIR

Fig. 125 (6), and the loop laid back over the winding and
threaded into slots 2 and 6, as in Fig. 125 (c). The half twist
is repeated, as in Fig. 125 (d), and the loop laid back in slots 2
and 6 for the second time, as in Fig. 125 (e). This second
half twist, in Fig. 125 (d), must be in the opposite direction to
the first one (Fig. 125 (6)), to bring the same side of the loop on
top. Otherwise a twist will be put in the skeins, which will

Slot Number
Main Winding

23 24

Distribution

1 1

Starting Winding
Distribution .

]

FIG. 126. Distribution of main and starting winding coils of a 4-pole motor
showing skeins overlapped.

Slot Number 1 23 i 5 6 78 9 10 11 12 13 14 16 16 17 18 19 20 21 22 23 24

Main Winding
. Distribution

Starting Winding
Distribution

FIG. 127. Distribution to avoid overlapping of skeins.

Slot Number

21J22

23 24

Distribution

1.1

1 2

Starting Winding
Distribution

1.1

FIG. 128. Distribution of an 8-pole machine.

Slot Number 1234 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Main Winding
Distribution

Starting Winding
Distribution

FIG. 129. Distribution of an 8-pole consequent wound machine.

make it hard to winfl smoothly, especially if it is looped back
and forth many times. The half twist in Fig. 125 (/) is made
in the same direction as that in Fig. 125 (6) By looping the
turns into slots 1 and 7 the winding of the skein is completed.
The winding of the second and subsequent coils is exactly the
same as the first. The completed winding for a four-pole,
24-slot machine is shown in Fig. 130 (at top).

In split-phase starting, squirrel-cage motors the starting
winding is connected across the line until approximately two-

REWINDING ALTERNATING-CURRENT MACHINES 189

thirds synchronous speed is reached, when its circuit is auto-
matically opened by a centrifugally-operated switch. The
center of the starting winding is between the pole centers of

FIG. 130. At top, complete main winding for a 4-pole, 60-cycle, split-
phase induction motor. At bottom, complete main and starting winding for
the same motor.

the main winding. Its distribution and the length of the
skeins must be determined by trial as explained for the main
winding. The starting winding is a resistance winding, con-
sequently it is very important that its re?istance in the

FIG. 130 (a). Dissembled view of a single-phase repulsion induction motor.

motor be the same as it was originally, which makes the length
of the skein important. The distribution of both the main
and starting windings and the number of times the skein is to

190 ARMATURE WINDING AND MOTOR REPAIR

be wound into each slot are shown in Fig. 130 (lower diagram).

After all the slots have been filled the coils should be forced
into position with a fiber drift, the insulating cells folded in
and fiber wedges driven in the slots.

Winding for a Repulsion- start Motor. For repulsion-
starting, induction-running motors the primary winding is

1234

FIG. 131 (a), (6), (c). At left, primary connections for a single-phase,
repulsion-starting induction motor. In center, connections for a 4-pole,
split-phase motor. At right, same as in center but for a series-parallel
connection of coils.

In the diagram at the left, N denotes neutral points. For a 220- volt circuit leads 2 and
3 should be connected together and leads 1 and 4 to the line. For a 110-volt circuit
leads 1 and 2 should be connected in parallel to one line and leads 3 and 4 to the other.
In the center diagram the coils are connected in series. To obtain clockwise rotation,
leads 1 and 2 should be connected to one line and leads 3 and 4 to the other. For counter
clockwise rotation, leads 1 and 3 should be connected to one line and leads 2 and 4 to the
other.

complete, as shown in Fig. 130, and the coils are connected
together, as indicated in Fig. 131 (a). Four leads arc brought
out, so that these motors may be connected externally for
either 220 or 110 volts.

Winding Small Motors
by Hand . A method
sometimes employed in
winding small motors of
the induction type is
known as hand-winding.

FIG. 132. Hand-winding method for main The Conductors are
coils of primary shown in Fig. 130.

at a time, beginning at the center of a pole, as shown in
Fig. 132. This method is mostly used when the number of
turns in the slots have no relatively large common divisor,
thereby eliminating the skein-winding method. The hand-
winding method should not be employed for the starting

REWINDING ALTERNATING-CURRENT MACHINES 191

winding of a split-phase motor, because the resistance of the
starting winding would not always be that required. A
skein winding should be used.

Windings for Odd Frequencies. It frequently happens
that small motors must be wound for odd frequencies, such as
125, 133, and 140 cycles on standard 60-cycle motor cores.
The number of poles of such motors is usually large, resulting
in a small number of slots per pole. Using a 24-slot primary,
wound eight poles, the slots per pole would be three and the dis-
tribution of the winding would be as shown in Fig. 128. Such
a winding is very difficult to wind and in a case of this kind,

1231

FIG. 133 (a) and (6). Connections for a single-phase, 8-pole, 24-slot, serie*
connected, consequent-pole, induction motor shown at left. At right, same
for series-parallel connection .

a consequent-pole winding would be used with a distribution
as shown in Fig. 129. This motor would be wound in the
same manner as shown in Fig. 125 (a) to (g), but all the coils
would be connected with the same polarity as indicated in
Fig. 133.

Connections for Main and Starting Windings. The coils
of each winding are connected in series, as shown in Fig. 134.
The leads of the starting winding are interrupted by a cir-
cuit opening device. Two circular stationary segments are
used in one design of motor insulated from each other and
from the frame and are mounted side by side on the bearing
housing. At the start they are short-circuited by the rotat-
ing shoes which close the circuit of the starting winding until
the speed reaches the point at which centrifugal force throws
them off. This is ordinarily at about half speed. The main

192 ARMATURE WINDING AND MOTOR REPAIR

Terminal Leads==i?

and starting windings are usually connected in parallel outside
the machine. The direction of rotation is determined by the
way the connections are made. To reverse the direction
of rotation, interchange either the two main winding leads or
the two starting leads; adopting the most convenient method.

Testing Out Small Induction
Motor Windings. The wind-
ings should be tested for
grounds with 1000 volts for
one minute. Each winding
should also be tested for short
and open circuits by applying
alternating current through a
wattmeter. An excessive read-
ing indicates a short circuit,
while no readings show an open
circuit. Polarity can be tested
by applying direct current to
the windings and testing with
a compass. % Adjacent poles
should attract opposite ends
of the needle. If the wind-

FIG. 134. Connections * for single- j are correct fog e nd COn:

phase motor windings. .

nections should be dipped in

a quick drying plastic varnish, which serves not only as an
insulating medium, but excludes all dust and serves to stiffen
the windings and make them moisture proof. After dipping,
the windings should be thoroughly dried in an oven.

Windings for Small Polyphase Induction Motors. The
method of winding a small polyphase motor is similar to the
winding of a single-phase motor. The slots are insulated with
fish paper and treated cloth cells, and the skeins, which are
also similar, are inserted into the slots as shown in the winding
diagrams of Figs. 135 and 136. Each skein or group of coils
overlaps the preceding one at the ends, as shown in Fig. 137,
the ends of the coils being separated from one another by a
layer of treated cloth. To produce a symmetrical winding, it
may be necessary to remove half of the first skein before the
last one can be put in place, in order to make them overlap

Line ' ,

REWINDING ALTERNATING-CURRENT MACHINES 193

in regular order. As the electrical characteristics of the ma-
chine would not be changed by so doing, the last skein can be
allowed to overlap two skeins, while the first skein is over-
lapped by two others.

Phase A

FIG. 135. Winding diagram and connections for stator of small 3-phasc

motor.

The skeins are connected according to the connection dia-
gram shown at the right in Figs. 135 and 136. The connections
determine the phase and polarity of the machine. For in-
stance, an armature having twelve skeins or groups of coils,

Phase A

Fia. 136. Winding diagram and connections for stator of a small 2-phase

motor.

can be connected for six poles, two-phase, or eight poles,
three-phase.

No starting winding is required for polyphase machines, and
since they have a good starting torque there is no need for a

194 ARMATURE WINDING AND MOTOR REPAIR

friction clutch on these machines. They should be tested
for breakdown between phases, and from each phase to ground,
and for short and open circuits, as described for single-phase
machines.

FIG. 137. Stator of a small 3-phase motor completely wound with skein coils.

II. WINDING SMALL INDUCTION MOTORS WITH
COILS IN PARTIALLY CLOSED SLOTS

FORMED

Like the armatures of direct-current machines, the stators
of small induction motors are made with partially closed and
with open slots. The coils used with partially closed slots
are usually either of a diamond shape or of a square shape
such as shown in Fig. 138. These
coils are wound on forms as in the
case of direct-current coils using wire
having a double-cotton covering. The
coils when made up are not insulated
as a whole, since the turns of the coil
must be inserted in the slot one at a
time. The insulation between coils
and the core must therefore be placed
in the slots before the coils are in-
serted. The character of insulation
required and the methods of placing
the coils in the stator of motors
having partially closed slots and for those having open slots
are outlined in what follows according to recommendations
by a writer in the Electric Journal, Vol. VII, No. 9.

FIG. 138. At left, a
basket or mush type coil.
At right a diamond type coil.

REWINDING ALTERNATING-CURRENT MACHINES 195

Fleh Paper Cell

I- Fttx* Strip

FIG. 139. Slot insula-
tion for a small induction
motor having partially
closed slots.

Insulation for Slots. Clean the slots with a blast of air to
remove all burrs and rough edges. Then cut a fish-paper cell
large enough to just line the slot up to the opening and so as
to extend at each end of the slot about three-quarters of an
inch. This fish-paper cell possesses sufficient mechanical
strength to protect the inner layer or cell of insulation which
should be of treated cloth that have high insulating strength
such as empire cloth or micanite. The inner cell should be
cut large enough to extend beyond the protecting cell and the
ends of the slot about one-quarter of an
inch and up through the opening in the
slot about an inch. The projections of
this cell serve as a guide when inserting
the coils and protect the wires from
any possible abrasion when being in-
serted. The winding and the core
should be insulated at the ends of the
slots by a fullerboard ring and a fiber
strip at each end. This will protect
the overhanging ends of the coils from coming into contact
with the core.

Basket Coils One Coil per Slot. Basket coils are wound
to a shape corresponding only approximately to their final
shape, as shown at the left, Fig. 138 It is important only
that the total length of the loop be correctly proportioned so
that the ends may be formed to the proper shape after the
wires are in the slot. When a suitable form is not available,
the coil may be wound around two pegs spaced the proper dis-
tance apart. Small pieces of tape should be fastened around
the coil at convenient points so that the wires may be
held together while they are being threaded into the slot.
Basket coils are largely used with a one-coil per slot wind-
ing. This means that each side completely fills one slot,
so that for a 48-slot core only 24 coils are required. In
this class of winding, one end of the coil has to drop below
the bottom of the slot in order to permit the adjacent coil
to pass over it, while the other side of the same coil remams
on a level with or only slightly below the top bore of the
slot.

196

ARMATURE WINDING AND MOTOR REPAIR

Winding a Three-phase Stator with Basket Coils. Assume,
for example, that a 48-slot core is to be wound for a four-pole,
three-phase machine with a coil throw (in slots) of 1 and 12.
Mark two slots to serve as a guide in placing the coils with
respect to their proper span or throw. Any slot may be
considered slot 1 and the other slot located by counting
the throw in a clockwise direction. In this case (Fig. 140) the
lower side of the first coil will go in slot 12. This side of the
coil must now be laid wire by wire, inside the insulating cell

in slot 12. The pro-
jecting edges of this
cell can then be cut
off close to the lami-
nations, the ends
folded over one an-
other and the slot

FIG. 140. Winding diagram for motor using closed by driving a
basket or loosely formed coils. tight-fitting fi b 6 r

wedge between the outer cell and the tips of the teeth. This
retaining wedge should not extend more than 1/2 i ncn beyond the
core on either side, as it is liable to curl up and rub against the
rotor. That part of the coil projecting beyond the slot should
now be taped with a layer of treated tape and a covering of
cotton tape for about half its length on each side of the core,
and formed to drop below the bottom of the slot. This is
known as the bottom part of the coil, as distinguished from
the other side which will remain on a level with the slots
known as the top part. In continuing the winding, the top
part should be left out for the present, of slots 1, 3, 5, 7 and 9,
since in completing the winding the coils in slots 2, 4, 6, 8
and 10, which drop below the slots, must be in place before
these top parts can be inserted. These coils which have one
side left out of the slot until the rest of the winding is com-
pleted, are known as throw coils. In the present case they
are coils, 1-12, 3-14, 5-16, 7-18, 9-20.

The first coil that can be wound into two slots as a top and
bottom coil is 11-22. The coil in slot 22, being a bottom coil,
is inserted in the same manner as the coil in slot 12 already
described, but is not taped or shaped until its other side is

REWINDING ALTERNATING-CURRENT MACHINES 197

threaded into slot 11. The ends can then be taped from iron
to iron with treated tape, in such a way as to overlap and seal
the projecting end of the insulating cell. This in turn should
be covered with a layer of cotton tape. The lower part of
the coil should then be shaped with a rubber or rawhide mallet
and fiber drift and treated with an insulating compound before

FIG. 141. In this illustration the armature winder is shown inserting the
turns of a basket coil into partially closed slots of a small induction motor
stator (Westinghuse Electric & Mfg. Company).

the next coil is put in place. The drift and mallet used in
shaping the coils should have their sharp corners smoothed
off, and in no case should the coils be struck directly with an
iron tool.

Fig. 142 is a partially wound stator, showing the finished

198

ARMATURE WINDING AND MOTOR REPAIR

shape of the coil ends. It is particularly necessary that both
tapings should be applied as tightly as possible, as a failure
to watch this point will result in the taping becoming loose
and baggy when the coil is shaped. This is especially true
if the slot is very deep as the process of dipping or shaping
the bottom coil below the bottom of the slot has a tendency
to pull the tape away from the laminations. To avoid this
tendency it is desirable to use glue on those turns of tape
that cover the projecting ends of the fish-paper cell.

FIG. 142. A three-phase stator partly wound with basket coils.

During the operation of taping, the beginning and end of
each coil should be brought out in such a manner as to lie on the
side of the winding farthest from the bore of the stator and so
placed that the beginning of one coil faces the end of the
adjacent coil in a manner convenient for connecting. The
method of winding just described should be continued until
all the coils are in place. The winding is completed by placing
the top parts of the throw coils in their respective slots.

Threaded Diamond Coil Two Coils Per Slot. The
threaded type of diamond coil is wound with insulated wire,
either on a form which finishes the coil to shape, or on a shuttle
which winds the coil first in the shape of a loop, after which
the loop is placed in a universal former or " puller." In
this device (see page 457, Chapter XVII), the two straight

REWINDING ALTERNATING-CURRENT MACHINES 199

sides are clamped between suitable jaws and are pulled
apart the required distance, the coil assuming the shape
shown on the right in Fig. 138. This type of winding is similar
to the basket winding, in so far as the slot is insulated instead
of the coil. On the other hand, the structure and the shape
of the finished coils is identical with that of an open slot
insulated coil. It is, therefore, not necessary to shape the
coil after it has been placed in the slot as in the case of the
basket winding. The slot insulation is practically identical
with that of the basket winding with the addition of a fiber
center strip to separate the upper and lower coils in the same
slot. In this case, however, the slots should be laid out in
groups according to a winding diagram, the number of groups
being equal to the product of the number of poles by the phase
of the motor. This grouping may be either uniform, alternate
or irregular, according to the design of the winding. When
the grouping is uniform, all groups consist of an equal number
of slots, and contain an equal number of coils. In alternate
grouping, every other group contains an equal number of
slots and coils. In irregular grouping there is no apparent
uniformity in the number of slots and coils per group.

Winding a Three-phase Stator with Diamond Coils. In
preparing to wind the machine, the slots forming the beginning
of each group should be marked to indicate that the coils
to be placed in them must be furnished with additional in-
sulation at the ends. This is necessary as these coils form a
boundary between phases, and consequently are subject
to phase potential, whereas the potential between any other
two adjacent coils is much less. Such coils are called the
" phase coils" of the winding.

Taking as an example a 72-slot, six-pole, three-phase motor
with a throw of 1 and 11. There will be 6 X 3 = 18 groups
of four coils each, with 18-taped "phase coils" in the winding.
The winding of the coils into the slots can be started by thread-
ing the bottom of the first coil into slot 11, the upper half
of the coil being left up as a throw coil. The bottom parts
of the next nine coils should be inserted in rotation, thus mak-
ing ten coils left up as throw coils. The succeeding coil should
be placed in two slots, 11 and 21. The rule to follow is,

200 ARMATURE WINDING AND MOTOR REPAIR

that no upper part of a coil can be placed into a slot, the lower
part of which is empty. After each bottom coil has been
threaded into its cell, both the insulating cell and the coil
should be raised to the top of the slot and the projecting
sides of the cell cut off as close to the laminations as possible.
The coil must then be forced back to the bottom of the slot.
The edges of the insulating cell should be folded over each
other and held in place by the fiber center strip. The latter
should be of a width to make a driving fit in the slot so as to
hold the lower coil and cell firmly in place, and should be long
enough to project out of the slot beyond the straight part
of the coil.

Every fourth coil, in the present case, is one of the special
" phase coils," and should be taped at the end from iron to
iron with an overlapping layer of insulating tape and a protec-
tive layer of cotton tape. The remaining coils may be left
without any special insulation at the ends.

The placing of the upper part of the coil in the slot requires
more care and skill, since the remaining space in the slot is
just large enough to receive the coil with its wires lying parallel
to each other. Fig. 139 shows the cross-section of a coil
three wires wide by four deep, the wire being threaded through
in such a way that wire 1 is first passed through the slot,
then 3 and finally 2. This should be continued with each
succeeding layer until the whole coil is in place. The cell
can then be folded in and the slot closed with a fiber wedge.
The two coils must completely fill the slot. If the insulation
already mentioned is not sufficient to do this, additional
filling strips of fullerboard, or treated wood, should be packed
into the bottom or sides of the slot as may be required, until
the coil is tight enough to prevent movement in the slot.
The upper and lower parts of the winding outside of the slots
must be carefully insulated from each other by a strip of
treated duck cloth. This strip should be wide enough to
extend from the fish-paper cell to the portion of the coil
farthest away from the slot, and should be threaded between
the ends of the coils as they are inserted in the slots.

REWINDING ALTERNA TING-CURRENT MACHINES 201

III. WINDING INDUCTION MOTORS HAVING OPEN SLOTS

The main difference between the coils used for a partially
closed-slot and an open-slot winding, is that the latter are com-
pletely insulated before being inserted in the slots. They
may, therefore, be wire or bar formed, and can readily be
insulated for any commercial voltage. Consequently, motors
of large size, or for voltages exceeding 550 volts are nearly
always of the open-slot type. The insulation over the coils

FIG. 143. Stator winding of a Fairbanks- Morse induction motor showing
the type of coil used.

consists of a cell of treated cloth or mica over the straight part,
and an overlapping layer of cotton tape over the whole coil.
Extra insulation for high voltages (see pages 163 to 172) is
made up by extra thickness or extra turns of the insulating cell.
The phase coils receive an extra wrapping of treated cloth tape
over the diamond ends in addition to the cotton tape, which,
as a distinguishing mark, may be of a special color on these
coils. After the final wrapping is completed, the coils should

202 ARMATURE WINDING AND MOTOR REPAIR

be dried in a moderate temperature in order to expel all mois-
ture, then, while still hot, they should be dipped in an insula-
ting compound and again subjected to moderate heat until the
compound is thoroughly dried. This compound serves to fill
up all the pores in the insulating materials and make the coils
dust and moisture proof. Before inserting the coils in the
slots the latter should be lined with cells of paraffined fish
paper cut wide enough to project out of the top of the slot when
folded, and to serve as a guide to the coils. These winding
cells sometimes called "slippers" furnish mechanical protec-
tion only to the coils.

Winding a Two -phase Stator Having Open Slots. Assume
a stator designed for six poles, 72 slots, two-phase, with a coil
throw (in slots) of 1 and 12. The total number of groups with
uniform grouping is equal to the product of the number of
poles by the phase of the motor, in this case will be 12, with six
slots each. The groups should be laid off in a clockwise
direction, the outside slots of each group being marked to
receive phase coils, of which there are two per group. The
coils can be inserted in regular order, beginning with the
bottom part of phase coil in slot 12, and inserting phase coils
wherever indicated. They should be driven into place by
means of a fiber drift and mallet. Paraffin may be used as a
lubricator, if necessary, as the coil should be a good driving fit
in the protecting cell. The top parts of the first eleven coils,
which are to be the throw coils will be inserted only temporarily
until the bottom coils have been inserted into these slots.
The remaining coils should be driven tightly into place, the
projecting edges of the winding cell trimmed off close to the
core and folded in, and the slot closed by driving a fiber retain-
ing wedge into the grooves at the top of the slot. In some
machines this wedge covers the entire face of the coil, while
in others short wedges are driven in at each side, leaving the
coil exposed at the middle of the core.

Testing the Windings. After all the coils are in place on
either of the types described, the winding must be carefully
inspected for mechanical defects. All coils must clear the
bore of the stator by at least one-sixteenth of an inch, and any
coil which obstructs the bore should be tapped down with a

REWINDING ALTERNATING-CURRENT MACHINES 203

mallet and fiber drift to give the allowed clearance. All cells
must be intact and sound, especially at the bottom of the
slots. If the punchings have been spread apart by driving in
the fiber wedges, they must be closed up again. The punch-
ings should also be inspected to see that no fragments project
into the bore of the stator where the rotor will rub against
them.

The winding should then be subjected to a break-down test
to make sure that the insulation is sound at those points which
in service will receive the most strain. These points are, from
coils to iron, and from phase to phase. Connect all coils of
the same phase together by a piece of copper wire. Then
connect the terminals of a testingtransformer between the dif-
ferent phases and from each phase to iron and apply the re-
quired voltage for a length of time depending upon the charac-
teristics of the machines. (See page 175, Chapter VII.) In
case this test punctures the insulaton of any coil or the insula-
ting cell in any of the slots, this insulation must be removed
and replaced. If the puncture occurs in a top coil, it may
sometimes be repaired by raising the coil out of the slot,
removing the punctured insulation, and carefully replacing
it with new material. If, however, a bottom coil or cell is
punctured, it must be removed by raising the tops of the over-
lapping coils. The same procedure must, be followed if a
coil has to be replaced, as in the case of a burn-out due to a
short-circuit. Such a process amounts practically to retracing
the operation performed in the original winding until the
injured coil is exposed.

Inserting a New Coil in a Winding. In repair work where
only one coil of a basket winding is burned out or damaged,
and in general if the coils have been painted and are stiffened,
rather than to remove all the throw coils, it is frequently
easier to thread in a new coil. To do this cut the damaged
coil at each side of the core and pull the wires out. New fish
paper and treated cloth cells must be inserted. In emergency,
f ullerboard or rope cement paper may be used in place of fish
paper. The treated cloth is variously known as treated
cloth, oiled linen, empire cloth, etc. (See Chapter VII.)
In this operation coils can be bent slightly to get them out of

204

ARMATURE WINDING AND MOTOR REPAIR

the way. Double cotton-covered wire of correct size and suit-
able length for rewinding must be used. The length and size
can be obtained from the old coil by counting the turns and
measuring the length of one. To make the wire thread in
easily it can be rubbed with paraffine. If the coil consisted
of two or more wires in parallel, that number should be used
in rewinding. The number of wires in parallel is the number
entering a given coil from any junction point of coil terminals.
Great care must be exercised during the threading process to

FIG. 144. Method of threading in a new coil to replace a damaged one.

see that all the wires lie parallel in the slots and are free from
kinks. Otherwise it will be impossible to get the full number of
turns into the slot. In case a large number of turns are needed
(more than 20 per slot) two or more lengths of wire may be
wound at the same time and connected in series after winding,
the joint being made outside of the slot. When the required
number of turns have been inserted, the edges of the cells
should be trimmed and folded in, the coil wedged in place and
taped at each end. It should then be painted with an insula-
ting paint.

REWINDING ALTERNATING-CURRENT MACHINES 205

Connecting the Coils. After the ground test, the coils
can be connected into groups, by joining together the begin-
nings and ends of adjacent coils until the group has but one
lead at each end unconnected, which form the leads of that

FIG. 145. The stator of an inductor motor with coils in place ready for
connecting the pole-phase-groups and making end connections of such
groups.

group. In doing this, the wires should be scraped clean
and a sleeve connector of tinned copper slipped over them.
They can then be soldered and taped. If suitable connectors
can not be obtained (as may occur in making repairs), the

206 ARMATURE WINDING AND MOTOR REPAIR

stubs may be wrapped with fine bare copper wire and then
soldered.

The terminals of the various groups should next be connected
into proper phase relations (see Chapter XI) with double-
braided rubber-covered wire or cable. The size of wire and
the grade of insulation will depend upon the current and vol-
tage. Joints in cables or wire larger than No. 6 should be
made by wrapping with fine bare copper wire. Great care
must be taken while connecting and soldering joints to pro-
tect the winding from the molten metal. After soldering,
all joints and splices should be rubbed smooth with emery
cloth and insulated with treated cloth tape, covered with
one or more layers of cotton tape, and saturated with an
insulating varnish or shellac. The connecting cables should
then be arranged over the end of the windings in such a manner
as to occupy the least possible space and yet keep them clear
from the frame or end brackets.

Points to Consider when Connecting Coils. The fol-
lowing points should be carefully watched in order to make a
good job in connecting induction motor windings.

1. All wires should, if possible, be tinned before the connectors are
put on. If this cannot be done they must be thoroughly cleaned by
scraping.

2. All soldering must be thoroughly done, making a smooth and
solid joint.

3. The wires of the joints should lap if space permits.

4. No acid flux should be used.

5. The wires or cables must be arranged in such a way as to occupy
the least space and not obstruct ventilation.

6. Wires or cables must be clamped or tied down to avoid vibration.

7. Joints must be carefully insulated.

8. After connecting, the winding should be tested for electrical
balance by causing a current to flow through it and comparing the
voltage and current readings of each phase.

The windings on the side on which the connections are made,
usually called the front side, are as a rule more rigid than those
on the rear side, due to the fact that the connections exert a
bracing effect. Since the windings in the rear lack this brac-
ing effect, it is usually necessary with diamond windings to

REWINDING ALTERNATING-CURRENT MACHINES 207

supply a supporting ring of insulated steel to which all the
coils are laced with rope made of six- or eight-ply waxed ends.
The inside diameter of this ring should just be large enough
to make a good fit over the winding. If it is much larger than
that of the winding, the coils will be under tension which
will in time tend to loosen them from the supporting ring.
Care should be exercised in lacing the coils to the ring, when
the rope is led through ends of the coils with a steel needle,
to see that no damage is done to the insulation.

Cleats and Terminals. The wires that tap into the winding
and extend to the outside of the frame are called the leads.
To avoid the possibility of these leads transmitting jerks
and vibrations to the soldered joints of the winding, they
should be fastened to the frame by iron or porcelain cleats.
Such cleats should also be used to support the leads from sag-
ging wherever they pass close to the iron or moving parts.

Painting. All of the windings, and especially the taped
joints, should be thoroughly brushed with shellac or a finish-
ing varnish. This material seals up any porous places in the
insulation, and excludes dirt and moisture (see page 176).

IV. INDUCTION MOTOR SECONDARIES

The squirrel-cage secondary is the simplest mechanically,
and at the same time is the most rugged and compact form
of moving element to be found in any electric motor. The
operating characteristics of a squirrel-cage rotor are dependent
on its resistance. A winding of low resistance will have
good efficiency and small slip, but will have poor starting
torque for a given maximum current. A winding of high
resistance, on the other hand, will have lower running effi-
ciency and large slip, but will give a high starting torque with
minimum current, and is suitable for mill or crane work
where starting under heavy load is frequent and operation
is for short intervals only. Where it is necessary, however,
to start a heavy load with small starting current, and operate
for long intervals with good efficiency, or wherever it is neces-
sary to vary the speed or operating characteristics of the motor
from time to time, a wound secondary should be used. The
windings have a low resistance, and are connected in star

208

ARMATURE WINDING AND MOTOR REPAIR

with the open ends connected to slip rings. Adjustable ex-
ternal resistance can then be connected in series when start-
ing, after which the rings can be short-circuited.

FIG. 146. Squirrel-cage rotor for a 350-hp., 3-phase, 440-volt, 1465 r.p.m.
induction motor (Crocker- Wheeler Company).

Squirrel -cage Secondaries. The core of a squirrel-cage
rotor is built up of laminated steel on a spider, and keyed
in place with a feather key on the spider and ring keys at the ends.

FIG. 147. Solid metal rotor "cage" winding of squirrel cage induction motor
(Fairbanks-Morse & Company).

This rotor cage is made by casting end rings of copper or brass around the ends of the
rotor bars which have previously been treated so that the metal of the bars is melted and
fused together with the metal of the ring into a solid mass. This makes a rotor cage
which is of one piece without joints. The iron core of the rotor was removed in order to
obtain this photograph of the winding only, to show the characteristic structure from
which the name "squirrel cage" is derived.

Ventilators are used on the rotors with wide cores. Rectangular
copper bars are usually used, cut from soft drawn bar stock.

REWINDING ALTERNATING-CURRENT MACHINES 209

The end rings are made of copper, brass, or various grades
of resistance alloys. The resistance of the ring, and conse-
quently the characteristics of the motor, depend upon both the
width and thickness of the ring, as well as its composition, and
may be varied over a wide range by changing these dimensions.

Phase-wound Secondaries. With wound secondaries
either diamond or basket coils can be used. Partially closed
slots are usually used and the slots are skewed, to prevent
humming in single-phase designs. Wire-wound coils can be
inserted in a manner as wire-wound, threaded-in coils for

FIG. 148. Large induction motor rotor partly wound with two-part strap
coils (Crocker-Wheeler Company),

direct-current armatures. The slots should be insulated with
fish paper and treated cloth cells. The coil throw is determined
by the number of slots per pole. The coils in each group
should be connected in series, and the groups of each phase
usually connected in series. The phases may be connected
in star. Wedges are required to hold the coils in the slots,
and the rotor should be banded and balanced in the same
manner as a direct-current armature.

In rotors of large machines, the coils are generally form
wound, of strip copper, and should be completely insulated
before insertion in the slots. For partly closed slots the com-

210 ARMATURE WINDING AND MOTOR REPAIR

plete coil may be composed of several strips, each completely
insulated with a wrapper of treated cloth and a winding of
cotton tape. In some types of machines, designed for espe-
cially heavy service, the insulation is composed of sheet mican-
ite, wrapped with cotton tape. A cell of fish paper should be
inserted in the slot for mechanical protection and the straps
threaded in one by one. The cell may then be folded in and
a wedge inserted. Wave windings are used for form-wound
rotors.

Collector rings are made of copper or brass. If mounted in-
side the bearings, they are usually provided with lugs which
are bolted through insulating washers to a ring in a small
spider. If mounted outside the bearings, the leads are
brought out through the hollow shaft and bolted to the rings.

V. WINDING LARGE ALTERNATING-CURRENT STATORS

The machines which fall under this heading include large
induction motors used in industrial plants, engine and water-
wheel driven generators, motor-generator sets, synchronous
motors and frequency changers. While in many cases differ-
ent types of windings are used, the method of winding the
stators is essentially the same in all cases. Particular refer-
ence will be made, however, in the winding details to the usual
construction and requirements of induction motors. The
recommendations for insulation of slots, make up of coils and
their insertion are those of a writer in the Electric Journal,
Vol. VII, No. 12, and represent good practice both for closed
and open slot construction.

Coils for Partially Closed Slots. When it is necessary to
replace the coils in large machines, they should be purchased
from the manufacturer since it is very difficult to properly
form and insulate a number of large heavy coils with the facili-
ties available in an ordinary repair shop or at the location
where the machine is operated, such as a generating station
or industrial plant.

The partially closed slot requires a form of coil which can
be either threaded-in through the slot opening or inserted
from the end. As it is ordinarily impracticable to insulate

REWINDING ALTERNATING-CURRENT MACHINES 21 1

the slots for the voltages commonly used on large machines
by the methods used for threaded-in coils, it is necessary
either to insulate each strand for the full voltage or to use
some form of winding in which the complete coil can be in-
sulated from ground and impregnated and then inserted into
the slot from the end.

The former method may be used where strap diamond coils
of a limited number of turns are to be used, and the voltages

FIG. 149. Large induction motor stator partly wound showing type of coil
used (General Electric Company).

are moderate. A standard application of this method which
is much used on large induction motors consists of a four coil
per slot winding. Two strap diamond coils for each slot are
completely insulated and impregnated. The width and depth
of the insulated coil are each equal to half the width and depth
of the slot. The width of the opening at the top is one-half
the width of the slot, and the winding process is the same as
for an open slot winding, except that there are twice as many
coils for the number of slots as in an ordinary winding.

Where a number of turns per slot are required with a partly
closed slot a concentric shoved through coil can be used.

212

ARMATURE WINDING AND MOTOR REPAIR

This may have as many turns as desired. The coils may be
formed from double cotton covered wires, round in the smaller
sizes and square in the larger. The wires should be cut off in
lengths equal to the total length of the coil plus enough to
allow for joints, and bound together in a long straight bar,
having the correct cross-section for the coil. This bar is

FIG. 150. Section of stator of water-wheel driven generator showing chain

winding.

then clamped by the middle in a forming machine, and the
ends are bent over suitable wooden forms to give the correct
shape to the finished end of the coil. The two free ends should
be left straight, so that they may be shoved through the slots.
Coils for Open Slots. Although the closed slot has some
advantages in the case of induction motors, the form of winding
necessary for more than four or at most six turns per slot, is

REWINDING ALTERNATING-CURRENT MACHINES 213

complicated and expensive to wind. The coils for an open
slot winding are easy and cheap to form, to insulate and to
install. A diamond winding can be readily chorded, and thus
a standard frame and core can be used for different windings.
In addition, a coil which can be completely insulated and im-
pregnated before insertion in the slots is more reliable. Open

FIG. 151. Partly wound stator of a 30-hp., 3-phase, 60-cycle, 685-r.p.m.
induction motor (Crocker- Wheeler Company).

slot windings are used, therefore, for generators and synchro-
nous motors, and for the large sizes of induction motors.

Both concentric (also called spiral or chain) and diamond
windings can be used with open slots Either type of coil
(see page 3, Chapter I) is formed at both ends, and can be
completely insulated and impregnated before assembling in
the core. The concentric winding takes up less room where

214 ARMATURE WINDING AND MOTOR REPAIR

the throw of the coils is great. All the coils in a group are of
different size and shape, however, and the adjacent groups
must be of different length. On account of the number of
different coils used, repairs are difficult to make and a larger
supply of extra coils must be kept in stock. For this reason
the diamond winding is quite generally used. A diamond
coil is of a simple form, easy to build and insulate, and one
form of coil is used throughout. Repair parts can thus be
reduced to a minimum.

Lap and Wave Connections. Either the lap or wave form
of diamond winding may be used. The end connections, by
which the coil ends are connected into groups and the groups
into phases, are very much more simple with the wave wind-
ing. This form of winding is, therefore, used whenever appli-
cable. The voltage between adjacent coil ends is much
greater, but as the coil must be insulated for full line voltage
under any conditions, this does not make any particular differ-
ence. Where more than one turn per coil is required, however,
or where a series-parallel combination is necessary, the wave
winding is more cumbersome than the lap winding. It is
accordingly used only with a one turn per slot coil, where all
the groups in a phase are connected in series. The lap winding
is thus much more common as most high voltage machines
require more than one turn per slot. The coils may be wound
from round or square wire or copper strap, depending on the
number of turns and the size of the conductor.

Insulation of Coils. The insulation to ground is practically
the same for a given voltage for all types of coils. The wire
in wire-wound coils should be cotton covered, and no other
insulation is needed between turns. When the coils are made
up of two or more layers of conductors, however, the layers
should be separated by drilling, cotton tape or treated paper,
according to the type of coil used. Strap conductors may be
insulated between turns with overlapping cotton or mica
tape. In most cases, however, the tape is used over the ends
of the coils only, the straight parts being insulated by inter-
weaving an insulating cell of cement paper and mica between
them. Where several straps are connected in parallel for
greater conductivity, they are ordinarily enameled, cotton

REWINDING ALTERNATING-CURRENT MACHINES 215

covered or taped to prevent eddy current loss. The several
turns, whether connected in parallel or series, may be bound
together with non-overlapping cotton tape and impregnated
before being insulated from ground. While drying, the straight
parts should be clamped in a press so that they will dry per-
fectly straight and without any interstices between the turns.
The hardened impregnating gums then bind the dried coil
into a compact unit.

The insulation to ground may consist of treated taping over
the whole coil, with a protective covering of cotton tape. This
material is widely used on the smaller machines, and for com-
paratively low voltages. For larger machines, however, and
for high voltages, the customary insulation consists of a wrap-
per of cement paper and mica on the straight parts and treated
tape or mica tape on the ends, with a protective layer of un-
treated cotton tape or mica tape on the whole coil, overlapping
on the ends and non-overlapping on the straight parts. The
entire coil may then be dipped twice in an insulating varnish,
and dried thoroughly in an oven after each dipping. The
requisite insulation for the high voltage machines can be se-
cured by extra turns of cement paper and mica wrapper.
Not over three and one-half turns is ordinarily used, however,
on account of the difficulty of properly impregnating such a
coil. Where this does not give sufficient insulation, two or
more separate wrappers may be used, and the coil dipped
twice in varnish and dried after the application of each
wrapper (see pages 163 to 172).

Inserting Shoved Through Concentric Coils. Windings
using coils of this type may be divided into two classes, de-
pending on whether the ends are bent down at one end or bent
down at both ends. The winding processes are practically
similar for both. Coils bent down at one end are used on
both two-phase and three-phase machines, the two windings
being practically identical with the exception of the end con-
nectors. In a two-phase machine the alternate groups are of
the same phase, that is, all the coils in each bank of coil ends
belong to the same phase.

In a three-phase machine, with a winding of this type, every
third group belongs to the same phase, and the groups of each

216 ARMATURE WINDING AND MOTOR REPAIR

phase alternate from one bank to the other. It should be
noted in this case that the groups of the same phase do not lie
adjacent to one another, but are separated by a distance
equal to the pole pitch and that there are only three groups of
coils per pair of poles. Where the groups of each phase have
one side adjacent to another group of the same phase, under
the same pole, with six groups of coils per pair of poles, it is
necessary to have three banks, to allow the end connections to
cross one another. In this case the coils of two of the banks
must be bent down on both ends. The ends of the third
bank are, therefore, bent down also, to secure uniformity of
the windings.

Considerable care is required in inserting coils of the shoved-
through type into the slots. The slots must be first cleaned
from all foreign matter. The coils should then be rubbed
with paramne, and the slots lined with paraffined fish paper,
cut and bent to an exact fit. The coil is thus made to slip
easily into position, and at the same time is protected from
damage by sharp edges of the iron. A tight driving fit is
absolutely essential with this type of winding. If, on trial,
the fit is too loose, strips of treated fullerboard or wood can
be placed in the slot, .or taped to the coils.

The smallest coil in each group should be placed in the
slots first, and worked into them both sides at once, until
the formed end comes within a short distance of the iron.
Wooden distance blocks of appropriate dimensions can be
used to secure uniformity. The other coils may then be
inserted in order. All the coils of one bank should be inserted
before the other bank is started. Where a two-bank winding
is used, and the coils are bent down at one end only, it is
necessary to insert the coils of the different banks from op-
posite sides of the core. By this method the winder forms
only the straight end of the coils, and his work is very much
simplified. When a three-bank winding is used, however,
the winder must bend down the ends of all the coils as he
connects them. In this case it is easier to insert all the coils
from one end. After all the coils are in place a fiber retaining
wedge should be driven in over the top of each coil to hold
it tightly in place, the fit being as close as possible without

REWINDING ALTERNATING-CURRENT MACHINES 217

Second Layer

Lead'

Connections when concentric coils
are used.

damaging the coil or spreading the lamination at the ventilat-
ing slots.

In forming the ends of the coils of both straight and bent
down type, wooden blocks may be used over which to bend and
shape the conductors. The inner layer should be connected
first, the scheme of connections shown in Fig. 152 giving the
best results. The connector consists of a copper sleeve,
which is soldered over the ends, the joints being staggered,
so that the coil ends
will not be unnecessar-
ily bulky. While making
the joint it is advisable
to protect adjacent con-
ductors from the heat

and Solder bv placing a
\

layer of cloth or mica
between them. The finished joint should be smoothed off
with a file or with emery cloth, and then insulated with
treated cloth and friction tape.

Each coil, during assembly should be bent to correspond
with the curve of the stator, so that no part of the finished
coil will extend above the bore. A piece of insulating material,
usually treated fullerboard, should be placed between the
respective layers, which are then bound together with treated
cloth and cotton tape. This serves both as insulation from
ground and as mechanical protection for the coils.

In the process of connecting, care must be exercised that the
ends of the same conductor are not joined together, thus produc-
ing a short-circuited turn. This can be prevented by testing
out with a test lamp. As a precautionary measure, however,
after the coils have been all connected, each one should be
tested out with a testing transformer. This device (see page
126, Chapter V for construction) should be placed over one
side of a coil and a thin piece of sheet steel over the other. If
there is a closed circuit any place in the coil, a heavy current
will flow, and the steel feeler will be strongly attracted to
the iron of the core.

Bar and Connector Winding. This type of winding consists
of solid copper bars which are completely insulated and shoved

218

ARMATURE WINDING AND MOTOR REPAIR

through the slots. One or two bars may be used per slot.
Within an inch or so of the end, the bar is uninsulated and the
bare ends are tinned. End connections of diamond or in-
volute form, insulated and provided with tinned ends, are
used to complete the coil. In order to facilitate the operation
of soldering, the ends of the connectors are slotted or drilled,

FIG. 153.- Section of the stator of a water-wheel driven generator showing
two-part bar coils, horn fiber slot insulation, wood wedges and ventilating
spaces in the core.

for without these openings it is difficult to force the solder
into the center of the joint.

In the case of a two-bar per slot winding, it Is necessary
that the bar next to the rotor be shorter so that the joints
at both ends of the connectors will be accessible to a soldering
iron. In this case, sufficient filling must be placed between
the two bars in the slot to allow the end connector to slip
over the top bar with sufficient clearance for taping. In case

REWINDING ALTERNATING-CURRENT MACHINES ' 219

there is not sufficient room in the slot to allow of this filling,
the top bars are cut away. The connector is shaped, as
shown in Fig. 154, so as to allow it to slip between the lower
bars and over the ends of the upper ones. All connectors
must be soldered in place and the joints taped. With this type
of winding repairs are very easily made, as any bar can be
removed without disturbing the bars in any other slot, It
is not adaptable to high volt-
ages, however, on account of
the limited number of possible
turns. Furthermore, the end
connections are very difficult to
brace adequately.

Diamond Coils. On large
machines a split frame con- FlG '
struction is necessary. After
the machine is completed and tested, the windings are re-
moved by the maker at two points, in order to allow the
frame to be disconnected for shipping. Stfme form of wind-
ing must, therefore, be used that can readily be disconnected
and reconnected on arrival at its destination. This condition
is most readily met by the open slot diamond coil. The
diamond winding is by far the easiest to put in place, to
remove when necessary without damage to the coil, to con-
nect into groups and phases, to brace at the ends, etc. It is,
therefore, much used on large or high voltage machines.

The assembly of wave and lap windings with these coils is
practically the same, and the processes on the larger machines
are similar to those employed on the smaller ones. The slots
should be cleaned and lined with paper protective cells.
The coils may then be inserted one after the other in order,
no attention being paid to grouping, as the coils are all alike.
On a large machine, the careful wedging of the coils in the
slots is especially necessary, as the mechanical stresses that
are brought to bear on account of the heavy currents in case
of short-circuit, are very large. Hence the coil must fit very
tightly in its place. If necessary, strips of treated cement
paper, fullerboard or wood should be placed in the sides and
bottoms of the slots to ensure a tight fit.

220

ARMATURE WINDING AND MOTOR REPAIR

When inserting the coils, the first half of each coil should
be driven tightly into the bottom of the slot with a fiber
drift and mallet. The coil which goes over it must also
be driven snugly into place. The fish-paper cell can then be
cut off even with the top of the slot and folded over the coil.
In order that this cell may not be torn by the wedge, a strip
of treated fullerboard, the full width of the coil, may be laid

FIG. 155. Stator of a 3-phase Fairbanks-Morse belted type alternator
wound with the coils shown in Fig. 117.

into the slot, so that the wedge will slide over it readily, and
at the same time compress the coil tightly into the slots.
The wedges may, for convenience, be divided into sections
6 to 8 inches long and driven into position by means of a
standard wedge driver or a blunt chisel. If the wedge is
a very tight fit, the coil should be driven down with a drift
just ahead of the wedge.

REWINDING ALTERNATING-CURRENT MACHINES 221

A four-coil per slot diamond winding used with partly
closed slot cores, is assembled the same as any other diamond
winding, with the exception that two coils lie side by side in
both the bottom and top of the slot. The shape of the slot
is such that no difficulty is encountered in inserting the second
coil. This is essentially an induction motor winding, being
used quite generally on both stator and rotor of large machines.

Double Windings. On large induction motors, two speeds
are sometimes secured by the use of two separate windings in
the same slots, connected for a different number of poles.
In this case the windings may be so designed that the coils
for one speed lie inside the coils for the second speed. The
two windings can be assembled at the same time, each pair
of coils being treated as if it were a single coil of a one speed
winding. Each slot thus contains four coils one above the
other. In such cases it is quite common to have the two wind-
ings of widely different capacities, depending, of course, on
the load to be carried at the separate speeds.

Testing Windings of Large Machines. After the coils
of the winding are all in place, their free terminals should be
temporarily connected together, and a test made for break
down to ground, with the standard voltage for the size and
type of machine (see page 175, Chapter VII). After the coils
have been connected into groups, and the groups connected
into phases they should be tested for break down between
phases, and for short-circuits between turns.

The break-down test consists of the application of the stand-
ard break-down voltage from the conductor to ground and
from phase to phase. The test for short-circuit consists in
placing an alternating-current magnet of the type shown in
Fig. 101, page 126 over one side of the coil or group, and hold-
ing a light piece of steel over the other side of coil. If current
flows in the coil, the piece of steel will be attracted. As no
current can flow under normal conditions, this test is a sure
indication of short-circuits. It cannot be applied where the
groups are connected in parallel but must be made, in this
case, before they are connected.

Connecting the Coils. The number of groups in any ma-
chine may be either one or two per phase per pair of poles,

222 ARMATURE WINDING AND MOTOR REPAIR

depending on the arrangement of the coils and groups. The
coils of each group are connected in series but the groups in
a phase may be connected in series, parallel or series-parallel,
although for voltages of 2200 or higher they are nearly al-
ways connected in series. The phases on a two-phase machine
are never connected together inside the machine.

Stator windings of alternating-current machines may be
connected either star or delta, the choice depending upon many
conditions. For machines of 2200 volts and over it is ad-
visable to connect the winding in star because the voltage
from any one terminal to ground is only (2200 -T- 1.73) of the
voltage between terminals and therefore the winding has a less
chance to break down to ground. However, this is not
a strict rule for sometimes it is more convenient to have the
stator delta connected for the reason that a smaller size of
wire can be used and a larger number of turns.

On a concentric winding the groups are readily distinguish-
able. In a diamond or involute winding, the number of coils
per group must be counted off, and the groups temporarily
connected by bending together the leads from a wire coil or
by slipping a copper connector over the stubs from a strap
coil. They should be permanently connected by soldering
the joints so made. On most machines, the connectors be-
tween groups consist of insulated copper strap, with suitable
openings at each end to slip over the group leads.
All joints after being carefully soldered, should be smoothed
up with emery cloth, so that no sharp-pointed edges are left
to damage the insulation.

The connectors between coils, ordinarily called stubs, should
be insulated with treated cloth tape, and a protective cover-
ing of untreated cotton tape. Large stubs are sometimes
covered with drilling caps, sewed to shape, and painted with
insulating varnish after they have been fastened in place.
Treated tape should be wound over the joints in the insulation.
The thickness of the taping and other insulation, depends on
the voltage of the machine. The connectors between groups
are ordinarily insulated and impregnated before they are put
on the machine. The joints between the connectors and the
group leads should be insulated in the same way as the stubs.

REWINDING ALTERNATING-CURRENT MACHINES 223

Where double windings are used they are usually arranged
for the connections to be made on opposite sides of the
machine. Each winding may be connected as if it were
entirely independent.

Bracing Needed for Heavy Windings. The stresses occur-
ring in the end connections of a large machine, due to magnetic
reactions between current carrying conductors, are quite
large. In addition, these effects are greatly magnified in
case of a short-circuit on the generator. Some method of
bracing these end connections is therefore necessary. This
ordinarily takes the form of an insulated steel ring, to which the
coils are tied with heavy twine. On a concentric winding
a separate ring is used for each bank of end connections. On
a diamond winding one ring at each end is sufficient. On a
large machine, however, it is usually necessary to brace this
ring by additional steel supports bolted to the frame.

VI. WINDING THE STATOR OF ALTERNATING-CURRENT
TURBO-GENERATORS

Although the alternating-current turbo-generator falls in
the previous group of large machines, on account of its winding
being more or less special, it will be dealt with separately.
There are special features of windings for this class of machines,
owing to the high speed at which they are operated and the
forced requirement that the inside diameter must be small on
account of the centrifugal strains of the revolving field rotor.
The generator core is therefore long requiring large coils fre-
quently more than 10 feet long with a span from 30 to 40
inches and weighing as much as 100 pounds. One of the best
discussions of the methods for winding an alternating-current
turbo-generator has appeared in the Electric Journal, Vol.
VIII, No 3. The winding details that follow have been taken
from that source.

Coils for A.-C. Turbo-generators. These machines have
open slots and usually employ a diamond coil built up in two
pieces for one or several turns per slot. The two-piece coil is
used where the throw of the coils is very great, as in large size
bipolar machines for 25 cycles. The one-piece coil is used

224 ARMATURE WINDING AND MOTOR REPAIR

on smaller machines, and in fact, wherever such a coil is not too
cumbersome to handle, provided the complete coil can be
passed through the bore. The coils of either type may be
formed from cotton-covered strap or cotton-covered wire.
Where a conductor of large cross-section is required, it is ordi-

FIG. 156. End of a large turbo-generator slowing method of bracing end
connections to resist short-circuit stresses.

narily built up of a number of square copper wires in parallel, to
facilitate bending and to obtain proper lamination. The
individual strands are usually cotton covered. Most turbo-
generators require more than one turn per slot in order to
obtain the desired voltage. These several turns are bound

REWINDING ALTERNATING-CURRENT MACHINES 225

together in a mechanical unit, insulated from each other and
insulated as a group from the frame of the machine.

The coils may have their end connections bent down at any
angle from zero to 90 degrees. The involute coils, of course,
lie flat against the face of the frame. Diamond coils are al-
ways bent down at the ends from 30 to 60 degrees, both to
provide ample clearance and ready access to the rotor, and in
order that they may be more suitably braced, as shown in
Fig. 156. Especial care should be taken to so shape the coil
ends that cool air can circulate freely through them. Coilp
of several turns sometimes have the individual turns insulated
separately after they leave the slots to give greater cooling
surface to the ends.

Forming the Coils. Except for conductors of large capacity,
the conductors are formed from a single copper strap. The
process of forming the coil is the same whether each conductor
consists of one bar or several bars or straps in parallel. Each
group of wires or straps which make up a conductor in the
latter case may be formed as a single strap and then bound
together with tape and treated as a unit. For a one- conductor
coil, designated as an open coil, a strap of suitable length may
be bent at the middle around a pin, forming a U with the sides
separated very slightly. This loop should then be mounted
directly in a mould. A steel pin through the bend of the U
will serve to hold the point of the diamond vertical while the
sides are bent to conform to the shape of the mould. Coils
of several conductors are sometimes built in this same way, by
forming the individual conductors in a group with a copper
strip of the same thickness as the insulation inserted between
them. The conductors should be separated and insulated
from one another, and then insulated from ground the same
as a one-conductor coil. The several conductors can be con-
nected in series when placed in the machine.

A coil of more than one conductor formed from a single
strap is known as a closed coil and requires several operations.
The strap should be first wound around pins set in a flat
table, to the required number of turns, and then given its
final shape by forming over a mould. Throughout the entire
operation the conductors must be kept apart by straps of

226 ARMATURE WINDING AND MOTOR REPAIR

metal to the same distance that they will be separated when
insulated and placed in the machine. Two-piece coils can
be formed from lengths of straight copper strap. The ends
should be bent at a suitable angle to the straight part of the
coil around two pins spaced a distance equal to the length of
the straight part. The conductor may then be clamped in a
mould and the ends bent into shape. The several conduc-
tors in a coil can be formed in the same mould, and separated
by strips of copper strap.

Insulation for Turbogenerator Coils. Since a turbogen-
erator may at times be subjected to very heavy overloads,
the insulation should be as nearly heat proof as possible.
For this reason, mica is much used. Where several con-
ductors per coil are required, the insulation between conduc-
tors usually consists of flexible mica tape and mica cells. The
individual conductors may then be suitably bound together.
All coils should be impregnated with an insulating compound
before the insulation from ground is supplied. The straight
parts may be clamped tightly in metal clamps and impreg-
nated, or may be impregnated without the clamps and placed
in clamps to dry. Either method results in a compact, uni-
form construction so that the coil may be fitted tightly into
the slot with minimum risk of damaging the insulation.

The insulation from ground usually consists of wrappers of
paper and mica over the straight part. The coil ends should
be insulated by layers of treated tape, with a covering of
untreated tape, and the whole coil treated with insulating
varnish. As in the case of engine type generators, if extra
insulation is required additional wrappers of the paper and
mica are frequently used, the coil being dipped and dried
after each wrapper is applied. This makes the building of a
high-voltage coil a long process, two weeks or more frequently
being required for the completion of a single coil.

Testing Turbogenerator Windings. Open coils may be
tested for short-circuits by applying a suitable voltage between
conductors. No short-circuit tests are usually made on the
individual closed coils, but an over-voltage test should be
applied after they are on the machine. All coils should also
be tested for insulation by wrapping the outside with tin

REWINDING ALTERNA TING-CURRENT MACHINES 227

foil, and applying from two to two and a half times normal
operating voltage across the insulation.

Inserting the Coils in a Turbogenerator. Except for the
great weight and large size of the coils which makes them more
difficult to handle, the winding of a turbogenerator is in all
essentials similar to the winding of an engine-type generator.
The straight part of the coil should be waxed and laid over
the slot opening, inside a paper cell with which the latter is

FIG. 157. Turbo-generator stator partly wound using one-piece coils.
The operator is binding the coils in position.

lined. In order to avoid bending the coil, a wooden drift
as long as the straight part may be used to drive it into the
bottom of the slot. The bottom halves of two-part coils
should be inserted first in all slots in the span of one coil,
the top halves being inserted later. One-piece co Is should be
inserted in regular order. After each top coil has been fitted
into place, the cell which lines the slot may be folded over,

228 ARMATURE WINDING AND MOTOR REPAIR

and retaining wedges in short sections driven in place over the
full length of the coil. Enough strips of extra material must
be used in the slot to make the wedges fit tightly over the coil.
The wedges may be driven in place by a tool of special shape,
operated by a compressed-air hammer.

In order to get the bottom half of the last few one-piece
coils into the slot, it is necessary to lift the top half of the
coils which go in the same slots. With a full pitch winding
this would mean one-half of the coils on a two-pole machine,
and one-fourth the coils on a four-pole machine. With a
"chorded" or fractional pitch winding, however, the number
of throw coils is less. The throw of the coils can be readily
seen in Fig. 157 which shows a generator partly wound with
one-piece coils.

In some cases, the improved space factors that can be se-
cured makes it worth while to use a one-coil-per-slot winding.
This winding can be inserted in the same way as the two-coil-
per-slot winding, except that alternate slots contain, respec-
tivelv, front and rear ends of coils.

Bracing for Windings. The instantaneous current which
flows in case of a short-circuit to a turbo-generator op, in fact,
any other generator, is very large and the magnetic stresses,
which vary as the square of the current and the sp^n of the
end windings are consequently enormous in the turbo-genera-
tor. For this reason adequate bracing of the coil ends is of
supreme importance. Greatest reliance is placed on metallic
braces securely bolted to the generator frame. For involute
coils, braces may be used that consist of U shaped clevises
which are thoroughly insulated with treated tape and bolted
to the machine frame at the ends through spacers. For dia-
mond coils, the braces take the form of malleable iron brackets
bolted to the frame. The coils should be rigidly secured
to these braces by nonmetallic clamps, reinforced by brass
plates, as shown in Fig. 156 and fastened to the braces by
insulated bolts.

Connecting the Winding. Turbine windings are almost
always connected with one group of coils per pole per phase.
The coils of each group are connected in series. In some
cases these connectors are riveted as well as soldered in place.

REWINDING ALTERNATING-CURRENT MACHINES 229

The groups per phase may be connected in parallel or series,
depending on the requirements of the machine. Ordinarily
the series connection is used on machines of 2200 volts or
higher. The end connections should be fully insulated and
supported on the rear of the braces.

Break-down Test. After all the coils have been placed
in the core, their free ends should be connected together and
the standard break-down voltage (see page 175, Chapter VII)
applied from the copper to the machine. A similar break-
down test should be applied between phases after the coils
have been connected into groups. No further tests need be
applied until the final load tests are run.

CHAPTER IX

TESTING INDUCTION MOTOR WINDINGS FOR MIS-
TAKES AND FAULTS

After the coils have been placed in the stator of an induction
motor and connected by the winder to give the required com-
bination' of phases, volts and poles, the next step is to connect
the motor to a test circuit and note whether or not the current
taken is the same in all the leads. If it is not, or the motor
does not operate properly in other respects, the winding must
be checked and tested for possible errors in connection of
coils and for other defects. For an experienced induction
motor winder, this is a simple job and one that is done very
quickly. For the repaiman who has had less experience in
connecting induction motors, the job is not so quickly per-
formed. Yet it is not a difficult one when the procedure is
correct. One of the most easily understood methods of pro-
cedure for checking induction-motor windings has been formu-
lated by A. M. Dudley and published in the Electrical Journal.
It is so simple yet complete that any repairman can with a
little patience discover and correct even the most puzzling of
errors and defects. It is given complete in what follows.

The errors which may occur in induction-motor windings
have been arranged by Mr. Dudley in the approximate order
of their occurrence as follows:

I. Short Circuit and Grounds.

1. Short circuits or grounds of one or more turns in an individual

coil caused by a break in insulation.

2. Short circuit of a complete coil caused by connecting together the

two ends of the coil. This is known by winders as " stubbing a
coil dead."

3. Short-circuiting a complete pole-phase-group.

4. Short-circuiting one complete phase of the winding.

II. Reversal of Part of the Winding.

1. This may be confined to a single coil or it may be a. pole-phase-
group or a complete phase.

230

TESTING INDUCTION-MOTOR WINDINGS

231

III. Open Circuits.

1. Caused by actually leaving the winding disconnected or open

at some point.
IV Placing One or More Coils too Many or too Few in a Pole-phase-group.

V. Using an Improper Group Connection.

1. For example, connecting in series when the winding should have
been in parallel or vice versa. This is called "connecting for
double or half voltage, respectively." A similar error would be
connecting delta instead of star.

VI. Connecting for the Wrong Number of Poles.

1. While this error comes under No. V, it is usually regarded as dif-
ferent and is sometimes difficult to locate unless it hasbeen noticed
that the speed of the motor when running light is not what it
should be.

Testing for Grounds and Short Circuits. Grounds of any
description are located by "ringing out" between the copper

Support at End
of Laminations

FIG. 158. Section and plan views of a short-circuit detecting device for
testing stator windings.

and the frame with a magneto or "lighting out" between the
same points with an ordinary ItO-volt lighting circuit test
lamp. They are comparatively simple to locate and easy tc
repair by the use of insulating material used with the job.
Short circuits of a few turns or a single coil become hot in a
short time if the motor is run light on normal voltage. Their
presence can be detected by feeling around the winding with
the hand immediately after starting the machine and noting
if any coils are much warmer than others.

232

ARMATURE WINDING AND MOTOR REPAIR

Short-circuit Detecting Device. A device that can be used
for detecting short circuits before the rotor is placed in the
stator and without applying voltage to the winding itself, is
shown in Fig. 158. It it somewhat similar to a large horse-
shoe magnet except that the iron part is built up of lamina-
tions. It may also be considered a core-type transformer
having the primary coil only and having one side of the iron
core missing. The coil is excited with alternating current of

a suitable voltage and then
the complete device is
passed slowly around the
bore of the stator as shown
in Figs. 159 and 160. The
laminations of the stator
core complete the magnetic
circuit of the testing de-
vice and an alternating
magnetic field flows in the
stator core as shown by the
dotted lines in Fig. 159.
When moving the device
around the stator bore, and
it passes over a short-cir-
cuited turn or coil, such
short-circuited turn or coil

immediately acts as a short-circuited secondary coil on a
transformer of which the exciting coil on the testing device is
the primary. As in -any short-circuited transformer an in-
creased current flows both in the primary and in the secondary
coil and can be detected by an ammeter in series with the
device or by the heating which immediately takes place in the
defective coil in the winding or by the attraction which the
other side of the short-circuited coil has for a strip of sheet
iron. By passing the device slowly around the core and ob-
serving its behavior from point to point, short circuits can
readily be detected. This refers particularly to short circuits
in individual turns or in one complete coil. A short circuit of
a complete pole-phase-group is more readily located by a com-
pass test, and a short circuit of an entire phase can be found

FIG. 159. Method of using the short-
circuit detecting device shown in Fig.
158.

TESTING INDUCTION-MOTOR WINDINGS 233

by a balance test. The balance test is made after the winding
has been checked for grounds and short-circuited turns and
has had the resistance of all the phases measured. If these
checks indicate the proper number of turns in series, a com-
paratively low alternating voltage is applied to the winding
of the stator without the rotor being in place. The current is
then read in all the phases and if it checks up the same or
balances as it is called, the machine is considered free of
grounds and short circuits.

FIG. 160. Testing stator winding for short-circuited coils by use of a small
alternating-current magnet.

Reversal of One or More Coils or Groups. It happens that
individual coils or sometimes entire groups are connected in
backward. If the error is confined to one coil it does not usu-
ally show up on the " balance " test and, of course, would not
be found on the resistance test since the resistance is the same
no matter whether the ends of the coils are interchanged or
not. Such reversed coils or groups can be located by means of
a polarity test with a compass. In this test the motor wind-

234 ARMATURE WINDING AND MOTOR REPAIR

ings are excited by a comparatively low direct-current voltage
so chosen that the current is limited to a reasonable value.
If the windings are so excited and a compass is placed inside
the bore and passed around slowly following the inner periph-
ery, the needle of the compass will reverse in passing from
a north pole to a south pole group and vice versa. If an indi-
vidual coil is reversed it will show a tendency to reverse the
compass needle when the needle is directly over that coil. If
an entire pole-phase-group is reversed the compass needle will
indicate the same direction of field on two successive groups.
Also if a coil is left out of circuit or is "dead," as already men-
tioned, it will indicate an irregularity at the instant the com-
pass passes directly above it. By checking the three phases
of a three-phase winding separately, in this manner and
marking the result inside the bore with chalk of different
colors, it is possible to check for the reversal of an entire phase.

Open circuits are best checked by " lighting out." This
test it made by connecting a 110-volt lighting circuit and an
incandescent lamp in series with the winding. If there is an
open circuit the lamp will not light.

Placing the wrong number of coils in two or more phase
groups can hardly be deteced otherwise than by actually
making a physical count. Since this is a simple matter, the
check is best made in that way.

Using an improper group connection has in general the
same effect as raising or lowering the voltage on a machine
and can best be checked by operating the motor on what
should be the correct voltage. If this voltage is considerably
too high as would be the case if the winding was connected
parallel when it should be series, or delta when it should be
star, the motor will emit more than the ordinary amount of
magnetic hum and will probably overheat in a short time even
when not mechanically loaded. If the voltage is much too
low, as would be the case if the winding was connected series,
when it should be parallel, or star when it should be delta,
the fact can be determined by trying a load on the motor.
When the load comes on the drop in speed will be too great
and the motor will pull out and come to a standstill on a load
much less than normal.

TESTING INDUCTION-MOTOR WINDINGS 235

Connecting for the wrong number of poles can most readily
be detected by checking the no-load speed. (If it has a wound
rotor, this must be short-circuited.) The no-load speed, being
approximately synchronous, is very nearly equal to cycles
X 120 -r- number of poles. If this gives a result differing
from that expected, the winding is connected for the wrong
number of poles.

Order in Which Tests Should be Made. The order in which
these various checks should be performed is usually as follows :
After the winder has completed the connection, the windings
are checked against the winding diagram. The coils per
group are counted and a visual inspection made for short
circuits, open circuits and reversed coils. A balance test is
made with low voltage to see if the separate phases show the
same result. A high voltage test is then made on the insula-
tion to ensure that the coils are not grounded between phases
nor on the iron core. The machine is then assembled and the
resistance of the completed winding is measured on all phases.
If these checks are satisfactory the machine can be passed for a
running light test without load. Sufficient voltage is put on
the windings to start the motor up. If it comes up to speed
without apparent distress or irregularity of any kind, the speed
is checked and the temperature of the winding is tested with
the hand all the way around the machine. If this is normal
the voltage is raised to its normal value and the no-load cur-
rent in all phases and the watts are read. If these values
check with previous calculations or tests on duplicate machines
the windings are considered to be correctly connected. If no
data is available on the no-load current it may be considered
reasonable if it does not exceed 40 per cent, and is at least 20
per cent, of full load current. The no-load watts running
light may be considered reasonable if they are roughly in
the neighborhood of seven or eight per cent, of the normal
rating of the motor. If the motor does not readily come up
to speed or the phases do not balance or there are signs of
unequal heating in the winding or other distress, the rotor
should be removed and the connections checked. If the error
is not apparent and a source of direct current is available the
compass test may be applied.

236 ARMATURE WINDING AND MOTOR REPAIR

Connections for Applying Direct Current when Exploring
A.-C. Windings with a Compass. The accompanying dia-
grams show the methods for making connections for the use
of direct current to excite the windings of an alternating-current
motor so as to explore the windings by the use of a compass.
When the winding is star connected, one lead of the direct-
current circuit is joined to A-B-C connected together and

D.C.Supgly>

FIG. 161. Method for checking 3-phase, star and delta connected windings
with direct current.

the other lead to the neutral point. When the compass is
held over the coils in the slots and moved around the stator,
there should be three times as many poles as there are poles
in the machines reversing alternately and spaced equally when
the winding is balanced. In case of a delta-connected winding,
it is only necessary to open the delta connection as shown in
Fig. 161 at the right and connect the two ends to the direct-
current leads.

CHAPTER X

ADAPTING DIRECT-CURRENT MOTORS TO CHANGED
OPERATING CONDITIONS

The changes in direct-current motors which the repairman
is most frequently called upon to make are: 1. Changes in
speed. 2. Changes in operating voltage. 3. Changes to
operate a motor as a generator and vice versa.

Changes in Speed. A change in speed of a motor from 10
to 15 per cent, can usually be made by increasing or decreas-
ing the air gap between the armature and fields. In general
the per cent, changes in air gap required will be three or four
times the per cent, change in speed. This will serve only as
a rough check for use when measuring the air gap and consider-
ing the change of speed in this manner. To make this change
with accuracy, it is necessary to have a magnetization curve
for the design of the particular machine dealt with. How-
ever if the change in speed desired is not too great, the change
in air gap will in most cases give desired results. The excep-
tions are those cases where the motor has been built to work
high up on the saturation curve where the iron is practically
saturated or very low down on the saturation curve where the
excitation is practically all used in the air gap. The speed is
increased by increasing the air gap and reduced by reducing
the air gap.

It is usually easier to reduce the speed by change in air gap
than to increase the speed. In the former case "sheet steel
liners or shims can be inserted next to the frame and different
lengths of air gaps secured over a considerable range depending
upon the length of the original air gap of the machine. If
the air gap cannot be increased by removing such shims, it
may be necessary to grind off the required amount from the
pole faces. Such a grinding should not be done more than
once if the operation will narrow the width of pole face to an
appreciable extent. When the increase in speed is more than

237

238 ARMATURE WINDING AND MOTOR REPAIR

can be secured by increasing the air gap, it may be possible to
add a resistance to the shunt-field circuit. In such a case it
will be necessary to check carefully the no-load and full-load
speeds.

In the case of a shunt motor, a compound characteristic
can be secured by increasing the air gap and then adding series
turns to the field poles until the full-load speed is the same as
it was before the change. For this purpose flexible copper
cable can be wound around the shunt coil and bound in place.
In reversing this change, that is, in changing a compound
motor to a shunt characteristic, the air gap should be reduced
and current shunted out of the series coils of the field until a
constant speed range is secured.

An inefficient and emergency method of reducing the speed
of a motor for a short time only is by the use of a resistance
in the armature circuit. When a variable resistance is
connected in series with the armature leads, it simply reduces
the impressed voltage on the armature winding by using up a
part of the voltage in the resistance. Since the speed of the
motor is proportional to the operating voltage, the desired
speed reduction can be thus secured.

In all cases where the speed of a motor is increased, atten-
tion must be paid to the armature construction so that the
increased centrifugal strains on the coils and the banding due
to the increase in speed will develop no serious defects, such
as rubbing of coils on the fields, high commutator bars and poor
commutation. When the increase in speed is 50 per cent, or
more, the manufacturer should be consulted for advice as
to the safety of the high speed for continuous operation.

Changes in Operating Voltage. The speed of a direct-
current motor varies directly with the operating voltage and
(theoretically) inversely with the flux of the fields. When
the operating voltage is increased on a motor, the excitation
of the fields is also affected but on account of the saturation
of the iron, the field flux is not affected in direct proportion
so that the speed of the motor on an increase of voltage follows
about in proportion to the increase. When the voltage is
not increased or decreased more than 25 per cent, over the
rated value for the motor, the correct speed can be obtained by

CHANGES IN DIRECT-CURRENT MOTORS 239

a change of the air gap. In such a change the operating
temperature of the field coils must be watched, since the
temperature will vary about as the square of the voltage.
A temperature higher than 160F. will injure the insula-
tion. To prevent this when the temperature of the fields
is found to be around this figure, it will be advisable
to rewind the field coils. In case the machine is operated
on under voltage with the air gap as large as possible, a re-
sistance can be used in the shunt field until the required speed
is secured. On a compound motor, no-load and full-load
speeds can be adjusted by changing the series field coils in
either of the cases mentioned.

Operating a Motor on One-half or Double Voltage.
It frequently happens that a 220-volt motor must be changed
to operate on 110 volts or a 110-volt motor on 220 volts.
In the case of a 220-volt motor, it is usually possible to connect
the shunt-field coils in two groups and then connect these
groups in parallel. In such a case when the motor is connected
to a 110-volt circuit, the voltage per coil will be the same so
that the field flux will be the same, but the speed will be only
one-half that on 220 volts. The speed can be brought up
to normal by increasing the air gap as much as possible and
using resistance in the shunt field as explained in a preceding
paragraph. The rating in horsepower of a motor so changed
will only be one-half of what it was before.

When a 110-volt motor must be changed to operate on 220
volts, the conditions are not so easy as in the previous case
for if the smallest air gap with the shunt fields in series cannot
be used, the field coils must be rewound. When operated
in this way the motor will have double the horsepower rating
on 220 volts that it did on 110 volts. The changes in the
motor which have been mentioned do not require a change
in the armature winding.

Changes in Armature Winding for Operating Motors on
One-half or Double Voltage. The most satisfactory way
of changing a motor to operate on a voltage which is one-half
or double the rated voltage is to rewind the armature to suit
the new conditions. This is also the most expensive way
and need not be resorted to when the change in voltage is

240 ARMATURE WINDING AND MOTOR REPAIR

not over 25 per cent. A direct-current armature winding can
usually be changed for operation on a lower voltage in two
ways. 1. By using new coils. 2. By reconnecting the old
armature winding. When the armature is to be rewound
for a different voltage, the number of turns in series between
brushes will vary directly as the old voltage to the new. The
cross-section of the wire for the coils will also vary inversely
as the old voltage to the new. For instance if a 110-volt
motor is to be operated on 220 volts, the armature conductors
will be doubled in number and their cross-section made one-
half, when the speed and horsepower rating will remain the
same as before. When the size of wire previously used is
known, one of one-half the cross-section will be three B. & S,
gauge numbers higher.

A practical way of making these changes in armature
windings in a repair shop is given on pages 243 to 260 as out-
lined by T. Schutter (Electrical Engineering, July and August,
1918) for the changes of one -half and double- volt age operation.

Size of Wire for D.-C. Armature Coils. Before winding
the coils needed in the repair of an armature it must be decided
whether a lap or a wave winding will be used. In the lap
winding there are as many current paths or circuits through
the armature winding as there are poles with the coils con-
nected in series in each of these paths or circuits. The wave
winding has only two current paths through the armature
winding regardless of the number of poles the machine has.
In each of these two paths the coils are connected in series.
Considering / the total armature current as given on the name
plate of the machine and i the current in each circuit of the
armature winding, the following formulas can be used to find
the maximum current that can be carried by the coils of
each circuit connected in series.

For a lap winding in which the number of circuits equals
the number of poles or p,

For a multiple lap winding, having two or more single lap
windings in parallel,

i = I -T- mp

CHANGES IN DIRECT-CURRENT MOTORS 241

In this case m is the number of single lap windings in parallel.

For a wave winding which has two circuits
i = I + 2

For wave windings having more than two circuits, usually
called multiple-wave or series-parallel windings

i = I -f- 2m

In this case m is the number of wave windings used in the
armature.

The permissible current density in armature conductors
varies from 1500 to 3000 amperes per square inch. The
cross-section or diameter of the wire for the armature coils
can then be found by dividing the current in each section
of the winding (i) by the allowable amperes per square inch.
The value to use in the case of small machines should not
exceed 3000; in intermediate sizes 2000; and in large sizes
1500.

When winding new coils in the repair of an armature it is
necessary to determine the available winding space in the
slot before deciding upon the number of turns and form of
coil to use. In case the armature is to be rewound for changed
conditions of speed or voltage reference should be made to
Chapters X and XI where details to be observed are given
for different changes in operating conditions.

Operating a Generator as a Motor and Vice Versa. It is
usually possible to operate a generator as a motor by setting

FIG. 162. Direct-current generator with armature rotating clockwise.

the brushes for the correct rotation of the armature and with
a backward lead instead of a forward lead when operating as
a generator. To successfully operate a motor as a generator
the air gap should be reduced to a minimum and the speed

242 ARMATURE WINDING AND MOTOR REPAIR

increased where the voltage as a generator is to be equal or
higher than the voltage as a motor. When the voltage is to
be lower, the same speed and air gap can usually be used.
When a direct-current machine is changed from a generator
to a motor, the current in both the armature and series field
reverses. If, therefore, the machine is operated as a cumu-
latively compounded generator, it will also operate as a differ-
entially compounded motor. In order to use a compound
generator as a motor, it is usually necessary to reverse the
connections of its series fields. The diagrams of Figs. 162
and 163 are self-explanatory.

Fio. 163. Differential motor with armature rotating clockwise and series
field opposing shunt field.

Adjusting the Air Gap on D.-C. Machines. When insert-
ing the armature in any machine in which the bearings are
independent of the frame, the air gap between the armature
core and pole faces should be checked up. Any inequality
of air gap will cause unnecessary friction and heating of the
bearings and an unequal heating of the armature iron. The
air gap may be adjusted horizontally in many cases by cross
beams and jack screws on the bedplate and vertically by
thin sheet liners inserted between the bedplate and the
yoke. The air gap can be gauged during the operation by
inserting a hardwood wedge on the front and back ends of
the machine at different points, noting the distance to which
the wedge enters each time. The adjustment should be con-
tinued until the air gap is the same around the entire circum-
ference of the machine.

Motor Speed when Reconnecting a D.-C. Motor Winding
Wave to Lap. When reconnecting a wave winding to form
a lap winding without changing the coils, or voltage, the speed

CHANGES IN DIRECT-CURRENT MOTORS 243

of a four-pole motor will be increased twice, and the speed
of an eight-pole motor increased four times. This is because
the wave winding is a two-circuit winding while the lap
winding has as many circuits as there are poles. The
speed, therefore, varies directly as the number of circuits in
the winding.

Change in Brushes when Reconnecting a D.-C. Motor
from a Higher to a Lower Voltage. When reconnecting a
direct-current armature from a higher to a lower voltage,
say 220 volts to 110 volts, the size of conductors must be
doubled as the current will be about double its original value.
The original brushes on the 220-volt machine must therefore
be changed to handle this increase in current without heating
and sparking. If twice as many brushes of the same size as
before cannot be used, the size of brush must be changed to
give twice the bearing surface. An increase in width of the
brush should be avoided, the length being increased to get the
proper bearing surface. A current density of more than 50
amperes per square inch will give trouble sooner or later in
a reconnected armature.

For a four-pole motor originally 220 volts, changed to
operate on 110 volts, the field coils should be connected two
in series and these groups connected in parallel in order to
get the same field current.

REWINDING AND RECONNECTING DIRECT- CURRENT ARMA-
TURE WINDINGS FOR A CHANGE OF VOLTAGE*

When a direct-current armature is to be rewound or re-
connected for a change from one voltage to another, the
number of turns in series between brushes as explained on page
239 will vary directly as one voltage to the other and the
cross-sectional area of the wire will vary inversely as the vol-
tages. To illustrate, take the case of a 2-pole armature con-
sisting of 24 coils, 20 turns per coil wound one wire-in-hand
using number 19 B. & S. gauge wire, which has a cross-sec-
tional area of 1290 circular mils. The armature has a 240-volt
winding and is to be changed so as to operate on a 120- volt

* T, Schutter, Electrical Engineering, July, 1918,

244 ARMATURE WINDING AND MOTOR REPAIR

circuit. This can be accomplished in two ways. First, by
rewinding the armature; second, by reconnecting the present
winding.

By comparing the original voltage with the new operating
voltage, it will be seen that the new voltage is just half of the
original voltage. If it took 20 turns per coil for 240 volts
then it will take half of 20 or 10 turns per coil for 120 volts.
As the machine is to do the same amount of work, it will
carry twice the current at 120 volts that it did at 240 volts.
For this reason the wire must have twice the cross-sectional area,
or 1290 X 2 = 2580 circular mil area. This is equal to a num-
ber 16 B. & S. gauge wire, still using the same number of coils
as before.

Original 240 volt connection

FIG. 164. A 240- volt lap winding which is reconnected for 120 volts as
shown in Fig. 165.

The other method is to reconnect the winding so that two
coils are in parallel, and bridge two commutator bars. This
will result in a winding of 12 coils, two in parallel, with 20
turns per coil. In Fig. 164 the original winding is shown, as
it was wound and connected to the commutator. The arma-
ture core contained 24 slots and there were two coil sides per
slot. The part of the slot which is occupied by a coil side is
called a winding space. The odd numbered winding spaces
are considered as being in the bottom of the slot and the even
numbered winding spaces being in the top of the slot. The
table for winding coils in Fig. 164 is given on page 245 and
the connecting table for the original winding as illustrated in
Fig. 164 on page 247.

CHANGES IN DIRECT-CURRENT MOTORS

245

By tracing the direction of flow of current through the wind-
ings from the positive brush to the negative brush, it will be
seen there are two paths or circuits, each consisting of 12 coils
in series, the two paths or circuits being in parallel with each
other. This winding operates on 240 volts.

TABLE FOR WINDING COILS ON ARMATURE OF FIG. 164

Coil number

Coils are wound

In spaces number

In slots number

1 and 24

1 and 12

3 and 26

2 and 13

5 and 28

3 and 14

7 and 30 .

4 and 15

9 and 32

5 and 16

11 and 34

6 and 17

13 and 36

7 and 18

15 and 38

8 and 19

17 and 40

9 and 20

19 and 42

10 and 21

21 and 44

11 and 22

23 and 46

12 and 23

25 and 48

13 and 24

27 and 2

Hand 1

29 and 4

15 and 2

31 and 6

16 and 3

33 and 8

17 and 4

35 and 10

18 and 5

37 and 12

19 and 6

39 and 14

20 and 7

41 and 16

21 and 8

43 and 18

22 and 9

45 and 20

23 and 10

47 and 22

24 and 11

246 ARMATURE WINDING AND MOTOR REPAIR

TABLE FOR CONNECTING COILS OP FIG. 164 TO COMMUTATOR

Beginning of coil
number

To bar number

End of coil number

To bar number

22 >'

PIG. 165. The 240-volt winding of Fig. 164 reconnected for a 120-volt
circuit by connecting two coils in parallel and using wider brushes.

Reconnecting a Lap Winding. In Fig. 165 the winding
shown in Fig. 164 has been reconnected so as to operate on
a 120-volt circuit.

CHANGES IN DIRECT-CURRENT MOTORS

247

As explained it will require one-half as much winding on 120
volts as it did on 240 volts. This can be accomplished by
connecting two coils in parallel, and using wider brushes;
that is, the brushes should be at least as wide as 1J/2 commu-
tator bars and not more than two commutator bars. If it
is possible to arrange for the use of the wider brushes, the
commutator can be bridged, as shown by the jumpers A, B, C,
etc., Fig. 165.

v 'I''

TABLE FOB CONNECTING COILS OF FIG. 165 TO COMMUTATOR

Beginning of coil
number

To bar number

End of coil number

To bar number

By tracing the direction of flow of current through the wind-
ing it will be seen that the current flows in four circuits from
the positive brush to the negative brush. In each path or cir-
cuit there are six coils in series, and the four paths or circuits
are in parallel. This reconnection is equivalent to rewinding

248 ARMATURE WINDING AND MOTOR REPAIR

the armature with twice the size of wire and one-half the
number of turns with which it was wound for 240 volts. Both
Figs. 164 and 165 show lap or parallel windings.

Reconnecting a Wave Winding. Fig. 166 represents a
four-pole wave or series winding, consisting of 31 .coils. The

Slot No.

Original240 v olt Connection

FIG. 166. Wave winding for a 240-volt, 4-pole armature.

armature core has 31 slots, and there are two coil sides per
slot. It is wound so as to be operated on a 240-volt circuit.
The coils in this Fig. 166 are placed so that the beginning of
each coil is placed in the bottom of the slot, and is represented
by the odd numbered coil sides.

Winding
Space No.

Slot No.

Reconnected for 120 Volts

- FIG. 167. The wave winding of Fig. 166 reconnected so that there are
four paths for current instead of two with four brushes instead of two. This
winding can now be used on a 120-volt circuit.

By tracing the current from the positive to the negative
brush, it will be seen that the winding is divided into two paths
or circuits. To change this winding so that it can be operated
on a 120-volt circuit, two methods can be used : First, reconnect
the winding so that it will consist of four paths or circuits,
and add another set of brushes as shown in Fig. 167. The

CHANGES IN DIRECT-CURRENT MOTORS
TABLE FOR WINDING COILS IN SLOTS OF FIG. 166

249

Coil number

Coils are wound

In spaces number

In slots number

1 and 16

1 and 8

3 and 18

2 and 9

5 and 20

3 and 10

7 and 22

4 and 11

9 and 24

5 and 12

11 and 26

6 and 13

13 and 28

7 and 14

15 and 30

8 and 15

17 and 32

9 and 16

19 and 34

' 10 and 17

21 and 36

11 and 18

23 and 38

12 and 19

25 and 40

13 and 20

27 and 42

14 and 21

29 and 44

15 and 22

31 and 46

16 and 23

33 and 48

17 and 24

35 and 50

18 and 25

37 and 52

19 and 26

39 and 54

20 and 27

41 and 56

21 and 28

43 and 58

22 and 29

45 and 60

23 and 30

47 and 62

24 and 31

49 and 2

25 and 1

51 and 4

26 and 2

53 and 6

27 and 3

55 and 8

28 and 4

57 and 10

29 and 5

59 and 12

30 and 6

61 and 14

31 and 7

additional brushes are marked AI and B\. Second, connect
two coils in parallel and still have two circuits in the winding.
This, however, will necessitate the dropping of one coil, as
shown by the heavy lines in Fig. 168. Only two brushes
will be required as shown, if they can be set so that they
cover at least 1J^ bars, the jumpers A, B, C, etc., can be
omitted.

250 ARMATURE WINDING AND MOTOR REPAIR

TABLE FOR CONNECTING COILS OF FIG. 166 TO THE COMMUTATOR

Beginning of coil
number

To bar number

End of coil number

To bar number

The winding as shown in Fig. 167, is a four-pole lap or
parallel winding, with four circuits or paths through the
winding. The second method of reconnecting Fig. 166 so as
to operate it on 120 volts is shown in Fig. 168. As previously
explained, each two adjacent coils will be connected in parallel,
and since there are 31 coils on the entire winding, one coil
must be dropped. In this winding (Fig. 168) coil No. 2, shown
with the heavy lines has no connection with the commutator

CHANGES IN DIRECT-CURRENT MOTORS 251

TABLE FOR CONNECTING COILS OF FIG. 167 TO THE COMMUTATOR

Beginning of coil
number

To bar number

End of coil number

To bar number

FIG. 168. The wave winding of Fig. 166 reconnected for a 120-volt circuit
by connecting two coils in parallel and using wider brushes. In this case one'
toil is dead as shown by the heavy lines.

252

ARMATURE WINDING AND MOTOR REPAIR

There are 31 bars in the original commutator, so that by taking
any two bars and putting a jumper across and considering
them as one bar, it will be reduced to 30 bars. The connections
will be made as shown by the following table:

TABLE FOR CONNECTING COILS OF FIG. 168 TO THE COMMUTATOR

Beginning of coil
number

To bar number

End of coil number

To bar number

Coil 1 is not con-

nected.

' 18

22
23

26
' 27

Reconnecting Duplex Windings.* The windings discussed
in what follows are wound with two wires in hand. The
* T. Schutter, Electrical Engineering, August, 1918.

CHANGES IN DIRECT-CURRENT MOTORS

253

same results can be accomplished by two windings, using one
strand (two simplex windings) at a time. In a winding of the
duplex type there are usually twice as many commutator
bars as there are slots and each of the two wires is connected to
separate bars. The brush, however, will cover at least 1J/2
to 2 commutator bars, as shown in Fig. 169, which is the wind-
ing and connections for a 120-volt lap-wound armature.

12 34 1 la 2 2a 33a

Slot No.

Original 120 Volt Connection

FIG. 169. Duplex lap winding for a 120-volt armature.

Each coil consists of two parts, which will be called Sec-
tion 1, and la, Section 2 and 2a, Section 3 and 3a, etc. The
following is the winding table for Fig. 169.

TABLE FOR WINDING COILS IN SLOTS OF FIG. 169

Coil number

Coils are wound

In spaces number

In slots number

1-la

1 and 6

1 and 3

2-2a

3 and 8

2 and 4

3-3a

5 and 10

3 and 5

4-4a

7 and 12

4 and 6

5-5a

9 and 14

5 and 7

6-6a

11 and 16

6 and 8

7-7a

13 and 18

7 and 9

8-8a

15 and 20

8 and 10

9-9a

17 and 22

9 and 11

10-10a

19 and 24

10 and 12

11-lla

21 and 2

11 and 1

12-12a

23 and 4

12 and 2

This 120-volt winding is to be changed so that it can be
operated on a 240- volt circuit. As shown in Fig. 169 Section

254 ARMATURE WINDING AND MOTOR REPAIR
TABLE FOR CONNECTING COILS TO THE COMMUTATOR OP FIG. 169

Start of coil number

To bar number

End of coil number

To bar number

lOa

*10a

lla

12a

1 and la of coil 1 are connected in parallel through the brush
connections, and by changing the connections so that Sec-

Slot No.

Reconnected for 2iO Volts

FIG. 170. The 120-volt lap winding of Fig. 169 reconnected for 240 volts by
connecting the turns of each coil in series and using smaller brushes.

tions 1 and la will be in series instead of in parallel, it will then
be possible to operate the winding on a 240-volt circuit. The

CHANGES IN DIRECT-CURRENT MOTORS

255

winding table for Fig. 170 will be the same as for Fig. 169
but the connecting table will be as follows:

TABLE FOR CONNECTING COILS OF FIG. 170 TO COMMUTATOR

Start of coil number

To bar number

End of coil number

To bar number

lOa

107

lla

12a

The brushes used on Fig. 170 should only be as wide as one
commutator bar. By tracing the current through the winding
in Fig. 169 it will be seen that there are three coils in series
in each of the paths or circuits from the positive brush to the
negative brush. In Fig. 169 there are six coils in each path or
circuit, or twice as many as before the reconnection.

If the winding in Fig. 169 had been connected to a 12-bar
commutator instead of a 24-bar commutator, the reconnecting
from 120 volts would have been somewhat different. In
Fig. 169 the numerals from 1 to 12, which are placed above
and between each two bars, will give an idea of how the con-
nections would look. For instance, Sections 1 and la of

256 ARMATURE WINDING AND MOTOR REPAIR

coil No. 1 would be connected across bars Nos. 2 and 3, instead
of Section 1 to bars 3 and 5, and la to bars 4 and 6.

Then to reconnect the winding from 120 volts to 240 volts
when only 12 commutator bars are used, Sections 1 and la
of Coil No. 1 would have to be connected in series, but not
through commutator connections as in Fig. 170. By omitting
the connections of commutator bars, No. 2, 4, 6, 8 and 10,
etc., and simply splicing the end of Section 1 to the beginning
of Section la of coil No. 1, the same results will be obtained
by using only 12 commutator bars.

Fig. 171 is a " duplex wave" winding for operation on a
120-volt circuit. This winding consists of 13 coils wound with
two strands of wire and connected to 26 commutators bars.
This winding could also be connected to a 13-bar commu-
tator. The winding table for Fig. 171 is as follows, each coil
being considered as two Sections, 1 and la.

TABLE FOR WINDING COILS IN SLOTS FOE FIG. 171

Coil number

Coils are wound

In spaces number

In slots number

1 and la

1 and 6

1 and 3

2 and 2a

3 and 8

2 and 4

3 and 3a

5 and 10

3 and 5

4 and 4a

7 and 12

4 and 6

5 and 5a

9 and 14

5 and 7

6 and 6a

11 and 16

6 and 8

7 and 7a

13 and 18

7 and 9

8 and 8a

15 and 20

8 and 10

9 and 9a

17 and 22

9 and 11

10 and 10a

19 and 24

10 and 12

11 and lla

21 and 26

11 and 13

12 and 12o

23 and 2

12 and 1

13 and 13a

25 and 4

13 and 2

The winding as it is now placed and connected will operate
on a 120-volt circuit. It is desired to change it so that it
will operate on a 240-volt circuit. There are two methods
by which this can be accomplished. The first is shown in

CHANGES IN DIRECT-CURRENT MOTORS 257

TABLE FOR CONNECTING COILS OF FIG. 171 TO COMMUTATOR

Start of coil number

To bar number

End of coil number

To bar number

lOa

lla

12a

13a

12 34 llo 22 a 3 3a

Original 120 Volt Connection

FIG. 171. A duplex wave winding for a 120- volt armature.

Fig. 172. In this case it will be seen that Section 1 of coil
1 is dropped, that is, it is not connected to the commutator.
This may be done so that a wave winding is possible. In-
stead of connecting the different sections of a coil in parallel

258 ARMATURE WINDING AND MOTOR REPAIR

through the commutator and brush connections, they are
now connected in series in the same way. The connecting
table is as follows :

TABLE FOR CONNECTING COILS OF FIG. 172 TO COMMUTATOR

Start of coil number

To bar number

End of coil number

To bar number

lOa

lla

12a

13a

It will be seen that bars Nos. 26 and 1 are bridged and are
acting as one bar. This is due to the dropping of Section 1
in coil No. 1 for reasons given above. By tracing the flow of
the current from the positive to the negative brush in both
Figs. 171 and 172, is will be found that there are twice as
many coils in series in Fig. 172 as there are in Fig. 171.
It is, therefore, possible to operate Fig. 172 on a 240- volt
circuit.

The other method of reconnecting Fig. 171 so as to operate

CHANGES IN DIRECT-CURRENT MOTORS

259

it on a 240- volt circuit is shown in Fig. 173. The coils in
this case have been placed the same as in Figs. 171 and 172,

1 la 2 2a 3 3

FIG. 172. The duplex wave winding of Fig. 171 reconnected for 240 volts by
connecting the coils in series through the commutator and the brushes.

12 3*

FIG. 173. The duplex wave winding of Fig. 171 reconnected by connecting
the turns of each coil in series and using one-half the number of commutator
bars. In this case two adjacent bars must be connected together and brushes
wide enough to cover two bars.

therefore the winding table will be the same. The connect-
ing table is as follows:

TABLE FOR CONNECTING COILS OP FIG. 173 TO COMMUTATOR

Start of coil number

To bar number

End of coil number

To bar number

, 15

lOa

lla

12a

13a

260 ARMATURE WINDING AND MOTOR REPAIR

From the above connecting table it will be seen that the
end of Section 1 is connected to the beginning of Section la
or, in other words, the two sections of a coil, which were in
parallel with one another in Fig. 171 are now connected in
series.

As already explained, the number of turns per coil will be
directly proportional to the voltage. In Fig. 171 assume
that there are 20 turns per coil with two wires in parallel. By
reconnecting as in Fig. 173, there will be 40 turns per coil,
using one wire. From this it will be seen that while there are
13 coils in each winding, Figs. 171 and 173, one has twice as
many turns per coil as the other. This fact makes it pos-
sible to operate Fig. 173 on a 240-volt circuit.

CHAPTER XI

PRACTICAL WAYS FOR RE- CONNECTING INDUCTION

MOTORS

There are certain changes in the windings of an induction
motor that can be made by the repairman to adapt a motor to
changed operating conditions. There are also certain changes
that should not be attempted. Details of the most practical
re-connections and their effects on motor operation have been
carefully outlined accompanied by diagrams, in the Electrical
Journal by A. M. Dudley. These details are summarized in
what follows.

The changes that can be made in the connections of the
induction motor may be divided into three classes. The first
class includes those which leave the motor entirely normal and
the performance in all essential respects the same as before
re-connection. Such changes, for example, are represented by
connecting the polar groups of a winding in series for 440 volts
and in parallel for 220 volts. The second class of changes
leaves the performance in some respects unchanged and alters
it in others. These may be represented by operating a motor
in star on 440 volts and in delta for 220 volts. In this case
there is little change in efficiency or power factor; the starting
and maximum torque, however, are only 75 per cent, of their
original values. In such a case the advisability of the change
depends upon the work that the motor must do. If the torque
at the altered values is sufficient to start and carry the driven
load easily, there is no objection to operating the motor indefi-
nitely so reconnected, since the motor will not run any
warmer than before and its efficiency and power factor may be
better.

The third class of changes leaves the motor operative in the
sense of producing torque enough to do the work required,
but so alters its performance as to heating, or efficiency or
power factor or insulation that it is undesirable to leave the

261

262 ARMATURE WINDING AND MOTOR REPAIR

motor operating indefinitely in such condition. Such changes
are represented by re-connecting a three-phase motor without
changing the coils for two-phase operation. This is equiva-
lent to operating the three-phase motor at 125 per cent, normal
voltage. In addition, the coils which should have extra in-
sulation where the phases change have only group insulation.
The iron loss and heating may be increased to a dangerous
degree and the power factor greatly decreased. Such changes
should only be used in an emergency and the proper permanent
changes made as soon as possible.

Points to Consider before Making Re -connections. Before
a repairman attempts to make a radical change in the connec-
tions of an induction-motor winding he should consider care-
fully the limitations of the design and the effects of the changes
The following points will serve as a guide.

1. Changes in voltage alone are the easiest and can usually be made.

2. Changes in number of phases alone can rarely be made satisfactorily
and are usually only makeshifts.

3. Changes in number of poles are limited, due to the mechanical form
of the coils.

4. Changes of frequency alone or in combination with voltage or phase
can sometimes be made if changes in speed are not objectionable.

5. Complicated changes should not be attempted except by persons
of some experience and should be handled with caution.

6. If the peripheral speed of the rotor (which equals rotor diameter
in feet X 3.14 X rpm.) exceeds 7000 feet per minute on any proposed
change, the maker of the motor should be consulted before making the
change.

7. In case of any doubt on any point, refer to the manufacturer of the
machine.

Diagrams for Different Changes of Connections. Two

kinds of diagrams are most used to indicate the connections
to be made in induction motors. These are shown in Figs.
174 and 175. The diagram of Fig. 174 is a three-phase, four-
pole winding connected in star on a 36-slot core. It represents
the coils as they would look if removed from the machine and
laid on a table with the actual connections made. The dia-
gram of Fig. 175 is a conventional sketch for the winding of
Fig. 174 showing the polar groups. The latter diagram is
much used,

RE-CONNECTING INDUCTION MOTORS

263

In the winding illustrated, since there are three times four or
12 polar groups and 36 coils, there are three coils connected
in series to form a polar-phase-group. It should be borne in

Full lines represent coils in tops of slots,
broken lines coils in bottoms of slots

FIG. 174. Complete winding diagram for a 3-phase, 4-pole motor having
36 slots, and connected in series-star.

mind that there are not actually 12 magnetic poles in the
machine, for the reason that three consecutive polar-phase-
groups unite to form orje magnetic pole by virtue of the phase
difference of the currents in the three phases. There are two
magnetic north and two magnetic south poles formed by this
winding at any instant, and these poles are equally spaced
around the air-gap like four mechanically projecting pole
pieces excited by four coils carrying direct current. Some-
thing of this conception is gained
if one imagines the armature of
a direct-current generator held
stationary and the field poles
rotated around it. At any given
instant the magnetic field can
be conceived to be the same as
the field which is formed by
the winding shown in Fig. 174.
The coils in slots 3, 6, 9, 12,
etc., are shown by heavy lines
to indicate that the insulation

on these Coils is heavier to FlG - 175 - Circle diagram for the
j.u j XT. connections shown in Fig. 174.

withstand the greater strain at

the points where the winding crosses or lies adjacent to coils
differing greatly in potential. This is called "phase insula-
tion," and may be put on the first coil in each group or it

264 ARMATURE WINDING AND MOTOR REPAIR

may be put on the first and last coil of each group where there
are a large number of coils in the group. This is one of the
reasons why a machine may not at times be reconnected
for another number of poles or phases. If such reconnections
were made the maximum differences in potential might occur
between two adjacent coils unprotected by this extra insula-
tion and a break-down result.

The conventional diagram of Fig. 175 represents the same
connections as Fig. 174, except that each pole-phase-group
as numbered 1, 2, 3, etc., in Fig. 174, is shown by a short arc
in Fig. 175. The numbers on the groups are identical with
Fig. 174, and the group connections. The arrows are shown
simply to indicate a method of checking up to insure the proper
phase relations. There is considerable danger, in a three-
phase connection, of getting a 60-degree relation between the
phases instead of a 120-degree relation or, as it might be ex-
pressed on the diagram, there is danger that the wrong end of
the B phase, for example, may be connected to the star point.
As a check against this, when the diagram is completed, the
current is assumed as going in at all three leads toward the
Y point. Arrows are put on each pole-phase-group as shown,
and when all three phases are traced through, the winding is
correct if the arrows on consecutive groups run alternately
clockwise and counter-clockwise. It may be argued that this
is an artificial assumption and that at no instant is the current
flowing toward the star in all three phases. It may also be
argued that in a correct winding, if the current be assumed as
flowing toward the star in two phases and always from it in
the third, the arrows will fall in successive sets of three in the
same direction and then three in the reverse direction. These
statements are true, but a little experimenting will show that
an incorrectly connected or 60-degree winding can in this way
be shown to give successive sets of three arrows and still be
wrong. There is but one exception to the correctness of the
check as shown in Fig. 175 where the current is assumed as
flowing toward the star in all three phases and the arrows
alternate in direction. This exception to the rule is the case
where the winding forms consequent poles or passes through
all the pole-phase-groups in a north direction instead of

RE-CONNECTING INDUCTION MOTORS

265

alternately north and south. Such connections are rarely used ,
and then usually on special motors wound for multi-speeds.
Diagrams for Three-phase Motors. Fig. 176 gives a com-

B C A

FIG. 176. Diagram for a 3-phase,
4%pole winding with parallel-star end
connections. Schematic equivalent
in center.

FIG. 177. Diagram for a 3-phase,
4-pole winding with 4-parallel-star
end connections. Schematic equiva-
lent in center.

(A-BXB-CXA-CV

Fio. 178. Diagram for a 3-
phase, 4-pole winding with series-
delta end connections. Schematic
equivalent in center.

FIG. 179. Connecting diagram for
a 3-phase, 4-pole, parallel-delta wind-
ing. A schematic equivalent is shown
in the center of the diagram.

bined conventional and schematic representation of a so-called
parallel star diagram, where the two halves of each phase are

266 ARMATURE WINDING AND MOTOR REPAIR

in parallel. If a given machine were connected, as shown in
Fig. 175, for a normal voltage of 440, it could readily be recon-
nected according to Fig. 176, and would then be suitable for
operation on 220 volts having the same performance in all re-
spects except that it would draw from the 220-volt line twice
as many amperes under a given load as it previously drew from
the 440-volt line. Similarly, if it had four poles, or a multiple
of four poles, it could still be paralleled again, or put in 4-
parallel star, as shown in Fig. 177, and operated on 110 volts,
and would still have the same performance at a correspond-
ingly increased current at the same load. '

Fig. 178 represents a variation in connection from the fore-
going, which is possible only with three-phase machines.
This is the so-called delta or mesh connection. If a machine
connected as in Fig. 174 for 440 vofts be reconnected as in
Fig. 178 it would be suitable for operation on a circuit having
a voltage of 440 -f- 1.73 or 254 volts.

COMPARISON OF MOTOR VOLTAGES WITH VARIOUS CONNECTIONS

If a motor connected originally as shown in any horizontal column had
a normal voltage of 100 its voltage when reconnected as indicated in any
vertical column is shown at the intersection of the two columns.

J
1

1
CO

1
4<

J
1

e!,

-
*

PH
^

_c
b
co

e!,

fi
<A

4
<N

3-Phase Series Star

100

3-Phase 2-Parallel Star. .

200

100

116

162

3-Phase 3-Parallel Star..

300

150

100

174

243

122

3- Phase 4-Parallel Star. .

400

200

133

100

232

116

324

163

108

3-Phase 5-Parallel Star..

500

250

165

125

100

290

145

405

203

135

100

3-Phase Series Delta

173

100

140

3-Phase 2-Par. Delta. . . .

346

172

116

200

100

280

140

3-Phase 3-Par. Delta....

519

258

174

129

105

300

150

100

420

210

141

105

3-Phase 4-Par. Delta....

692

344

232

172

140

400

200

133

100

560

280

188

140

112

3-Phase 5-Par. Delta

865

4-30

290

215

175

500

250

165

125

100

700

350

235

175

140

2-Phase Series

125

100

2- Phase 2-Parallels

250

125

146

200

100

2-Phase 3-Parallels

375

188

125

219

110

300

150

100

2- Phase 4-Parallels

500

250

167

125

100

292

146

400

200

133

100

2-Phase 5-Parallels

625

313

208

156

125

365

183

122

500

250

167

125

100

RE-CONNECTING INDUCTION MOTORS

267

Use of Table of Connections. Different reconnections or
conversions are shown in the Table on page 266 where the
problem just shown may be worked out by selecting Three-phase
Series Star in the horizontal
column (first line) and reading
across to the vertical column
headed Series Delta where the
figure 58 appears. This means
that if 100 volts was normal on
the series star connection and a
change is made to series delta
the corresponding voltage is 58.
By multiplication, if 440 was
the series star voltage, the series
delta voltage would be 4.4 X 58
= 254, as noted above. Figs.
179 and 180 show a parallel
and 4-parallel delta connec-
tion, respectively, and bear
the same relation to Fig. 178 that Figs. 176 and 177 do to Fig.
175.

Two-phase Diagrams. In Fig. 181 a development of a two-
phase winding is shown similar to Fig. 174, except for the differ-
ence in the number of phases. An inspection of the coils
represented in heavy lines and a comparison with the coils in

B c A

(,B-C)(A-C)IA;B)

FIG. 180. Connecting diagram
for a 3-phase, 4-pole, 4-parallel-
delta winding.

Full lines represent Coils in tops' of slots, broken lines coils in bottoms of slots

FIG. 181. Complete winding diagram for a 2-phase, 4-pole motor series
connected. The coils from x to y form one pole-phase-group.

Fig. 174 indicates at once what is meant by the " phase coils"
or phase insulated coils being differently situated. This also
explains one of the good reasons why two-phase motors should

268 ARMATURE WINDING AND MOTOR REPAIR

not be reconnected for three-phase, or vice versa, without
changing the position of these " phase coils.
Fig. 182 gives the conventional and schematic equivalent cf

FIG. 182. Connecting diagram for
a 2-phase, 4-pole winding with series
connections.

FIG. 183. Connecting diagram for
a 2-phase, 4-pole winding with parallel
connections.

Fig. 181. The arrows shown in the three-phase diagrams are
omitted here, for the .reason that the two phases are not inter-
connected, and the only effect
of reversing one phase is to
reverse the direction of rotation
of the motor. This is readily
corrected by reversing the two
leads of one phase at the motor
terminals. Figs. 183 and 184
give parallel and 4-parallel, two-
phase connections and bear the
same relation to the series con-
nection as the three-phase star
and delta diagrams. From these

n l n 2 l z

FIG. 184. Connecting diagram 2-parallel and 4-parallel connec-
for a 2-phase, 4-pole winding with tions it may be readily seen that

4-parallel connections. . , , r i

where the number of poles is a

multiple of three, as 6, 12, 18, etc., there is a possible analogous
3-parallel connection; also where the number of poles is a
multiple of five, such as 10, 20, etc., there is a corresponding

RE-CONNECTING INDUCTION MOTORS 269

possible 5-parallel diagram. These are the connections which
are indicated in the Table on page 266 as "3-parallel" and
" 5-parallel."

Fig. 185 shows a possible three-phase connection which
may be made from a two-phase winding by a method similar
to the Scott transformer connection. It is a connection which
should be used only as a temporary expedient until better
arrangements can be made.

-Coils removed
, from circuit

IBs

FIG. 185. The T-connection by which a 2-phase motor may be operated on
a 3-phase circuit.

Meaning of the Term Chord Factor. It is well known that
the span of the coil must in general be somewhere near the
quotient of the bore periphery divided by the number of poles.
It is not so generally understood that changing the span of the
coil within limits has an effect similar to increasing or decreas-
ing the number of wires in the coil. If the coil is exactly pitch,
i.e., spans exactly from the center of one pole to the center of
the next, the turns of wire in that coil are producing their
maximum effect upon the magnetic field. The coil is then
considered to span 180 electrical degrees. It is customary to
wind the coil in slots so that it spans something less than a
full pole pitch. The effect of the turns in the coil is then some-
what less than the maximum. The effect of the turns in the
coil varies as the sine of half of the angle in electrical degrees
which the coil spans.

To illustrate, if there are 72 slots in an eight-pole machine,
the coils would be exactly " pitch" if they lay in slots 1 and 10,

270 ARMATURE WINDING AND MOTOR REPAIR

or in other words, if there were eight slots between the two
slots in which the two sides of any coil were located. Such a
coil would span 180 electrical degrees. Half of 180 degrees is
90 degrees and the sine of 90 degrees is 1 ; therefore the effect
of the turns in such a coil is one. Suppose instead the coil lies
in slots 1 and 8. It would then t span 140 degrees elec-
trically, since 72 -f- 8 = 9 slots represents 180 degrees, and
one slot therefore represents 20 degrees. The sine of half of
140 degrees, or 70 degrees, is 0.938. It follows that the effect
of the turns in this coil is less than that of the full pitch coil
by the ratio of 0.938 to 1. This is of interest in the present
problem, because it is often possible in making changes to
change at the same time the span of the coils by one slot,
more or less, by springing the coil mechanically, and so im-
prove the performance of the machine under the new con-
ditions. The point becomes of vital importance, immediately,
when changing the number of poles without changing the
throw of the coils. Referring again to the 72- slot motor,
assume that the coils are wound in slots 1 and 8. For
an eight-pole connection these coils will have the effect of
0.938, as explained above. If the connections are changed
for six poles the effect is entirely different. 72 -r- 6 = 12 and
180 -:- 12 = 15, or each slot represents 15 electrical degrees.
A throw of 1 and 8 covers seven complete slots, or
7 X 15 = 105 degrees; the sine of half of 105 or 52.5 degrees
= 0.79, which means that when connected for six poles
the coils have an effect of only 0.79, as against 0.938 when
connected for eight poles. It is possible to avoid using the
sine of half the angle and secure a factor which is sufficiently
accurate practically by using the expression,

(Number of slots per pole) 2 2 (Number of slots dropped) 2
(Number of slots per pole) 2

Using the same eight-pole example above, the number of
slots per pole is 72 -r- 8 = 9 and the pole pitch is 1 and 10.
When the coil is wound 1 and 8 it spans seven slots and there
are 9 7=2 slots dropped. The expression then becomes

= A ^ = 0.948

RE-CONNECTING INDUCTION MOTORS 271

and similarly for the six-pole,

(12) 2 - 2(5 2 )
12 2

which agrees roughly with the other method. A coil should in
no case be chorded more than half of the pole pitch, as second-
ary disturbances of the magnetic field are occasioned by chord-
ing which become prohibitive at that point. The expression,
"sine of half the angle spanned by the coil," is given the name
chord factor, and it should be considered in the work of re-con-
necting. For example, if the poles are changed from 8 to 6,
as in the example above, and the chord factor changes from
0.938 to 0.79, the new line voltage should be 0.79 *- 0.938
times the old, neglecting the effect of other changes which are
being made. If nothing else was undergoing change and the
-normal voltage was 440 in the first place, it should be 370 after
the change is made or, expressing it another way, if it was still
operated at 440 volts after the change, the motor should be
thought of as operating at about 18 per cent, over voltage.

Phase Insulation. It is the practice of many manufac-
turers to put heavier insulation on the coils at the ends of the
polar groups which are mechanically adjacent to one another
and are also subjected to the voltage between phases, which
may be the maximum voltage between supply lines. Such
coils are illustrated at Nos. 3, 6, 9, 12, 15, etc., in Fig. 174.
By comparing this diagram with Fig. 181 for two-phase
connection, it appears at once that both the number
and location of these so-called "phase coils" should be
changed at the time the machine is re-connected from two- to
three-phase, or vice versa, assuming that the voltage can be
changed so that a phase change is permissible. Also in
changing the number of poles, the number and location of the
"phase coils" must also be changed. In fact, whatever re-
connection is attempted the "phase coils" should be checked
and re-arranged, since this is comparatively easy and adds
considerably to the protection of the machine from break-
downs of insulation.

272 ARMATURE WINDING AND MOTOR REPAIR

RE-CONNECTING MOTORS TO MEET NEW OPERATING
CONDITIONS

Re-connections Frequently Made. The following changes
are made on account of changes in operating conditions or the
service conditions on the circuits from which the motors are
operated.

1. Changes to operate on a different voltage.

2. Change for operation on a different phased circuit, three- to two-
phase, etc.

3. Changes to operate on a different frequency.

4. Change in number of poles of the motor.

In case of change No. 4, the re-connection may be independent of all
other changes to secure a faster or slower speed or it may follow as a re-
sult of change No. 3 in order to keep the same speed on a driven
machine when the motor is operated on the new frequency.

Procedure when Considering a Re-connection of Windings.

The procedure in checking up a machine to see whether or
not it can be re-connected is as follows : First, ascertain the ex-
isting connection and the throw of the coils in order to know
what the possibilities are in the way of number of turns and
throw. Second, if it is a phase or voltage change, find di-
rectly from the Table on page 266 what connections will give
approximately the proper new voltage and new phase. If
any one of these connections is possible with the number of
poles in the machine, select it as the new connection and ar-
range the phase coils properly at the beginning or ending of
the groups, or at each end of the groups if there are enough of
them in the old winding. Since the speed has not changed, the
horsepower should remain approximately the same, and the
current in the coils themselves will remain somewhere near the
original. If the frequency is to be changed either independ-
ently or in conjunction with a phase or a voltage change, the
applied voltage should be changed in the same direction and
by the same proportional amount as the frequency is changed,
or if the voltage is to remain unchanged the number of turns in
series in the coils should be changed in the opposite direction
to the frequency and by the same amount. For example, if a
25-cycle motor is to be run on 30 cycles, it should have the volt-
age increased 20 per cent., or else have the groups re-connected

RE-CONNECTING INDUCTION MOTORS 273

so that there will be 20 per cent, less turns in series and run on
the same voltage.

If the number of poles is to be changed, and consequently
the speed, check first the effect of the coil throw or chording
with the new number of poles. Then think of the motor
winding as generating counter emf. and bear in mind that
with a constant field a higher speed will generate more emf.
and a slower speed less emf. Converted into voltage this
means that with a higher speed a higher voltage should be
applied in direct proportion and with a lower speed a lower
voltage should be applied. If the voltage can not be changed
try to change the diagram of group connections so as to vary
the number of turns in series in the right way, that is, if the
voltage should be increased, the same effect can be obtained
by decreasing the number of turns a like amount. In all
these cases it is the voltage per turn or per conductor which
counts, just as in a transformer, and a careful consideration
of the effect of different connections will show whether the
desired change in voltage per conductor is being accomplished.

Practical Example for Reconnection. Assume a 25-hp.,
four-pole motor operating on 40 cycles, two-phase, 220
volts. It is desired to know whether it can be re-connected
to operate on 60 cycles, three-phase, 550 volts at the same
speed and horsepower. An inspection of the machine shows
that it has 72 slots and 72 coils and that any individual coil
lies in slots 1 and 15, also that the groups are connected in
parallel. Since there are 72 -r- 4 = 18 slots per pole, each
slot is 180 -r- 18 = 10 electrical degrees and 14 slots = 140
electrical degrees. (The throw of 1 to 15 means spanning
14 slots.) The sine of one-half of 140 degrees or 70 degrees
= 0.94 = chord factor, or figured by the formula without trig-
onometry, since there are 18 slots per pole and a throw of
1 to 15 means dropping four slots from exact pitch, the chord

/18 2 2 X 4 2
factor = -\l ^g 2 - = 0.948. The synchronous speed

of the motor on 40 cycles as it stands is 4800 -r- 4 = 1200
rpm. To get this same speed on 60 cycles it is evident the
motor will have to be connected for 7200 -r- 1200 = six poles.
If the throw of the coils be left 1 to 15 they will throw two slots

274 ARMATURE WINDING AND MOTOR REPAIR

further than full pitch, since 72 -f- 6 = 12 slots per pole and

1 to 13 would be exact pitch. Throwing the coil over pitch
has the same effect as throwing it under pitch so the new chord

/122 2 X 2 2

factor on six poles = A/ 22 - = 0.97, or sine of one-
half of 150 degrees = 0.98. Taking into account the changes
in phase, poles, frequency and chording, the new applied volt-
age per phase should be ~ X | X ~ x ^? = 305 volts.

o o 40 0.94

The explanation of this expression by terms is: The first
term, (880 -r- 3) comes from the change in phase from 2 to 3.
Since the original connection was in parallel and was for two-
phase, the voltage across one phase in series would be 2 X 220
= 440, and the voltage across both phases in series would be

2 X 440 = 880 volts. If the winding is divided into three
separate phases not interconnected, the applied voltage on
each phase would be (880 -f- 3). The next term, (4 -r- 6) is
due to the change in poles. A motor with six poles would run
slower on the same frequency than a motor with four poles and
would generate less counter emf. Consequently, the applied
voltage should be decreased in the same proportion. This
should not be confused with the fact that the frequency is
being changed in this case and the speed kept the same be-
cause a separate factor is introduced to take care of the fre-
quency. The pole change should be considered as an item
separate from the frequency change. The next term, (60 -4-
40) is due to the change in frequency and is the application
of the rule to change the applied voltage directly as the fre-
quency is changed. The last term, (0.98 -f- 0.94) is due to the
difference in chord factor. With a throw of 1 to 15, the coils
are more effective to generate counter emf. on the six-pole
than on the four-pole connection by the ratio of the chord
factors 0.98 to 0.94, hence the applied voltage should be raised
with the counter emf.

The value of 305 volts means that if the winding was divided
into three separate phases not interconnected in any way, the
voltage should be 305 volts across each phase. If the three
phases are connected in series star> as in Fig. 175, the applied
voltage should be 1.73 X 305 = 530 volts. Since this is only

RE-CONNECTING INDUCTION MOTORS 275

about 3.5 per cent, off from the 550 volts which is to be used,
the motor will operate satisfactorily. This calculation for
voltage so far neglects the difference in the so-called " distri-
bution factor" between three-phase and two-phase, but this is
immaterial. This factor acts the same way as the chord
factor, and is about 0.955 for any normal three-phase windings
and 0.905 for any normal two-phase winding, so that the ap-
plied voltage should really be 530 X I QQC) = 5(30 volts,

which is almost exactly what is required. This motor could
then have its phase coils re-arranged for six poles and be
connected series star and would be satisfactory for the new
conditions. The changes involved do not materially effect
the slip, so that no change is required in the rotor winding.
This example illustrates a rough calculation that can be made
to see what the possibilities of re-connection are.

Changes in Voltage Only with all Other Conditions Re-
maining the Same. This is the simplest change which can be
made in an induction-motor winding and in principle is the
same as that of a transformer coil in which the number of turns
of wire in series must be varied in exact proportion to the vol-
tage applied. Practically all commercial motors are arranged
so that they can be connected for two voltages, say 110 and
220, or 220 and 440. This is accomplished by putting the polar
groups in series, as in Figs. 175 and 182, for the higher voltage,
and in parallel, as in Figs. 176 and 183, for the lower voltage.

Assume a case in which the motor as it stands is connected
for 2200 volts and is connected in series star as in Fig. 175.
It is desired to re-connect it for 440 volts for the same horse-
power, phase, cycles and speed. Four hundred and forty
volts is 20 per cent, of 2200. Refer to the Table on page 266
and use the horizontal column marked 3-PAase, Series Star.
Since a re-connection is desired to give 20 per cent, of the
original voltage, read along the horizontal line until the figure
20 occurs This is found first under the vertical column
marked 3-Phase, ^-Parallels. This is obvious, of course,
because if the number of poles in the machine is divisible by
five, it could be re-connected in five parallels and operated on
2200 -T- 5 = 440 volts. But suppose the number of poles

276 ARMATURE WINDING AND MOTOR REPAIR

is not divisible by five. Look still further along the hori-
zontal line and the figure 19 appears under the vertical
column headed 3-Phase, ^-Parallel Delta. In other words, if
the number of poles in the machine is divisible by three it can
be put in three-parallel delta and operated on 2200 -r- (3 X
1.73) = 424 volts, which is near enough to 440 to give
perfectly satisfactory operation.

If the number of poles on the machine is not divisible by five
or by three, it is evident from the table that it is not possible
by any ordinary three-phase connection to approach closer than
550 volts with a four-parallel star connection, or 330 volts using
a four-parallel delta connection. The relation between 550 volts
and 330 volts as just given is not quite the theoretical 1.73
which would be expected, but this is due to the table being
made up to the nearest integral figure without using frac-
tions. The error in this instance is three per cent, which is
immaterial.

A point which is brought out by the example just cited is
that, so far as insulation is concerned, a motor may always be
re-connected for a lower voltage for instance, a 2200-volt
motor may be re-connected for 440 volts but, on the contrary,
a motor originally designed for 440 volts may not be run on 2200,
even if the re-connection is possible so far as number of turns is
concerned, because the insulation will not stand the dielectric
strain. It may be stated generally that practically all manu-
facturers use two classes of insulation up to 2500 volts, one
class good up to 550 volts and the second good from 600 volts to
2500. This should be carefully considered and a motor never
re-connected from the lower into the higher class, although
the change from the higher to the lower is permissible from
the insulation standpoint.

Change of Phase Only. The most common problem which
presents itself is the change from two- to three-phase, and
vice versa. Theoretically, for the same voltage there should
be about 25 per cent, more total turns in a two-phase winding
than in a three-phase winding. Then, if a three-phase motor
be re-connected for two-phase at the same voltage and with the
same coils, it will exhibit all the symptoms of a motor operating
at 25 per cent over voltage and usually would overheat to a

RE-CONNECTING INDUCTION MOTORS 277

dangerous degree after a short period of operation. Con-
versely, a two-phase motor re-connected and run on three-
phase at the same voltage with the same coils will show all the
signs of a motor operating at 20 per cent, under voltage. In
this case there are too many turns in the machine, One-fifth
of the total coils might be dead-ended to secure the proper volt-
age on the remaining 80 per cent. The dead coils should be
distributed as symmetrically as possible around the machine to
balance the voltage as nearly as possible on all phases. Parallels
in the winding should be avoided, as they give a chance for un-
balanced, circulating local currents, which may cause excessive
temperatures. Since the normal full-load current on a three-
phase motor at any given voltage is about 12.5 per cent, greater
than the two-phase full-load current at the same voltage, it
follows that the three-phase horsepower will have to be cut
down about 12.5 per cent, from the two-phase in order to keep
the current density in the winding as it was on two-phase.
Unless the current density is kept approximately the same
greater heating will result. Another makeshift, shown in Fig.
185, is the so-called "Scott or T connection for operating a two-
phase motor on three-phase. By this scheme 14 per cent, of
the coils in one phase of the two-phase machine are omitted as
symmetrically as possible around the machine. One end, BI
of this phase, is then connected to the middle of phase Ar~A 2 .
The resulting voltages between the points Ai-Az-B^ are prac-
tically 4n a balanced three-phase relation. This connection
would give fairly good results if the coils between AI and BI
were so situated on the machine that they would be acted
upon by the magnetic field in exactly the same manner as
the coils between BI and A 2 . Practically, as motors are
wound nowadays, this is rarely possible, and if the usual wind-
ing is connected in T there are practically always unbalanced
currents in the three phases. The current in the high phase
will be about 20 per cent, greater than the current in the low
phase. This results in a poorer performance in torque, power-
factor, efficiency and heating. The efficiency on the T con-
nection is 1.6 per cent, lower, the power-factor 5.2 per cent,
lower, the starting torque 38 per cent, lower, the maximum
torque 4 per cent, lower and the temperatures from 8 to

278 ARMATURE WINDING AND MOTOR REPAIR

13.5 higher than on the normal three-phase winding. This
shows that changing from two-phase to three-phase, and
vice versa, is at best very unsatisfactory. It is better to re-
wind with normal three-phase coils and avoid the troubles
which may follow.

One essential in any phase re-connection is to go over the
winding and re-arrange the " phase coils/ ' or coils having
heavier insulation, so that they will come properly at the ends
of the groups where the voltage is highest. This is illustrated
in Figs. 174 and 181.

One case of voltage and phase change which works out very
well is the change from three-phase 550 volts to two-phase
440 volts, or vice versa. This uses all the turns in the winding
for either connection, since the two-phase voltage should be
about 80 per cent, of the three-phase, and since the higher volt-
age on the three-phase cuts down the current, which would
otherwise be higher than the two-phase circuit. If the phase
coils are re-arranged there is practically no objection to such a
re-connection and the motor will give essentially the same per-
formance on either connection.

The table on page 266 shows the possibilities of interphase
connections, as well as the different voltage changes. For
example, in the case just cited, follow the horizontal line
marked 2-Phase Series to the first vertical column headed 3-
Phase Series. The figure is 125. This means that a motor
originally connected two-phase series, if re-connected three-
phase series, should be operated on 125 per cent, of the original
voltage. Or, if the two-phase voltage was 440 the three-phase
would be 1.25 X 440 = 550 volts. The convenience of the
table is demonstrated for phase changes, as well as voltage
changes, or for combinations of both.

Changes in Frequency. The occasion often arises for chang-
ing 25-cycle motors to 60-cycle and 60 to 25. There is also
some changing done from 60 cycles to 50 and 50 to 60. Occa-
sionally 40-cycle motors are changed to 60, but these changes
are infrequent.

In all cases of changed frequency the question that first
arises is: How is the resulting change in speed to be taken
care of? The synchronous speed of any motor (which is only

RE-CONNECTING INDUCTION MOTORS 279

a few per cent, higher than the full-load speed) is given by the

Alternations per Minute .

general expression - Number of Poles This would be

3000 , oe 7200 ,

XT r -vm f r 25 cycles, ^ u r> ^ f r 60 cycles,

Number of Poles ' Number of Poles

etc. If then the frequency is changed and the number of
poles left the same, the resulting rpm. will vary directly as
the frequency. This immediately brings up two questions:
First, is the mechanical design of the rotating part adequate
to allow such a change in speed? Second, can the speed of
the driven machine be adjusted to suit the new speed on the
motor?

Consider first the case where the frequency is changed and
the number of poles remain the same. The resulting change
in speed in this case is taken care of either by applying the
motor to a new load or by changing the pulleys on the old load
so as to keep the same rpm. on the driven machine. The
next thing that must be considered is the necessary change
in the voltage applied to correspond to the change in frequency,
or the other way about, if the new circuit at the new frequency
has the same voltage as was used with the original frequency,
how can the coils in the motor be re-connected so as to get the
proper voltage on each coil?

The easiest rule to remember is to vary the applied voltage
on the motor in exactly the same way as the frequency is varied.
If this be done the magnetic field in t*he iron will remain the
same and the current in the stator and rotor coils will remain
the same, if the motor is working against the same torque.
This is another way of saying that if the frequency and voltage
are varied together, the motor will develop the same torque at
all times and have flowing in it approximately the same cur-
rent. If the torque remains the same, the horsepower de-
veloped will vary directly as the applied frequency. For
example, a 60-cycle, 50-hp. motor operated on 25 cycles at
41.6 per cent, of its original voltage would develop the same
normal full-load torque, which would be 20.8 hp.

Changing from 25 to 60 cycles. A change from 25 cycles to
60 cycles, can often be made by impressing twice the voltage
on the coils on 60 cycles as on 25 cycles. A 220-volt, 25-cycle

280 ARMATURE WINDING AND MOTOR REPAIR

motor operated on 440 volts, 60 cycles, will have about double
the horsepower. Theoretically, this should be 60 -r- 25 = 2.4
times the voltage, instead of twice, and the resulting horse-
power would be 2.4 times. In this case suppose the motor
was connected in series star for 440 volts on 25 cycles and it is
desired to run it on 440 volts, 60 cycles. It should then be
connected in parallel star and run on 440 volts, which would
have the same effect as impressing 880 volts on the original
series connection. On 60 cycles the motor would then run
2.4 times as fast and develop about twice the horsepower.

Sixty-cycle Motors on 5Q-cycle Circuits. Sixty-cycle motors
are often run on 50 cycles without change. From the rule
above, that the voltage must vary with the frequency to keep
the same magnetic densities, it will be noted that the densi-
ties on 50 cycles at the same voltage will be six-fifths of the
60 cycles densities. The motor will then operate as if it had
120 per cent, of normal voltage impressed. This will result in
increased iron losses, which makes the motor hotter, and the
decreased speed on 50 cycles with the same number of poles
also makes the ventilation poorer, so that the output of the
motor in horsepower should be reduced to keep down the cop-
per losses.

Another point that should be watched in changing frequency
if the motor has a squirrel-cage rotor, is to make sure that the
rotor winding has enough resistane to give the proper starting
torque. As the frequency is raised the resistance of the short-
circuiting rings at the ends of the rotor winding should
be increased to keep the same relative value of starting torque
to full-load torque. As long as the motor starts its load satis-
factorily no change is necessary, but if trouble is experienced,
the short-circuiting rings may have to be changed for ones of
higher resistance. Conversely, when decreasing the frequency
the resistance can be reduced to advantage, thereby cutting
down the rotor copper loss and the heating.

Change in Frequency with Same Speed. In this case the
number of poles must be changed in the same ratio as the
frequency, or as nearly so as possible. For example, if a
motor has four poles and is operated on 25 cycles, it will
have a synchronous speed of 3000 -f- 4 = 750 rpm. If the

RE-CONNECTING INDUCTION MOTORS 281

motor is to have the same speed on 60 cycles, the nearest
possible pole number is 10 and the synchronous speed will be
7200 -r- 10 = 720 rpm. It is apparent that in very few cases
of this kind is it possible to re-connect the same winding.
The main reason for this is in the throw or pitch of the coil.
In the four-pole winding the individual coil spans approximate
one-fourth of the stator bore, and in the 10-pole winding
normal coils should span about one-tenth of the stator bore.
In the paragraph on "chorded windings" (page 269) it was
pointed out that the coil throw has an effect on the generated
counter emf proportional to the sine of one-half the electrical
angle spanned by the coil. This consideration makes hardly
possible such a condition as connecting a winding for 10 poles
when the individual coils have a four-pole throw. When
reducing the frequency the number of poles should become
smaller to keep the same speed. This introduces another
difficulty in the magnetic circuit. In re-connecting the wind-
ing the object is to keep the total magnetic flux in the machine
the same as it was originally. This keeps the magnetic
density in the teeth constant. This total magnetic flux is
divided up into as many equal parts or circuits as there are
poles The iron in the stator core between the bottoms of the
slots and the outside of the core has to carry the flux for each
magnetic circuit. Consequently, if there are 10 poles and
10 magnetic circuits, the core iron below the slots has to carry
at a given cross-section one-tenth of the total magnetic flux.
With the same total magnetic flux, if there arc only four poles
and four magnetic circuits, the same cross-section of core has to
carry one-fourth of the total magnetic flux, which it is probably
unable to do. This is the reason why the rotor diameter and
stator of a 25-cycle machine are smaller than those of a 60-
cycle machine of the same horsepower and speed, although the
outside diameter may be nearly the same. It is to get a
larger cross-section behind the slots for the passage of the
magnetic flux, since the total flux is divided into fewer parts,
owing to the smaller number of poles. From this it follows
that a machine may in general be re-wound or re-connected for
a larger number of poles, but that great caution is required in
re-connecting for a smaller number of poles.

282 ARMATURE WINDING AND MOTOR REPAIR

It is easier to re- wind or re-connect 25-cycle machines for 60
cycles than it is to re-connect 60-cycle machines for 25 cycles.
This follows logically from the physical fact that there is more
copper and more iron in 25-cycle machines for the same horse-
power, voltage and rpm. than in 60-cycle machines. It is
always easier to make changes where there is a larger supply
of material available. Another condition that is against chang-
ing the number of poles on a squirrel-cage motor is the cur-
rent in the short-circuiting rings of the rotor winding. These
rings are in nearly the same relation as regards current that
the primary core is as regards magnetic flux. That is, the
total secondary amperes, which remain nearly the same if
the re-connection is done properly, are divided into as many cir-
cuits as there are poles, and it follows at once that the smaller
the number of poles the larger must be the cross-section of the
short-circuiting rings, although the total secondary amperes
remain nearly the same. Altogether, the possibility of re-con-
necting for different numbers of poles when changing frequency
is usually a matter for the designing engineer to investigate.

Changes in the Number of Poles, all Other Conditions
Remaining the Same. The need for such changes comes
from the desire to speed up or slow down the driven machine
to meet new requirements. It might be broadly stated that
there are many cases where a change of two poles is permis-
sible, as for example, changing from four poles to six, or from
ten to eight and the like. The changes would consist in re-
arranging the phase coils to agree with the new grouping and
checking the chord factor, to note its effect on the voltage.
It is often possible to get a fair operating half speed by con-
necting for twice the number of poles. Practically all re-
connections involving pole changes give only a fair operating
performance. <^-

Testing a Re-connected Motor. After a motor has been re-
connected or after any change is made in the winding, it should
be started up slowly and the load gradually thrown on, ob-
serving carefully to see if there are any signs of distress, such as
sudden heating, noise or mechanical vibration. If the motor
seems to operate normally read the amperes in each phase
and the voltage across each phase to see that they are balanced

RE-CONNECTING INDUCTION MOTORS 283

and are reasonable in amount. The full-load current for three-
phase 550 volts is somewhere near one ampere per horse-
power for normal motors of moderate speeds between five and
200 horsepower. At other voltages this will be inversely as
the voltage, that is at 440 volts, three-phase, about 1.25 am-
peres per horsepower. On two-phase the current per phase
is about 87 per cent, of the corresponding three-phase value.
If the readings as above look reasonable a thermometer should
be placed on the stator iron and another on the stator coils and
read at 15-minute intervals for an hour, and at half -hour in-
tervals thereafter, until the temperature is constant. The
speed should be checked at intervals. If the rpm. shows a
tendency to decrease rapidly or fall below 90 per cent, of syn-
chronous speed, it may be suspected that the rotor has too
much resistance and is getting hot. By making all these
checks, reasonable assurance may be had that the reconnec-
tion is satisfactory.

Effects of High and Low Voltage on Motor Operation.
All changes in alternating-current motors whether of phase,
voltage, poles or frequency, may be considered as voltage
changes and reduced to such terms. In making such calcula-
tions and comparing the results, it is advisable not to apply a
voltage that differs from the rated voltage by more than plus
or minus 10 per cent. The general effect of high and low
voltage may be expressed briefly as follows:

Effect of High Voltage:
a. Increases magnetic density.
6. Increases magnetizing current.

c. Decreases "leakage current" (leakage reactive component).

d. Increases starting torque and maximum torque.

e. Decreases slip or change in speed from no load to full load.
/. Decreases secondary copper loss.

g. Increases iron loss.

h. Usually decreases power-factor.

i. May increase or decrease efficiency and heating, depending upon
the proportions of primary copper loss and iron loss in the normal
machine and also the degree of saturation in the iron.

Effect of Low Voltage:

a. Decreases magnetic density.

b. Decreases magnetizing current.

284 ARMATURE WINDING AND MOTOR REPAIR

c. Increases leakage current.

d. Decreases starting and maximum torque.

e. Increases slip.

/. Increases secondary copper loss.

g. Decreases iron loss.

h. Usually increases power-factor.

t. May increase or decrease efficiency and heating, depending upon
the proportions of primary copper loss and iron loss in the normal
machine and also the degree of saturation in the iron.

Operating Standard Alternating-current Motors on Differ-
ent Voltages and Frequencies. When a motor is to be oper-
ated on a different voltage or frequency than the motor was
designed for, there should be a corresponding change made in
circuit to which the motor will be connected. That is, if there
is to be an increase in frequency or voltage, an equal decrease
in voltage or frequency of the circuit will bring about normal
results. Present-day designs of motors can be used in most
cases without excessive heating when the variation either up
or down of frequency or voltage is not more than 10 per cent.

For an induction motor the following conditions result with
a change of frequency or voltage:

1. Pull-out torque and starting torque vary as the square
of the voltage and inversely as the square of the frequency.

2. The copper loss in the primary varies as the square of
the current. The current varies inversely as the voltage, but
is not affected by a change in frequency, except to the slight
extent produced by changes in magnetizing current. The
secondary copper loss and slip tend to vary inversely as
the square of the voltage, but this tendency is modified by
the changes in primary IR drop and magnetic leakage. The
secondary copper loss and slip remain constant with change
in frequency.

3. The iron loss is composed of hysteresis and eddy current
losses. The hysteresis loss varies as the 1.6 power of the flux;
the eddy current loss varies with the square of the flux; and
the flux varies directly as the voltage and inversely with the
frequency. The magnetizing current varies directly with
the flux except for modifications produced by saturation of the
magnetic circuit.

4. The power-factor is usually decreased by an increase in

RE-CONNECTING INDUCTION MOTORS

285

voltage or a decrease in frequency and vice versa, but the total
change is small.

5. The efficiency is not materially altered by a change in
either frequency or voltage.

The accompanying curve, Fig. 186 shows the operating
voltage on which a standard motor can be used when con-

450
400

ocn

300
250
200
150
100
50

-X

5 10 15 20 25 30 35 40 45 50 55 60 65 70
Frequency Cycles per Sec.

FIG. 186. Standard motor frequency and voltage curve.

This curve indicates the voltage which, if employed in connection with the respective
corresponding frequencies, will result in the operation of apparatus at approximately
uniform core densities. By adherence to the relations betweeen frequency and voltage
indicated, the range of application of standard apparatus can be broadened and the
required number of different designs minimized. Allowance is made for the use of
somewhat reduced densities at the higher frequencies, as indicated by the drooping
ciiaracter of the curve (R. E. Hellmund, Electric Journal, September, 1910, page 691).

nected to a circuit of a different frequency from that for which
it was designed. The accompanying table also shows the
effect on speed and horse-power rating when a motor is used
on a different voltage and frequency.

The voltage and frequency of a motor should never be varied
in opposite directions at the same time. In general any change
from normal frequency should be accompanied by a change in
voltage proportional to the square root of the frequency. In
the case of 400-volt, 60-cycle motor operated on a 66% cycle
circuit, the voltage should be 422. That is V(66% -5- 60)
X 400 = 422. In case of decreased frequency, the motor
should be operated on less than normal voltage on account of
the increased current and temperature.

286 ARMATURE WINDING AND MOTOR REPAIR

EFFECT ON SPEED AND HORSEPOWER WHEN MOTORS ARE OPERATED
ON DIFFERENT VOLTAGES AND FREQUENCIES

Rating of motor

Voltage and fre-
quency of circuit

Speed

Hp. rating

220 volts, 25 cycles
440 volts, 25 cycles
(a) 440 volts, 60 cycles
440 volts, 60 cycles
(6) 220 volts, 60 cycles

250 volts, 33 cycles
500 volts, 33 cycles
220 volts, 25 cycles
220 volts, 33 cycles
220 volts, 50 cycles

Increased 33 to 25
Increased 33 to 25
Reduced 60 to 25
Reduced 60 to 33
Reduced 60 to 50

Increased 33 to 25
Increased 33 to 25
Reduced 60 to '25
Reduced 60 to 33
Reduced 60 to 50

(a) Where good power-factors are essential it may be advisable to use 550-volt, 60-
cycle motor or a 220-volt, 25-cycle circuit and reduce the rating according to the heating
between 35 to 45 per cent, of the rating at 60 cycles.

(6) Standard 60-cycle motors of liberal rating can be used on a 50-cycle circuit with
the same rating. Best results are secured however, when 220-, 440-, and 550-volt
motors are operated on 200-, 400- and 500-volt circuits at 50 cycles.

Factors which Limit a Change in Number of Poles of an
Induction Motor. The principal factors limiting a change in
the number of poles of a squirrel-cage induction motor are
given as follows by a writer in the General Electric Review:

(a) The number of turns in series per phase. These must remain the
same since the applied voltage is to be unchanged.

(6) The insulation between the conductors of different phases. Of
this there must be sufficient to not reduce the factor of safety against
breakdowns after the regrouping of the conductors has been carried out.

Because the designs of induction motors vary widely with
different manufacturers and also in the product of each maker
(for the purpose of supplying motors for various types of
service), it will be impossible to make other than very general
statements regarding the expected change in characteristics
of the motor when running at the higher speed. Furthermore,
the following statements must not be expected to hold true
when the number of poles has been decreased sufficiently to
raise the normal speed more than say 25 per cent.

After the reconnection,

(a) The normal speed will be equal to approximately the original
normal speed times the original number of poles divided by the new
number of poles.

(6) There will be a somewhat higher torque per pole exerted, due to
the slightly increased flux per pole that arises from the shortened pole
pitch, so that the total motor torque might be expected to be decreased
but little by the change.

RE-CONNECTING INDUCTION MOTORS 287

(d) The starting torque will probably be slightly decreased.

(e) The power-factor might be expected to be somewhat higher.

(/) When the power-factor is higher the rating of the motor can be
increased about in proportion to the square root of the increase in speed
with the same heating.

(flf) The efficiency will be practically the same as before the change.

Single -circuit Delta and Double-circuit Star Connections.

The use of a one-circuit delta and a double or two-circuit
star winding for a three-phase motor, as pointed out by Henry
Scheril (Electrical Record, March, 1919) depends upon operat-
ing conditions or design to meet certain requirements. One
of the best illustrations is offered in the winding of a phase-
wound rotor of the induction motor. Suppose that the rotor
has been wound with a two-circuit star winding and the voltage
between rings on open circuit is 220 volts. Let us assume that
the full-load current in the rotor is 200 amperes. Since the
winding has a double circuit, each circuit will take care of
100 amperes. The voltage per phase will be 127 volts.

Manufacturers, in general, standardize the brush rigging
used in connection with machines, using a certain number of
brushes per ring. If it is found, for instance, that 200 amperes,
as in this example, brings the current density in the brush to
too high a value, it would mean that the number of brushes per
ring must be increased. Since this is not practical nor eco-
nomical, a change in the connections of the winding may bring
about the desired results, that is, reduce the current density
in the brush to within the allowable value.

If the winding were then reconnected from two-circuit star
to single-circuit delta, then the voltage between terminals
will be increased from 220 volts to 254 volts and the current
per ring would be reduced from 200 amperes to 173 amperes,
or a reduction of 13.5 per cent. This reduction may just be
sufficient to decrease the brush density to the desired value.
Since controllers used in the rotor circuit are made to take care
of voltages of reasonable variations, an increase in voltage
from 220 volts to 254 volts will not require a change in the
controller and therefore the 220 volts controller can be used
on the 254-volt circuit.

288 ARMATURE WINDING AND MOTOR REPAIR

Cutting out Coils of an Induction Motor. In case coils
burn out in an induction motor, they can be cut out and the
motor operated for a time or until it can be repaired. It is
advisable to cut the entire coil and tape up the two ends so
that they can not come together. There is a limit to the
number of coils that can be cut out as this is equivalent to
raising the voltage on the motor which will cause heating.
Where more than two coils must be cut out the motor should
be repaired at once or the same number of coils in each phase
cut out, evenly distributed around the stator.

PROCEDURE WHEN CONNECTING THE COILS OF AN
INDUCTION-MOTOR WINDING

At the end of this chapter typical diagrams for connecting
the windings of polyphase alternating-current motors are
shown as prepared by A. M. Dudley. 1 When an armature
winder or repairman knows the number of poles and phases
for the winding, and has been supplied with the necessary
coils of the proper throw, the problem of inserting the coils
and then connecting them up can be easily understood for
each of the diagrams referred to if the fundamental procedure
for any one is understood. This procedure is explained by
the author of the diagrams as follows:

The diagrams are not dependent on the total number of
slots in the machine nor upon the number of coils per group,
nor upon the throw or pitch of the coils, but are general for
all machines of the same number of phases and poles. Each
one of the small arcs in each diagram represents the ends of
the coils in a single pole-phase-group in the winding. This is
illustrated in Figs. 187 and 188 showing a stator in three stages
of being connected. In Fig. 187 (A) a machine is shown in
which the coils have simply been placed in the slots by the

1 The diagrams shown have been selected from a series of eighty-one
devised by A. M. Dudley, Engineer industrial division of the Westing-
house Electric & Mfg. Co., to be incorporated in an excellent book by
him on "Connecting Induction Motors." (McGraw-Hill Book Co.,
New York.) Mr. Dudley's diagrams begin with two-pole windings and
give practically all possible combinations for two- and three-phase, star
and delta and series-parallel connections.

RE-CONNECTING INDUCTION MOTORS

289

winder and no connections have been made. The wires
which are the beginnings and endings of the coils are sticking

,0 O

a g^

^ rfl CD

a ra 05

M 'grt

c3 8

OO O O

"-3
s +? *

put at random. In Fig. 187 (5) the coils have been connected
into several distinct groups and the remaining wires, which

290

ARMATURE WINDING AND MOTOR REPAIR

protrude radially toward and away from the center of the
machine, form the beginning and the end of each pole-phase-
group.

Connecting Pole-phase-groups of a Winding. The opera-
tion which has been performed between Fig. 187 (A) and Fig.
187 (B) can be described in this way : Suppose, for example, that
there are 96 total coils in the winding and that it is to be con-
nected for three phases and four poles. There will then be 3 X

FIG. 188. Stator of a 100-hp., 3-phase, 8-pole, 220-volt, 60-cycle, 120-
coil, 120-slot, induction motor completely wound with pole-phase-groups
properly cross connected.

4 = 12 pole-phase-groups, and 96 -f- 12 = 8 coils in each group.
Starting at any arbitrary point, the winder connects the first
eight coils in series by connecting the end of coil No. 1 to the
beginning of coil No. 2, and the end of coil No. 2 to the begin-
ning of coil No. 3, etc., until eight coils are in series. The
beginning of coil No. 1 is then bent outward and left long and
the end of coil No. 8 is bent inward and left long. Between

RE-CONNECTING INDUCTION MOTORS

291

these two are seven short " stubs" or coil-to-coil connections,
which are shown taped up in Fig. 187 (B) . The winder then
proceeds to connect coils No. 9 to No. 16 in series in the same
manner to form pole-phase-group No. 2, and so on around the
machine until he has completed 12 pole-phase-groups and
used all the coils. The winding then looks as shown in Fig.
187 (B).

In case the winding has certain coils provided with heavier
insulation on the end turns to take the strain of the full voltage

FIG. 189.

FIG. 190.

FIG. 191.

FIG. 192.

FIG. 193.

FIG. 194.

FIG. 189. Thirty-two coils connected into 8 pole-phase-groups for a
2-phase winding.

FIG. 190. Same as Fig. 189 with pole-phase-groups connected according
to direction of arrows.

FIG. 191. Same as Fig. 190 with B-phase reversed.

FIG. 192. Forty-eight coils connected into 12 pole-phase-groups for a
3-phase winding.

FIG. 193. Same as Fig. 192 with pole-phase-groups connected according
to direction of arrows.

FIG. 194. Same as Fig. 193 except leads are brought out from different
groups.

of the machine where different phases are adjacent, the opera-
tion is slightly different. In this case the number of coils per
pole-phase-group must be determined before the coils are
inserted in the slots, and the specially insulated phase coils
placed on both ends of each group. In this case the location

292 ARMATURE WINDING AND MOTOR REPAIR

of the pole-phase-groups is definitely determined by the winder
before he starts connecting the coils together.

The next step is to mark the pole-phase-groups A-B-C-A-
B-C, etc., around the machine and then to connect all the
groups together in the proper manner to form a three-phase
winding as indicated by the diagram of connections. The
completed winding will then appear as shown in Fig. 188.

General Theory on which Connection Diagrams are Con-
structed. Simple methods by which any winding may be
checked for phase polarity are shown in Figs. 189 to 194, in-
clusive. In Fig. 189 a winding chosen at random is shown
"stubbed" into pole-phase-groups for a two-phase connection,
and in Fig. 192 stubbed for a three-phase connection. To de-
termine the proper connections for the pole-phase-groups in a
two-phase winding, the rule is to mark on the groups arrows
alternating in direction in pairs. That is, on two successive
groups the arrows are clockwise and on the two immediately
adjacent, the arrows are counter-clockwise. Such arrows,
for example, are shown in Fig. 189 just above the windings.
If now one end of any group in a phase is chosen as a "lead"
and all the groups are followed through and connected as indi-
cated by the arrows, the connection will be correct. Such a
connection is shown in Fig. 190. However, suppose the arrows
had alternated in pairs, but started with a different group, as
shown just below the windings in Fig. 190. The result is shown
in Fig. 191, which is just as correct as Fig. 190, except that the
motor would run with the opposite direction of rotation.
Since the rotation can be changed by reversing the two leads
of either phase outside of the motor, it is evident that the rule
using the arrows alternating in pairs is correct in all cases.
It should also be noted that it makes no difference from what
group the lead is taken, provided all the groups are followed
through with the arrows.

In the three-phase machine the method is even more simple.
The rule in that case is to put arrows on the groups alternating
in direction from group to group, as shown in Fig. 192. Any
group may then be chosen as a "lead" group or a "star"
group so long as the arrows are followed in passing from the
lead to the star in each phase. Figure 193 shows one arrange-

RE-CONNECTING INDUCTION MOTORS

293

ment and Fig. 194 another equally correct, and there might be
an indefinite number more, simply by choosing the lead from
another group and following the arrows through to the star
in each phase. Although shown for a developed four-pole
winding only, these diagrams may be considered as strictly
general, as additional groups may be added to make six, eight,
or any other number of poles, and the current passed through
them in any order, so long as the phases are kept in the correct
rotation and the current in the right direction as indicated
by the arrows.

In case of a delta connection instead of a star, check the
connections through as for a star and then connect the A star

FIG. 195. Method of checking
a delta connection from a star
connection.

FIG. 196. Winding diagram for
2-pole, 2-phase motor with series
connections of coils.

to the B lead, the B star to the C lead, and the C star to the A
lead, as shown in Fig. 195; or connect the A lead to the B
neutral, the B lead to the C neutral, and the C lead to the A
neutral. The three motor leads will be taken from the corners
of the delta so formed.

Determining Number of Poles from Slot Throw of Coils.
As an example assume a 96-slot stator whose coils span 12
slots. Then number of slots -t- span of coil = number of
poles, or 96 -f- 12 = 8 poles. Suppose however, the coils
span 10 slots. The quotient is then 9.6 which is an impossible
number of poles. This indicates a chorded winding and the

294 ARMATURE WINDING AND MOTOR REPAIR

correct number of poles is probably the next lower even
number which will again be 8. This is not an invariably
correct rule. A further check is as follows: Divide the
number of slots by the number of phases. If this number
is divisible by the number of poles obtained as above, it may
be safely assumed that the correct number of poles has been
determined.

Typical Circle Diagrams for Connecting Induction Motors.
On the following pages, 295 to 300 typical winding diagrams
are shown by which induction motors may be connected for
a variety of operating conditions.

RE-CONNECTING INDUCTION MOTORS

295

FIG. 197. Winding diagram for
2-pole, 3-phase motor with series-star
connection of coils.

FIG. 198. Winding diagram for
4-pole, 2-phase motor with series con-
nections of coils.

FIG. 199. Winding diagram for
4-pole, 3-phase motor with series-
star connections of coils.

FIG. 200. Winding diagram for
4-pole, 3-phase motor with series-
delta connections of coils.

296

ARMATURE WINDING AND MOTOR REPAIR

CAB

A C

FIG. 201. Winding diagram for FIG. 202. Winding diagram for

4-pole, 3-phase motor with 2-parallel 4-pole, 3-phase motor with 2-parallel
star connections of coils. delta connections of coils.

FIG. 203. Winding diagram for FIG. 204. Winding diagram for

4-pole, 2-phase motor with 4-parallel 6-pole, 2-phase motor with series con-
connections of coils. nections of coils.

RE-CONNECTING INDUCTION MOTORS

297

FIG. 205. Winding diagram for
6-pole, 3-phase motor with series-star
connections of coils.

B C

FIG. 206. Winding diagram for
6-pole, 3-phase motor with series-
delta connections of coils.

B C

FIG. 207. Winding diagram for
6-pole, 3-phase motor with 2-parallel
star connections of coils.

ABC

FIG. 208. Winding diagram for
6-pole, 3-phase motor with 2-parallel
delta connections of coils.

298 ARMATURE WINDING AND MOTOR REPAIR

ABC

FIG. 209. Winding diagram for FIG. 210. Winding diagram for

6-pole, 3-phase motor with 3-parallel 6-pole, 3-phase motor with 6-parallel

delta connections of coils. delta connections of coils.

FIG. 211. Winding diagram for FIG. 212. Winding diagram for

8-pole, 2-phase motor with series con- 8-pole, 2-phase motor with 8-parallel
nections of coils. connections of coils.

RE-CONNECTING INDUCTION MOTORS

299

ABC ABC

FIG. 213. Winding diagram for FIG. 214. Winding diagram for
8-pole, 3-phase motor with series-star 8-pole, 3-phase motor with series-
connections of coils. delta connections of coils.

i ...

13 X v
.9
17 11
15 j

-1-,-^

1 l H - t - f -^~ l

4 '

H? .

i
i

: ; . ^

FIG. 215. Winding diagram for FIG. 216. Winding diagram for

10-pole, 2-phase motor with series 10-pole, 3-phase motor with series-
connections of coils. star connections of coils.

300

ARMATURE WINDING AND MOTOR REPAIR

ABC

FIG. 217. Winding diagram for FIG. 218. Winding diagram for

10-pole, 3-phase motor with series- 10-pole, 3-phase motor with 5-parallel
delta connections of coils. star connections of coils.

B C

FIG. 219. Winding diagram for
10-pole, 3-phase motor with 5-parallel
delta connections of coils.

ABC

FIG. 220. Winding diagram for
10-pole, 3-phase motor with 10-par-
allel star connections of coils.

CHAPTER XII
COMMUTATOR REPAIRS

Commutator troubles are more easily located than faults in
an armature, but a repair job on an old style motor, many of
which are still in use, is sometimes a trying and tedious
operation.

Causes of Commutator Troubles. Sparking at the brushes
is generally the first symptom of commutator trouble. One
of the most frequent causes of sparking is a rough or pitted
commutator. This may be due to many irregularities, such
as overload; brushes out of line; not set at neutral points in
regard to load; poor contacts; current density per square inch
of brush contact too great; open circuit; weak magnetic
fields; commutator out of round; high or low bars or high mica.
The most common cause of sparking is high mica which causes
the brushes to chatter and to make poor
contact. This condition results in a
rapid blackening and burning of the
bars, sometimes to the extent that the
copper is eaten away leaving the mica
segments standing out above the surface
of the commutator. Some motors seem
to be particularly subject to this
trouble, due to the fact that the mica
is too hard a grade, or the copper too
soft. When there is no time to turn
down the commutator the high mica and how they are held by
can be removed by grinding down the clam P in s rin g s -
with a piece of sandstone, and using fine sandpaper for
smoothing.

Troubles Resulting from High Mica. High mica, while
it may seem a small matter at first, is often the cause of more
serious complications. The commutator may become so

301

SEGMENT

FIG. 221. Section of a

302 ARMATURE WINDING AND MOTOR REPAIR

hot from the poor brush contact afforded, that the solder
will be melted and thrown out, resulting in short-circuits
between bars and open-circuits due to the armature leads
becoming disconnected. About the only permanent relief for
sparking at brushes due to high mica is undercutting the mica.
This remedy is recommended when it is reasonably certain that
the high mica is caused by the natural condition of the copper
or mica. If it is not, then the real cause must be found,
otherwise, undercutting the mica would probably improve
the running condition somewhat, but would fail to remove the
cause.

The mica should not be cut too deeply, a depth of ^ 2 to KG
inch below the surface of the copper being sufficient. Care
must be exercised to remove the mica the full width of the
segment, for any thin slivers left flush with the surface will
often defeat the purpose of the undercutting. (For details
for undercutting mica, see page 320.) This method has
corrected some stubborn cases of sparking and if the job is
properly done, all that will be necessary to preserve sparkless
commutation is to keep the slots clean and well below the
surface of the copper.

Remedy for High or Low Bars. A new or repaired motor
may have a commutator that is not " settled." That is, the
clamping ring has not been drawn up as tightly as it should be.
When the mica end rings are "cut they are only slightly flexible
due to the shellac in them and cannot be made to fit perfectly
when cold. When the commutator is hot, the shellac in the
mica will soften and allow it to move under the strain of the
centrifugal force of the bars when the machine is running.
This movement of the mica allows the bars to move and is
frequently the cause of high or low bars. To remedy this
trouble, the machine should be run until its normal operating
temperature is attained and then shut down. The clamp-
ing ring can then be tightened. This process may be neces-
sary several times, or until the commutator is perfectly solid.
Care must be taken not to tighten the bolts too much while
the commutator is warm.

In the case of a high bar, it should be tapped down until it
rests firmly against the mica end rings. It can then be filed

COMMUTATOR REPAIRS

303

even with the rest of the bars. A low bar can be raised by
prying up, and inserting a narrow strip of mica beneath it,
but in the majority of cases this makes a poor job. Usually
the only alternative is to turn the commutator down to the
level of the low bar.

Burn-out Between Bars. Probably the most frequent com-
mutator trouble is a burn-out between bars. It occurs often
on the corner of the bars, and is not infrequently caused by

FIG. 222. Mica segments taken from damaged motors showing the effects
of short circuits in the commutator. Two tools are also shown made from
hack saw blades for use in plugging a commutator.

oil working along the shaft from the bearing and up onto the
commutator. This oil collects dust and dirt and finally
causes current to leak from one bar to the other. The mica
then becomes carbonized, and a short-circuit results. This
is one of the causes of burned-out armature coils. Sometimes
the short-circuit will burn itself clear, and no harm will be
caused except to burn a hole in the mica. However, it may

304 ARMATURE WINDING AND MOTOR REPAIR

continue to arc across and burn a good sized hole in the bars
also.

Plugging a Commutator. When mica segments are burned
but not too deep, the holes can be cleaned with a thin knife
blade and plugged with some kind of filling. If a good filling
compound is used, and the commutator kept free from oil, it
will hold for a year or possibly longer. It is always advisable
to save the wearing surface in this manner whenever it can be
done, for every time a commutator is turned down in a lathe
on an average of three years of its useful life is lost.

A good filling compound for commutators can be made as
follows: Two parts plaster-of-paris; one part powdered mica;
and enough glue to make a thick paste. This, when applied,
will dry quickly, and assume about the same degree of hard-
ness as the mica segments.

When a segment becomes burned deep down into the com-
mutator, a new one must be inserted. Before attempting to
do this, the armature should be thoroughly blown out with com-
pressed air in order to remove all dust that may have accu-
mulated. This is essential, for it is an easy matter for small
particles of foreign matter to work in under the back end be-
tween the bars and the sleeve when the commutator is loose.
Determine just which mica segments must be taken out, and
number the bars at each burn-out, since the bars may have to
be removed also in order to get the mica segments out. If
the segments were not shellaced when the commutator was
built, the chances are they can be lifted out without disturb-
ing the bars. Otherwise the bars will have to be taken out
with the segments.

Removing Bars and Mica Segments for Repairs. Remove
the bolts that hold the clamping ring in place. Mark the
ring so that it may be put back just as it was taken off. Tap
the end of the ring lightly with a hammer. If the mica ring
does not loosen from the commutator, it will have to be heated,
as the ring is probably stuck fast with shellac. Heat the
commutator with a torch to an even temperature all around.
This will expand the copper and cause it to bulge out from the
end ring. Tap the ring again lightly, and it will be found to
work loose. Then it can be pulled out.

COMMUTATOR REPAIRS

305

Pry the bars apart slightly at one of the burned places to
see if the mica segments are stuck to the bars. If they are
not, it is a simple matter to remove them from the commu-
tator. If they are held fast, the leads from one bar adjoining

Keep all Dirt and
Moisture out

Leave Mica extending
One-sixteenth Inch

Seal thoroughly against
Dirt and Moisture

FIG. 223. Section of an assembled commutator of a railway motor.

the burned segment should be unsoldered and the bar lifted
out. Proceed in the same way with the remaining bad places.
New segments can be marked off by using one of the old bars
as a guide. The bars should be scraped and filed clean^
and all rough corners rounded off.

Repairing a Burned Commutator Bar. Frequently it
happens that there is a good-sized hole burned in the commu-

Fio. 224. Repair of a burned place on a commutator bar.

tator bar. This should be repaired before it is used again.
As there is usually no stock of exactly the proper size on hand
in a small repair shop out of which to make a new bar, a good
repair is the next best thing. In Fig. 224 (A) shows the end of
20

306 ARMATURE WINDING AND MOTOR REPAIR

a bar with a burned place and (B) the method of repair. In
this case the bar was cut down enough to remove the burn,
and a piece of copper strip carefully squared and soldered in
the cut. It was then riveted with a small copper rivet, the
location of which is shown at (X). The rivet was used to
prevent the patch from flying out should the commutator for
any reason become hot enough to melt the solder. Such a
patch should be filed down to the dimensions of the bar. The
cut for a patch of this kind can be quickly made if a milling
machine or shaper is handy, otherwise a good sharp fils will serve
the purpose.

Replacing a Repaired Commutator Bar. Before replacing
a commutator bar that has been repaired, an inspection should
be made of the back mica ring, to be sure that no dust or solder
has lodged there. The mica segments should be replaced
first, and then the bar pushed in. Shape the commutator
as nearly as possible into a circular form and replace the end
ring. Tighten the clamping nuts as much as possible while
the commutator is cold. It is a good plan to paint the end
of the commutator with shellac, in order to fill up any cracks
that may exist between the bars and the ring.

Tightening up a Repaired Commutator. When a commu-
tator has been taken down in a repair shop and assembled
again, all lock nuts and screws should be first set up hard
and the commutator baked in an oven. Then the lock nuts
and screws can be tightened up again since the heating causes
the copper to expand and put pressure on the mica which,
combined with the heat, drives out all traces of shellac in the
mica. Then as the commutator cools and the copper becomes
normal the bolts can be taken up. This proces's should be
repeated and finally the commutator cooled quickly by a fan,
and the nuts tried again. If they seem tight the commutator
is ready for assembly on the armature shaft. It requires some
experience to determine just how tight the bolts can be drawn
on a commutator without injuring the mica end rings. The
one precaution is not to draw up the bolts finally until the
commutator has cooled. The commutator surface can now
be turned down in a lathe. After doing this a short-circuit
test should be made with a test lamp from bar to bar. Failure.

COMMUTATOR REPAIRS

307

308 ARMATURE WINDING AND MOTOR REPAIR

of the lamp to light indicates that the commutator is free from
short-circuits. See also Chapter V, pages 123, 126 and 135.
A test for grounds should also be made at this time (see pages
131 and 175).

Baking Commutator with Electric Heat. In those cases
where an oven is not available for baking a rebuilt commutator
and it is too large to heat with a torch, the electrical method
shown in Fig. 226 has been successfully used (Albert Krause,
ElectricalWorld, July26, 1919, page 190) A satisfactory heat-
ing element can be made up by using a layer of ^{Q-UI. asbestos
paper around the commutator and winding over this about

FIG. 226. Uniform temperature produced by heating element wound around

commutator.

90 ft. of No. 22 Nichrome resistance wire. By applying
250 volts to the terminals of this resistance wire a uniform
heat in the commutator can be produced to permit evening up
the insulating segments and clamping the end rings. Ordi-
narily it would be advisable to have a variable resistance in
series with the heating element so its temperature can be kept
at a desirable value. By covering the armature after the
commutator is heated in this manner varnish may be applied
thereto and baked in by the heat conducted to the armature
through the leads from the commutator.
Removing and Repairing Grounds in a Commutator. The

COMMUTATOR REPAIRS 309

ground frequently occurs between the sleeve ring and the end
of the bars. A small hole is generally burned through the
mica ring or taper cone. Some times the mica ring on the rear
end is punctured. In that event a number of bars in the
neighborhood of the ground will have to be taken out. The
burned mica should be cut out
and a patch put on. When the
trouble occurs on the front end
of the commutator, remove the
ring and cut out the bad mica.
The patch can be made as shown

TV rkr*>-r mi . FIG. 227. Patch on a mica end

in Fig. 227. This new mica must ring of a commutator .
be a trifle thicker than the original

mica removed, for it will squeeze together somewhat when
the ring is drawn up tight and the commutator heated.

After the repair, test the commutator for grounds with a
proper voltage (see page 175) . Place one terminal on the shaft,
and move the other completely around the surface of the com-
mutator. Freedom from grounds will ,be indicated by no
sparking at the terminal moved over the surface. A lower
test voltage should be used on low- voltage commutators and
a higher voltage for those of high voltage machines.

On some of the old-style motors, the sleeve nut is on the
back of the commutator. When this happens, the armature
leads will have to be disconnected and bent back out of the
way in order to work on the commutator. The best proce-
dure in such a case, if there is time, is to remove the com-
mutator and reverse the sleeve, as both ends are usually bored
to the same diameter. By reversing the sleeve in this manner
in order to get the nut where it can be easily reached, much
time and work will be saved when future repairs must be made

Turning Down a Commutator without Removing Armature
from Machine. The method sometimes used in turning down
a commutator on a repair job where the armature is too large
to remove, is to leave one or two pairs of brush arms on and
run the machine from these at as low a speed as the field regu-
lation will permit or possibly with a water rheostat in the arma-
ture circuit. This method will do where no other means of
turning the commutator is available. There is always bad

310 ARMATURE WINDING AND MOTOR REPAIR

sparking and burning at the point of the cutting tool due
to its short-circuiting the bars when it crosses the mica. The
tool has to be sharpened frequently and a job is seldom good
even where the greatest care is exercised.

For these reasons it is always preferable to belt the machine
to a separate motor and turn down the commutator with
the fields unexcited. In connecting up a motor for driving
the armature the speed should be made as low as possible,
preferably not over 75 revolutions per minute.

Temporary Cover for Use When Turning Down a Commu-
tator. In cases where it is necessary to turn down the commu-
tator of a direct- current machine without removing it from

ir First Cord on Cover

2r Cover Stretched up Over Leads.

a- Lost Tie on Cover 4i- Last End Pulled under

Cord Winding.

FIG. 228. Steps in applying a cotton cover over the end of an armature
before turning down the commutator to prevent copper chips falling behind
the bars to cause short circuits.

the frame, H. S. Rich has made use of a muslin cover (Elec-
trical Record, December, 1918) tied over the open leads so that
copper chips cannot find a way down behind the bars to cause
a short-circuit. This cover is made and applied as follows:
Cut a strip of muslin wide enough to reach across the commu-
tator and well past the leads, and long enough to go around
one and one-half times. Tie this very securely with fine
strong cord, with the muslin laid over the bars as shown in
Fig. 228.

Draw the covering and stretch a little at a time all around
and up over the open leads far enough back to allow of at least
two separate cords to be tightly wound around and very se-

COMMUTATOR REPAIRS 311

jfe
curely tied on the core body. Turn the armature slowly to

see that the covering does not interfere with any brush holder.
If so, the holder should be shifted. The surplus edging, all
loose strings and threads should now be trimmed off all around.

By turning back the muslin over the leads the first cord
tied around the bars is covered neatly. The cords wound
around the core body can be secured without knots by wind-
ing over the first end a few turns, and then by winding over
a short extra loop, the last end can be jerked under all the
turns and cut short so that a knot is not needed. For a
permanent armature covering, shellac should be applied all
over it which seals the cords and stiffens the muslin.

Refilling a Commutator. When a commutator is to be
refilled, disconnect the armature leads and remove the com-
mutator. A simple device for accomplishing this is shown
in Fig. 229. Two long bolt rods are screwed into holes tapped
into the sleeve ring, and a bar of heavy iron placed across the
end of the shaft, with bolt rods coming through, as shown.
By tightening the nuts evenly, the commutator can be pulled
off. Next count the number of bars carefully, and enter
this in a note book for future reference. Remove the sleeve
and if possible save the mica rings. If these are in good
condition they can be used again. Carefully caliper the
diameter of these rings and enter this in the note book also,
as the new commutator will have to be bored to fit these rings.

With a micrometer caliper,
measure the thick and thin
edges of one of the bars in
thousands of an inch; also the

thickness Of the mica Segment. Fl " 2 29.-Device for removing
It is advisable to Order the a commutator from an armature

bars and mica segments from

the motor manufacturer sawed to the proper size, as in all
probability this can be done cheaper than in an ordinary repair
shop not equipped for this work. When ordering new bars
a detailed drawing should be sent, giving all necessary
dimensions of the old bar for boring and turning purposes.
Hard drawn copper is usually used as it wears at about the
same rate as the mica segments (see page 319).

312 ARMATURE WINDING AND MOTOR REPAIR

Use of a Commutator Clamp. When assembling the com-
mutator, a clamp will be necessary to hold the bars to-
gether while boring. Several makeshift methods are available,
but it will pay any repair shop to have suitable cast-iron
clamps, such as shown in Fig. 230. The clamp should be
smaller than the diameter of the commutator, so that when
it is drawn tight, there will be a space of about % inch between
the sections. When using this clamp as shown in Fig. 230
at the right, wooden blocks (C) can be employed to hold the
clamp about midway of the commutator. D is an iron face
plate. The clamp (B) should first be placed on the plate as
shown, and the bars and mica segments stacked in a circular
form within it. Care must be exercised to make sure that a

FIG. 230. At the left, a clamp for holding commutator bars together when
being assembled. At the right, the use of this clamp is shown.

mica segment is placed between each copper bar. Count
the bars carefully, so that their number corresponds with the
number of bars in the original commutator.

Take several pieces of copper wire (about No. 9 B. & S.
gauge) and remove the insulation. Place these around the
commutator near the top and lower ends to act as band wires,
and twist them tight. The clamp may then be removed,
and the commutator straightened. Bring out the mica
segments even with the surface of the bars by holding the
fingers against the inside edge of the segments and tapping
the bars on the outside with a small mallet. Place a square
or steel scale on the face plate and tap the bars on the outside
with a small mallet. Place the square or steel scale on the
face plate and see that the bars line up perpendicularly with
one edge of the square. If they do not, a gentle pressure one

COMMUTATOR REPAIRS 313

way or the other on the top end of the commutator with the
palm of the hand will bring them in line. See that each bar
and segment is down flat against the surface of the plate,
since that end will be fastened to the face plate on the lathe
when facing off the ends of the bars. Tap each bar and seg-
ment down solid with a square ended punch, a little narrower
than the thickness of the bar. When this has been done, the
band wires can be drawn a little tighter, and the surface of
the commutator, where the clamp will fit, should be filed
to remove any protruding mica, and present a smooth surface
for the clamp.

Replace each section of the clamp about the commutator
again using the wooden blocks mentioned before. Draw
the clamp tight, being sure to leave the same amount of
space between each clamp section. A small gas burner,
or some other source of heat should be handy, and the com-
mutator placed over it and heated. When it is good and
hot to the hand, tighten the clamp, allow it to cool, and again
tighten.

Boring out the End of a Commutator. The next thing to do
is to bolt the commutator to the face plate of the lathe and
center it. The same wooden blocks can be used again for
supporting the clamp. Face off the end of the bars, and then
groove out the end for the taper rings to the same diameter
as the mica rings on the sleeve. Take one of the old bars,
and with a bevel protractor determine exactly the taper used
on the old commutator. The usual taper employed is shown
in Fig. 231. A small groove (shown at A) should be cut below
the intersection of the two tapers to allow room for the edge
of the mica ring. The other end should be treated in the same
manner.

When the boring has been completed, a close inspection
must be made of all turned surfaces, to make certain that no
copper has been dragged over the mica to form a short-
circuit. The corner of the groove A (Fig. 231) should be
carefully gone over, as it is here that drag-overs most fre-
quently occur. Scrape and wipe the mica rings on the sleeve
clean; also wipe out the inside of the commutator with a soft
rag. Place the sleeve in the commutator again, and draw

314

ARMATURE WINDING AND MOTOR REPAIR

up on the end nut. On a refilling job, when the old mica
rings are used again, cracks may be found between the band
ring and the commutator copper due to the irregularity of
the mica ring caused by the shape of the old commutator.
These cracks can be filled by pushing in thin sheets of mica.
Shellac should be used liberally on the ends.

Bake the commutator in an oven for about one hour until
it becomes thoroughly hot, and tighten the nuts. Reduce
the heat somewhat, and bake at a low heat until the shellac

Ill

FIG. 231. I shows taper of commutator bars. II is a template for laying
out mica end rings. Ill shows a scheme for laying out the taper E, of the
template shown in II.

becomes hard. Tighten the nuts again and allow the com-
mutator to cool. When cold give the nuts a final setting up.
The clamp can now be removed.

In order to finish the commutator, use a mandrel to fit
the bore of the sleeve, and take the final finishing cuts. After
finishing, a bar to bar test with the test lamp outfit should
be made to be sure there are no short-circuits between bars.
A test for grounds should also be made by holding one wire
of the test lamp on the iron sleeve and passing the other from
bar to bar around the commutator. A high-voltage test for
grounds should be made also later.

COMMUTATOR REPAIRS

315

The slots for taking the armature leads must next be cut.
It is well to cut them a little wider than the diameter of the
lead wires. These slots should be made in a milling machine
if one is at hand, otherwise a hack saw can be used. Some-
times two blades fastened together will give the required
width of slot. The commutator can now be placed on the
armature shaft.

Mica Used in Commutators. The selection of mica for the
insulation of commutator bars and for V-rings is usually based
upon the type of motor being repaired, but built-up mica has
been found to be the most satisfactory. India mica is gener-
ally too hard and domestic mica too brittle for use between

FIG. 232. Mica segment (B) cut from sheet using bar (A) as pattern.
Such a segment is cut large at top and at ends so as to turn down evenly
with copper bars when commutator is finally surfaced.

bars. Domestic mica also has cracks and fissures which fill
up with dirt and cause short circuits. Amber mica gives
good service but it comes in such odd sizes that it is too
wasteful for the average repair job. Hard-baked built-up
mica is found best to be used between bars and the unbaked
best for building up the V-rings. The unbaked mica is
usually obtained in sheets 0.025 inch and 0.030 inch thick.
Heavier mica would be difficult to bend and to cut in shape,
The same sizes of hard-baked mica with the addition of
0.02 inch and 0.035 inch have been found sufficient stock to
fill almost any size of commutator. In cutting up the un-
baked mica for the V-ring insulation it is safer first to fit a
piece of paper around the ring and then use this as a template.

316

ARMATURE WINDING AND MOTOR REPAIR

Shaping Mica End Rings. When it is impossible to save
the old mica end rings, new ones must be made. These can
be made of flexible mica usually 0.030 inch thick. It is
best to make the rings in three pieces to the layer on small
commutators; three layers constituting a ring. When heated
slightly, this mica can be bent around the sleeve to the desired
shape. Trim the edges square on each piece and fit them
tightly in the bore of the commutator. The taper rings
should be fitted in the same manner with a slight space being
left between abutting ends.

Templet for Making Mica End Rings. A good shop method
for making a template to lay off mica end rings is as follows :

FIG. 233. Parts of a medium sized commutator.

<.o) and (c) Mica end rings. (6) Iron clamping V-ring. (d) Mica bushing sleeve
for inside of commutator, (e) Commutator bars and mica segments assembled.
Sample bars and mica segments are shown at (/) and (0).

Mark off on a piece of cardboard an arc of a circle, the radius of
which is equal to the diameter A in Fig. 231. Strike another
arc from the same center with the radius of the dividers made
smaller by the distance C. This will give a templet for the
band D.

A templet for the taper E can be prepared as illustrated in
the following example : Assume the diameter B to equal seven
inches. Then in Fig. 231 lay off one-half this diameter, or
inches from A on the edge AB as shown at 0. If the

COMMUTATOR REPAIRS 317

taper is 30 degrees, take a 30- by 60-degree triangle and square
its lower edge with the line AB, having its apex, or point
touching the point at 0. Draw the line OC, which will be
the radius of the arc of a circle D, shown dotted. Strike
another arc from the same center with the radius made smaller
by the distance F, Fig. 231. This will give a templet for the
taper E.

Micanite as a Commutator Insulation. For use in insulating
commutators, the Mica Insulator Company, New York City,
has developed a mica laying machine which successfully
constructs large sheets of micanite from thin mica laminae
with cement uniformly applied between the layers. These
sheets are finished by drying and baking under high pressure
after which the sheets are packed through milling machine and
finished to accurate thickness. The company gives the follow-
ing directions for using micanite as a commutator insulation.

If micanite segments for insulation between copper bars
of a commutator are to be cut from full size sheets of micanite
plate, a fine tooth band saw with moderate set should be used.
Bookbinders' shears or foot power cutters are poor tools to
use as the edge of the micanite is likely to be bruised and this
may result in dragging of the copper over the micanite when
the commutator is turned in a lathe. Short circuits, or near
short-circuits, are often caused in this manner. It is not neces-
sary to coat the surface of either the copper bars or the mica-
nite segments with shellac before assembling the commutator.
After turning the V's for the end rings carefully examine all
surfaces and scrape clear any places where the lathe tool has
dragged the copper. Before placing micanite rings, carefully
remove all dust from both commutator and rings. Short-
circuits are sometimes caused by overlooking these details.
After the commutator is assembled on the shells, with the
rings in place it should receive a baking at a temperature of
about 150C. (302F.). This baking is for the purpose of
embedding the end of the segments and setting the cement
in the rings.

Precautions when Tightening a Commutator. It is essen-
tial on account of the expansion of the copper, that the commu-
tator is not screwed up too tight before it is baked. The best

318 ARMATURE WINDING AND MOTOR REPAIR

results are obtained by tightening up the commutator gradu-
ally after baking as it cools, giving a final tightening when nearly
cold. Many of the troubles met with in commutators is due
to excessive tightening. In arch-bound commutators, it is
important that the assembled segments are not tightened too
hard in the arch before the V parts are turned, as the result-
ing stresses from the expanding copper, distort the parts and
often are the main cause of high bars. Similar care should be
taken not to exert more pressure than necessary on the end
rings and whenever possible means should be provided to
measure the pressure. Cases have been known where the
angles of the clamp rings in small commutators have been
changed by excessive pressure. Approximately 1500 Ib. per
sq. in. on the projected area of the inside cone or taper rings
is a satisfactory medium pressure.

Making Micanite End Rings. It is often necessary for
the commutator builder, particularly the repair man, to make
up his own rings. For that purpose a different kind of plate
known as No. 1 micanite or moulding plate is furnished by
the Mica Insulator Company. This plate is made from white
mica selected with same care as for micanite commutator seg-
ment plate and contains the proper proportion of cement to
produce satisfactory moulding characteristics. Micanite is
not ductile and will not stretch uniformly like sheet metal. It
is, therefore, necessary to cut the pattern for micanite com-
mutator rings into the exact development of the ring before
moulding. The pattern should be long enough to allow for a
tapered splice at the ends. For cementing the splice, use a high
grade shellac of about 2 Ib. to the gallon of solvent. When the
No. 1 micanite is placed on a steam table or metal plate
heated to about 140C. (284F.), it becomes plastic and can
be moulded readily into the desired shape. The micanite
should not remain on the hot plate any longer than necessary
as the cement might set, making it less pliable. It is advisable
to turn the micanite over so as to heat uniformly from both
sides. After placing the micanite in the mould or in the com-
mutator, it should be under pressure while cooling.

Causes of Excessive Commutator Wear. Warren C. Kalb
has pointed out (Power, March 18, 1919) that commutator

COMMUTATOR REPAIRS 319

wear may be caused by the use of abrasive brushes. It may
also result from the same mechanical or electrical causes
that produce brush wear. Mechanical conditions of this
nature are: Rough commutator due to poorly prepared
surface; rough commutator caused by burning; vibration from
any source such as the pound of a direct-connected engine,
belt lacing, improper mounting, loose bearings or commutator
out of true; high peripheral speed, causing brushes to chatter
or creating an excessive temperature at the brush faces due
to friction ; type of brush-holder or angle of operation, resulting
in the chattering of the brushes.

Copper Used to Make Commutator Bars. The bars for
commutators can be made from drop forged, cold rolled,
hard drawn or cast copper bars. The drop forged copper is
usually considered the best and cast copper the worst for bars.
Probably most commutators are made of cold rolled or hard
drawn copper. Cast copper is poor on account of the many
pin holes and larger that are liable to be in the structure of
the bar. These show up at every turning and require filling
to prevent carbon dust from accumulating. If they are near
the edge of the bars the carbon is liable to bridge the mica
between the bars and cause a short-circuit and burn a hole
in two bars. This calls for a patching of the bars and the
insulation which adds a weak spot to the commutator and
simply hastens the time when it must be rebuilt. A good
grade of copper for bars is economical from points of service
and reduction of repair costs.

However, when hard-rolled, or drop-forged copper is not
available for the repair man, cast copper makes a good substi-
tute. If the castings are soft they have a minimum conduc-
tivity of 60 per cent, and are more free from blow holes; also,
they wear from 75 per cent, to 90 per cent, as long as the hard-
drawn or drop-forged copper bars. The castings are soft but
tough, and the low conductivity is due to the fluxes used in
melting the copper. A reduction in the amount of flux used
increases the conductivity of the copper, but it also tends to
make the metal more brittle, more porous and liable to have
blow holes below the surface. The 60 per cent, soft-copper
castings do not need filling or machining on the sides. They

320 ARMATURE WINDING AND MOTOR REPAIR

can be easily straightened and any projection flattened out on
a surface plate by the aid of a steel flattener and a hammer.

Test for Oil -saturated Mica in a Commutator. When the
mica between commutator bars appears to be saturated with
oil the following test can be applied. Wet a rag or piece
of waste with gasoline and wipe the surface of the commutator
dry. Then with a torch heat the commutator at the point
where the mica appears to be saturated. If oil has worked
down between the mica and the bar it will ooze out in small
bubbles. In such a case new mica segments must be inserted.

Blackening of a Commutator at Equally Spaced Points.
This is usually caused by an open circuit in the armature
winding. The open circuit may be at the commutator necks
or at the rear-end connections of the winding when two-piece
coils are used. When these spots are noticed on a commutator
it should be tested out for open circuits at once before further
commutator troubles develop.

UNDERCUTTING MICA OF COMMUTATORS

The object of undercutting the mica of commutators is to
clean out the mica between the copper segments to a depth of
about J^2 to %2 mcn s that the copper bars will wear evenly
free from the poor commutation that often results when the
mica is hard and does not wear down evenly with the copper
segments. The undercutting operation should be done after
the armature has been rewound, soldered and banded. The
commutator should also be trued up and all excess solder
removed from the neck and face.

Tools for Undercutting Mica. Special motor-driven or
belt-driven saws are available for doing this work which can be

FIG. 234. Hand tool for undercutting mica that can be connected to a small
motor by a flexible shaft.

used when the armature is centered in a lathe. The saws
are usually clamped on an arbor which is mounted on a head

COMMUTATOR REPAIRS

321

that moves on slide rails. By means of a hand-operated lever
or foot pedal controlled by the operator, the revolving saw
is carried over the face of the commutator. A shaper equipped
with a special tool can also be used. Frequently a milling
machine will be found convenient for undercutting the smaller
sizes of armatures. A less expensive device consists of a motor-
driven circular saw which is mounted in such a way that the
saw can be guided over the commutator face by hand. Two
such motor driven tools are shown in Figs. 234 and 235.
Size of Circular Saw Required. For the cutting tool
a circular saw or miller about to 1J inch in diameter

FIG. 235. Motor operated hand tool that can be connected to a lighting
circuit for undercutting the mica of a commutator.

with from 15 to 30 teeth seems to give the best results. A
small diameter saw must be used in order to cut the slot to
the proper depth and at the same time not cut into the neck
of the commutator. The saw should be driven at approxi-
mately 1500 rpm. and be about 0.005 inch thicker than the
mica in order to remove the mica completely. If the com-
mutators are of large diameter, one or two spacers of the same
thickness as the bars may be used between the saws and two
or three slots cut at the same time. The cutting edge of the
saw should revolve in a direction toward the operator while
cutting the mica. When the hand-operated motor-driven
21

322 ARMATURE WINDING AND MOTOR REPAIR

tools are used they should be drawn toward the operator,
in order to properly guide the tool.

It is advisable to have a jet of compressed air or a fan so
located that the particles of mica and copper will be blown
away from the armature to prevent this material falling in
behind the commutator and at the same time make it easy
for the operator to see the slot when the saw is throwing the
particles toward him on the face of the commutator.

Finishing Slots and Commutator Surface after Undercut-
ting. After the slots are sawed it is advisable to go over them
with a sharp hand tool to remove remaining particles of mica
and thin strips along the edges of the slot. This can be done
with a sharp knife, or V-shaped tool, which can also be used to
slightly bevel the sharp edges of each bar and remove all
burrs. Several hand tools that can be used for undercutting
mica when motor operated tools are now available are described
in the following paragraphs.

After the undercutting operation has been completed the
commutator should be stoned and polished with fine sandpaper
to remove all burrs of copper.

Brushes for Use on Undercut Commutators. The removal
of mica permits the use of brushes without abrasive qualities
and has resulted in the development of very low friction
brushes. In fact some of the very lowest friction brushes
manufactured are hard and do not contain any graphite.
Such brushes give to a slotted commutator the brown gloss
that shows perfect operation. The life of a hard non-abra-
sive brush is several times that of a graphite brush. Some
criticism has been made of slotted commutators which has
been traced to the fact that the brushes used were not suitable.
Often when the same brushes are used on an undercut com-
mutator that were used before, the commutator will be ridged
and worn rapidly and the slots filled with copper dust with
short-circuits and burned out coils the result. The fault in
such cases is in the brush not in the undercutting process.

It should also be remembered that the brush tension on
an undercut commutator can often be much lower than before.
Brush authorities recommend lj^ to 2J pounds per square
inch of cross-section on stationary motors, from 3 to 5 pounds

COMMUTATOR REPAIRS

323

per square inch for crane motors and from 4 to 8 pounds per
square inch on railway motors when the commutators are
undercut.

Another factor in good operation of undercut commutators
is cleaning at regular intervals. If the motor is operated in a
place where the accumulated dust and dirt is dry, it can be re-
moved with a blast of compressed air. Even in such cases
the slots should be scraped occasionally. A lubricant should
never be used on an undercut commutator nor any other when
operating conditions are correct.

Hand Tools for Undercutting Mica of Commutators. An
easily made hand tool for undercutting the mica of com-

FIG. 236.

-Short saw blade clamped in a vulcanized fiber holder for use in
undercutting mica.

mutators is shown in Fig. 236 as devised by William H. Watson
(Power, Oct. 1, 1918). The slot in the holder is made as deep
or deeper than the width of the hacksaw blade, with small bolts
to clamp the blades at each end. Any depth of cut desired can
be made by adjusting the blade in the holder, but there sel-
dom is occasion to move the blade. The handle part is cut
away just enough to allow the fingers to pass over the commu-
tator without rubbing. In undercutting commutators care
must be taken that no rough edges are left after the mica is
cut, but it is sometimes hard to avoid this, especially if it is an
old commutator. These rough corners can be smoothed, after

324 ARMATURE WINDING AND MOTOR REPAIR

the mica has been undercut, with a V-shaped tool made of
hardwood or vulcanized fiber, run between the segments as
shown at A, and it does not spoil the bar. The front corner
of the undercutter, which is made of fiber, can be used to
smooth these raw edges.

Another hand tool recommended by T. M. Sterling (Power,
Nov. 5, 1918) for undercutting mica and as useful in other
ways by the repairman, is illustrated in Fig. 237. In the
illustration A represents a steel straight-edge about ^ inch
thick and slightly longer than the length of the commutator
segments and square at one end to butt up against the head of
the commutator. B is the undercutting tool, about 12 inches

FIG. 337. Hand tool for undercutting commutator mica.

long, made from a piece of %Q hexagonal tool steel forged flat
for about one-fourth of its length to a thickness at point D
equal to the thickness of the mica between the segments. The
"half-heart" shaped lobe C affords a bearing place for the
fingers in using the tool.

When using this tool lay the straight-edge on the commu-
tator with its edge in line with the mica between the segments
and the square end against the head of the commutator.
Hold it firmly with the left hand. Then, holding tool B in
the right hand, draw the point D along the mica the same as
drawing a pencil along a ruler in ruling paper. This operation
will start a nice groove in the mica without burring the edges
of the copper on either side. After one or two passes the tool
can be turned endwise with point D down and the mica taken
to any desired depth or a saw blade may be used to finish the

COMMUTATOR REPAIRS 325

groove. The disadvantage of the saw blade is that the set
of the teeth burr the copper more or less, while with this tool
all chance of a burr is eliminated.

The tool will also be found handy in raising coil leads out
of the slots in the head of a commutator when unsoldering
them preparatory to removing the commutator or making
coil repairs. After heating the ends of the coils until the solder
starts to melt, drive the point of the tool into the slot under
the ends of the wires, with lobe C down, for a fulcrum, on the
bottom of the slot. The ends of the coils can then be easily
raised by a downward pressure of the hand. Square edge
E can be used as a scraper for taking the surplus hot solder
off the ends of the wires as they are raised.

CHAPTER XIII

ADJUSTING BRUSHES AND CORRECTING BRUSH

TROUBLES

Many of the troubles which are charged to brushes of motors
and generators can be traced to improper application and
adjustment or to other defects of the machine that show
up in sparking at the commutator. A careful selection of
brushes is important but no more so than carefulness in ad-
justment and frequent inspection in operation, for proper
care of brushes and brush rigging results in good commutation
and prolonged life of both brushes and commutator. The
repairman should, therefore, give due consideration to brush
adjustment when overhauling a machine.

Fitting or Grinding-in Brushes (Instruction Book, Westing-
house Electric & Mfg. Co.). When it becomes necessary
to install a new set of brushes, sand paper or garnet paper
may be used that is long enough to go around the commutator.
It should have a lap of several inches and be so mounted on the
commutator as to preclude the lapped end butting against the
brushes when the machine is rotated for grinding. With
many commutators the friction between the commutator bars
and sandpaper (if the paper is taut) will suffice to keep the
paper from slipping, especially if, when starting, the paper is
given a pull in the direction of rotation by the operator.
If the paper persists in slipping use a little glue to stick the
under end of the paper at the lap to the commutator. All
traces of the glue must be removed from the bars before
putting the machine into service.

A second method is to remove the middle brush on each arm
and bind the paper to the commutator by running tape or
string entirely around the periphery. If necessary, also bind
the paper at the inner and outer ends of the commutator.
After the paper is anchored the commutator may be rotated

326

ADJUSTING BRUSHES AND CORRECTING TROUBLES 327

by hand, or in any convenient way. Great care must be
exercised in " grinding in" brushes in this way, as the cutting
is very rapid, especially with soft brushes, and much of the life
of the brush may be ground away in a very few revolutions.
If after the brushes have been surfaced in the above manner,

FIG. 238. Grinding-in brushes of a direct-current generator with a strip of

sandpaper.

the trailing edge shows a poor seat on account of having had
to mount the ridge due to the lap in the paper, a final surfacing
should be done with very fine sandpaper by hand. The
foregoing method is particularly desirable when a machine
has very hard brushes, or a large number of soft brushes.

328 ARMATURE WINDING AND MOTOR REPAIR

It is a great time-saver after the knack of applying and anchor-
ing the paper is understood.

With slip rings, if carbon or graphite brushes are used, the
same general scheme is applicable, but if metal graphite
brushes are used, emery cloth, or its equivalent, is preferable.

There are those who advocate and those who practice
pulling the grinding paper in the direction of rotation in
order to more accurately surface the brush. This seems
logical enough on first thought, but inasmuch as the com-
mutator is not in motion when the brushes are surfaced in
this manner there is no assurance that the brush will bear
the same relation to its holder (and therefore to the com-
mutator) when the commutator is rotating and, consequently,
no assurance that the contact will remain fixed. This perhaps
explains why brushes frequently show perfect contact when
idle, but poor contact when in service. This method of
fitting the brushes, therefore, is not necessarily more depend-
able than other methods, though in some instances it can
be recommended.

After the brushes have been fitted all dust should be care-
fully blown out of the commutator by compressed air. The
pressure should be about, 50 to 80 pounds per square inch used
with a J^-inch nozzle. A higher pressure may injure insula-
tion and is not necessary.

It is important that the brushholders, whether for direct
or alternating-current use, be neither too close to nor too far
from the commutator or slip rings. A suitable distance is
from three-sixteenths to one-quarter of an inch.

If the brushes are copper-coated, the coating should be
not allowed to come in contact with the commutator or slip
rings. This means that as the brushes wear the copper coating
should be scraped back and not allowed to extend below the
brushholder box. With proper shunts there should be no
need for copper coating from an electrical standpoint, although
for mechanical reasons it may be desirable as a protection
for soft or structurally weak brushes.

If the toe of the brush is very sharp it is good practice to
"nose off" the knife edge. This reduces breakage and mili-
tates against gouging into the commutator.

ADJUSTING BRUSHES AND CORRECTING TROUBLES 329

Adjustment of Brushholders (Instruction Book, Westing-
house Electric & Mfg. Co.). If the commutator or slip-ring
speed is low, a holder having a sluggish spring and slow action
will give entirely satisfactory results but such a holder on a
high-speed machine may fail miserably. High speeds require
quickness of action on the part of the brush. This means
a sensitive quick-acting spring and a holder designed to aid
the spring. A high-speed holder, therefore, may give excellent
results on low-speed machines, but a low-speed holder is not
likely to approach even fair results on high-speed machines.
This is a point a repairman must bear in mind when changing
the speeds of direct-current machines.

The box, or part holding the brush, should be as nearly a
full box as possible. Whether the rotation is against the toe
or against the heel, both the toe and heel sides of the holders
should have approximately equal areas for the support of the
brushes. It is obvious that the shorter the box, the greater
will be the shifting movement of the brush on the commutator
due to looseness in the holder; and the deeper the box, the
less will be the shifting. From this it follows that the boxes
should not be too short, nor should the brushes be too loose
in the boxes.

It is an established fact that with even very slight play in the
holder the changing brush friction due to load, temperatures,
etc., will result in a changing relation between the brush and
the commutator, and thereby cause changing and misleading
brush drops from day to day. This accounts for the difficulty
in verifying, on different days, brush drops, although the loads
may be identical. The neutral does not change, but the
relation of brush contact to commutator does change, and
the brush drops vary accordingly. Another explanation for
the changes that take place in brush drops is that when
brushes have been freshly " sand-papered " or " ground in"
the surfaces have unglazed or soft faces, the brush drops
being influenced by the porous condition of the faces, and the
lubricating values of the particles liberated during the time the
brush face is seating itself to the commutator. During this
period the commutation, with certain kinds of graphitized
brushes, may be much better than after the glaze or permanent

330 ARMATURE WINDING AND MOTOR REPAIR

surface forms on the brush. Drops taken during this time may
not be the same as drops taken after the brush is "faced up"
and may prove seriously misleading. Definite or reasonably
permanent drop values are, as a rule, to be had only after
the brush faces have reached a fixed condition.

From the foregoing it is obvious that brushes should fit
snugly, but not tightly, in the holders; also that there are
certain relative dimensions which are preferable between
the brushes and holders and which must not be ignored if the
best results are to be obtained. For instance, the side of the
holder in the direction of rotation should not have less area
than the lagging side. In fact, the side against which the
brush presses should have preferably a greater area than
the opposite side certainly not less area. With a limited
area to oppose the brush, movement of the brush in the holder
due to looseness, oscillation and imperfect relations between
the brush and commutator, accentuates the change in the
angle of contact, and results in poor commutation, as well as
in mechanical and electrical damage to that part of the brush
in contact with the holder.

If a brush is thin, its length and holder contact should be
greater than if it is thick, for the reason that the thinner the
contact area on the commutator, the less stable will be the
brush on the commutator, and the greater will be the need
of support from its holder. To illustrate : If the brush con-
tact is a knife edge, serious chattering may result and there
may be a maximum of instability, but if the knife edge be
gradually removed the brush will become increasingly stable
as the thickness of the brush contact increases, and if the
brush angle is correct the maximum stability will be reached
when the brush has its greatest thickness in contact with the
commutator. It stands to reason, therefore, that the greater
the area due to thickness, the more stable will be the brush
on the commutator, regardless of the holder, and that as a
consequence a holder for a thick brush might not prove at all
suitable for a thin brush, the speeds being equal. If, however,
there is the same amount of lost motion in the holders for a
given length of thin brush and thick brush, and the brush
angle is such that the thick brush moves its maximum in the

ADJUSTING BRUSHES AND CORRECTING TROUBLES 331

holder, then the thick brush will show a greater change in its
relation to the commutator than the thin brush. The thick
brush, however, will not be so likely to shift its maximum on
account of its bearing surface being greater than the bearing
surface of the thin brush. In other words, it rests on a
greater area.

Causes of Rapid Brush Wear. Warren C. Kalb (Power,
March 18, 1918) has found that the following electrical con-
ditions cause rapid brush wear: Sparking, from any cause
whatever; glowing of brushes; pitting of brush faces.

Glowing results from excessive current density and may
be local or may cover the entire contact end of the brush.
Common causes are unequal collection of current by different
brushes of the same polarity and very heavy short-circuit
currents in the coils undergoing commutation, which, added
to the load current, bring the temperature of the carbon
at the face of the brush up to the glowing point.

Pitting may result from glowing in small spots on the brush
face. It is also caused at times by particles of copper becom-
ing attached to the brush face. This causes a heavy current
to localize in a small area, disintegrating the carbon and
forming a small crater in which the copper embeds itself.

The increased brush wear caused by increase in current
density is due to the higher temperature so created at the
brush face and the consequent more rapid disintegration of
the carbon at that point. Any factor tending to further
increase the temperature at the brush face will add to the
rapidity of wear. Some of the things having this effect are:
High coefficient of friction ; higher contact drop than is needed
for commutation, especially on machines of low voltage and
high current capacity; contact drop too low for sparking
voltage, permitting heavy currents to flow in the short-
circuited coil; lack of carrying capacity or very high current
density.

Higher current densities are permissible where there is no
sparking voltage than where there is. This is because the
heating effect of commutation current is absent and higher
load currents can be applied before the same temperature
rise is attained.

332 ARMATURE WINDING AND MOTOR REPAIR

Average volts per commutator segment may be as high as
15 to 20 volts without difficulty being encountered. Or
the reactance voltage may be that high and still be neutralized
by interpole flux or fringing field within sufficiently close
limits to secure good commutation. But the sparking
voltage that is, the resultant between the reactance voltage
and the other electromotive forces generated within the coil
undergoing commutation should be less than the contact
drop of positive plus negative brush to attain perfect com-
mutation. Inasmuch as 3 volts is about the highest brush-
contact drop obtainable, it will be seen that the sparking
voltages mentioned would not be neutralized within 12 to 17
volts. Such a voltage would set up excessive currents in the
short-circuited coils and result in rapid burning away of the
brush faces even with no-load current whatever carried by
the machine.

Methods for Locating the Electrical Neutral in Setting
Brushes. The following methods have been suggested by

V
FIG. 239. Pilot-brush method of locating brushes on the electrical neutral.

T. F. Barton (Power, July 16, 1918) for setting brushes on
the electrical neutral in direct-current machines.

With the machine running at no load, normal speed and
voltage, shift the brushes until a low-reading voltmeter shows
no deflection when connected to points just inside the heel

ADJUSTING BRUSHES AND CORRECTING TROUBLES 333

and toe of a pilot brush. The pilot brush must be the full
size of the brushes used on the machine and can be made of
fiber or wood with holes drilled through it to allow contact
on the commutator at the desired point, which should be
at the center of two adjacent commutator bars as indicated
in Fig. 239.

Operate the machine as a shunt-wound commutating-pole
motor, checking the speed in both directions of rotation,
holding the same value of armature voltage and shunt-field
current in each case. The brushes are on neutral when the
speed is the same in each direction of rotation.

It is preferable to locate the brushes on the electrical neutral
and make the adjustments on the commutating- and series-
field windings to give the desired results. Brush shift can be
used, provided the final brush position is in a satisfactory
commutating zone. The capacity of a machine is not re-
duced by shifting the brushes if the rated voltage and current
can be successfully obtained.

Angle at which Brush is Set. When machines are not
subject to reversals, the proper angle for the brush will depend
upon whether the rotation is against the toe of the brush or
against the heel. It also depends upon the friction coefficient
of the brush. If set against the toe, the angle should approxi-
mate 35 degrees. Above 35 degrees the toe becomes very
sharp and mechanically weak. If set against the heel, the
angle may be from 12 degrees to 25 degrees. In instances
where the brush touches more bars than required, the sharp
toe can be beveled off thereby reducing the brush contact
and the short-circuiting current under the brush. This also
increases the mechanical strength of the brush. Whether
running against toe or heel, it is good practice to round off the
sharp or knife edge of the toe. This reduces the possibility
of the knife-like edges digging into the commutator in case
the brushes jam or become tight in their holders and at the
same time protects against brush breakage.

The friction coefficient of a brush is a variable one, influenced
by the characteristics of the brush, by the glazing of the
contact face of the brush, the varying brush temperature due to
load and spring pressure and the condition of the commutator.

334 ARMATURE WINDING AND MOTOR REPAIR

whether it is smooth or undercut, hot or cold, in good or poor
mechanical condition. The poor mechanical condition of the
commutator can be eliminated by proper attention. The
other influences are permanent except as they may be modified
by finding the most suitable brush angle.

Checking Brush Setting. To check the spacing of brushes,
place a strip of paper completely around the commutator under
the brushes and mark the position of each set. The strip can
then be removed and the distance between marks measured.
If the distance varies the brush holders should be moved into
the proper position. The degree of accuracy with which
brushes should be set depend upon the type of machine.
With interpole machines, the spacing should be more accurate
than with other types of machines. The brushes on any given
stud should be staggered with respect to those on adjacent
studs so that the entire commutator will be covered by the
brushes except a slight space at each end. This prevents
grooves wearing in the commutator.

Brush Pressure. This depends upon the character of the
brush, the commutator speed in feet per minute, and the
mechanical condition of the commutator. It varies from two
to five pounds per square inch, the pressures generally used
being from 2.5 to 3.5 pounds on commutators and from three to
five pounds on slip rings. If there is a tendency for the com-
mutator to burn, this can at times be corrected by increasing
the pressure, and thereby increasing the abrasive effect of the
brush sufficiently to scour out the burning and maintain a
polished surface. When the mica is not undercut the brush
should have sufficient abrasive effect not only to scour out
possible burning of the commutator bars but also to wear and
keep the mica flush with the commutator bars.

Common Brush Terms. The following terms are frequently
used in describing brush characteristics and their performance.

Contact Drop. This refers to the drop in voltage between
the brush and the commutator. It differs with different
materials, different brush pressures and different loads, being
between one and two volts. For a given material and given
load the contact drop decreases with increased pressure and
increases with the load. It varies little with changes in speed

ADJUSTING BRUSHES AND CORRECTING TROUBLES 335

above 2500 feet per minute and but little with temperature
changes under normal conditions.

Brush Friction. The coefficient of friction depends upon
the brush material, brush angle and commutator speed. With
a given material and brush angle, the coefficient of friction
increases with load and brush pressure and in general decreases
with increased commutator speed. Increased temperatures, in
actual practice, justify the conclusion that the friction coef-
ficient increases with increased temperatures.

Specific Resistance. The ohmic resistance of a cube having
one-inch sides.

Current Density or Carrying Capacity. This is based on the
maximum continuous current capacity per square inch, with-
out glowing, honeycombing, undue heating or sparking.

Glowing. A brush is said to "glow" when it becomes red or
incandescent in spots in proximity to the commutator. This
is due to the short-circuiting current under the brush, or to the
short-circuiting current in combination with the working or
load current. It may also be due to a lack of homogeneity
of the brush (hard or soft spots of the same or different mate-
rials having characteristics foreign to the material in the body
of the brush); to incorrect brush position; to improper brush
selection; to selective commutation; to the brush covering too
many bars; to high mica; to the machine having inherently
poor commutating characteristics due to improperly shaped
or spaced main poles or commutating poles; to bad commutat-
ing pole adjustments, or to a poor distribution of the armature
windings. In the event the glowing can not be stopped by
correcting such of the brush or mechanical troubles as may
exist; or by shifting the brushes, or readjusting the commutat-
ing poles, or both ; a brush having higher current density values
and less susceptible to the effects of high voltages between the
commutator bars should be installed. If a suitable brush
can not be had, the design of the machine itself may be subject
to modification.

Honeycombing. When brushes gradually burn away, form-
ing small craters in their faces, they are said to " honey comb."
This is due to continuous sparking of a more or less hidden
nature and may be of either a slow or rapid nature. If the

336 ARMATURE WINDING AND MOTOR REPAIR

growth is slow it is sometimes possible to correct it by increas-
ing the brush pressure, thereby decreasing the contact drop
and the contact arcing, and at the same time increasing the
abrasive effect of the brush and grinding away the minute
craters as they form. Honeycombing is due to the same
causes, on a reduced scale, as glowing, and at times both may
take place simultaneously on the same machine. Honey-
combing has been traced directly to high mica, in combination
with loose commutators; also to improperly spaced main
poles ; and unequal air gaps; and in non-commutating-pole
machines, to an abnormal shifting or distorting of the mag-
netic flux. The trouble is most frequently corrected by sub-
stituting a more highly refractory brush possessing abrasive
characteristics.

Hardness. Brushes have widely varying physical densities
some being very hard, others very soft. Very hard brushes
usually carry a large amount of abrasive and have a low current
density or carrying capacity. The softer brushes are more
highly graphite, carry abrasive to a limited extent only and
have a high current density. The hardness or scleroscope
reading is, therefore, indicative to a limited degree of the char-
acter of the brush.

Brush Inertia. This has to do with the weight of the brush.
The lighter the brush, the more readily will it follow the irregu-
larities of the rotating element, and the more promptly will
it respond to a given spring pressure. Again, the wearing of
the commutator may not prove so great. If a light and a
heavy brush prove equally satisfactory electrically on a ma-
chine, the lighter brush is preferable. If the materials from
which brushes are made are heavy, it is advisable to have an
increased number of small brushes, rather than a limited num-
ber of large brushes; also the spring pressure should be in
excess of the pressure on lighter brushes having the same
dimensions, and used for the same service.

Refractory. Any material which resists the ordinary meth-
ods of reduction is said to be refractory. A brush, therefore,
which resists wholly, or to a marked degree, high tempera-
tures, such as the heat generated by an electric arc, is said to
be highly refractory.

ADJUSTING BRUSHES AND CORRECTING TROUBLES 337

Peripheral Speed. This is the speed in feet per minute of
the commutator or slip ring. Peripheral speeds which vary
greatly require brushes having different characteristics. Gen-
erally speaking, carbon brushes are not as suitable for high
peripheral speeds as graphite brushes.

Procedure for Locating Causes of Brush Trouble. Spark-
ing at the brushes of a motor or generator can usually be traced
to one or more of the following causes.

1. Too low brush pressure.

2. Incorrect spacing of brushes.

3. Unequal air-gaps or defective fields.

4. Brushes not operating on electrical neutral.

5. Incorrect thickness of brushes.

6. Using brushes of wrong characteristics.

The above causes are given in the order in which they should
be checked up in the machine. All but two of these defects
may cause abnormal short-circuit currents under the face of
each brush or between two or more studs of the same polarity.
This point should be kept in mind when searching out the
trouble. The procedure for investigating each of the causes
has been outlined by E. H. Martindale of the National Carbon
Company as follows:

Before starting an investigation, eliminate the question of
high mica. If the commutator runs nearly true, stone it
enough to grind the mica even with the bars. If, however, the
commutator is not true, or there are flat bars, turn or grind the
commutator before looking further for the seat of the trouble.

Too Low Brush Pressure. Low pressure will cause poor con-
tact between the brushes and the commutator, and force the
current to pass from one to the other through a small arc.
This will burn both the brushes and commutator and produce
high mica. The heat of the arcs between brushes and commu-
tator will also increase the temperature rise. Further, poor
pressure produces a high contact drop, which will also heat
the commutator and brushes. This high contact drop and
arcing caused by too low pressure frequently heats the brushes
to a red heat, or, as we say, "to the glowing point, " which is
always accompanied by pitting or disintegration of the faces

of the brushes. This again reduces the available contact area
22

338 ARMATURE WINDING AND MOTOR REPAIR

and increases the current density through the balance of the
brush. The bad effects of too low brush pressure are greatly
aggravated when the commutator is slightly elliptical or runs
unbalanced.

Whether the pressure is high or low, care should be exer-
cised to get all the brushes on one machine at a uniform pres-
sure. If the pressure is unequal, the brushes with the highest
pressure will carry the highest current. In many cases some
brushes on a machine may carry three or four times as much
current as other brushes, due to this variation in pressure.
This excessive current may be enough to burn off the pigtails,
overheat the brushes, and cause glowing and pitting of the
brush faces. The best value of brush pressure should be deter-
mined for any machine by trial, as it is influenced greatly by
local conditions. A pressure of from 1.75 to 3 pounds per
square inch of contact surface may be used for motors and gen-
erators, and from four to seven pounds per square inch for such
machines as crane motors and railway motors where vibration
is comparatively severe. However, these values are given only
as an indication of best average practice, and are not always
the best to use. See table on page 342, paragraph 6 of head-
ing No. I.

Incorrect Spacing of Brushes. By spacing of brushes, is
meant the distance between brushes on adjacent studs meas-
ured around the commutator. Thus, in a four-pole machine
the brushes on each stud should be at a distance exactly one-
fourth the circumference from the brushes on the adjacent
studs. This may be checked up by counting the total number
of commutator bars, then dividing by the number of poles, and
setting the brushes that number of bars apart. A better way,
however, is to obtain a piece of wrapping paper of the same
length as the circumference of the commutator and of the
same width as the commutator. Measure off the correct
brush spacings and draw lines across the width of the paper.
Replace this on the commutator, and with one set of brushes
set so as to toe one mark, set the brushes on all the other studs
exactly on their respective marks. If, then, the sparking
can not be entirely eliminated by shifting the brush yoke either
forward or backward, examine the field coils and air gaps.

ADJUSTING BRUSHES AND CORRECTING TROUBLES 339

Defective Fields. The field coils on the machine are usually
connected in series so that a part of one coil may become short-
circuited without becoming apparent except by test. Obtain
a voltmeter with a suitable scale and test the voltage drop
across each coil. The drop should be the same for each field.
If one coil has higher drop than the others, the coil has been
incorrectly wound. If one coil is lower than the other, the
coil is either incorrectly wound or a portion of it has become
short-circuited. Either case will require a new coil.

Another more common defect in compound machines is a
reversal of one of the series field poles. For this test use a
compass and check the shunt as well as the series coil. First
pass current through the shunt coils. Next, bring the compass
near one coil; mark it either South or North, depending on
which end of the needle is attracted to the field pole. If the
bearings of the compass are not perfectly free, use care not
to bring the needle suddenly too close to the pole, as it is
easy to reverse the magnetism in the needles of moderate-
priced compasses. After marking one pole, proceed around the
machine and test in the same way each pole. The needle
should reverse direction at each pole. If two adjacent poles
are found to attract the same end of the needle, one coil is
reversed.

Field coils can also be tested for polarity by two ordinary
iron nails as described on page 136.

After all the shunt fields have* been tested, current from
some source should be passed through the series fields only,
care being taken to see that this current flows in the same
direction as when the machine is in operation. Nearly all
compound machines are cumulative compound, in which case
each series field should attract the same end of the needle as the
shunt field attracts. The differential compound winding is
used only when it is desired to have a rise of speed with an
increase of load, and is not in extensive use. In such machines,
however, the series fields produce poles of opposite polarity
to those produced by the shunt fields.

Unequal Air Gaps. Unequal air gaps may be the cause
of serious trouble, particularly in "lap" or "parallel wound"
machines. Poor centering of the armature when the machine

340 ARMATURE WINDING AND MOTOR REPAIR

was assembled, may be the cause or it may result later from
worn bearings. The wear may be either at the bottom of
the bearings, due to the weight of the armature, or on a belt-
connected machine, at the side of the machine, due to the
pull of the belt. This may be checked mechanically by
measurement with thin sheets of metal or fiber. For lap
wound machines a more accurate method, however, is to
disconnect the bus bars or leads which connect the studs of
the same polarity. Then, without any load on the machine,
each stud will be independent, and with a low-reading volt-
meter any difference between the voltage generated under
the various poles can be detected. This is valuable on a lap
wound machine, but on a wave wound machine the variations
are equalized and the mechanical method is more reliable.

After having correctly spaced the brushes, checked the
drop across the field coils, the polarity of the field poles and
the uniformity of the air gaps, it should be possible to shift
the brushes to a point where no sparking will appear with
a steady load on the machine. If the sparking reappears
with a variation in load, either the brushes are of incorrect
thickness or the neutral field that is, the distance between
adjacent pole tips is too small.

Incorrect Thickness of Brushes. The best way to investigate
this sparking is with a voltmeter reading about 5 volts.
Take two ordinary lead pencils and trim down one side of
each pencil as much as passible without exposing the lead.
Near the top of each pencil cut a groove deep enough to
expose the lead and attach a piece of lamp cord or other
flexible wire to each pencil. Remove one brush and hold
the pencils in the brushholder with the points on the com-
mutator and the flat side of one pencil against the front edge
of the brushholder and the flat side of the other pencil
against the back edge of the holder. This will give the volt-
age generated across the brush, or, in other words, the
"commutation voltage/' Now shift the brushes to a point
where they do not spark and read the " commutation voltage."
Increase the load until the brushes spark and again read the
" commutation voltage." If it has decreased, the brushes are
too narrow that is, the period of commutation is too brief.

ADJUSTING BRUSHES AND CORRECTING TROUBLES 341

The remedy is to use thicker brushes or to shift the brushes with
each change of load. This condition, however, seldom obtains.

An increase in the commutation voltage indicates that the
brushes are too thick that is, they span too many bars or the
neutral field is too small. This thickness of the brushes may
easily be changed by cutting down the face of the brush with a
hack saw or file, leaving the body of the brush the proper size
to fit the brushholders. This will help the trouble, and if the
sparking does not entirely cease, it may be necessary to widen
the neutral field by increasing the distance between adjacent
pole pieces. This can be done by filing the edges of the pole
faces.

Brushes of Wrong Characteristics. If none of above remedies
cure the sparking, the machine is being operated with brushes
not adapted to the service. The next step should be to give
a manufacturer full details regarding the machine and ask
for recommendation of a brush best adapted to the machine
and service, since carbon brush manufacturers maintain
an engineering department capable to give valuable advice
along this line.

As a guide in the location of other brush troubles, the ac-
companying table will be found useful.

Possible Causes of Brush Troubles on Motors and

Generators and Their Remedies Compiled by

E. H. Martindale of the National

Carbon Company

I. CAUSES AND REMEDIES FOR SPARKING AT BRUSHES

1. Brushes off electrical neutral. Shift to neutral by trial, or. set on
neutral by means of voltmeter.

2. Brushes spanning too many bars. Trim down faces of brushes for
short distance back from end or, if holders are clamp type, order thinner
brushes.

3. Brush studs not parallel with the commutator bars. Bend the brush
studs or grind or shim under the bolts which fasten the studs to the yoke.

4. Incorrect brush spacing. Check the spacing by counting the num-
ber of bars between studs or by placing a strip of paper around the com-
mutator with divisions marked off equal to the number of studs, and
correct the spacing by rotating the brush studs or the brushholders on
the studs.

342 ARMATURE WINDING AND MOTOR REPAIR

5. Brushes tight in brush holders. Clean the holders with gasoline and
if brushes are still tight, sandpaper them down or file out the holders
carefully.

6. Brush pressure too low. Pressure should be 1% to 2% Ib. per square
inch cross-section for stationary motors and generators, 23^ to 4 Ib. for
elevator and mill motors, 3 to 5 Ib. for crane motors, 4 to 7 Ib. forrail way
motors.

7. Too low contact drop of brush. Consult a brush manufacturer.

8. Insufficient abrasive action of brushes. Use a commutator stone or
more abrasive brushes.

9. High mica. Use abrasive brushes, a commutator stone or undercut
the mica.

10. Chattering. See heading No. VII for remedies.

11. Poor adjustment of interpoles. Consult the manufacturer of the
machine.

12. Overloads. Undercut the mica, use low-friction brushes and check
up all causes for short-circuit currents (see heading No. IV) to reduce
temperatures as much as possible.

13. Open circuit in armature coil. Rewind that part of the armature.

14. Loose end connection. Scrape and resolder all defective connections .

15. Worn bearings. Shim or renew the bearings.

16. Unequal air gaps. Shim the short poles or grind off the faces of
the long poles, or if from worn bearings, see paragraph above.

17. Short-circuit currents be'ween brush studs caused by unbalanced
armature winding. Consult the manufacturer of the irachine.

18. Eccentric commutator on high-speed machine. Turn or grind.

19. Poor belt lacing. Re-lace or still better use a continuous belt.

20. Pound of reciprocating engine driving the machine.

21. Unstable foundation.

22. Cross currents between generators operated in parallel driven by
reciprocating engines due to variation in angular speed of engines. Use
heavier flywheel.

II. CAUSES AND REMEDIES OF FLAT SPOTS ON
COMMUTATOR

1. Any form of sparking. See heading No. 1.

2. High bar. Tighten the commutator bolts and turn or grind the
commutator.

3. Low bar. Use commutator stone or turn or grind the commutator.

4. Eccentric commutator on high-speed machine causing the brush to
jump from the commutator at the high spots. Turn or grind the com-
mutator.

5. Surges of load current due to short-circuit on the line or an instanta-
neous high peak load.

6. Mechanically unbalanced armature. Place on balancing ways and
add weight at lightest point.

ADJUSTING BRUSHES AND CORRECTING TROUBLES 343

7. Difference in hardness of commutator bars. Undercut the mica and
use non-abrasive brushes.

8. Difference in hardness of mica. Undercut the mica and use non-
abrasive brushes.

III. CAUSES AND REMEDIES FOR BLACKENING OF
COMMUTATOR

1. Sparking. See heading No. I for causes and remedies.

2. Too much lubricant. Clean commutator with gasoline.

IV. CAUSES OF HEATING IN A MOTOR OR GENERATOR

WITH REMEDIES

1. Severe sparking. See heading No. I.

2. Short-circuit currents.
(a) Brushes off neutral.
(6) Faulty brush spacing.

(d) Unequal air gaps.

(e) Crooked brush studs.

(/) Too low contract drop of brushes.
(g) Unbalanced armature.
For remedies of items a to g see same causes under heading No. 1.

3. Too high or too low brush pressure. See paragraph 6 under heading
No. I.

4. High-friction brushes. Undercut mica and use low friction brush.

5. Commutator too small. Consult manufacturer of machine.

6. Too high a ratio of brush area to commutator surface. Use fewer
brushes of higher carrying capacity and lower friction.

7 Overloads. See paragraph 12 under heading No. I.
8. Chattering of brushes. See heading No. VII.

V. CAUSES AND REMEDIES FOR HONEY-COMBING OF

BRUSH FACES

1. Short-circuit currents. See paragraph 2 under heading No. IV.

2. Too low brush pressure. See paragraph 6 under heading No. I.

3. Brushes of insufficient carrying capacity. Consult a brush manu-
facturer.

VI. CAUSES AND REMEDIES FOR BRUSHES PICKING
UP COPPER

1. Heavy short-circuit currents. See paragraph 2 under heading No. IV.

2. Sand under the brush faces. Wipe brush face carefully after sand-
papermg either brushes or commutator.

3 Commutator not thoroughly cleaned after turning. Finish the surface
with a commutator stone after turning.

344 ARMATURE WINDING AND MOTOR REPAIR

4. Collection of copper dust by lubricant in abrasive brushes. Undercut
the mica and use non-abrasive brushes.

5. Electrolytic action. Change the grade of brush ; better consult a
brush manufacturer.

VII. CAUSES AND REMEDIES FOR BRUSHES CHATTERING

1. High-friction brushes. Change the grade or pressure.

2. Rough commutator. Use a commutator stone.

3. Dirty commutator. Clean with gasoline.

4. High mica. Use a commutator stone or undercut the mica.

5. Wide slots with thin brushes. Fill the slots with commutator
cement.

6. High bars. Tighten the commutator bolts and turn or grind the
commutator.

7. Flat spots. Use commutator stone unless the flat spots are too
large for stoning, in which case turn or grind the commutator.

8. Brush operating at the wrong bevel, frequently found where brushes are
operating in a stubbing position with angles of less than 20 degrees. Change
the grade of brush or angle of operation. Better consult a brush manu-
facturer or the manufacturer of the machine.

VIIL CAUSES AND REMEDIES FOR LOOSENING OF
BRUSH SHUNTS

1. Poor workmanship in attaching shunts.

2. Insufficient carrying capacity. Consult a brush manufacturer.

3. Heating. See heading No. IV.

4. Vibration. See heading No. VII.

5. Combination of heating and vibration.

6. Loose terminal screws causing unequal distribution of load.

7. Unequal brush pressure causing unequal distribution of load. See
that all brush pressures are uniform and conform to recommendations
given in paragraph 6 under heading No. I.

8. Heavy short-circuit currents between different brushes. See b, d, e, g,
of paragraph 2, under heading No. IV.

CHAPTER XIV

INSPECTION AND REPAIR OF MOTOR STARTERS,
MOTORS AND GENERATORS

On account of the variety of troubles that may be corrected
in a simple way at one time but require more extensive repairs
at others, it is a very difficult matter to lay down hard and fast
rules that will always work out on every repair job when fol-
lowed to the letter. The ability to know when temporary
repairs will suffice and when a permanent job must be done
at once, comes largely from experience. The term repairman
has been used throughout this book as a title that a good engi-
neer can bear with pride when he measures up to all the quali-
fications of the man who in the majority of cases knows what
to do and just how to do it and seldom guesses without a good
percentage of the probabilities of being right in his favor.
The main difference between the designer and the repairman
is that the former must know what to do while the latter must
know -how to do it. A capable repairman combines both
qualifications through years of experience. One of the best
ways of saving time for the young engineer entering the repair
field, is to serve as an apprentice with a large electrical manu-
facturing company. In this way through association with
those who design and those who build and test, a great deal
of information is absorbed that only years of experience from
one job at a time will make possible. This fund of information
is essential in an intelligent discussion of any repair job or the
presentation of rules or suggestions for looking for the trouble
and then actually making the repair.

In the accompanying pages there are presented suggestions
for the repairman in the inspection and repair of motors and
generators prepared by H. S. Rich, at the suggestion of the
author and published in the Electrical Record, October, 1918
to April, 1919. These suggestions are accompanied with ex-
ploded views of the device discussed, so that the repairman can

345

346 ARMATURE WINDING AND MOTOR REPAIR

form a mental picture of the work described which will approxi-
mate as closely as possible the impression that would be se-
cured by actually doing the work.

Cost of Repairs for Polyphase Motors. For the purpose
of furnishing estimates for repairing polyphase motors that
will guide the repairman who has had a limited experience of
repairwork, George A. Schneider has compiled the accompany-
ing table (Journal of Electricity, May 1, 1917) from costs of
actual repair jobs which he has handled. The table as pre-
sented has been revised to take into consideration the cost
of materials entering into repairs for the year 1917. The
data refers particularly to 60-cycle, two and three-phase,
squirrel-cage motors wound for the standard voltages of 110
to 550.

For most of the sizes listed the costs were arrived at by
taking the average cost of repairs for a given frame and then
applying this cost to the various ratings built in that frame.
This will be apparent by comparing the costs for the different
ratings. Take for example, frame G. The cost of rewinding
the stator is $34.75. This figure has been applied to the
following ratings all of which are built in that frame : 1 horse-
power, 900 revolutions per minute; 1.5 horsepower, 1200 revo-
lutions per minute, and 3 horsepower, 1800 revolutions per
minute. The frame sizes specified do not apply to any par-
ticular line of motors, but were arbitrarily chosen for the
purpose of this table. However, the relative output of a
given frame at the different speeds will be found to agree quite
closely with several lines of induction motors on the market.

These estimates may also be used equally well for motors of
other frequencies by taking the figures applying to a 60-cycle
rating built in the same frame. This comparison can be
easily made by referring to the manufacturer's rating and
dimension sheets for thaf particular line of motors. The
tables may be further applied to slip-ring or phase-wound
motors, since the cost of rewinding the rotor of such a machine
will not differ materially from the cost of rewinding its stator.
On this basis the cost of completely rewinding a 10 horsepower,
1800 revolutions per minute slip-ring motor built in frame /
will be $119, or $59.50 for the rotor or stator separately.

INSPECTION OF MOTORS AND GENERATORS 347

COST OF REPAIRS FOR GO-CYCLE POLYPHASE MOTORS BASED ON A LARGE
NUMBER OF REPAIR JOBS

Horse-
power

Syn-
chronous
speed in
rpm.

Frame
size

Rewinding
stator

Re-
soldering
rotor

Bearing
linings,
per set of
two

Paint-
ing

Re-
crating

0.50

1200

$26.25

$2.50

$1.35

$1.00

0.50

1800

24.25

2.25

.35

1.00

.00

0.75

1200

28.00

3.00

.85

.00

0.75

1800 ,

24.25

2.25

.35

.00

1.00

900 '

34.75

4.00

.10

.50

1.00

1200

28.50

3.00

.85

.25

.00

1.00

1800

26.25

2.50

.35

.00

1.50

1200

34.75

4.00

3.10

.50

1.50

1800

28.00

3.00

1.85

.00

2.00

1200

34.75

4.00

3.10

.50

2.00

1800

28.50

3.00

1.85

.25

.00

3.00

900

53.50

6.50

5.25

.50

3.00

1200

48.50

6.75

3.55

.50

3.00

1800

34.75

4.00

3.10

.50

5.00

900

73.75

8.75

8.05

.75

2.00

5.00

1200

53.50

6.50

5.25

.50

1.50

5.00

1800

48.50

4.75

3.55

.50

1.50

7.50

900

70.75

12.00

7.85

.00

2.50

7.50

1200

59.50

7.00

6.60

.75

2.00

7.50

1800

53 . 50

6.50

5.25

.50

1.50

10.00

900

75.00

13.25

7.85

2.00

2.50

10.00

1200

70.75

12.00

7.85

2.00

2.50

10.00

480

59 . 50

7.00

6.60

1.75

2.00

15.00

720

93.75

15.50

10.25

3.00

4.00

15.00

900

71.25

14.25

10.25

3.00

4.00

15.00

1200

75.00

13.25

7.85

2.00

2.50

15.00

1800

73.75

8.75

8.05

1.75

2.00

20.00

600

156.25

19.00

12.10

3.25

6.00

20.00

900

93.75

15.50

10.25

3.00

4.00

20.00

1200

71.25

14.25

10.25

3.00

4.00

20.00

1800

75.00

13.25

7.85

2.00

2.50

25.00

600

156.25

19.00

12.10

3.25

6.00

25.00

720

156.25

19.00

12.10

3.25

6.00

25.00

900

143.75

17.75

12.00

3.25

6.00

25.00

1200

93.75

15.50

10.25

3.00

4.00

35.00

600

18.7 . 50

20.50

19.95

3.50

6.25

35.00

720

156.25

19. CO

12.10

3.25

6.00

35.00

900

156.25

19.00

12.10

3.25

6.00

35.00

1200

143.75

17.75

12.00

3.25

6.00

50.00
50.00

600
720

V
V

218.75
218.75

21.75
21.75

30.85
30.85

3.50
3.50

6.25
6.25

50.00

900

187.50

20.50

19.95

3.50

6.25

50.00

1200

156.25

19.00

12.10

3.25

6.00

Points to Consider when Estimating Cost of a Motor Repair
Job. The estimates for rewinding the stator or resoldering
the rotor given in the accompanying table do not include any

348 ARMATURE WINDING AND MOTOR REPAIR

preliminary work required to put the stator structure in fit
condition to receive the new winding or work required on the
rotor before the actual resoldering can be started. In other
words, the figures cover only the actual rewinding or resolder-
ing, as the case may be. However, this preliminary work is
frequently necessary and must always be considered in making
up estimates. It is due to a number of causes.

For example, the motor bearing linings may have worn
down sufficiently to allow the rotor to rub against the stator.
If the motor has operated very long in this condition the lami-
nations of either or both stator and rotor will probably be
damaged, which may require considerable work to put them
into their original condition. Again, a defective or broken
bearing may injure the shaft. Sometimes this damage will be
serious enough to require a new shaft. New bearing linings will
probably be required in either case. Burned-out windings
may also be accompanied by fusing of parts of the stator lamina-
tions. These fused portions must necessarily be removed
before actual replacement of the coils can be commenced.

In a rotor which has been badly overheated, allowing the
melted solder to be thrown out, arcing is frequently set up
between the rotor bars and the end rings causing serious
burning. When this occurs, new end rings are often needed
either for one or both ends of the rotor or perhaps part of the
bars will have to be replaced. With bolted end ring construc-
tion there is also liability of trouble. The expansion of the
end rings caused by the excessive heat, tends to snap the bolts
between the rotor bars and the rings, producing the most
favorable conditions for arcing. Burnouts of this kind,
for either soldered or bolted construction, are quite common
in connection with motors which have been started from
time to time under loads which have required heavy starting
torque with long periods of acceleration. Two- or three-
phase motors allowed to operate single-phase for a considerable
length of time may also develop troubles of this kind. Very
often the rotor will be badly damaged while the stator has
only been only slightly overheated. Conversely, in some
cases the stator will be burned out while the rotor is uninjured.

From these points it will be clear that estimates for repairing

INSPECTION OF MOTORS AND GENERATORS 349

motors should not be made until after the motor has been
given a careful inspection for otherwise there is liable to be a
wide discrepancy between the estimated and the actual cost
of making the repair. In furnishing an estimate under
conditions where a detailed inspection is not possible, details
of what the estimate covers should be given with a notation
of additional repairs that may be necessary after an inspection.
The data of the table on page 347, is based on a large number
of repair jobs -and is conservative for labor and material
conditions in 1917.

I. INSPECTION AND OVERHAULING OF DIRECT -CURRENT
MOTOR STARTERS

The following points to be considered by the repairman
when inspecting and repairing a motor starter are given by
T. H. Reardon (Electrical Review, May 25, 1918).

FIG. 240. A starting rheostat with no voltage release for use with shunt and
compound direct-current motors.

In direct-current motor service the ordinary rheostat
in which an arm moves over contacts arranged on the arc of a
circle is the most convenient starting device. All such rheo-
stats are provided with an electromagnet which holds the
arm in place after it has been brought up to the last notch
or full running position. This magnet loses its power if the

350 ARMATURE WINDING AND MOTOR REPAIR

current goes off the line and the arm flies back to the starting
position. The arm is returned to the starting position by a
spring, which is always acting in opposition to the pull of
the magnet that tends to keep the arm in last-notch position
as long as the magnet is energized. If the pull of the spring
is too great, which somtimes occurs when an inexperienced
man makes certain changes in the way of correcting things,
the pull of the magnet will not be sufficient to hold the arm
in place when it is brought up and the arm will fly back.
Conversely, if the i^rength of the spring becomes reduced,
which often happens when motors are placed in damp places,
or worse still, where dampness and certain bleaching agents
(such as chlorine and oxides of chlorine) act jointly in bringing
about metallic deterioration in apparatus, steel springs will
be found corroded to such an extent that they do not possess
their original strength and elasticity. As* a preventive
measure, springs as well as all other metal parts exposed to
corrosive action should receive a drop or two of a mineral
oil or be slightly smeared with vaseline occasionally to prevent
such deterioration. The pull of the spring and the pull of
the magnet are balanced against each other, but there are
other conditions that must be taken into account in making
adjustments.

After the current ceases to flow in the low-voltage release-
magnet coils, which will not be until the motor comes to
rest (the motor acting as a generator will maintain this current
until the motor stops), the magnet will still exert a considerable
pull on its armature due to the relatively large amount of
residual magnetism that remains in the core of an electro-
magnet after the current has ceased to circulate and before the
armature is pulled away from the magnet poles. An arm
that will not fly back to starting point when the motor stops,
if moved back by hand and then brought up again, will not
stick. In certain cases, however, it may stick and when it does
do so it is not due to any pull exerted by the magnet but it will
probably be found that the sliding contact on the bottom of the
arm bears so hard against the contacts that the strength
of the spring is not sufficient to overcome this braking effect
due to the friction of the shoe moving over the contacts.

INSPECTION OF MOTORS AND GENERATORS 351

Sometimes this friction effect may exist only on one contact,
one contact standing higher than the rest, and if this particular
contact happens to be next to the magnet, the movement
of the arm will be retarded before the arm has a chance to
acquire any momentum that would carry it back provided
it once got started.

If the magnet pull due to residual magnetism after current
circulation ceases should be responsible for the arm sticking
after the motor stops (this will very rarely be the case, how-
ever), the keeper can be lightly tinned over with a soldering
iron, thus placing a certain amount of reluctance in the mag-
netic circuit of the magnet and its keeper. A drop of oil
occasionally on the stud on which the arm is pivoted will
help in securing free movement.

II. INSPECTION AND OVERHAULING OF AUTO-STARTERS FOR
A.-C. MOTORS

When examining and testing an auto-starter to locate
troubles, T. H. Readron (Electrical Review, May 25, 1918)
mentions the following points as possible causes of the troubles :

Auto-starters are generally used for starting alternating-
current motors of the induction type when the motors are
above five horsepower. Small motors are usually thrown di-
rectly on the line without any starting device and, although
the starting current is four or five times greater than the
normal current, the fluctuation, as a rule, is not serious enough
to necessitate the use of an auto-starter. Usual practice is as
follows: Polyphase induction motors up to 5 hp. are thrown
directly on the line. Motors of 7.5 to 30 hp. are started
by means of a star-delta switch which is not an induction
starter but simply a switch operating in oil and changing the
connections of the motor from star on starting to delta on
running. The motor in that case is provided with six termi-
nals to accommodate the change. Auto-starters or auto-
transformers are mostly used for squirrel-cage induction
motors of considerable size or over 30 hp.

The auto-starter being adapted for alternating-current
work, differs from the rheostat in that it possesses not only

352

ARMATURE WINDING AND MOTOR REPAIR

some resistance but considerable reactance in damping
current flow, while the rheostat possesses resistance only.
Moreover, the rheostat is a step-by-step starting device;
its arm is to be moved slowly over the contacts, while the
auto-starter has but one step from starting to running and
the handle should always be thrown promptly from one posi-
tion to the other position.

The accompanying diagram (Fig.
242) shows the plan of the auto-starter
as usually constructed. Six wires are
brought to the rocker cylinder of the
switch, which is moved by the switch
handle. One wire is a feed or line
wire, the wire next goes to the motor,
the next wire is line and the next one
motor again, etc. When the switch
is thrown to the starting position, the
six contacts on the switch cylinder
meet six contacts on the starting
block and the current from a line
contact on the switch cylinder passes
to the contact on the starting block
through one coil of the reactance and
back to the next contact on the switch
cylinder, and thence to the motor.

When the handle is thrown to the
running position, only three contacts
on the switch cylinder make contact

these are the motor contacts on the switch cylinder and they
meet three contacts on the running block, which contacts are
directly connected to the line wires either through fuses or
overload relays.

The line contacts on the switch cylinder, when the cylinder
is thrown to running position, do not connect with anything
they stand clear or dead-ended. These contacts on the switch
cylinder can be identified by using a lamp bank and making
contact with them, the switch handle being in off position.
A light will be obtained between line and line.

The first trouble to look for is burned or imperfect contacts.

FIG. 241. An auto-
starter or compensator for
use with alternating-current
squirrel cage induction
motors.

INSPECTION OF MOTORS AND GENERATORS 353

After taking the oil pan off the switch contacts can be inspected
without any trouble.

If the contacts are burned or rough, they should be taken
off and filed smooth or replaced with new ones. The same
applies to the fingers that meet the contacts. The handle
should be thrown to one position and then the fingers should
be tried to see that they press firmly and evenly against the
contacts on the switch cylinder. If they do not, throw the
switch handle to off position and they can easily be bent in-
ward sufficiently to make a firm contact. It will be well not
to bend them too much at once, for if this is done, they will
not slide over the switch cylinder contacts properly.

I |l^Finish

Side View ^Start Y- Connected

FIG. 242. Connections for a common form of auto-starter for alternating-
current motors.

In regard to broken wires, the wires that are attached to the
switch cylinder, six in number, are bent slightly every time
that the switch handle is moved. These wires now and then
break off. The other six wires that enter the auto-starter
go to fixed immovable contacts and rarely cause trouble.
To save labor, test everything out as far as possible with the
lamp bank or some other way equally good, depending upon
what kind of testing apparatus there is at hand.

It will be well to bear in mind that a ring obtained with a
magneto or a light obtained with a lamp bank is not always
conclusive. Instances are quite common where a wire 01
cable breaks inside of the insulation and there afterward re-
mains a sheath of metallic oxide or smudge that will pass cur-

354 ARMATURE WINDING AND MOTOR REPAIR

rent enough to give a dull light on a lamp bank or a feeble
ring on the magneto. When it is decided that a certain wire
is broken, take out the two screws that hold that particular
contact to the switch cylinder and pull down the terminal
with a pair of pliers.

If the wire is intact, it will resist a strong pull. If the wire
is broken and is held by the insulation, it will easily pull
apart. If it pulls down through the switch cylinder suffi-
ciently, attach a piece of new wire or cable to it and use it
for a snake to draw the new wire into place, pulling at the top
where the wires enter the auto-starter. If this does not work,
draw the broken wire out at the top at all events and get a
piece of stiff brass or steel wire, about No. 14 gauge. Bend a
smooth small loop on the end of it and pass it down from
above. Have a helper with a small hook wire watch for it and
hook it above the switch cylinder so that it can be passed
through the proper hole in the switch cylinder. When this is
done, the terminal can be soldered on at the lower wire and the
terminal drawn back and wire spliced at the stop of the auto-
starter. It should always be the aim to make such a repair
if possible without disassembling the parts of the auto-starter
as such parts go back with difficulty.

It may be advisable or even necessary to use a wire slightly
smaller than the original one in order to get through tight
places. There will be no decided objection to doing this as the
cross-section of copper is always sufficiently ample to justify
a slight reduction when such a reduction in size is really
necessary.

The following method of insulating leads on auto-starters
will prevent siphoning of oil:

(a) Remove insulation from each lead just above the highest oil level
for a distance of two inches.

(6) Sweat the strands of the cable thoroughly together so as to close
up all spaces between the conductors for a distance of one inch.

(c) Insulate the lead with treated cloth tape, wrapping the tape tightly
around the conductor and brushing each layer with insulating varnish
while wrapping.

(d) Extend the wrapping to tape with three overlapping layers at
least one inch on the insulation at each end of the bare section.

INSPECTION OF MOTORS AND GENERATORS

355

III. INSPECTION AND OVERHAULING OF DRUM TYPE
CONTROLLERS

The drum type controller is used with variable speed shop
motors, on trolley cars, elevated, subway and railway trains,
on cranes, and on some types of elevators. This controller
may be mounted in any position so as to permit handy con-
trol. On trolley and railway cars, it is placed vertically with
the lever on top. For shop use it is sometimes mounted on a
machine or on the wall upside down, being well out of the way
and yet having its lever within reach. On some elevators it is

2
RESISTANCE. 6

CONTROLLER.

FIG. 243. Gridiron resistance used with drum type controller shown at
the right for variable speed motors. (Figures refer to numbers of para-
graphs of text.)

installed horizontally and operated with a steel cable running
over a grooved sheave wheel. Some drum type controllers
have only a single arrangement of segments with the reversing
connections operated by a separate lever, while others have
duplicate segments, half of which are arranged to operate on
reverse.

1*. The hard usage to which a controller is subjected usually
will loosen some of its screws or connections, so these should
be looked over often and kept tight. The line connections

* The following paragraph numbers refer to the parts in Figs. 243 and
244.

356 ARMATURE WINDING AND MOTOR REPAIR

to the fingers and the segments themselves require the most
attention. A finger generally carries a copper block of various
bevelled shapes, riveted to a stiff phosphor bronze strip which
provides the necessary tension and is regulated by an adjusting
screw. The fingers should be provided with flexible stranded
copper pig-tails which carry the current. If the spring carries
the current, it is liable to heat and loose its tension. Set the
fingers so that they can not be jammed under a segment instead
of riding upon it. If any are badly worn renew them, like-
wise overheated springs and defective pig-tails.

2. The primary segments for induction motor control and
the field segments for direct-current motors do not carry as
much current as the armature or slip ring segments and are
thus subject to less wear. The ones mostly worn are in series
with the slip rings or the direct-current brushes, thus if con-
tacts are very poor the motor will not operate or it may run
slowly if on alternating current. If the copper segments are
badly worn, replace them with new ones having the proper
curvature and with counter sunk screws tightly fastened down.
A slight lubrication of vaseline is good for segments.

3. The direct-current controller should supply the motor
shunt field with current on the first point, likewise the alter-
nating-current controller should close the circuit to the induc-
tion motor primary winding either before or at the same instant
that the armature receives current. If these contacts are not
properly made at the first point, fuses are liable to blow as the
result of no field.

4. Insulating partitions are often provided between the
segments to prevent arcing over. If these are badly burned,
renew them.

5. Controller diagram connections which are generally
pasted inside of the cover are worth saving. If constant arcing
is liable to deface the diagram, either provide a duplicate for
reserve or remove the one from the cover and mark on it the
controller number, etc., for identification.

6. Test with a magneto on a dead controller, or with a test
lamp on a live one for any possible stray ground to the casing.
It should test free, as a ground would likely bother the operator.

7. The controller resistance is usually of the gridiron type

INSPECTION OF MOTORS AND GENERATORS

357

which allows any speed continuously. This resistance has no
connection with the primary winding or line segments. See
that no grids are broken, as they are quite frail. A few extras
in stock are handy. Test the resistance for a ground to its
own frame. A perforated sheet iron hood should be used for
its protection.

8. The conductors from the controller may be carried in
conduit to the motor, and the mo'tor, conduit and controller

Forward Reverse

12345678 57 6W I

" H) Connections of drake
Coils when Used. If Brake

Series.

CONTROLLER CONNECTIONS.

FIG. 244. Wiring connections for a drum type controller.

should be well connected to a good ground pipe to protect the
operator against sneak currents.

9. The different points or stops on the controller are deter-
mined by a notched star wheel just underneath the top cover.
If its spring is broken or very weak, provide the good tension
needed for positive operation, as placing the drum between
contacts will cause trouble.

.10. If the motor fails to run when the controller is on
the second or third point, shut it off and test for current at the
line contacts with a test lamp. Further testing of the motor

. -.

358 ARMATURE WINDING AND MOTOR REPAIR

primary connections at the controller contacts may reveal an
open circuit. If the motor still refuses to run, test the resist-
ance for open circuit and look carefully over the slip-ring
brushes for poor contacts. Exposed wiring between the motor,
resistance and controller may get broken and thus stop its
operation One open slip ring connection will cause slow
speed.

11. The sudden and repeated Operations to which a con-
troller is subjected 'demands that it be tightly bolted down.
For this purpose cast lugs are provided to take the bolts.

12. To keep dust from getting in and sparks from getting
out, the front case should be kept fastened on by the finger
nuts provided on both sides.

13. The control of the speed is accomplished by means of
the stepped segments which cut out the resistance gradually.

14. The ON and OFF positions should be plainly marked
for the operator's guidance.

15. See that a limit stop is provided either in the shape of a
cast projection or an extra long tooth on the star wheel, so that
the segments can not be rotated beyond a whole circle.

16. It is advisable to provide a circuit breaker in the line
with the controller so that in case it is left on part speed with a
dead line, no damage can be done to either the motor or its
driven machine when the current returns.

17. One phase connection open will prevent the motor
from starting, and it may smoke while trying to start.

IV. OVERHAULING A LARGE COMPOUND D.-C. MOTOR

All motors should be inspected and given a general over-
hauling at least once a year. Large sizes should not be slighted
in any detail just because they look rugged and appear to stand
everlasting work. Close observation of all parts and con-
nections will often reveal some surprising defects. However
small, they should not be allowed to go unattended.

When repairing a motor, it is advisable to remove all fuses
and put them into the tool kit or your pocket so that by no
chance current can be switched onto the motor when not
expected.

INSPECTION OF MOTORS AND GENERATORS

359

In case of a belt drive, shift the motor, slip off the belt and
tie the latter up to something near by. If direct-connected,
open the coupling by removing all the bolts. Replace the
nuts on them and tie the whole bunch together to save losing
them.

If the motor is suspended, it would be well to lower it with
two cham tackles and disassemble it on the floor, as its parts
are too heavy to carry down a ladder and rather large to step
over on a scaffold. If on a shelf or platform, slinging it to
the floor is often not very difficult.

FIG. 245. Essential parts of a compound direct-current motor. (Figures
refer to number of paragraphs of text.)

1 *. Remove the key and pulley, first marking or measuring
the latter's location,- which will help in replacing it. If
direct-connected, remove the commutator end shield to allow
the armature shaft to draw open the coupling. Mark location
of the coupling on the shaft. If the key is very tight, remove
the set screws and pour kerosene into the holes, which will
work around the key, then some motor oil may be dropped
in to lubricate the key-way. By lightly tapping proper shaped
drifts, the key can often be started in a few minutes without
a lot of battering and damage. When the key is removed if
the pulley is tight on the shaft, more kerosene will help.
Also by using an 18-inch monkey wrench on the rim, it

* The following paragraph numbers refer to the parts shown in Fig. 245.

360 ARMATURE WINDING AND MOTOR REPAIR

can often be started. Do not hammer on the rim, it may
crack. If made of paper it will flatten. Strike a hard wood
block butted against the hub or drive thin wedges behind. Tie
the key and set screws to the pulley when removed for security.

2. Before removing commutator end shield, mark the
position of the rocker arm which may get shifted.

3. With the pulley removed, try to lift the shaft a trifle
in the bearing lining to learn if it is worn out of true. A year's
wear may cause the armature to strike the lower pole shoe.
It is well to observe this before disassembling.

4. Mark or tag the armature leads connected to the brush
rigging, then disconnect and remove the brushes.

5. Drain the oil wells and remove both end housings also
the brush rigging. Close up the oil wells before the screw
plugs get mislaid. Be careful of the glass gauges.

6. Remove the armature to a pile of blocking straight
ahead. This can be done with one lift by slipping a large
iron pipe over the pulley shaft, which will carry the armature
clear through and two feet beyond without dropping it.
A suspended motor can be disassembled to best advantage
on the floor. Set the armature on some burlap and not
where any metal clips will imbed themselves into the insula-
tion. Clean the armature thoroughly, especially the air-
ducts.

7. If the commutator has enough stock left and is rough or
grooved, turn it down in a lathe, taking only very fine cuts
with a diamond pointed tool. Then polish with oil and No. 00
sandpaper, but not emery. If the" commutator is badly
worn down, a new one should be on hand and replaced, having
the same number of segments.

8. Blow the dust out of the field frame and from around the
field coils, or scrape with a stick but use no metal bar or knife.
Wipe out clean with waste and some benzine. Examine the
insulation of the field coils, shunt and series field leads, and if
any abrasion is found use tape, thick armature varnish or
shellac for insulation. If the field coils are loose, tighten
the bolts usually found outside the frame. If a coil is burned
out, remove and re- wind Weigh the wire that is stripped off
and replace the same size and weight carefully wound on and

INSPECTION OF MOTORS AND GENERATORS 361

well insulated. In replacing a field coil, its shunt or series
polarity are liable to be uncertain. To determine this,
properly connect to the field windings three or four cells of
dry battery and with a magnetic compass explore the polarity
of each pole, marking it with chalk for reference. For both
the series and the shunt windings, the two markings should be
similar on any one pole, then the fields coincide. The newly
wound coil should have an opposite sign to the coil on either
side; if not, change its connections and test repeatedly to
make sure, for the operation of the motor depends on the
field polarities. Use armature varnish or shellac on all
coils and leads.

9. Do not paint the inner surface of pole shoes.

10. If the commutator leads are loose solder solidly and
neatly, being careful not to let hot solder drop down
behind the commutator as it is liable to cause a short circuit.

11. If the armature has been striking the pole shoes, the
banding wires will likely be polished in spots and partly worn
through. Even if they are only loose, repair them now by
slipping under plenty of shellaced mica.

12. If any armature coil is burned out, slip in a new one and
thus put it in good shape with full power.

13. Caliper the armature shaft at both ends. If badly
worn or cut a new one should be fitted in so that new bearing
linings will fit perfectly.

14. If the bearing linings are worn enough to allow loose
play of the shaft, put new ones in the end shields. Extra
linings should be kept in stock the year round as it will save
hours delay when one wears out suddenly from poor oil
feed. Remove the set screw and strike against a hard wood
block on the outer end of the lining but avoid damaging the oil
rings. Usually each end has a different sized lining so that the
proper pair are needed. If the oil rings get damaged true
them at once. Try the linings on the shaft to see if they fit
the bore. If linings are split they are easily removed. Fit
the new ones exactly into place under their set screws no
matter how long it takes, for if a lining ever turns over out
of place it will run hot and be ruined. If the old linings are
re-babbitted on the premises, have oil grooves and channels

362 ARMATURE WINDING AND MOTOR REPAIR

cut in for lubrication and scrape the insides for a perfectly
snug fit.

15. Flush out the oil wells with gasoline or benzine and wipe
them out dry with cheese cloth, as waste is liable to leave
threads to entangle the oil rings.

16. Clean the brush rigging especially around the insulators
on the holder bars. Renew short brushes and bevel them
as near as possible. Renew broken or weak tension springs.
See that the holder bars are bolted tightly or they may turn.
Clean out the brush holders and allow the brushes free move-
ment, with the pig-tails fastened tightly for good current
connection.

17. Now replace the brush rigging, put in the armature and
put on both end shields. Connect the leads to the brushes.
Set the rocker arm in its original neutral position and oil the
bearings enough to allow the shaft to be rotated by hand. If
it binds, look for the cause at once, as it should turn easily
with the oil rings in their grooves. Turn the armature by
using a monkey wrench on the shaft pushing against the key
covered with tape to avoid marring.

18. Set the brush holders one-eighth inch above the com-
mutator. Bevel all new and old brushes by drawing under
them strips of No. 1 sandpaper until a good and full contact
surface is assured. Have the tension on all springs similar,
but not too strong. Smear the commutator with a little
vaseline; it will not carbonize like oil. Replace the pulley
and key and then rotate the armature and see that no pig-tails
interfere.

19. Replace the motor and bolt it down firmly. Oil the
bearings fully and run the armature with current for a few
minutes with the belt off. Look sharply for trouble especially
see that the oil rings are lubricated properly.

20. Slip on the belt, tighten it up, and run motor under some
load. If direct-connected have the coupling well insulated
and securely bolted. Look especially for sparking as the
brushes may have imperfect contact, too light spring tension
or be off neutral. Watch the motor for a few hours underload.
Motor brushes should be set back of neutral point, or given a
"lag" as it is termed.

INSPECTION OF MOTORS AND GENERATORS 363

21. If a compound motor persists in sparking badly under
nearly a full load, it may be that during the repairing process
the shunt and series fields were connected opposed to each
other, whereas their polarities should be similar at each pole.
Opposed fields really cause a still weaker field at increased
loads and will result in sparking and flashing, but with the
series .and shunt fields magnetized in harmony there should be
no sparking at the brushes if they have good full contacts and
are neutrally placed.

22. If the direction of the armature is wrong, reverse the
brush leads.

V. OVERHAULING A 60-HORSEPOWER INDUCTION MOTOR

Although an induction motor does not frequently call for a
complete overhauling, the details given in what follows cover
such a case in order to bring out the point which should be
considered when the motor is completely taken apart and then
reassembled after repairs.

1*. Remove all fuses and kill the line.

2. Slip off the belt and try turning the rotor by hand. If
it will not turn, then one trouble has been located.

3. Remove the key and pulley by driving wedges behind the
latter.

4. Drain the oil wells and remove both end shields which in
this size of motor are probably in half sections.

5. Remove the rotor straight out ahead onto a pile of block-
ing. Slip a pipe over the short shaft so that the rotor may be
carried through with one lift. If it seems to be polished on
its periphery, this is evidence that at least one bearing lining
is worn badly and has allowed the rotor to drop. Clean the
rotor of all dust and dirt.

6. Clean the primary winding with waste and benzine and
then look for any possible abrasion on the winding. If nec-
essary apply some thick shellac. Coat all the winding with
armature varnish, which will dry in a short time.

7. Remove both bearing linings if they are worn enough to
allow loose play on the shaft. If they are split they may be

* The following paragraph numbers refer to the parts shown in Fig. 246.

364 ARMATURE WINDING AND MOTOR REPAIR

re-babbitted on the premises, but must fit snugly and have
oil channels properly cut in them.

8. Clean out both oil wells as dirt is liable to settle and stop
the rings.

9. Replace the rotor and oil the bearings. Try to turn the
former by hand, and see that the oil rings turn freely.

10. If the motor seems to be in good shape put on the pulley
and key it securely.

Relay Plunger -

'SetScre*-

White Mark, ....

Calibrating Point.

Piston Rod

Overload
Calibrating
"-Scale ^

Dash Pot
OVERLOAD RELAY

ROTOR

FIG. 246. Squirrel cage induction motor installation showing arrange-
ment of starter and type of relay that can be used instead of fuses. (Figures
refer to numbers of paragraphs of text.)

11. Open the compensator case and look for trouble. Most
compensator cases after being unbolted should be let down care-
fully as they contain oil in which the contacts are submerged.

12. Pull the operating handle over to the starting side and
see that the contact fingers are making good connections.
If any are burned short renew them. If their cables or leads
are not all tight, one or more may be found which has the con-
ductor broken inside of the insulation. These must be re-

INSPECTION -OF MOTORS AND GENERATORS 365

connected solidly. Now throw the operating handle sharply
to the running side and examine the other contact fingers and
their connections. After all repairs are made to the fingers
and contacts replace the oil case and put in the fuses for a trial
spin of the motor. It ought to run nicely now.

13. Try the no- voltage release by opening the switch. If
it fails to work instantly see if the solenoid core is free to move.
If dirt is obstructing it or if the rod is bent, remedy it and
adjust the stop nut for a certain distance. See that the
solenoid leads are properly connected to one phase of the
motor line behind the compensator. This circuit energizes
the no-voltage release coil.

14. See that the fuses fit snugly in the clips, as a loose fit
is liable to disconnect one phase. See that the line wires are
tightly connected to the cutout and switch. Loose connec-
tions on a 50-hp. line are liable to heat very quickly. The
fuses for a 50-hp. compensating starter ought to be not over
150-amp. size with knife-blade clips. This allows for about
20 per cent, overload at starting. If there is much load on
the motor when being started, the fuses may be increased to
175 amp., but if they are too large there is no overload protec-
tion afforded to the motor.

15. For a motor of this size, it is worth the expense of
installing overload relays in place of fuses, as they can be
time-set for overloads of dangerous duration. The small
spring contact on top of each relay is connected in series with
the no-voltage release at the side of the compensator and also
in series with one phase from the motor. Thus an overload
on the motor raises the relay plunger which opens one or both
of the top contacts, which in turn kills the no-voltage release
solenoid and by its gravity drop the controlling handle is
tripped to dead center thus cutting off the power. The coil
of each overload relay is connected in series with each phase
of the motor so that both are protected. The oil dash pot
needs to be filled with relay oil and screwed up tightly.
The plunger may be set at any amperage and locked with a set
screw. In the piston cup is a plate having a few holes through
which the oil is forced when the plunger raises. This plate
may be set for any time, either slow or fast. It is well not

366 ARMATURE WINDING AND MOTOR REPAIR

to have the plunger work too quickly as an induction motor
may be called upon to carry a heavy load for about a minute,
when it would be bothersome to have the relays kick. But
more than a minute allows a large current to heat up the
motor. These relays are trustworthy, and when encased in a
steel box they work silently and promptly just when con-
ditions demand their action. When they kick out and the
power goes off, the relays will reset themselves, but they will
act again at the very next overload.

After all the operations outlined in items 1 to 15 have been
carefully completed, the job may be considered finished
and the motor ready to run.

VI. OVERHAULING A 26-HORSEPOWER SLIP-RING MOTOR

A machine like a bull-dozer in a forging shop, a crane or a
dredge is likely to have a motor of the slip-ring type with an
external resistance control. Such a motor draws current from
the line with no heavy rushes and will start with not more than
one and one-quarter times its normal full load demand. This
is a good feature as it is easy on the generator, fuses and line
voltage. Otherwise there might be the annoyance of stopping
and starting every few minutes. These motors, however,
need repairing at times like all other motors and the following
procedure is given for such repair :

1*. Assume a case that upon removing the belt and trying
to turn the rotor by power, it is found that it runs very slowly,
suddenly blowing a fuse or the overload release opens the
line and the motor stops.

2. Remove all fuses, drain the oil wells and remove both end
shields.

3. Remove the bearing linings from the end shields and
try them on the shaft. If they are worn loose renew or re-
babbitt them to fit snugly, and cut in oil grooves.

4. Remove the armature and rest the shaft on some block-
ing. Examine the slip rings and if they are worn down thin
or rough, either renew them or turn them down very smoothly
in a lathe. They are liable to roughen from poor brush

* The following paragraph numbers refer to the parts shown in Fig. 247.

INSPECTION OF MOTORS AND GENERATORS 367

contact. If renewed see that they are properly insulated
from each other with one-quarter-inch fiber board, and tightly
bolted up.

5. See that the band wires are tight on the armature. If
they are worn thin renew and solder them all around.

6. Coat all the armature winding with a special black
varnish, avoiding the core disks, for magnetic conduction.

BRU5H RIGGING.

25-HP. SLIP RING MOTOR.

( Leads -to Armature
'(Winding Secondary.

WIRING.

FIG. 247. Parts and external resistance control for a slip ring alternating-
current motor. (Figures refer to numbers of paragraphs of text.)

7. Clean all the dust and dirt from out of the primary
winding, but use no sharp metal as the insulation is easily
damaged. Wash with benzine, and coat with armature
varnish.

8. Clean the brush rigging of all metal dust and gummed oil.
Renew the short brushes and the broken and weak tension
springs. All brushes should have good sized pig-tails to carry
the current. The brush contacts should be perfect on the ring
surfaces.

368 ARMATURE WINDING AND MOTOR REPAIR

9. Replace the armature, brush rigging and both end shields.
Smear vaseline on the slip rings and adjust the tension springs
fairly strong.

10. See that the leads to the brushes are not loose or dis-
connected as^ the armature will turn very slowly and heat
badly if only one conductor is open. It makes no difference
to which ring any lead is connected coming from the external
resistance.

11. Examine the rheostat and see that each lead is con-
tinuous from its own resistance section and that no coil
is open or loosely connected. The sliding arms must bear
simultaneously on successive buttons. If any are worn or
badly burned they should be renewed.

12. The motor primary leads must be connected each to the
proper leg of the three- or four- wire supply line.

13. See that the line is securely bolted to th^ switch, and have
the latter's clips clean and capable of a tight contact, as cur-
rent for a 25-hp. motor is liable to burn any loose connections.

14. Overload relays or fuses may be used with this motor.
Relays save time as they can be re-set, and limited to any
reasonable load.

15. Oil the bearings fully and run the motor. See that the
oil rings are turning properly.

16. When operating a slip-ring motor be careful to move
the rheostat lever slowly to avoid a rush of current. For
speed regulation the controller has many speeds forward
and reverse, with twenty or more contact fingers and blocks
which have to be renewed at times because of arcing.

17. If a non-reversing motor runs the wrong way, merely
exchange the leads of one phase of its line. Changing the slip-
ring connections has no effect.

18. A slip-ring motor should have its brushes and rings
protected from dust and dirt, and the wear on the bearings
watched just as closely as those of an ordinary induction motor.

VII. OVERHAULING A SINGLE-PHASE COMMUTATOR MOTOR

On account of the rather special construction of the single-
phase commutator motor, there are certain points to which
the repairman should pay especial attention when it becomes

INSPECTION OF MOTORS AND GENERATORS

369

necessary to overhaul such a motor or make repairs to it.
These points are outlined in detail in the following paragraphs
which refer to numbers of Fig. 248.

1.* Laminated field frame of the motor.

2. Main field winding which produces the main flux. The
auxiliary field regulates the power factor and is connected
permanently to one auxiliary brush, having a switch contact

FIG. 248. Essential parts of a single-phase commutator motor. (Figures
refer to numbers of paragraphs of text.)

for the other auxiliary brush. Both of these windings should
be coated with black armature varnish to improve their
insulation.

3. Commutator end shield.

4. Oil-well cover. If this cover is missing, a new one should

* The following paragraph numbers refer to the parts shown in Fig. 248.
24

370 ARMATURE WINDING AND MOTOR REPAIR

be fitted to keep dust from settling where it may cause the oil
ring to stop.

5. Oil gauge. When overhauling a motor take apart and
clean out the gauges as any obstruction may produce a false
oil level. The bearings should also be drained and flushed
with gasoline.

6. Oil plugs. See that they are tight or the oil may drip
from them and waste.

7. Brush-holder studs. Bolt all of these up tight to avoid
tipping the brushes from the commutator by their accidental
turning.

8. Carbon brushes with copper pig-tails, in the variable
speed compensated type of single-phase motor, remain on the
commutator all the time, similar to direct-current motors.
With the repulsion type they are short-circuited at the start
but are removed by a centrifugal governor when up to speed.
Give all brushes good contact with the commutator by drawing
under them No. 00 sandpaper against the carbon. A little
vaseline will stop the squeak of brushes.

9. Short-circuiting connection on the back of the brush
yoke for the main brush set or energy brushes.

10. Brush holders. See that they are cleaned of all gum
and dirt to allow free action of the brushes. Renew weak
springs.

11. The brush yoke is sometimes of moulded composition
and sometimes of cast iron. This should be well cleaned to
avoid short circuits.

12. A fan is used to keep up a circulation of cool air against
the windings.

13. Laminated armature with slot windings. The squirrel-
cage winding is usually placed in the bottom of the slots, and
the compensating winding near the surface, and between the
two windings a steel bar separator. All coils should be well
insulated to be moisture proof. See that band wires are
tightly soldered, with plenty of mica under them. If the
bearing linings are worn, the surface of the armature will
show polished places.

14. Commutator. To it are connected the leads of the
commutated type of armature winding. In the repulsion

INSPECTION OF MOTORS AND GENERATORS 371

type the commutator bars are all connected at full speed and
the armature runs like an induction rotor. See that all com-
mutator leads are well soldered in. If commutator is rough,
turn it down in a lathe, but ever so little at a time with a
diamond-point tool. Polish with No. 00 sandpaper and oil.
Never use emery cloth.

15. Shaft. In some makes if it becomes worn or damaged
it can be easily withdrawn and a new one replaced. If its
bearing surfaces are cut they should be turned down smoothly
before new bearing linings are fitted on.

16 and 17. Motor terminals to the line. Some designs allow
for 110-volt operation by connecting the terminals in parallel,
and 220-volt operation with them in series.

18. Terminals from the auxiliary or compensating field
winding.

19. Pulley end shield.

20. Bearing linings. If they are worn enough to allow loose
play of the shaft, new ones should be put in or the old ones
re-babbitted with oil channels cut in from end to end.

21. Oil rings. See that they are not bent out of shape and
when re-assembling be careful to get them into the slots.
This is very important and a light might better be used
than to guess at their location. Also turn the armature
by hand and see that the rings turn also. Oil well before
running.

22. Pulley should be tightly keyed on and fastened with a
set screw. Only a balanced motor pulley should be used.

23. Set screw for motor pulley.

NOTE. No rheostat is needed, although speed controllers are
used. To reverse the armature rotation, interchange the
leads to the compensating or auxiliary brushes.

VIII. OVERHAULING A DIRECT-CURRENT ENGINE TYPE
GENERATOR

At least once a year the generators of any manufacturing
plant should be thoroughly overhauled, cleaned, painted, and
minor repairs made, even though not badly out of order.
Such an inspection does no harm and often reveals a weak
part that may later cause a break-down at a busy time.

372

ARMATURE WINDING AND MOTOR REPAIR

A good job can be done in two or three days by two good
repairmen on machines up to 150 kilowatts.

1 . * If belt driven mark the position of the sliding base before
setting back with the belt screw.

2. On both belt-driven and direct-connected machines,
mark the position of the rocker arm, assuring the return
of the brushes to the neutral point.

3. Avoid if possible, disturbing the brush holders.

FIG. 249. Parts of an engine type direct-current generator that should be
regularly inspected. (Figures refer to paragraphs of the text.)

4. With a traveling crane or a tripod rigging, remove the
outboard bearing and pedestal, then the whole brush rigging
"en masse." Be careful to mark the armature leads for
proper re-connection.

5. The armature shaft should be blocked up level to its
bearing before the latter is removed, otherwise the poles will
have to support the weight and the other bearing sleeve
may crack.

* The following paragraph numbers refer to the parts shown in Fig.
249.

INSPECTION OF MOTORS AND GENERATORS 373

6. Remove either the pulley or the coupling key from the
other end of the shaft, and by properly slinging the armature
it can be withdrawn far enough out past the poles to allow
any small work to be done. No hitch or sling should be made
around the commutator, even if it is covered with burlap.
Support all weight by the shaft only.

7. Blocking under the armature core should be ready to
support all of the weight when withdrawn, and so arranged as
to prevent its rolling sideways.

8. Now with either an air hose, hand bellows or brush
remove all dust, dirt, and lint from around the poles, in the
air ducts, and from inside the frame, first covering the engine
crank, if direct connected. The armature winding should
be blown out neatly, and where any dirt is inclined to adhere
use a narrow flat stick to avoid damaging the insulation.

9. After a good dusting all around a washing with gasoline
will remove any remaining gummy dirt or oil.

10. Black armature varnish may be applied to all windings
and leads, but not to the pole shoes, or armature core disks,
because the distance between such parts is usually small
and too much varnish may "freeze" the armature tight
against the lower poles after a shut-down.

11. The armature winding and leads may be painted up
to the commutator bars but not beyond. Plenty of it between
the commutator open leads helps to insulate them. The
band winding at the outer end of the commutator and its
bolted end plate may also be painted.

12. If the commutator is very rough or badly cut it should
be turned down either in a lathe or by a turning tool attached
to the generator frame. After turning off ever so little copper
No. 00 sandpaper should be applied with oil for a polish,
but use no emery. If the commutator is badly worn down and
beyond repair, a new one should be on hand, having been
ordered weeks in advance, giving the manufacturer all the
specifications possible from the name plate. Furthermore,
when the commutator arrives examine the bore, diameter
length and number of segments, otherwise somebody's mis-
take may not be discovered till the machine is taken apart.
Do not pound on the commutator bars as they may sink.

374 ARMATURE WINDING AND MOTOR REPAIR

13. If the shaft shows any marks of cutting have its surface
trued before any bearing sleeve is repaired.

14. If the bearing sleeves are worn decidedly, have them
re-babbitted to fit snugly but not too tightly, and grooved
for oiling.

15. Be careful not to damage the oil rings, especially when
replacing the bearings. Also see that all sediment is removed
from the oil wells, and the glass gauges cleaned out.

16. If the armature band wires or mica insulation are
loose, slip under some sheets of mica and re-solder the bands
securely.

17. See that no field coils are loose on the poles; if so apply
more wedging of wood strips.

18. If the brushes are dirty remove them, one set at a time,
to avoid any mix up and clean the holders with gasoline, allow-
ing the brushes to move freely. Replace short brushes with
new ones; renew weak springs; tighten the pig- tails; or if of
copper gauze they may be set ahead. Give each brush the
proper and same tension, not too loose and not so tight as to
cause cutting. Allow about one-eighth inch under the holders,
and tip brushes at proper angles. Get the proper bevel
on each brush to fit the commutator arc by drawing No.
or No. 1 sandpaper under it, cutting the carbon away while
it rests on the commutator. Vaseline applied lightly to the
commutator will help the lubrication of the brush contacts
and prevent cutting and sparking.

19. Connect the main leads to their proper bolts tightly,
and if belt driven slip on the pulley with its key and set screw,
and with plenty of oil in the bearings turn the armature by
hand a number of times and watch all parts for interference.
Especially see that the oil rings are turning.

20. A timely warning do not coat the generator with
aluminum paint just for looks, as it is a metallic conductor.
Leaky voltage and shorts may result. White enamel radiates
heat the best, although it is some trouble to keep clean.
Dark green is the usual color.

A. All these operations take some time when properly done,
but what a pretty running machine one has when they are
honestly completed.

INSPECTION OF MOTORS AND GENERATORS 375

B. If any insulation is found peeled off or cracked open,
use a little mica, tape, shellac, or armature varnish.

C. When ready to assemble the generator, put all bolts
possible into place before tightening any. Close all oil cocks
and fill up the bearings. Keep the armature leads away from
the rotating armature.

D. When all is ready, slip on the belt, slide back the base
to its former position and start the engine slowly, watching
carefully the oil rings.

CHAPTER XV
DIAGNOSIS OF MOTOR AND GENERATOR TROUBLES

The electric motor and generator are sturdy pieces of ap-
paratus when given the proper attention and operated within
their normal ratings. The troubles which they develop can
usually be .traced to two main causes. 1. Faulty operating
conditions. 2. Mechanical and electrical defects.

Faulty operating conditions may either bring to light exist-
ing mechanical and electrical defects or create new ones.
Under normal conditions, therefore, correct operating con-
ditions will reduce to a minimum, the troubles due to the
second cause. The factors in faulty operating conditions may
be classified as follows:

1. Lack of proper cleaning.

2. Operation in damp places.

3. Exposure to acid fumes and gases.

4. Lack of frequent inspection and replacement of worn
parts.

5. Operating temperatures too high.

Lack of Proper Cleaning. All machinery having moving
parts requires lubrication and however well the provisions are
made to confine the oil to the parts needing the lubrication, it
will find its way in time to adjacent parts of the machine.
In motors and generators the oil which is allowed to accumu-
late on the windings has a detrimental effect on the quality
of the insulating materials used and is a frequent cause of
short circuits and grounds. The presence of oil also invites
the accumulation of dust and dirt which quickly fills up small
spaces that have been provided for ventilation of windings
and core. This results in an increase in temperature of the
parts involved. The only remedy for troubles resulting from
this cause on a repaired machine, or when a new one is sub-
stituted, is frequent cleaning at regular intervals. Motors

MOTOR AND GENERATOR TROUBLES 377

operated in especially dusty places, as in cement and flour
mills, should be blown out daily with compressed air or a
hand bellows. Care must be taken not to blow the dust into
the windings. An air pressure of not more than 50 pounds
per square inch with a J^-inch nozzle should be used to avoid
injury to insulation.

Operation in Damp Places. The materials employed in
the insulation of windings of motors and generators consist of
special papers, cotton tapes and cloth, and mica. With
the exception of mica, all of these materials absorb moisture
more or less or are said to be hydroscopic. In damp places
where the operation of the machine is not continuous enough
for the generated heat to keep moisture out, the insulation
may absorb enough moisture to reduce the insulating strength
to the rupture point resulting in all sorts of troubles. Motors
and generators should not be installed in damp places except
they be provided with windings having a special moisture-
proof insulation which manufacturers can furnish when the
conditions under which the machine is to operate are known.
In case a machine has been exposed to dampness or been
soaked with water in a fire without having the insulation
damaged by the fire, it should be thoroughly dried out before
being placed in operation. (For details of drying out a ma-
chine, see pages 181 to 185.)

Exposure to Acid Fumes and Gases. As in the cases of oil
and water, acid fumes destroy the insulating strength of most
insulating materials. They also attack the metal of the
machine and thus bring on commutator troubles. In chemi-
cal works and in industrial plants where acid fumes are present,
small motors should be enclosed in a fume-tight case which
can be ventilated. Where this is impossible specially in-
sulated windings must be provided. The coils needed in the
repair of such a machine should be secured from the manufac-
turer and in most cases it will be advisable for the manufac-
turer to do the repair work. Machines operated should
be cleaned more frequently than under other conditions and
the exposed windings painted regularly with an acid-resisting
paint or a good linseed oil paint. The commutator should be
wiped daily with a little vaseline. The metal parts of the

378 ARMATURE WINDING AND MOTOR REPAIR

machine should also receive a coat of acid-resisting paint at
frequent intervals.

Lack of Frequent Inspection and Replacement of Worn
Parts. Short circuits and grounds are frequently traced to
rubbing of the motor winding on field poles caused by a bent
shaft or worn bearings. A worn or bent shaft may also cause
the core of the armature to strike the pole faces and jam the
punchings of the core enough to cut the coils and cause a
ground in the winding. A loose fit of the commutator on the
shaft, when it has been replaced after a repair job, may result
in sufficient movement to break off the leads to armature coils
at the necks of the commutator. The re-soldering of such a
defect does not cure the trouble. An inspection should be
made to locate the cause and eliminate it. In this case it-
will probably call for a new shaft. A mixture of dust and
foreign material in the bearing oil cups or not enough oil is the
start of many bearing troubles which can be eliminated by
frequently draining the bearings and cleaning with kerosene.
When oil is used that has stood for some time in a receptacle
or been pumped from barrels to a tank, it is advisable to filter
it before using it in motor or generator bearings. Wear of
bearings should be inspected frequently when a large tight
belt is used on the motor.

Roughness of the commutator should be guarded against.
When it occurs and stoning does not remedy the trouble, the
real cause should be searched out. The same applies to high
mica. The air gap of the machine should also be frequently
checked by thickness gauges for this is a good check on the
condition of the bearings. A record made of all of these tests
and inspections for reference when other troubles develop
will save much time and trouble.

A frequent check for proper end play of the armature shaft
in its bearings will prevent grooving of the commutator that
otherwise might be suspected to be caused by unsuitable
brushes.

Operating Temperatures Too High. All insulating material
such as paper and cotton materials have a maximum tempera-
ture limit above which they will not retain their insulating
strength and will simply char and eventually crumble, causing

MOTOR A*ND GENERATOR TROUBLES 379

electrical faults in the windings. The materials which have
the highest maximum operating temperature are miea, and
synthetic resins like bakelite. Where surrounding tempera-
tures are known to be high as in the case of motors used to
drive blowers in a boiler room or the scrapers on an economizer,
special heat-resisting insulation must be provided instead of
that ordinarily used.

For the ordinary design of industrial motor the Standardi-
zation Rules of the American Institute of Electrical Engineers
recommend the following as maximum temperatures for the
different kinds of insulating materials specified:

For cotton, silk, paper and other fibrous materials not so treated as to
increase the thermal .limit, 95 C. (203F.). When measured by ther-
mometer, 80C. (176F.).

For cotton, silk, paper and other fibrous material treated or impreg-
nated and including enameled wire, 105C. (221F.). When measured
by thermometer, 90C. (194F.).

For mica, asbestos, or other material capable of resisting high tempera
tures, in which the previous materials if used, are for structural purposes
only and may be destroyed without impairing the insulating or mechan-
ical qualities, 125C. (257F.). When measured by thermometer, 110C.
(230F.).

All parts of electrical machinery other than those whose temperature
affects the temperature of the insulating material, may be operated at
such temperatures as may not be injurious in any respect. But no part
of continuous-duty machinery subject to handling in operation, such as
brush rigging, shall have a temperature in excess of 100C. (212F.) for
more than a very brief time.

NOTE. When a thermometer is used to measure the maximum (hottest
spot) temperature, a correction of 15C. is added to the thermometer
reading in order to allow of the impossibility of locating the thermometer
at the hottest spot, which will be beneath the insulation next to the metal
of the conductor.

The hand is a very poor thermometer and should not be
relied upon except as a general indicator of operating tem-
peratures. Any temperature above 120F. is very uncomfor-
table to the touch but is well within the range of safe opera-
tion. Until a surface temperature of about 176F. is reached
there is little danger to injury of insulation. The motor
should however be closely watched at this temperature and
not allowed to go much beyond it except for a very short
time.

380 ARMATURE WINDING AND MOTOR REPAIR
CAUSES AND REMEDIES FOR TROUBLES IN DIRECT-CURRENT MACHINES

Faults

Cause

How most readily
detected

Remedy

1. Too high

1. Too high speed

1. Voltmeter reads

1. Slow the engine.

voltage.

of engine.

greater than standard

and lamps burn with

undue brilliancy.

2. Too strong mag-

2. Same.

2. Introduce more

netic field.

resistance in shunt

field.

2. Too low

1. Too low speed of

1. Voltmeter shows

1. Increase speed of

voltage.

engine.

lower than standard

engine.

and lamps burn dimly.

2. Too weak mag-

2. Same.

2. T a k e out resist-

netic field.

ance in shunt field.

3. Brushes not prop-

3. Same.

3. Rock brushes

erly set.

sack and forth till

highest voltage consist-

ent with sparkless

commutation is shown.

3. Excessive

1. In a generator,

1. By too high read-

1. Cut out necessary

current.

too many lamps burn-

ing of ammeter for

number of lamps. Re-

ing or motors running

capacity of machine.

duce load on motor

By excessive sparking

circuits. In this case

of dynamo brushes

none of the motors

and too high reading

may be doing too

of dynamo ammeter.

much work, but there

may be too many in

dynamo circuit.

2. In a motor, too

2. By excessive

2. Reduce the load

much mechanical

sparking of motor

on the motor.

work being done by it.

brushes nad too high

reading of motor

ammeter.

3. Short circuit; leak

3. By excessive

3. Locate and re-

or ground in external.

sparking of brushes,

move leaks or grounds.

circuit.

and heating of whole

armature.

4. Short circuit in

4. By heating of

4. Stop machine.

armature coil.

short-circuited coil

Locate coil. If en-

more than the others.

tirely burned out must

be renewed.

5. Grounds in arma-

5. Same as 4.

5. Locate the

ture. Two grounds

grounds. Re-insulate

to the core amount to

the coils containing

a short circuit.

them.

6. Due to excessive

6. By sparking of

6. File away pole

friction in bearings or

brushes. By sound

pieces or re-center arm-

by armature striking

of armature striking

ature. Clean and oil

pole pieces. In gen-

while running. By

journals, or re-fit bear-

eral any cause tending

heating of motor bear-

ings.

to slow motor.

ings.

MOTOR AND GENERATOR TROUBLES

381

CAUSES AND REMEDIES FOR TROUBLES IN DIRECT-CURRENT MACHINES

(Continued)

Faults

Cause

How most readily
detected

Remedy

4. Excessive

1. Excessive current;

1. Same as given un-

1. Same as given

sparking at

therefore due to any

der "Excessive cur-

under "Excessive cur-

brushes.

of the causes given

rent."

under that head.

2. Brushes improp-

2. By taking brushes

2. Fit and set ac-

erly set.

out of holders and ex-

curately, then shift

amining rubbing sur-

the brushes backward

face. By measuring

or forward till spark-

the peripheral dis-

ing is reduced to a

tance between brush

minimum.

sets.

3. Brushes make

3. By sighting un-

3. Sandpaper the

poor contact with

derneath between

brushes and adjust the

commutator.

brushes and commu-

spring tension until

tator.

they rest evenly on

commutator with light

but even pressure.

4. Rough, non-con-

4. A rough commu-

4. Smooth commu-

centric commutator.

tator can be detected

tator with fine sand-

by lightly touching

paper. If eccentric-

finger nail to it whi'e

ity is due to uneven

- ',

running; an eccentric

wear of bearings, re-

commutator by the

new or re-line them.

regular rise and fall

of the brushes.

5. "High" or "flat"

5. By the jumping

5. Same as above, or

bars in commutator.

or vibrations of the

turn down the commu-

brushes.

tator in lathe. Slot

out the mica to a

depth of Ho or \$Q in.

6. Broken circuit in

6. Commutator

6. Locate coil by

armature or commu-

flashes, and nearest

drop of potential

tator.

the break is cut and

method. If in com-

burnt. Flashing con-

mutator bridge over

tinues when armature

the break. If in arm-

is slowly turned.

ature coil, it must be

renewed.

7. *Veak fie d mag-

7. Dynamo fails to

7. Short circuits or

netism, caused by

generate full emf.

grounds are easily lo-

broken circuit in field

, If very weak, motor

cated and remedied if

winding or short cir-

runs very slow, taking

external to the wind-

cuit in same; two or

a current many times

ings. If internal,

more grounds in wind-

full load current.

faulty coil must be re-

ings; reversal of one

wound or repaired if

or more field coils.

only grounded. A re-

versed coil will lower

the voltage instead of

increasing it, and it is

remedied by reversing

the connections.

382

ARMATURE WINDING AND MOTOR REPAIR

CAUSES AND REMEDIES FOR TROUBLES IN DIRECT-CURRENT MACHINES

(Continued)

Faults

Cause

How most readily
detected

Remedy

8. Unequal magnet-

8. One brush sparks

8. Only remedied by

ism.

more than the other.

re-shaping pole pieces.

9. Dirty commuta-

9. Flashing around

9. Clean commuta-

tor, causing brushes

commutator.

tor. (Methods given

to vibrate, particular-

later.)

ly if of carbon.

10. Poor brushes,

10. By ragged ap-

10. Renew brushes.

especially if of his;h-

pearance of brushes

Try different grades of

resistance carbon,

around edges and for-

hard and soft brush.

hard blisters forming

mation of hard spots.

on them.

11. Vibration, espe-

11. By a humming,

11. Reduce cause of

cially of brush hold-

singing sound of

vibration or give the

ers, causing rapid vi-

brushes.

brushes a little greater

bration of brushes.

pressure on commuta-

tor.

4 (a) Excessive

12. Wrong interpole

12. With low field

12. In motor, pro-

sparking in inter-

polarity.

excitation, examine

gressing in direction of

pole machines.

field polarity with a

armature rotation, po-

compass, the armature

larity should be N-n-

being first removed.

S-s, etc. In genera-

tor, progressing in di-

rection of armature

rotation, polarity

should be N-s-S-n, etc.

13. Interpoles not

13. By inspection.

13. Adjustable when

exactly over commu-

poles are bolted to the

tation belt.

frame.

14. Brushes not set

14. Trace out by

14. Usual setting is

so that coils under-

following up coil ends.

in geometric neutral.

going commutation

Set for minimum

are under interpole.

sparking under aver-

age load.

15. Interpole air gap

15. See if all inter-

15. Adjustable when

too long or too short.

pole gaps are equal.

poles are bolted to the

frame. Weaken inter-

pole strength by

shunting the interpole

winding.

5. Heating of

1. Excessive current

1. Sfeme as given

1. Same as given

armature.

through it and there-

under "Excessive cur-

fore due to any of the

rent."

causes given under

that head.

2. Eddy currents

2. Core becomes

2. This can be cor-

and hysteresis in core.

hotter than armature

rected by improving

coils after running for

ventilation by special

a short time.

fans or air guides.

MOTOR AND GENERATOR TROUBLES

383

CAUSES AND REMEDIES FOR TROUBLES IN DIRECT-CURRENT MACHINES

(Continued)

Faults Cause

How most readily
detected

Remedy

3. Conduction from

3. Other parts con-

3. Locate source of

other parts as from

nected to armature,

heat by thermometer

commutator or bear-

as commutator, shaft

or feel by the hand

ings, the heat being

or bearings, hotter

and correct , it by

conveyed to armature.

than the armature.

cleaning and lubrica-

tion.

6. Heating of

1. Too great pres-

1. By feeling the

1. Reduce pressure

commutator.

sure of brushes, fric-

commutator with the

by adjusting spring.

tion causing heat.

hand.

2. Excessive spark-

2. Same.

2. Discover the

ing.

cause of sparking and

correct it according to

the particular cause

given under sparking.

3. Excessive cur-

3. Same.

3. Discover cause of

rent.

excessive current and

correct according to

particular cause al-

ready given.

4. Conduction from

4. Same.

4. If from bearings,

other parts.

lubricate orre-fitthem.

7. Heating of

1. Excessive cur-

1. Too hot to bear

1. Locate the par-

field coils.

rent in field circuit

the hand. If exceed-

ticular coil in which

due to short circuits

ingly hot, by smell of

fault lies and repair or

or grounds.

burning shellac or var-

r e-wind. Methods

nish or charring cotton.

given later.

2. Eddy currents in

2. The pole pieces

2. Only remedied by

pole pieces, heat being

are hotter than the

better design, use lam-

conducted to the coils.

coils after a short run.

inated pole shoes.

8. Heating of

1. Lack of lubrica-

1. By feeling with

1. Fill oil cups; clean

bearings.

tion.

hand. Oil cups emp-

feeding pipes.

ty or feeding pipes

clogged.

2. Dirty or gritty

2. By feeling with

2. Remove cap and

bearings.

hand.

thoroughly clean.

3. Bearings out of

3. Unequal wear of

3. Bearings must be

line.

bearings, and shaft

lined up or shells re-

will not turn freely by

babbitted. If very

hand.

serious, new bearings

will have to be made.

4. Rough or cut

4. Shaft will show

4. Turn down shaft

shaft.

the roughness in the

in lathe, or scrape the

bearings.

5. Shaft bent.

5. Unequal wear in

5. Shafts can only

bearings and arma-

be straightened by

ture will wobble. Very

disconnecting from

hard to move by hand.

armature and re-heat-

ingand re-forging.

384

ARMATURE WINDING AND MOTOR REPAIR

CAUSES AND REMEDIES FOR TROUBLES IN DIRECT-CURRENT MACHINES

(Continued)

Faults

Cause

How most readily
detected

Remedy

6. Oil rings stuck.

6. Inspection.

6. Adjust rings in

grooves.

9. Too low

1. Too much load.

1. By speed indica-

1. Reduce the me-

speed (referring

tor; heavy sparking,

chanical load.

to motors).

heating of all parts

and bearings.

2. Any of the causes

2. Same, and same

2. Discover particu-

given under " Heating

as given under " Heat-

lar cause and remedy

of bearings," causing

ing of bearings."

same as given under

excessive friction.

" Heating of bearings."

3. Short circuit or

3. By motor taking

3. Same as under 5,

grounds in armature.

excessive current

"Excessive current."

without load as shown

by ammeter or heavy

sparking and heating.

4. Too low voltage

4. By motor volt-

4. By increasing the

at terminals.

meter or speed indi-

line voltage.

cator. By heavy

sparking and heating.

10. Too high

1. Too light load (in

1. By noticeable in-

1. Increase load.

speed (referring

series motors).

crease in speed.

to motors).

2. Weak field shunt

2. Same.

2. Strengthen field.

motor.

3. Too high voltage

3. Same.

3. Correct line vol-

at terminals, due to

tage by remedies 1 and

high voltage of dyna-

2 under "Too high

mo.

voltage."

11. Dynamo

1. Too weak residual

1. Very little attrac-

1. Send a current

fails to generate

magnetism, caused by

tion by the pole pieces

through field from a

emf.

a jar or reversal of

when tested with a

few cells or from a

current not sufficient

piece of iron.

running dynamo.

to reverse magnetism.

2. Short circuit

2. Magnetism very

2. Locate the

within machine, or

, weak.

grounds or short cir-

grounds in field wind-

cuits and correct them.

ings.

3. Reversed field

3. All poles should

3. Make polarity op-

coils.

have alternate mag-

posite by reversing the

netism; if a coil is re-

connections of the coil.

versed it will show

Each pole should be

magnetism, but may

opposite to the one on

not be of opposite po-

each side of it.

larity.

4. Series and shunt

4. Voltage falls as

4. Reverse connec-

windings connected up

load is increased, the

tions of either field,

opposite to each other

external circuit being

but not both

closed showing that

they are working

against one another.

MOTOR AND GENERATOR TROUBLES

385

CAUSES AND REMEDIES FOR TROUBLES IN DIRECT-CURRENT MACHINES

(Continued)

Faults

Cause

How most readily
detected

Remedy

5. Brushes not prop-

5. Magnetism and

5. Find central posi-

erly placed.

emf. increased by

tion by experiment or

shifting the brushes.

from drawings of con-

nections 1 .

6. Open circuit in

6. Test circuits with

6. Set up on all

field or armature.

magneto.

connections. Press

Brushes not making

brushes on commuta-

good contact with the

tor to start building

commutator. Loose

up.

connections.

7. Too much resist-

7. Voltage does not

7. Cut all resistance

ance in the shunt field

exceed that due to

out of the shunt field

circuit, i.e., greater

residual magnetism.

circuit. Reverse the

than the " critical " re-

The voltage due to

shunt field.

sistance. Shunt field

residual magnetism

bucks the residual

drops when the shunt

magnetism.

field circuit is closed.

12. Motor fails

1. Too much load.

1. No motion and

1. If motor does not

to start.

fuse in circuit melts

start at once, turn off

or circuit-breaker acts.

current and search for

See if motor runs all

cause. Reduce load

right when light.

on motor.

2. Excessive friction,

2. Same, and motor

2. Remedies same as

due to any causes giv-

hard to turn when not

given under " Heating

en under heading

loaded, and with no

of bearings."

"Heating of bear-

current.

ings."

3. Short circuit of

3. Motor refuses to

3. If connections are

field or armature or

revolve, though shows

made wrong, consult

among connections.

signs of strong mag-

maker's diagram and

netism. Will turn

correct them. Test for

easily by hand if un-

continuity and short

loaded and with no

circuits as given later.

current. If current is

very great, it is indi-

cation of short circuit.

If fault is in field,

magnetism will be

weak.

4. Open circuits due

4. Weak magnetism

4. Turn current from

to field switch open,

shows a loose connec-

motor, and search for

fuse melted, loose or

tion in field circuit; no

cause of discontinuity;

broken connections,

magnetism, that field

examine all switch

or some fault at gen-

switch is open. May

fuses and connections,

erator.

be heavy current in

tightening all. Test

armature. If there is

for continuity in ma-

no armature current

chine circuits and re-

there will be no spark

pair broken or burnt-

at brushes when

out coils.

raised.

2.1

386

ARMATURE WINDING AND MOTOR REPAIR

CAUSES AND REMEDIES FOR TROUBLES IN DIRECT-CURRENT MACHINES

(Concluded)

Faults

Cause

How most readily
detected

Remedy

13. Flickering
of lamps.

14. Noise.

1. Uneven running
of engine, probably
due to governor fail-
ing to properly func-
tion.

2. Loose connec-
tions, either on ma-
chine, switchboard or
external circuit.

1. See third fault
(6), fourth fault (4, 5,
9, 10, 11), eighth fault
(3, 4, 5), and tenth
fault.

2. Armature running
against the brushes.

1. By flickering of
lamps or vibration of
voltmeter indicator.

2. Same.

2. By unusual noise.

1. Overhaul engine,
specially governor.

2. Examine all con-
nections and see that
they are firm and
make good contact.
Look for arcs.

2. Correct direction
of rotation.

Electrical Defects. Troubles in motors and generators
due to electrical defects show up in a variety of Ways. Their
location, however, is more often a simple than a difficult matter
and depends upon a few testing devices and a great deal of
patience and persistence. An experienced operator or re-
pairman learns to diagnose electrical troubles by a process
of elimination where there may be several causes. For
such an analysis of common electrical troubles in motors
and generators the table on pages 380 to 386 will be found
useful. It was compiled by the authors of the Naval Elec-
tricians Text Book and has been taken from that most practical
work.

CAUSES AND REMEDIES FOR TROUBLES IN ALTERNATING-
CURRENT MACHINES

Induction Motor Troubles. The most common troubles
with induction motors show up by the machine failing to start
or by stopping while connected to the line. When the ma-
chine refuses to start and the fuses are not blown, the cause
may be in too large a load. The remedy in such cases is to either
reduce the load or use a clutch for starting since the squirrel-
cage motor has a limited starting torque but can often carry

MOTOR *AND GENERATOR TROUBLES 387

a much larger load while running. When investigating a
trouble of this kind it is important to see that the voltage is
normal and that the bearings are in good condition. If none
of these conditions seem to cause the trouble a test should be
made for open circuits in the windings. (See Chapter IX.)

An excessive current at starting may be due to a high volt-
age applied or too great a load. When the voltage is high
an auto-transformer with suitable taps must be used for
starting. When an auto-transformer is used and the starting
current is excessive, a test should be made for proper connec-
tions. Too low taps on the starter should be avoided.

An overload on an induction motor will usually show up
by the mo tor coming to rest or taking about 10 times more than
normal current. When fuses and circuit breakers do not
operate under such heavy current, the machine will burn out.
It is therefore important that fuses and settings of circuit
breakers be used to prevent such burn-outs. Since the torque
of an induction motor varies with the square of the voltage,
low supply voltage may be the cause of the machine stop-
ping. Also worn bearings will allow the rotor to drop on the
stator and thus block the former and stop the machine. For
other details in testing and inspection of induction motors see
Chapter XIV.

Locating Troubles in Winding of Induction Motors.
When the cause of trouble with an induction motor in starting
or in operation has been traced to the stator windings, details
of locating the fault are outlined in Chapter IX. When
the stator has been repaired and again placed in service its con-
nections should be carefully looked over. The points to be
observed when inspecting such a stator are outlined on page
363. In case the connections of an auto-starter are reversed,
the starting current will be excessive and insufficient torque
with the switches in the running position will show up. If
the motor refuses to start the cause may be due to a defect in
the change-over switch or to a loose connection or open
circuit in the auto-starter.

Mechanical Adjustments. In those cases where induction
motors are in continuous operation regular inspections should
be made of the air gap to see that it is not appreciably reduced

388 ARMATURE WINDING AND MOTOR REPAIR

at the bottom of the rotor on account of wear of the bearings.
When the air gap is reduced a test should be made before the
trouble is attributed definitely to wear of the bearings. It is
also important that an armature have a small play endwise in
the bearings. This should be about Me inch. Excessive
heating or excessive belt tension may be the cause of a great-
deal of bearing trouble on an induction motor which is belt-
driven.

Troubles Due to Electrical Faults. When errors have been
made in winding an induction motor or in re-connecting the
coils, trouble will usually show up in heating of the coils, a
peculiar humming sound for large and unbalanced currents
in the different phases. Before testing for faults of re-connec-
tion of coils, it must be determined that the connections of the
winding to the supply circuit are correct and that none of the
connections of the phaso coils have been accidentally jammed
together so as to short-circuit them. Troubles in rotors of
induction motors are rare, however, in squirrel-cage types
trouble may be located in soldered joints which have melted
or corroded. High resistance at joints lowers the efficiency,
increases the heating and also increases the starting torque of
the motor. Heating or operation at reduced output or speed
may be due to a three-phase motor running on one or two
phases.

Troubles with Synchronous Motors. When a synchronous
motor fails to start the trouble is generally due to an overload.
In testing out for such a condition the motor should be started
light. If the operation is not satisfactory the load should be
reduced. Poor starting may also be caused by reduced volt-
age or open or faulty connections in starting apparatus. An
open circuit is usually indicated by no current flowing in a
particular phase. An excessive current usually indicates a
short circuit but may be due to grounds. An excessive cur-
rent in a synchronous motor is a dangerous condition and the
trouble should be located.

When a synchronous motor is used for power-factor correc-
tion overheating indicates an excessive current. The machine
can usually be temporarily operated by reducing the load or
reducing the amount of leading current. When a synchronous

MOTOR AND GENERATOR TROUBLES 389

motor is operating satisfactorily the current in the armature
phases should be about equal when the rotor is turning slowly.

Trouble in the field winding such as an open circuit causes a
shutdown or excessive armature heating. When the field cur-
rent seems excessive, a test should be made for polarity of the
armature coils and reversal of connections. When a syn-
chronous motor fails to show normal starting torque and will
not carry the load, the trouble will frequently be found in the
field circuit in the form of an open circuit, short circuit or
reversal of one or more field windings.

Causes of A.-C. Motor Fuses Blowing. In addition to
overload on a motor, many other things cause fuses to blow.
The following causes and symptoms have been formulated by
Henry W. Zeuner, Milwaukee (Wis.) Electric Light Company
(Electrical World, Sept. 20, 1919).

1. Operator throwing starting switch of compensators
from starting to running position too quickly.

2. Operator throwing switch into running position without
touching the starting position at all.

3. Motor winding becoming grounded.

4. Excessive current due to low voltage, short circuits
in stator windings, single-phase operation, etc.

5. Starting switch being in running position when service
comes back on line after interruption.

Wound-rotor motors and squirrel-cage motors which are
not protected by a no-voltage release may be shut down at
any time because of the last-mentioned cause. To overcome
this trouble a group of such motors is often protected by an
oil switch equipped with no-voltage release. The arrange-
ment has the disadvantage, however, that whenever the oil
switch opens it becomes necessary for some one to go around
and open each individual motor switch before the oil switch
is closed again.

A ground may only blow one fuse ajid leave the motor
operating from one phase of the line. When a polyphase
motor is running single-phase it not only gets hot but it
makes a growling noise which is especially noticeable under
heavy load. By the time these symptoms indicate the prob-

390 ARMATURE WINDING AND MOTOR REPAIR

able source of the trouble and it is decided to shut down the
motor, a second fuse is sometimes blown by the excessive
current per phase. If only the first defective fuse found is
replaced, the motor may be started from the line side of the
switch which is connected ahead of the fuses. When the
switch is thrown over into the running position the motor is
again operating on single phase. Because the motor con-
tinues to heat up the user often thinks there must be
something wrong with the motor itself and calls for help,
whereas if he had only tested all the fuses in the first place and
replaced all of the burnt-out fuses, instead of only one of
them, he could have put the motor back in service without
assistance and with a minimum loss of time. Another source
of trouble which may be detected at the service fuses is low
voltage. This causes excessive heating and often burns
out stator coils. The voltage can be checked roughly with
test lamps.

Inspection of Motor-starting Devices. From the fuses it is
customary to proceed to examine the motor-starting switch,
compensator or controller. Contact fingers are often found
bent or burned so that they do not make contact and the
connecting leads are sometimes burned off. The two-minute
starting resistance supplied by the manufacturers for starting
slip-ring motors is often burned out by being used for speed-
regulating duty which requires a much heavier resistance.

Testing Motor for Grounds. When testing an alternating-
current motor for grounds, a magneto or high-voltage testing
transformer should be used as the line voltage will seldom
show a ground unless it is making very good contact.
National Electrical Code rules require that motor frames be
grounded, but this requirement is not always carried out.
When the motorframe is insulated from the ground the motor
can be kept in operation with one phase grounded to the
motor frame, but tlae defect should be remedied at the first
opportunity. A second connection between the stationary
windings and the motor frame will burn out the coils which it
short circuits, and if not given prompt attention the two
grounds may result in the burning out of the entire winding.

MOTOR AND GENERATOR TROUBLES 391

When both the supply system and the motor frames are
grounded, as is usually the case, one ground on the stator
winding will blow one of the motor fuses unless the fuse is
too heavy, in which case the ground may burn out some of the
stator coils. However, every ground which blows a fuse
does not occur in the motor winding. If the motors test
clear, the wiring leading to the motor may be found grounded
to the conduit.

It is a good plan to run the defective motor while testing
when it is practicable to do so. This makes it possible to
conduct additional tests and very often to observe symptoms
which do not appear when the motor is standing still. Under
such operating conditions the speed should be tested. Low
speed and inability to pull the load are usually an indication
of bad connections between the rotor bars and the rotor end
rings.

Hot Stator Coils. Another condition to observe while the
motor is operating is whether the stator coils are hotter in
one place than in another. This is one of the signs of bearing
trouble. When the bearings become worn the stator coils
around the section of reduced air gaps get hotter than the coils
adjacent to the section where the air gap has been increased.
If the bearings are not renewed before the rotor begins to
rub on the stator, the motor windings may be seriously in-
jured, especially if the fuse is too large and does not blow.
When it is impracticable to operate the motor the belt should
be taken off and the bearings examined to find out if there
is too much play in them.

Tension of Belts. Much belt and bearing trouble is caused
by the belt being too tight. This is especially true with
vertical belt drives from squirrel-cage motors to high-speed
machines. When bearing trouble is experienced with in-
stallations of this description it is usually because the pulley
surface is too small to accelerate the driven machine as fast
as a squirrel-cage motor comes up to speed, and consequently
the belt comes off. In attempting to keep the belt on the
pulley the belt is usually made so tight that the excessive
tension soon wears down the bearings. This trouble can

392 ARMATURE WINDING AND MOTOR REPAIR

sometimes be overcome in installations of small motors by
using larger pulleys, a larger motor, or by changing from a ver-
tical belt drive to a horizontal belt drive with the slack side
of the belt on top. With large belted motors driving high-
speed machines the slip-ring type of motor should be used in
order to bring the driven machine up to speed more slowly.

Troubles in Rotor Windings. A bad contact in the rotor
winding is not so easily found when the end rings are cast
on to the end of the rotor bars. This cast construction was
originally adopted to overcome the bad joints which developed
with screwed and soldered connections in the rotor windings.
The change has not entirely eliminated the trouble, however,
because the molten metal does not always unite with the end
of the copper bars when the end rings are being cast. When
this is t"he case trouble due to the poor contact often develops,
especially if the motor is subject to much dirt and vibration.
With a little experience a defective joint between the cast
end rings and the rotor bars can be found by tapping the end
rings with a hammer and noting the difference in sound at
various places. A bad joint in this type of construction can
be repaired by welding the bar and end rings together. Welded
rotor windings have been very successful, and it is unusual
to find bad joints in rotor windings which have been put
together in this manner.

Examination of Stator Winding. The stator winding
should also be thoroughly examined while the end shields
are off the motor. If one or two coils are burned out, the
other coils can in most instances be raised out of the slot and
new coils slipped into place provided that the coil insulation
is flexible and in good condition. Should this be attempted,
however, after the insulation on the coils has become brittle,
there is a grave danger of damaging the insulation to such an
extent that the entire stator winding may have to be renewed.
When the condition of the insulation is doubtful and only one
or two coils need replacing, it has been found a good plan
to cut away the burnt-out coils and thread new wire through
the slots turn by turn, without disturbing any of the other coils.

The damage from a local short circuit in the stator winding

MOTOR AND GENERATOR TROUBLES 393

can be limited to the one or two coils affected and the motor
still kept temporarily in service by promptly cutting out
the damaged coil with a jumper. However, if the trouble
is neglected it will spread to other coils in the same phase, and
then the windings of the other phase or phases will also burn
out.

Sparking at Slip Rings. This trouble in wound-rotor mo-
tors occurs when the tension spring has not been reset from
time to time to keep the brushes in close contact with the slip
rings. This is the most frequent cause of the trouble, and
it can usually be eliminated by adjusting the tension of the
spring and where necessary turning up the slip rings and re-
newing the brushes.

CHAPTER XVI

METHODS USED BY ELECTRICAL REPAIRMAN TO
SOLVE SPECIAL TROUBLES

When called upon to locate troubles in electrical apparatus
many electricians and repairmen find themselves in a state of
wondering just what to do first. The mystery supposed for
so long to be associated with the operation of electrical appa-
ratus seems to discourage many able electrical men to attempt
to search out an electrical trouble when it seems to be a compli-
cated one. By the use of the same good sense that makes
such men successful in the work with which they are familiar,
most electrical problems in repair can be located and a satis-
factory remedy applied. In a serious case where expert help
is reasonable and within easy and prompt access, it may not
always pay for an inexperienced man to spend the necessary
time to handle a difficult trouble but in the ordinary run of
plant operation such troubles are few. What is most needed
is a little clear thinking aided by a few testing instruments.

One of the ways by which a repairman can develop con-
fidence in his own ability to search out trouble and prepare
himself to properly diagnose troubles, is to read and profit
by the experiences related by those who have worked out
puzzling troubles and have told in the columns of electrical
trade journals how they proceeded. The author has made
a selection of such experiences covering a wide range of operat-
ing troubles which are liable to come up at a time or under
circumstances which make them difficult to handle. The
details given will furnish suggestions as to methods of pro-
cedure in many other cases than those described.

Sparking at Commutator Caused by Poor Belt Joints. In
operating motors or generators with laced belts, it is important
that the joints be flexible and the ends neatly laced close
together to present as nearly as possible a continuous surface
to the pulley. Where this is not done laced belts very fre-

394

METHODS USED TO SOLVE SPECIAL TROUBLES 395

quently cause sparking due to any one of three causes: (1)
The laced belt when passing over the pulley may cause a jerk
or mechanical vibration of the armature which is severe
enough to throw the brushes from the commutator and cause
sparking due to the broken circuit. (2) When the bearings are
loose in their housings or when the armature shaft is loose in
its bearings, the jerk of the lacing may pull the armature to
one side. This develops a higher voltage in a generator or a
higher counter electromotive force in a motor in the part of the
armature which is pulled closest to the pole faces. This will
cause sparking due to a pulsating current or in severe cases
due to a short circuit current between studs of the same
polarity. (3) The belt may slip while the laced ends are
passing over the pulley, which will permit a variation in the
speed of the armature. This will cause a sudden change in
the current and may produce sparking due to the inductance
of the armature coils.

Plugging a Commutator. When commutator bars show a
tendency to blister and bead at the outer ends, " plugging"
the commutator often gives relief. To do this, proceed as
follows: Grind all of the "set" from the teeth of a piece of
hacksaw blade. With this piece of blade saw the commutator
side mica to the outline indicated at (a) in the accompanying
diagram, Fig. 250. Next drive in mica wedges or "plugs" so
called, to entirely fill and
perfectly fit the grooves
made by sawing. The
best results are to be ob-
tained by loosening the
commutator before driv-
ing in the Wedges, but FlG ' 250.-Method of repairing mica seg-
; ment by sawing at (a) and plugging.

this is not absolutely

necessary. For other details of commutator repair sec

Chapter XII.

Knock in a Motor Armature Due to Band Wires Being Too
High. In the banding of many types of direct-current arma-
tures, the bands are laid onto a layer of fish paper or of thin
fiber which affords smooth bedding for the band wires and which
also provides additional insulation between the band wires and

396 ARMATURE WINDING AND MOTOR REPAIR

the armature coils. On many of the later designs of armatures
the tension used in banding has been greatly increased, and as
a further precaution against the bands cutting into the insula-
tion of any coils that may be high in the slot, a layer of thin
tin is interposed between the band wire and the insulating
band. That it may not be advisable to adopt this improved
method on machines that have not been designed for it, is
illustrated by the following experience (Electrical Record,
June, 1918).

A 20-hp. direct-current motor which had just been repaired
and shipped, was returned within a few days with the com-
plaint that the armature "had a knock in it." At the time of
testing the motor, a slight vibration had been noticed, but
leveling of the motor had eliminated the vibration. Careful
inspection of the motor this time, however, revealed that one
of the tin clips of an end band had been striking one of the
bottom pieces. Investigation disclosed that a contributory
cause of the knocking was that one of the bearing linings had
been bored a little out of center, but even after this had been
corrected a piece of % 4 -inch fiber could not be inserted between
the armature core and the bottom pole pieces in certain
positions of the armature The real cause of the trouble was,
then, that the layers of insulation and tin under the band
wires, projected the wires and their holding clips too far into
the air gap. As the air gap was thinner than ordinarily found
on such motors, the remedy was to remove the tin and install
band wires of smaller diameter and with less tension, and to
use thin copper for the band wire clips. These changes re-
sulted in a %2-inch air g a P- The testers were deceived as to
the cause of the vibration because the tilting of the machine
floated the armature to a position where the high clip cleared
the pole piece it was striking.

Heating of an Armature Traced to Poor Soldering of Com-
mutator Connections. E. C. Parham has described (Electrical
Record, March, 1919) an interesting case where it was necessary
to replace the armature of an exciter that had been in opera-
tion so long that the commutator bars had worn so that the
pressure of the end rings was beginning to buckle the bars
up in the middle. An extra armature was obtained because

METHODS USED TO SOLVE SPECIAL TROUBLES 397

the exciter could not be spared from service. This armature
was installed in place of the old one and the old one was
repaired by re-filling its commutator. About a year later a
switchboard short circuit which started by a stroke of light-
ning, so badly burned the new exciter commutator that it was
necessary to take a cut off it. The repaired armature was
taken from storage and installed but it could not be used,
because it heated all over as soon as a field was put on the
exciter. As the armature was the only equipment part
that had been handled, the trouble evidently was in it. It
was removed and the commutator disconnected for testing.
The commutator was found to be perfectly clear but the
armature leads where they were connected to the commutator
were found to be a mass of short circuits due to the criminal
carelessness of the man who had done the soldering. After
picking out the solder, trimming the leads and re-connecting
by a workman who knew that class of work, the operation
was normal.

One lesson that stands out very prominently from this
experience is this: When the armature of an indispensable
machine must be replaced on account of the necessity of
making repairs, after the repairs have been completed, re-in-
stall the repaired armature at once for test so that the armature
that is known to be right, may be kept as the spare and with
a fair degree of certainty that it will be all right when oc-
casion arises for using it. Another lesson is that every
shop or station that pretends to do connecting and soldering,
should have access to a short-circuit test for locating just
such trouble as described.

How a Commutator was Repaired under Difficulties. In
an instance related by R. L. Hervey (Electrical World, Feb.
26, 1916) the commutator of a generator had on a number
of occasions given trouble because of the mica breaking down
between the segments and at the end rings. After a member
of the engine-room crew had failed to relieve the trouble,
an electrician was called in. Two segments were located
that had been hot enough to melt the solder out of the joint
between the bars and the risers. This appeared to be the
seat of the trouble.

398 ARMATURE WINDING AND MOTOR REPAIR

Before the commutator was opened it was blown out with
an air blast at the front and back to remove the copper and
carbon dust and prevent it from falling into the commutator
and giving trouble when the latter was opened. One of the
bars that showed signs of being hot had welded to the rear
end ring, forming a ground. As this bar passed from one
brush to the other the ground changed from the positive to the
negative bar. causing the ground lamps to flicker. Consider-
able oil had crept along the shaft
and found its way into the com-
mutator, starting the trouble by
FIG. 251. Section of patched causing the binder in the mica ring

to disintegrate. Four segments were

removed so that a patch could be put in the mica ring, as shown
in Fig. 251. The old ring was measured for thickness, and
two pieces of special ring mica were put in, with a thin piece
of clear mica covering each of the four joints. A high-grade
shellac was used to hold the pieces in place until the commutator
was tightened and also to fill up the small openings in the
mica and keep oil out.

The risers used on this machine were poorly attached
to the commutator bars, being merely soldered in a shallow
slot in the center of the bar. Before the heated segments
could be replaced it was found to be nesessary to scrape the
thin coating of solder from the riser and out of the slot in
order to make a good joint. If a current is passed through a
soldered joint while the solder is hot enough to flow, the
structure of the solder is so changed that the resistance of the
joint is very much increased and will continue to give trouble
if worked near its original carrying capacity, although to all
appearance a good joint has been made.

When the commutator was built amber mica was used
between the bars. About every two years it has been neces-
sary to remove a few pieces, and these had been replaced by
micanite, which is supposed to be able to withstand the bad
effects of oil better than the former. At this time 10 pieces
of mica that showed signs of pitting were taken out. Two
pieces were of the original amber and the rest were micanite,
some of which had not been in service more than two years.

METHODS USED TO SOLVE SPECIAL TROUBLES 399

This case does not, however, afford a fair comparison between
the two grades of mica. If the leads had been poorly soldered
into the riser or the riser into the bar, a high-resistance joint
would have resulted, causing imperfect commutation with
creepage of current across the mica between bars. This
creepage will in time break down the mica segment and thus
short-circuit the bars.

In order to test for a short circuit, after the leads were
soldered in and the commutator drawn tight the engine was
started and the field switch closed. If a short circuit had

FIG. 252. Method of attaching turning rig to machine showing its operation
with tool upside down.

existed, it would have caused one or more coils to get hot
in a few minutes, but such a condition was not observed in
this case.

Since it had been necessary to turn this commutator so
often, the owners had purchased a turning rig which was
supposed to fit this machine. A common difficulty of turning
devices is that they seldom fit a machine unless made specially
for it. In this case if the rig had been put on the proper side of
the pedestal it would have been necessary to leave the bear-
ing cap off or to have the tool considerably above the center
of the commutator. In the latter instance, in case the tool
caught, it would have dug into the commutator. The only

400 ARMATURE WINDING AND MOTOR REPAIR

other way by which the rig could be attached was to place it
so as to cut with the tool upside down. Although the rig
was not so steady as it would have been if worked in the cor-
rect position, no difficulty was encountered.

The turning rig used on this job was very easily attached.
It was unnecessary to remove a single brushholder in order
to put it in place. The screws at A in Fig. 252 held the main
part of the rig to the bearing cap. The two screws B were
screwed against the side of the pedestal to steady the end of
the rig. By loosening the screw C the transverse-feed length
was readily adjusted. The end of the transverse-feed arm
was supported by the long screw D resting on a brushholder
arm. To prevent any lateral motion of the armature, the
screw E was held in the center of the shaft by the adjustable
arm F.

Holder for Sandpapering Commutators. In Fig. 253
there is shown a holder for use in sandpapering commutators
that makes it unnecessary to shape a block to fit the curved
surface of the commutator. This device as made by Peter

J. M. Clute (Popular
Science Monthly, June,
1919) consists essentially
of a handle which is
broadened at its lower
extremity, and has two
blocks, 2 by 2j- in., with

an arc of a circle on each

, on, ui i

face. These blocks are

pivoted, and will adjust
and accommodate them-

FIG. 253. Convenient device for sand- i

papering a commutator. selves to an y COrnmu-

tator curvature. Two or

more of the blocks may be used, depending upon the size of
the commutator.

Use of a Portable Electric Drill to Undercut Mica of Com-
mutator. An interesting method of undercutting a commu-
tator on which the mica had come to the surface aftter the
commutator had been turned, has been devised by R. H. N.
Lockyear, of Trail, B. C. (Electrical Record, June, 1918).

METHODS USED TO SOLVE SPECIAL TROUBLES 401

Mr. Lockyear describes as follows the circumstances under
which the electric drill was successfully used for this purpose.

FIG. 254. Special arrangement to use a small portable electric drill to
undercut mica on a commutator.

FIG. 255. Special rigging for the use of an electric drill shown mounted on a
generator to undercut the commutator mica.

"The writer had occasion to undercut a commutator on a
500-kw. generator, the mica having come to the surface.

402 ARMATURE WINDING AND MOTOR REPAIR

after the commutator had been turned. Work had com-
menced by hand but it was found difficult to make head-
way. We had a machine for undercutting railway commuta-
tors having a width of about three inches. This machine was
rebuilt with longer guide rods and a bracket made for mounting
it on the brush shifting yoke of the generator. This machine
was connected with a flexible shaft but soon discarded because
of the vibration. It then occurred to the writer to direct
connect a Van Dorn type DA-OO electric drill, which was
available, to the spindle of a milling machine and undercut the
mica with this equipment. The scheme proved most effective
and since that time we have used it on 26 generators with a
great saving of time, which is an important factor in a power
plant when repairing large generators. In the tool we employ,
the cutter has a speed of 1650 rpm., a diameter of %-in., and a
thickness of 0.003 in."

Jerky Operation of New Commutator Traced to Burred Com-
mutator Bars. In some instances where it has been necessary
to temporarily reduce the speed of a series motor and there
have not been available suitable resistances for connecting in
series with the armature, the desired result has been obtained
by connecting a resistance across the brushes. The resistance
so disposed not only diverts part of the line current from the
armature, but also provides an external path through which
the motor armature can act as a generator, thereby producing a
braking effect. The extreme of this condition is active when
a railway motor flashes over from brush to brush. The cur-
rent generated through the short circuit formed by the arc
is so great that the suddenly imposed load checks the speed of
the armature, the sudden checking of the speed constituting
what railway men call " bucking. " An action which is similar
in kind but which is milder in degree, occurs when an armature
coil is short-circuited. The coil becomes 'a short-circuited
loop moving in a strong field, and the local current of the coil
is so heavy and the drag imposed by it so great that the re-
mainder of the armature may be unable to support continuous
motion. Under this condition the armature will turn in
jerks, because the resisting drag of the short-circuited coil
is greater in some positions than in others.

METHODS USED TO SOLVE SPECIAL TROUBLES 403

In a certain case (Electrical Record, June, 1918) a new com-
mutator was installed on a direct-current armature which had
operated normally up to the time of removing the old com-
mutator. After installing the new commutator, on trying
to start in the usual manner, the armature would turn very
slowly and in jerks, and if the starter were advanced half of
its travel the breaker would blow. An ammeter was connected
in series with the armature, and the current was found to be
of almost full-load value on the first notch of the starter.
A bar to bar test was made around the commutator and nearly
all of the armature coils appeared to have zero resistance.
Just at this time, full-load current being in the armature, which
was blocked so that it could not return, an explosion accom-
panied by a flash occurred on the commutator. Investigation
disclosed a burn between two bars in the groove that is
turned in the rear end of a commutator. Close inspection
showed that the groove appeared as a continuous copper
band extending entirely around the commutator. In cutting
the groove the tool had been allowed to drag the copper over
from bar to bar all round the commutator. On clearing
the mica bodies of the copper bridges, operation became
normal.

Why Brush Studs Heated on an Eight-pole Machine. The
accompanying diagram shows the paths of the current through
a multiple connected arm-
ature that is used in a four-
pole machine, the two brush
studs of the same polarity
being connected by means
of copper busses with ends
strung onto the studs that
are to be connected. The
incoming current arriving

at brush a divides among FIG. 256. Diagram of paths of cur-

three paths in order to reach rent through armature *

brush d to which the other end of the external circuit is con-
nected. One path, ad, is directly through one-quarter of the
armature to d; the second path is through one-quarter of the
armature to brush b and thence by way of bus bd to brush d

404 ARMATURE WINDING AND MOTOR REPAIR

and out; the third path carries as much current as the two
others combined; it runs through bus ac to c where it divides
equally, half passing through the upper right hand quarter of
the armature directly to d and out and the other half through
the upper left hand quarter of the armature to brush b and
thence by way of bd to d and out. With the busses of negli-
gible resistance and the sets of brushes equally spaced on the
commutator all of the studs would carry approximately equal
currents.

On an eight-pole machine there are more current ramifica-
tions because there are more studs and more busses and, there-
fore, there are more paths for the current.

It was on such a machine in a particular plant, that two of
the brush studs would get so hot as to burn the insulating
bushings by means of which the studs were insulated from the
yoke. Investigation disclosed that the machine was abusively
overloaded and that all studs heated above normal but that
only two got hot enough to burn bushings. An inspector
^suggested that the heating was due to the generally poor con-
dition of all contacts and especially to the poor contacts be-
t"-een the stud threads and the threads of the nuts by means of
which the busses, the yoke, the washers and the studs were
held together.

Acting on this theory, V-shaped pieces of copper strap were
formed and so installed on the studs as to form a conducting
bridge 1 from the outside nut and washer over the yoke (with-
out touching the yoke) to the inside nut and washer. This
bridge short-circuited all contacts with which the threaded
parts of the studs were involved. The scheme worked so well
that similar bridges were installed on all of the studs. After
giving the whole stud construction a good cleaning there was no
further trouble.

An Accident Due to Incorrectly Set Brushes. In a case
where an isolated plant was furnishing a meat packer with
electric power at 110 volts for a 10-hp. motor, in order to
provide emergency service, connections were made to the
central station's 220-volt mains. The higher voltage required
the installation of an additional motor with its separate starter.
When the new 220-volt motor was installed no electrical test

METHODS USED TO SOLVE SPECIAL TROUBLES 405

was made until the work had been completed and the belt
put on the pulleys. With all the machines connected to the
line shaft, the motor was started with the following results.
(R. L. Hervey, Electrical World, Jan. 13, 1917). An operator
observed that the speed was excessive and ran to the switch
box, missing by about six inches a piece of the rim of a 36-in.
iron pulley; another piece went through the roof. Before the
motor stopped the line shaft was bent and the boxes thrown
out of line by the unbalancing of the pulleys. The dealer who
sold the motor was held responsible for the damage, and there-
fore had an examination made. The connections were found
correct, but the rocker arm was loose and the brushes were
about 60 electrical degrees back of the neutral position.

Before the motor was started the brushes were set as closely
to the neutral as could be by tracing out the armature leads.
The line and armature wires were taken off the starter to test
the magnetic switches, which were found to be working prop-
erly. The line wire was then re-connected and the shunt field
tested and found complete by observing the flash when the
switch opened and by holding an iron bolt against the field
poles to test the pull of each pole. With the belt off, the
brushes set approximately correct and a current in the shunt-
field, the armature connection was made and the switches
closed, the motor started and ran as nicely as could be wished
for. This little bit of carelessness in not testing the motor
before putting on the belt caused a shutdown for seven hours,
an expenditure of $24 in repairs and endangered the life of one
man.

Wrong Setting of Brushes for Direction of Rotation Caused
Motor to Flash. If line voltage be applied to a shunt- wound
motor without using any starting resistance in the armature
circuit, the immediate result is likely to be the blowing of a
fuse or the opening of a circuit breaker, because the armature
resistance is so low compared to the resistance of the field
winding that the latter is unable to get current for producing
the field on which starting depends. In the case of a small
motor, the fuse may hold and the motor may start without
demonstration because the maximum current involved is small.
A large motor is apt to flash and the brushes to burn owing to

406 ARMATURE WINDING AND MOTOR REPAIR

their being unable to handle an abnormal current. Shunt-
wound motors operate better with their brushes set a little
back of neutral than they do with the brushes set forward of
neutral or on neutral, because with a slight backward shift
field distortion helps commutation.

In one case mentioned by E. C. Parham (Electrical Record,
October, 1918), a 20-hp. motor had been overhauled and
was to be given a light running test. As there was no
starting box available, it was necessary to start the motor
from full voltage through a switch and also a circuit breaker.
On closing the breaker and then closing the switch, there was
a heavy flash at the brushes followed by blowing of the circuit
breaker, but the motor failed to start. Thinking that the
flashing might be due to full voltage being applied to armature
and field connected in parallel and without a starting resist-
ance, the motor was re-connected so that its field became
energized on closing the breaker, the armature receiving cur-
rent on closing the switch after closing the breaker. An
attempt to start again caused flashing and blowing of the
breaker, but this time the armature actually began to turn
and it was observed to have started to turn in the wrong
direction. On reversing the field connection, there was no
further demonstration. All of the trouble was due to the
brushes having a forward shift when the motor was connected
to run in the wrong direction.

Proper Adjustment of a Reaction-type Brush-holder.
The angle at which a reaction-type brush-holder is set on a
commutator plays an important part in the life of the brushes
and commutator. When the brush is set against the direction
of rotation of the commutator, as shown in Fig. 257, the correct
setting of the brush is of greater importance than when set in
the opposite direction. If the brush is set too straight, the
friction will pull it away from the holder and cause a chattering
and sparking. If it is set too flat, there will be a tendency for
the brush to wedge between the holder and the commutator.
Some machine manufacturers using this style of brush-holder
furnish a templet by which the holder can be accurately set,
and in such a case it should be carefully used.

R. L. Hervey (Electrical World, Jan. 22, 1916) relates a case

METHODS USED TO SOLVE SPECIAL TROUBLES 407

where he was called upon to locate the cause of sparking on a
generator using this type of holder. Upon making inquiries
it was learned that the engineer had purchased a sheet of high-
grade carbon, so called, and made his own brushes. To save
carbon and work, both ends were made with the same angle
and the holder was turned to suit the brush. The brush
length had been made so great
that the arm pushed the brush
away from the holder instead
of against it. The length L
(Fig. 257) should never be so
great that end of arm B will be
above the line AO drawn per-
pendicular to the face and
through the arm center 0. The
manufacturers of this holder
recommend that the holder be
set as shown; nevertheless, there
are a .great many installations
working in the opposite direc-
tion.

Heating of Brush-holders Traced to Defective Contact
Springs and Remedied by a Flexible Shunt. R. L. Hervey
(Electrical World, July 1, 1916) relates the following conditions
in a case where his attention was called to the discoloring and
heating of brush-holders on several generators. The holder
was of the parallel-motion type, as shown in the sketch,
Fig. 258. To prevent the burning of the hinge joints of
the parallel motion, phosphor bronze springs 0.010 in. thick
were used to carry the current around these points. As long
as the contact between these thin springs and the moving
arms was perfect no trouble occurred. But as soon as any
part of the contact became defective an overload was thrown
on the remaining contacts, causing overheating. The total
contact area of the springs on the holders of the same polarity
was 0.360 sq. in. fo'r 250 amp. A properly designed contact to
carry this amount of current should have 1% sq. in. of surface
To relieve the trouble, the springs were taken out and replaced
with a flexible shunt, as shown in the sketch. In such a case

FIG. 257. Diagram showing
proper setting for reaction-type
brush-holder.

408

ARMATURE WINDING AND MOTOR REPAIR

it should be mentioned that unless the soldering of the flexible
wire into the lug is very carefully done the solder will get into
the wire and stiffen it. In order to prevent this, the shanks
of the lugs were flattened with the wire in it, forming a very
strong electrical joint.

The box type of construction of this holder afforded a sub-
stantial contact between the holder and the brush. Due to the
weight of the moving box and its supporting arms, however,
this brush-holder was displaced by a stationary holder and
sliding brush with a light spring. With a high commutator
speed, the heavy brush and holder does not follow the irregular-

Woren ff/re Shunf-.

Sen*

Spring

FIG. 258. Substitution of woven wire shunt for springs of parallel motion

brush-holder.

ities of the commutator as well as the lighter brush. The
brush that can move in its holder is better able to follow the
grooves of the commutator as the armature weaves backward
and forward in the bearings. Although the brush-holder is
usually considered a minor part of the motor or generator, it
has held the attention of the designing engineer for a good
many years and still needs the attention of the operator and
repair man.

Simple Scheme for Banding Armatures. The method
shown in Fig. 259 has been found convenient in one repair
shop (G. H. Vescelius, Electrical World, July 19, 1919, page
132) where a banding lathe and tension device was not avail-
able and only one man to do the work. In using this scheme
the free end of the banding wire is soldered to the armature

METHODS USED TO SOLVE SPECIAL TROUBLES 409

anywhere and the latter rotated by means of a "dog" and
pipe as indicated in the illustration until the wire has filled
all of the banding grooves. In performing this operation no
tension is exerted on the banding wire. Next, the remaining
end of the wire is soldered to the armature and a weighted
pulley hung on a loop of the banding wire nearest this end.

By selecting a weight twice as heavy in pounds as the de-
sired tension in pounds per square inch the wire will be drawn
up to the correct tension by rotating the armature backward.
At any time when it is necessary to adjust the position of the
banding wire this can be done easily by the person turning
the armature because the tension will not be relieved as would
happen when banding in the ordinary manner.

FIG. 259. Weighted pulley in place for producing tension in banding wire.

Use of a Crane to Band an Armature in an Emergency.

In a large railway substation it was necessary to replace the
banding wire on the armature of a 1000-kw. rotary con-
verter. The bands required about 20 turns of steel wire. The
scheme tried at first was to mount the steel wire reel so as to
turn- freely and pass the wire through a tension regulator con-
sisting of two fiber-lined blocks bolted together. The wire
was then wound on the armature as evenly as possible by
turning it with bars, a method which was found slow and
clumsy. The rotary was of the type that is brought up to
speed with an induction motor with its rotor mounted on the
shaft of the rotary armature. The idea was therefore con-
ceived of removing the stator of the induction motor and using
the rotor as a drum or pulley to rotate the rotary armature.

410 ARMATURE WINDING AND MOTOR REPAIR

Three turns of a 1.25-inch rope were wound on this rotor and
one end fastened to the hook of a traveling crane. The crane
was then run the full length of the station with one man feeding
the rope to the drum and keeping it taut. In this way a
smooth, tight band was wound on the rotary armature with
considerable saving in time over turning it by hand.

Method Used to Band a 2000-Hp. Rotor. The banding of
armatures and rotors, when done in a factory where such jobs
are an every-day occurrence, has been more or less reduced to a
science and the methods of procedure are many. When,
however, the rotor to be banded reaches the 2000-hp. mark,

Banding
Wire

Wire Tens/on Device. Spool.

- Motor.

Crane
Trolley

'-Flywheel.

FIG. 260. An arrangement devised to band a 2000-hp. motor armature.

it is out of the question to think of transporting it to the repair
shop. Such a problem came up in a steel mill at one time when
a 2000-hp. motor which operated the rolls became seriously
grounded. The following method was decided upon (Maurice
S. Clement, Electrical Record, April,. 1919) : The entire trolley
of a traveling crane was lowered to the floor and placed in line
with a huge fly-wheel which was On the same shaft as the rotor.
The cable drum of the crane trolley was converted into a belt
drum by covering it with canvas which was held in place by
means of friction tape. This made a more or less flat surface
and eliminated both the cable grooves and the grease from the
face of the belt.

As the lower half of the fly-wheel was in a pit, it was nee-

METHODS USED TO SOLVE SPECIAL TROUBLES 411

essary to place a roller at the edge of this pit as a belt runner.
Another roller was placed on a frame somewhat resembling
a large brace, both ends of which were secured to the floor,
halfway between the crane trolley. The fly-wheel was ar-
ranged so as to ride the belt and serve as a belt tightener. The
banding wire tension device used in this case was similar in
principle to those used on wire nail machines, the sheaves
being adjustable in the slots so as to tighten or slacken the wire
at a moment's notice.

An Improvised Method Used to Turn a Commutator.
In a case where a rush order was received by a repair shop
to turn the commutator of a generator at a large country
house, R. L. Hervey (Electrical World, June 10, 1916) de-

FIG. 261. Method of using lathe cross-feed in turning a commutator.

scribes the following method of doing the work when the small
turning rig which was taken along was found too small. In
the garage, however, a large engine lathe was found. This
started an examination to determine whether the armature
could be taken out and swung in the lathe. This promised
impossible. While looking around the shop an angle plate
was uncovered. This plate and the cross-feed of the lathe
were bolted together by a special screw of the rig. The plate
was then bolted to the end plate of the generator frame, as
shown in Fig. 261. There was but one bolt holding the angle
plate on the generator frame, which made the equipment look
very unsteady. A hole in the angle plate was selected so that,
if the plate turned around the bolt, the tool would swing
away from the commutator and not into it. The screw of the

412 ARMATURE WINDING AND MOTOR REPAIR

cross-feed was too short to take a cut across the commutator,
therefore the tool had to be re-set for each cut. This was a
little troublesome as the tool had to be tapped into position
with a hammer. The commutator had grooves in it about
% inch deep, requiring considerable cutting. However, the
job was satisfactorily done with this arrangement.

The generator was direct connected to a low-speed horizontal
gasoline engine. While working on the commutator the valves
were taken out to prevent compression, and the turning effort
provided by a 1-hp. motor belted to the fly-wheel. The elec-
trical equipment of the plant consisted of the generator and a
storage battery of 58 lead cells. These supplied current for
the lights, a motor-driven ice machine, laundry, water-pump
and machine tools in the garage.

Cause of Motor Reversing its Direction of Rotation on
High Speed. R. L. Hervey has described the following pecu-
liar performance of a motor (Electrical World, Jan. 20, 1917) :
A 2-hp. 220-volt direct-connected compound motor driving a
band saw had been in service for approximately 18 months
without giving the slightest trouble until the operator
wanted to work the saw at its highest speed, which was ob-
tained by using a field rheostat. When all of the resistance
was cut in the field circuit the motor slowly came to a stop and
started in the opposite direction. This was tried several
times, and resulted in blowing the fuses most of the times.
There could be but one condition that would cause this re-
versal of direction of rotation, namely, the compound winding
being in opposition to the shunt field and overcoming it when
the shunt field was weakened. By reversing the compound
field leads the trouble was removed. This motor had been in
operation for a year and a half without any indication of
anything being wrong. The brushes and commutator were
in excellent condition, which speaks well for the design of the
machine.

Checking Connections of an Inter pole Motor. The polarity
of the interpole field coils should be the same as that of the
preceding main poles in a direction against the rotation. The
polarity of the interpole field with relation to the main fields
for both directions of rotation of motors and generators is

METHODS USED TO SOLVE SPECIAL TROUBLES 413

shown m Fig. 262. To determine the proper connections of
interpoles proceed as follows:

As a rule one end of the interpole field and one armature
lead are permanently connected together, and since the correct

FIG. 262. Polarity of interpole fields with relation to main fields for direc-
tions of rotation of motors and generators.

polarity is determined by the manufacturer, there is no oc-
casion to change it. To reverse the direction of rotation the
armature and the interpole field must be reversed as a unit.
Reversed interpole connection is not noticeable at light load.
As load is applied, a motor with reversed interpoles will spark

Line

Series Ft'efd * 7/tf Field

Armature Armature

FIG. 263, Connections for testing out correct connections of interpole
windings for proper armature rotation.

badly and drop in speed, while on a generator difficulty will be
experienced in maintaining rated voltage. Assuming that one
end of the interpole field and one armature lead are perma-
nently connected, wire the machine temporarily as a series
motor with the shunt field disconnected. Care should be taken

414 ARMATURE WINDING AND MOTOR REPAIR

to apply the current only momentarily to prevent too high
an armature speed.

The connections for this test are shown in Fig. 263. If direc-
tion of rotation is wrong, interchange terminals 3 and 4 or 1
and 2. Having secured proper direction of rotation with the
series connection, insert the shunt field and open the series field
as shown in Fig. 263. Upon starting up again as a shunt motor,
the direction of rotation should be the same as it was with the
series field. If it is not, interchange terminals 5 and 6 and
make up permanent connections. The correct polarity of the
interpoles may then be checked with a compass or by bridging
a main pole and interpole with a piece of iron, bearing in mind
that unlike poles attract and like poles will repel it and check-
ing with the diagrams in Fig. 263.

Heating of Field Coils Traced to Wrong Type of Starting
Box. On motors that are designed for field control of speed,
E. C. Parham has pointed out (Electrical Record, February,
1919) that the shunt winding is likely to be lighter than that
of a motor where the shunt winding is to be subjected contin-
uously to full operating voltage. Therefore if the shunt field
of such a variable speed motor is connected to full voltage con-
tinuously, it will heat beyond the usual temperature guarantees.
If the line voltage happens to run well above normal, as it often
does, the temperature will reach a value liable to injure the
insulation and ultimately to short-circuit the winding.

The owner of a machine shop at one time asked that a man
be sent to find out why the speed of his motor was so low and
why the motor got so hot after several hours of working. As
this was all the information obtainable, the inspector took a
voltmeter and an ammeter that could be used on either direct-
current or alternating-current circuits, only to find out later
that no instruments were required. The motor in question
proved to be of the variable speed, field control, type. There
was no name plate on the motor, but the type was inferred from
the type of resistance box which had been furnished with it
but which was not being used. The motor was being operated
as a constant speed motor on voltage that was 15 per cent,
above normal. Of course the maximum speed obtainable was
that due to full field hot. The field coils heated too much but

METHODS USED TO SOLVE SPECIAL TROUBLES 415

did not heat sufficiently for their increased resistance to give
the speed required. In order to bring the speed to the nec-
essary value and at the same time relieve the field coils, the
resistance box, which had many taps, was connected in series
with the field coils and different tap wires tried until one was
found that gave approximately the proper speed. As the
load of the motor was comparatively light there were no ob-
jectionable features attending the starting of the motor on a
weakened field, so the connection was made permanent with
the understanding that the operator would immediately get a
starter that was adapted to the work.

Safe Operating Temperature of Portable Desk Fans.
A well-known manufacturer of fan motors points out that
warm weather always brings a number of complaints in re-
gard to heating of small motors. These complaints are usually
from dealers or users who are possessed with the idea that if
the motor-body does not feel nice and cool to the hand, the
windings are in imminent danger of burning out. The latest
standardization rules of the American Institute of Electrical
Engineers provide that, in motors with the class of insulation
used in reliable makes of motors, the temperature as recorded
by a thermometer, should be within a limit of 80 C. This
is equivalent to 176F. No one would care to place his
hand in contact with the frame of a motor at anywhere near
that limit.

Let us say, for instance, that a small motor has a tempera-
ture rise of 40F. when operated continuously. Such a motor
operating on a day when the normal air temperature is 70,
would reach a temperature of only 110, which would feel only
comfortably warm to the hand. The same motor, operating
on a mid-summer day when the thermometer runs from 90
to 95, will attain a temperature between 130 and 140 F.
Any temperature over 120 is quite uncomfortable to the
touch and gives rise to alarm on the part of the inexperienced
motor user.

However, a motor with a surface temperature of 140
is in no danger of injury from overheating, and motor
users who make complaint of the heating effect under such
conditions should be advised to continue operating their

416 ARMATURE WINDING AND MOTOR REPAIR

motors until it is apparent that the windings are being dam-
aged. Unless there is some odor of burning insulation, the
motor can not be considered in danger.

An Adjustable Shunt for Series Fields of Exciters. In
adjusting the compound characteristics of three exciters for
parallel operation in conjunction with a Tirrill regulator, it
was found impossible to get the necessary series-field shunt
adjustment with the shunt taps provided by the manufac-

' 0.010 'German Silver

Boned to Facilitate Ventilation

Toffacn/ne. To Conduit
To^Series Field terminal^ \

Section cut from one
Strip to obtain Fine
Adjustment-..

Series Shunt which was Pep faced

FIG. 264.- Construction of series-field shunt showing bowed formation to
facilitate ventilation.

turers. Consequently each shunt was replaced by a very
simple -and easily adjustable shunt like that shown in Fig. 264,
(Sydney Fisher, Electrical World, June 10, 1916). By means
of these the characteristics of the machines were made practi-
cally identical. The shunt was made of strips of German
silver slotted to permit the bowed formation shown. Rough
adjustments were made by adding or removing strips, and
finer adjustment obtained by varying the cross-section of

METHODS USED TO SOLVE SPECIAL TROUBLES 417

one of the strips. The free passage offered to air through the
shunt permits good ventilation with resulting low temperature
from no load to full load. The construction of the shunt
permited using the minimum number of connections, therefore
they can be made much more readily than where more con-
nections are required. This is an important consideration
since the resistance of poor contacts becomes appreciable,
the series-field resistance being small.

A Peculiar High-speed Motor Trouble. On one occasion
R. L. Hervey (Electrical World, Oct. 28, 1916) explains that
a 500-volt 10-hp. direct-current motor, running at a speed of
3400 rpm., continued to open its circuit breakers every time
an attempt was made to start it. The wiring contractor,
after assuring himself that the connections were correct and
that the trouble was inside the motor, refused to give more
time to the job. The local representative of the machine
company claimed that every machine turned out by his
factory was given a 24-hour test, therefore the trouble
could not be in the motor.

The motor was a shunt-wound, bipolar design, with two
commutating poles. The three leads were marked A, F,
and C, which were assumed to mean armature, field and line.
After testing out the wiring between the starting box and the
motor, the connections were made as stated above. Since
the inertia of the machine parts to be started was large and
the speed quite high, a large starting current was expected,
so that the circuit breakers were set for their maximum current.
An effort was made to start the motor to test the connections
and also to observe its action when the voltage was applied, an
ammeter having been connected in the circuit so that the
starting current could be measured. When the starter
handle reached the third button the current was 50 amp.
The circuit breakers opened and the starter was smoking
badly. These results indicated an open-field circuit. The
internal connections of the motor were checked with consider-
able difficulty. The lead marked A was found to be the line
and C the armature. After making this change the motor
was started again with the same results. The machine
representative objected to the motor being taken apart, but

418 ARMATURE WINDING AND MOTOR REPAIR

yielded to persuasion. A small resistance tube connected in
the shunt-field circuit was found in the bottom of the frame.
While this tube could not be reached for examination, it was
suspected of being the cause of the trouble, so a jumper was
connected around it. Another effort to start the motor was
successful. The motor speed was 3350 rpm. with a line
voltage of 575. The resistance, after being patched, was con-
nected in the field circuit when the speed measured 3730 rpm.
As the motor was built for 3400 rpm. at 500 volts, it was
thought that the factory test was made at that voltage and the
resistance put in the field to bring up the speed. Since, how-
ever, the voltage was 575 at the place of installation, the speed
of 3350 rpm. was high enough and as the resistance had open-
circuited once, it was decided to leave the resistance out to
prevent future trouble.

Ways that End-play Variations Show Up. The principal
purpose of having end-play in the armature of a motor or of a
generator is to keep the brushes from tracking in the same
path and thereby wearing a groove in the commutator or in the
collector rings as the case may be. To prevent such wearing
of grooves, the armature end-thrust clearances are so disposed
that when the armature is running in its normal zone, which is
governed by the pull of the pole-pieces on the armature core,
the thrust clearances at the two bearings are equal. Assuming
the correct gear or pulley alignment to the connected load, the
armature will never run sideways far enough to knock the
bearing on either side; because when it has run over a certain
distance, the field will pull it back toward and past the mag-
netic center. The same action will then be repeated but in
the opposite direction. If, however, the end-thrust is un-
equally distributed for any reason, or if the machine as a
whole is not level, knocking will occur which in the case of
small, high-speed machines may take the form of serious
vibrations that are likely to pound down the bearing
linings.

E. C. Parham (General Electric Review, January, 1916) has
mentioned a case where four motors of fractional horsepower,
which were direct connected to tool grinding wheels, were
installed and three of the outfits gave entire satisfaction, but

METHODS USED TO SOLVE SPECIAL TROUBLES 419

the fourth one vibrated so badly that it shook the whole
supporting structure. The rotating elements of two of the
units were then interchanged and the vibrations remained
with the same rotor. Spinning the rotating element of
the troublesome machine in a high-speed lathe proved
the shaft to be perfectly straight. Close inspection dis-
closed that the end-play was not distributed on the faulty
outfit in the same manner as on the faultless ones. All
vibration was eliminated by loosening the thrust collar so
that the rotor when running could center itself, then starting
the motor and letting it slow and stop, and then tightening
the collar.

Connections for Two 220-volt Motors When Operated on
440 Volts. It is practical to operate two 220-volt 7-hp.

Mofor-pu//ey

fields are to be separ-
ate, disconnect at X and
run f. to field on start i,

''-Driver

Driven-

Fields

FIG. 265. Belt arrangement for two motors operating in series and diagram
for connecting two shunt motors in series.

motors in series on 440 volts and secure satisfactory results,
providing that the motors are approximately the same as
regards windings and speed. If the drive is by a belt and the
motors run at the same speed a good plan is to set the motors
with the pulleys in line and run a short belt from one to the
other. That is, first belt the motors together by a short belt
and then put the regular belt over the short belt as shown in
Fig. 265. Another way is to couple the two armature shafts
together with a suitable coupling and attach the driver pulley
on one of the armature shafts.

The connections for two shunt-wound motors in series are
shown in Fig. 265. The connections for two compound-wound
motors in series and the connections for one shunt-wound

420 ARMATURE WINDING AND MOTOR REPAIR

motor and one compound-wound motor in series in Fig.
266. (Frank Hoskinson, Electrical Record, September, 1917.)
In case one of the motors is higher in speed than the other
one, it may be advisable to use different sizes of pulleys so as
to secure the proper speed.

Starting Box
Ling Field .Arm.

Starting Box

Field\ ^ 7 -

Shunt Fie/cfs

Shunt Fields. Series Fields

Shunt Motor Compound Motor

FIG. 266. At the left, diagram for connecting two compound motors
in series and, at the right, series connections for one shunt motor and one
compound motor.

Cleaning Motors with Compressed Air. When using com-
pressed air to clean out dirt and dust from motor windings
the pressure should not be higher than 100 Ibs. gauge. The air
can be applied by the use of a rubber hose about J^ inch in
diameter fitted with a short piece of ^-inch iron pipe to act
as a nozzle and direct the air into the windings so as to blow
out accumulated dirt.

Most motor windings are impregnated or painted with some
sort of insulating varnish and this usually presents such a
smooth surface that air, even at extremely high velocity, is
not likely to lift the edges nor tear the insulating fabric. If
the dust that accumulates in the motor is of an abrasive char-
acter, it is by all odds more advisable to use a higher pressure
than mentioned and get rid of the dust than to let the dust pile
up until it stops ventilation of the windings and perhaps even
cuts the insulation.

Testing out Phase -rotation. Manufacturers of alternating-
current generators generally have a standard direction of rota-
tion clockwise, for example and likewise there is adopted
a standard direction of phase-rotation. This is necessary,
because alternators that are to be operated in parallel must
have their phase-rotations the same. The adopted phase-

METHODS USED TO SOLVE SPECIAL TROUBLES 421

rotation bears a definite relation to the numbering of the
alternator terminal blocks. For example, the blocks of a
three-phase alternator would be numbered 1, 2 and 3 and the
standard phase-rotation would be in the same order. The
phase-rotation of every alternator produced is tested by means
of a phase-rotation tester. The tester is virtually an induction
motor in fact, a standard motor could be used as a phase-
rotation indicator. The rotor of the tester, however, is simply
a vane of iron free to turn under the influence of the stator
magnetism. The stator winding of a three-phase indicator
has three terminals, which are marked 1, 2 and 3. These
terminals are connected to corresponding terminals of the
alternator the phase-rotation of which is to be tested. Re-
duced voltage is applied to the stator of the alternator and
the direction of rotation of the vane of the indicator shows the
phase-rotation of the alternator.

When an induction motor is used as a phase-rotation indica-
tor it should be connected first to the supply end of one ma-
chine and then to the other. If the motor rotates in the same
direction it shows the phase-rotation of the two machines to
be the same. If the rotation reverses, one phase of one of the
alternators should be reversed.

An Induction Motor Trouble Due to Wrong Stator Con-
nections. The importance of checking up connections of
induction motor windings is shown by the following trouble
experienced by A. C. Hewitt (Electrical World, Jan. 16, 1916).
In a cement plant where a 50-hp., three-phase, 440- volt,
60-cycle squirrel-cage induction motor operating at 450
rpm. was used to drive a 33-inch Fuller mill for pulverizing
limestone the motor on several occasions was overloaded
by feeding the material to the mill too fast and in large sizes.
This overloading, combined with fluctuations of voltage as
much as 30 per cent, below normal rating, caused the motor to
heat and finally called for the replacing of several stator coils.
The motor was repaired by the plant electrician and again
placed in operation, but still it ran hot, with the mill showing a
smaller output than a duplicate installation where the motor
was running cooler.

When called to investigate the installation the repairman

422 ARMATURE WINDING AND MOTOR REPAIR

noticed that the upper half of the motor was much hotter than
the lower half. An inspection of the connections seemed to
show that they were correctly made and that there were no
grounds or open circuits. The fuses were in good condition
and the auto-transformer motor starter operated satisfactorily.
All of the connections for the starter windings checked with the
diagram furnished with the motor.

The winding was a two-circuit delta arrangement and
the motor had 16 poles. In checking over the number of
poles in each of the two circuits it was finally found that the
winding was divided in a horizontal plane through the stator,
and that instead of each circuit having eight poles the upper
circuit had only seven and the lower had nine poles. This
explained why the upper half ran hotter than the lower. The
electrician had made a miscount when connecting up the
windings, with the result that the load was unevenly distrib-
uted. The connections were changed so that each circuit
contained an equal number of poles, and the motor operated
without the heating trouble previously experienced.

Stalling of Wound Rotor Induction Motor Explained.
A complete open circuit in the wound rotor of a polyphase
induction motor, will prevent its starting because under such
conditions there can be no secondary current. Such a com-
plete open circuit may be caused by two of the rotor brushes
being so stuck in their holders as to make no contact with the
collector rings. An exception might obtain in the cases of
fractional horsepower motors which are so small that the
rotor may start by virtue of the eddy currents induced in
the rotor laminations. However, the torque due to such re-
actions could support no load. The function of energy circuit
of a repulsion induction single-phase motor is similar to that of
the wound circuits of the rotor of a polyphase induction
motor, in that both are the seat of the induced current by
virtue of which the motor is able to do its duty Therefore
a complete break in the energy circuit of a repulsion induction
motor, will render the motor unable to start. That an in-
complete open circuit in the energy circuit, may produce
different results, is illustrated by the following experience.

A butcher complained that his motor-driven meat grinder

METHODS USED TO SOLVE SPECIAL TROUBLES 423

which had been working satisfactorily for months, was develop-
ing a tendency to slow down when coarse stock was fed into
the hopper. With the finer stock, the motor apparently could
do its work all right. On operating the motor with the end
cover removed, vicious sparking was seen. At first this was
thought to be caused by armature trouble but a close inspec-
tion disclosed that the sparking was due to the energy brushes
being stuck in the brush-holders. The brushes had worn so
short that the brush shunts were jammed into the boxes and a
few more jobs of grinding probably would have burned and
worn the brushes entirely out of contact with the commutator.
Even with the arcing, the energy current had been sufficient
to support the lighter loads, although after installing new
brushes, it became evident that the motor had been operating
at reduced speed at all loads.

Loose Bearing Caused Induction
Motor to Fail to Start. H. Wilson
(Power, August 5, 1919, page 231)
describes the following experience
after repairing a large three-phase
induction motor, when an attempt
was made to put it back into
service. On closing the compensa-
tor switch to the starting position,
the motor failed to start, although
it was evident from the sound of
the machine that it was getting
current through its winding. The
first thing that suggested itself was
an open circuit, so the starting com-
pensator was tested to see if the
current was coming through single-
phase only. In order to do this
quickly, we disconnected the motor
lead at the machine, and while the
switch was held on the starting
position a test lamp was connected across the different leads,
as in Fig. 267. This showed current on each phase, at a
reduced voltage, of course. We then put the switch on the

3-Phase

Motor

FIG. 267. Method of testing
for open-circuit.

424

ARMATURE WINDING AND MOTOR REPAIR

running position and the test
lamp burned brightly across
each phase, showing that the
motor was getting the current
all right.

The rotor seemed to be rub-
bing a little on the stator, as
it was somewhat hard to turn,
so the clearance was adjusted
by means of two draw-bolts
on the bearing housings. This
took some time, since when the
bearings were moved one way
a little, the rotor would bind;
then we would shift it back
slightly until finally getting it
into a position where the rotor
turned freely. Another at-
tempt to start motor met with
no better result than the first.
The next step was to trace out
the winding connections, which
were found to be apparently
correct. The winding being
connected two parallel star
made it somewhat complicated
to trace out. However, by
looking to see if the connec-
tions went under or over the
winding, I would mark an
arrow on the group of coils to
show the polarity. We went
around each phase in this way,
starting with the outside lead
in each case. The arrows on
each group pointed alternately
in opposite directions, show-
ing the connections to be cor-
rect, as in Fig. 268.

METHODS USED TO SOLVE SPECIAL TROUBLES 425

The next suggestion was to test the polarity of the winding
with a compass, when direct current was flowing through the
coils. As a source of direct current an automobile starting
and lighting battery was pressed into service. The three
motor leads, A, B and C, Fig. 268, were joined together and
connected to one terminal of the battery. A wire from the
other battery terminal was taken to the common or neutral
point on the winding, where the three phases connected
together, such as A*, B* and C*, Fig. 268. This connection
was not attached permanently, as it would have run the
battery down, owing to the low resistance of the circuit through
the winding, but was attached each time only while the swing
of the compass opposite each group was obtained. This test
proved the connections to be correct, as the compass needle
pointed in opposite directions on alternate groups as it was
moved around the winding.

Since the polarity of the coils was undoubtedly correct, it
left us all puzzled as to the cause of the trouble, as we had
tested for grounds in the winding, open circuit, polarity,
tested for the power at the motor terminal, and also mechanic-
ally, to see that rotor was free.

We had about decided to give the job up and send the
machine to the manufacturers to have it fixed up, when one
of the engineers came in on the job and after hearing our
story remarked that the motor had given similar trouble,
two or three years previous, after being repaired, due to the
bearing letting the rotor rub on the stator. This gave us an
idea that there might be a slight looseness in the bearing,
sufficient to let the rotor lift up and rub on the stator. Acting
on this, the bearing bolts were adjusted until the rotor would
just clear the bottom of the stator. This proved to be the
source of all the trouble; on closing the compensator the motor
started up and ran satisfactorily, greatly to the relief of every-
body concerned.

Three-phase Motors used on Single-phase Lines. It
often happens that utility customers have a supply of three-
phase motors when only single-phase service is available, or
that a customer is asked to purchase single-phase motors until
the load becomes large enough to justify a three-phase line

426

ARMATURE WINDING AND MOTOR REPAIR

extension. This difficulty can be overcome by a system pat-
ented by Professor Arno in which the inherent characteristics
of three-phase motors are developed from a single-phase source
of supply.

The system, which is used extensively in Australia, calls for
the use of a three-phase master motor in addition to the three-
phase power motors. The master motor is a standard machine
of either the squirrel-cage or wound-rotor type, and the larger
the better. The lower size limit of this motor is in practice
about 10 to 15 per cent, of the total load connected to the sys-
tem and is at least double the size of the next largest motor.

MASTZRHOTOR S^IKRIL CA6f HOTOK5 iUPKlNOHOTOK SOUIRXLCAGtJIcm

FIG. 269. Method of operating 3-phase motors from single-phase line.
The four motors to the right are power motors, and the one to the left is a motor
which has its third phase connected to the third phase of all the other motors. The mas-
ter motor is started up as a split-phase induction motor and runs light. Its function is
to supply auxiliary current to the third phase of the other motors, and by this means
the 3-phase power motors are given practically the same characteristics as if they
were operated from a 3-phase line.

All motors, including the master motor, are connected to the
single-phase supply mains, and the third phases, which are
not connected to the source of supply, are connected together.
The master motor is started up under no load with special
starters equipped with an auxiliary winding similar to that
used for the starting of split-phase induction motors. The
master motor should as a general rule run unloaded as a phase
giver. It is permissible, however, to load the master motor
up to about 25 per cent, of its normal rating in special cases,
but even when the master motor is running mechanically
loaded, it may be electrically overloaded. Any of the power
motors becomes a master motor immediately its load is thrown

METHODS USED TO SOLVE SPECIAL TROUBLES 427

off, and if the master motor cannot properly perform its func-
tion, a small unloaded power motor may become overloaded
in helping out the master motor.

Through the voltage induced in the master motor's third
phase, which is not connected to the supply mains, such motors
at full speed are able to supply auxiliary current to the loaded
motors during starting and overload periods. As long as the
master motor is running the overload capacity of the loaded
motors is raised, as compared with purely single-phase induc-
tion motors, and the three-phase motors are able to start
up from the single-phase line with practically the same start-
ing torque as they would have when connected to a polyphase
line. Should all the motors including the master motor be
loaded to the same extent in proportion to their rated capacity,
no current would flow in the third-phase connections. After
the master motor has been started up, the working motors can
be started one after another as regular standard three-phase
motors. The gain in efficiency by operating the individual
motors as three-phase units is practically offset by the losses
in the master motor, so that the overload efficiency of the
whole installation is about the same as that of a straight single-
phase system. The normal output of the power motors when
operating in this manner is about 75 to 80 per cent, of their
standard three-phase rating.

An Apparent Overload Trouble That was Traced to a Defect-
ive Fuse Block. R. L. Hervey (Electrical World, June 24,
1916) relates the following trouble at a plating plant where a
5-hp. single-phase, induction motor, from apparent overload
continued to blow fuses for several days and then refused to
start. The motor had been in daily use for six years and the
repairman, after examining the equipment, said that new
bearings were needed as the rotor was rubbing the stator
laminations. The shaft was also cut and grooved, which
required it to be turned down so that the bearings could be
properly fitted. After the motor had been replaced and in
service for a few days the fuses started blowing again.

The repairman found one of the bearings hot, which was
taken out and " eased up" a little. A few days later another
shutdown occurred. This time the repairman reported that

428 ARMATURE WINDING AND MOTOR REPAIR

the motor was overloaded, and a larger motor was required.
As the plant was being operated exactly as it had been for three
years without a shutdown, the owner of the motor would not
accept the report, and asked that a thorough examination
be made to locate the trouble.

When putting in new reinforced fuses it was noticed that
the fuse clips and block had been quite hot, for they were
discolored and loose. The clips were closed up until they
held the fuse tightly. A close examination of the motor and
starter showed them to be in first-class condition. An amme-
ter was connected in the line and the motor started with no
load. The current taken indicated that the winding was not
defective. The load of two buffing wheels was then put on
and the needle of the 50-amp. meter went off the scale. A
further examination ' of the fuse block showed two loose
connections. It was the heat generated by this high contact
resistance that was causing the trouble with the fuses. It is
very probable that while the rotor was rubbing the stator
the motor was taking an excessive current and injured the
fuse block. This instance shows the need of a careful inspec-
tion of all wiring when a repaired motor is again connected
in circuit.

Cause of Noise in a Three-phase Motor Driving an Exhaust
Fan. The frequency of the current that is applied to the
stator of an induction motor in commercial operation, ordinar-
ily is constant within narrow limits if the speed of the prime
mover to which the current is due, is as constant as it should
be. The frequency of the current that is generated in the
rotor of the motor, however, varies with the speed of the rotor,
being greatest when the rotor is at rest and least when the
rotor is operating at maximum speed. This is because the
frequency of the rotor current depends on the difference of the
speeds of the rotor and of the rotating field. In other words
the frequency of the rotor current depends on the slip of the
rotor. The principle of operation of frequency indicators of
the vibrating type is, that the pendulum properties of each
reed are such that it will respond synchronously to the im-
pulses of but one value of frequency. On the same principle
if a miscellaneous lot of pieces of iron and steel be subjected

METHODS USED TO SOLVE SPECIAL TROUBLES 429

to the flux of an alternating current of varied frequency, some
of the pieces will take up vibration at one value of frequency
and other pieces will not be effected by that value of frequency
but will take up vibration at some other value of the frequency.

In one case a large three-phase induction motor that was
connected to an exhaust fan, was complained of on account of
noise emitted and transmitted to all parts of the building when
the motor was being started. The noise was described as a
" clatter. " An investigation disclosed that the noise occurred
between narrow speed limits obtainable on the second notch
of the controller and the sound was similar to that of rubbing a
tin can with a piece of sandpaper. By throwing the controller
back and forth between the first and second notches, there
was obtained an average speed at which the noise was main-
tained almost continuously. By listening with the ear applied
to different parts of the machine, and at the same time touch-
ing the machine here and there with a lead pencil, the source of
the trouble was located in one of the slots in which slide the
bolts that hold the motor frame to the sliding base. On each
side of the machine extending from one bolt to the other, a
steel strip was provided in order to keep the head of the bolt
from turning when there was occasion to turn its nut with a
wrench. The strip was so " bellied" that there was about J^g
inch clearance between the middle of the strip and the bottom
of the machine which effectively constituted a reed supported
at both ends, and the time element of the construction was
such that the strip vibrated in response to rotor current fre-
quency that obtained on the second notch of the controller.
On removing the strip and re-bending so as to bring the bulged
part up against the bottom of the rail, no further trouble was
experienced with noise.

Cause of a Burned Out Starting Winding in a Single-phase
Motor. Unless of the commutator type, a single-phase
motor, in order that it may be started without manual assist-
ance, must include some phase-splitting device that is effective
in producing the rotating field so necessary to self -starting
of such motors. The phase-splitting device usually takes the
form of an auxiliary starting winding and of an arrangement
of resistance and inductance or of resistance and reactance,

430 ARMATURE WINDING AND MOTOR REPAIR

so disposed that the currents in the main stator winding and
in the starting winding shall be out of phase with each other.
As the starting winding is intended to be active for only a
second or more, it is not proportioned to stand continuous
application of the voltage. Accordingly, a centrifugal switch
that turns with the rotor, is provided for automatically cutting
out the starting winding as soon as the rotor approaches full
speed. If for any reason the starting winding is left in circuit
for appreciable time, the winding is likely to be injured.

In a certain instance (Electrical Record, August, 1918) one
of these motors stopped a few minutes after having been
started for the first time and, as the owner expressed it, "the
motor had no more life in it." The motor was of the resistance-
inductance self -starting type, in which the phase splitting was
accomplished without the use of any external resistance or
reactance. Examination of the motor disclosed that its
failure had been due to the main stator winding being in
series with the automatic switch which should have included
only the starting winding. The motor would start all right,
but the opening of the centrifugal switch would cut out the
main winding, leaving the starting winding connected across
the line. It developed that the motor had been bought at a
low price because it was in bad condition and had been re-
wound by a local repair man, who did o good job of winding,
but a poor job of connecting.

Cause of One Motor Failing to Start While Another was
Running on the Same Circuit. At an industrial plant where
440- volt, three-phase, 60-cycle service was used to operate
motors of many different sizes, a peculiar trouble was experi-
enced by A. C. Hewitt (Electrical World, Jan. 8, 1916) when
one of three fuses was blown on a feeder circuit. One feeder
circuit supplied some four or five motors, one of which was
rated at 75 hp. and the others at 10 hp. each. Each motor
was equipped with a knife switch, fuses and starting compensa-
tor, and the feeder circuit was fused at the distributing switch-
board. The 75-hp. motor was delta-connected, while the
10-hp. motors were connected " Y" (Fig. 270).

The 75-hp. motor was carrying a load of about 50 hp. Only
one of the 10-hp. motors was in use, and it was driving a 7-hp.

METHODS USED TO SOLVE SPECIAL TROUBLES 431

load at the time the trouble developed. Both motors had
been running for about four hours when it became necessary
to stop the small motor for a few minutes. Upon trying to
operate it again it refused to start. The electrician tested
the motor fuses with two 250-volt lamps in series, without
removing the fuses from the circuit, and found them good.
The lamp test seemed to show, however, that the voltage be-
tween the middle and either outside wire was not as high
as it should be. He then tested the fuses on the 75-hp. motor
while the motor was running and found tho same conditions

r Leads to Other Motors--*

Fuses

Starting
Compensotor

lOHp
Motor

75 Hp Motor
FIG. 270. Connections when 10-hp. motor failed to start.

as with the 10-hp. motor. Evidently something was wrong,
and with reduced voltage between the middle and outside
wires it was natural to suppose that there was a high-resistance
connection somewhere in that feeder. The large motor was
stopped, and the fuses were tested as before, but this time no
voltage at all was found between the middle and the outside
wires. Then the electrician tried to check up the first results
by starting the large motor, but since it was a squirrel-cage
three-phase induction motor it would not start on a single-
phase circuit.

He next tested the main-feeder fuses and found one of them
had blown. A new feeder fuse was put in, and everything
started off all right. The reduced voltage between the middle
and the outside wires was just one-half of the line voltage,
or 220 volts, since the middle wire amounted to a 50 per cent.

432

ARMATURE WINDING AND MOTOR REPAIR

tap on the winding of the large motor when it was running
single-phase. This case seemed quite a puzzler at the time,
and it was thought at first that the reduced voltage on the
middle wire was being generated in the large motor. How-
ever, a little study of the conditions soon made plain the real
reason as explained.

Cause of Synchronous Motor Failing to Start. E. C.
Parham (Power, June 24, 1919) explains the following trouble
with a synchronous motor in a large mill when it was necessary
to repair the winding which had been damaged in a fire.

FIG. 271. Pole-phase group shown reversed at X.

A 'complete set of coils was available and the rewinding was
done by local repairmen. After completing the job, however,
connecting the motor to its source of power failed to turn
the rotor, although it had started very promptly before being
damaged. Higher compensator taps were tried, but to no
avail. An expert alternator winder was called from the
factory and found that one pole-phase group of coils had been
reversed as at X, Fig. 271. The interchanging of the leads
of this group, as in Fig. 272, restored normal starting torque.
It will be noticed that by tracing each phase through to the
star connection in Fig. 272, the arrows point in opposite direc-
tions on adjacent pole-phase groups, which is the correct
condition. In Fig. 271 the wrong connection is indicated
by three adjacent arrows pointing in the same direction.

The existence of such a condition may be proved by con-
necting ammeters into each phase conductor and observing
the intake currents while the rotor is turned by hand; a re-

METHODS USBD TO SOLVE SPECIAL TROUBLES 433

versed pole-phase group of coils will greatly unbalance the
currents of the three wires. It is essential that the rotor be
rotated so as to equalize impedances due to different phases
including different amounts of iron in their magnetic circuits.

FIG. 272. Correct connection for a 4-pole series-star winding.

Effect of Decreased Frequency on Operation of an Induc-
tion Motor -generator Set. The immediate effect of operating
a generator below its rated speed is to increase the amount of
field current required in order to maintain normal voltage.
This is due mainly to the fact that the maintenance of normal
voltage under given load conditions requires that certain
number of field lines of force be cut each second by the arma-
ture conductors, and if the necessary rate of cutting can not
be obtained from the existing conductor rate of motion through
the existing field, the field strength must be increased until
the increase makes up for the lack of conductor speed. If
the field strength can not be increased in proportion to the
speed deficiency, voltage can not be maintained at its normal
value.

One effect of decreasing the frequency of the voltage that is
applied to an induction motor is to decrease proportionally
the speed of the motor and of its connected load. It follows,
then, that if an induction motor is the driving member of a
motor-generator set of which the driven member is a con-
tinuous-current generator, the effect of low frequency will be
to reduce the speed of the set. If the speed decrease is too
great, the generator field will not have sufficient margin to
permit of maintaining normal voltage at full load the load

434 ARMATURE WINDING AND MOTOR REPAIR

itself producing further speed reduction incident to the in-
crease of the amount of motor slip.

E. C. Parham (Electrical World, Jan. 1, 1916) has referred to
such a situation where an operator on a 50-cycle system
specified, in ordering a small motor-generator set, a 60-
cycle motor because he knew that ultimately his service would
require 60-cycle operation. In ordering the 60-cycle set
for temporary 50-cycle operation he had considered only the
motor, and had felt safe in assuming that it could stand the
larger current incident to operating on the lower frequency.
The safety of such an assumption would largely depend on how
nearly the motor would operate at its normal rating under the
prospective load conditions and with the proper frequency.
The point of interest in the present instance, however, is that
when the 60-cycle set was put into service it was found
impossible to maintain the generator voltage at normal value
even with the generator field rheostat resistance all cut out.
In order to get out of the difficulty it was necessary to install
upon the generator a set of shunt-field coils that had a greater
number of ampere-turns.

Alternating-current motors are designed and are rated to
stand a reasonable departure from their voltage, current and
frequency specifications, but before abusing them with eyes
wide open it would be advisable for operators to ascertain the
permissible limits of abnormal use.

Simple Rules for Re-connecting Alternating-current Motors.
The repairman should devise simple rules that he can re-
member easily so as to be able to handle simple changes in
motors without having to study up the connections or ask his
neighbor to explain them. A few such suggestions are given
here as used by Maurice S. Clement (Electrical Record, October,
1918) for re-connecting a motor to suit changes in voltage,
speed, etc. For example, if a three-phase, four-pole motor
is to be changed to two-phase, four-pole, at the same time
retaining the same speed, the grouping must be changed.
A three-phase, four-pole connection has 12 groups; that
is, one group to each pole, of which there are four, and
four poles to each phase, giving 12 groups. Therefore, the
numDer of poles multiplied by the number of phases givos

METHODS USED TO SOLVE SPECIAL TROUBLES 435

the number of groups. By this rule a two-phase, four-
pole machine has eight groups. The next element to be
taken into consideration is the number of coils in each
group.

In a three-phase, four-pole machine with 48 coils there
would be four coils per group; with 36 slots, three coils per
group. In a 48-coil, four-pole, three-phase machine there are
12 groups of four coils each. To change this to a two-phas?,
four-pole winding without change of speed, the following is
the rule:

(Number of coils) + (Number of poles X Phases)

= 48 -T- (4 X 2) - 6.

Therefore the new grouping will be arranged with six coils
per group.

In changing the grouping of a 36-coil machine from four-
pole, three-phase to four-pole, two-phase, the result becomes
a trifle puzzling, as it apparently gives four and one-half
coils per group. Since we know that cutting a coil in half
to suit a particular grouping can not be done, the grouping of
coils will be in the following order: 4, 4, 5, 5, 4, 4, 5, 5. This
method accounts for the half coils and will be found to be
evenly balanced. If the machine is to be changed from two-
phase to three-phase, simply reverse the operation that is
described above.

For general details for connecting alternating current motors,
see Chapter XI.

Changing 440-volt Motor for 220-volt Operation. The
accompanying illustrations show how a 440-volt, 4-pole,
2-phase motor was reconnected for 220-volt operation by
reconnecting the coils in parallel. The diagram of Fig. 273 (a)
shows the original connections of the motor, and Fig. 273 (6)
indicates the connection after the necessary alterations had
been made. The changes indicated in the diagram, Fig.
273 (6), were made as follows:

Terminal Xi was joined to the beginning of coil group
A i and end of group A 4 . Then the beginnings of groups A s
and A 4 and ends of A\ and A 2 were connected, Terminal X z

436

ARMATURE WINDING AND MOTOR REPAIR

was joined to the beginning and end of groups A 2 and A s
respectively. The 5-phase coils were connected in a similar
manner.

The foregoing alterations give very well-distributed end
windings, which are considered essential to obtain a uniform
flux distribution; but the impedance of the parallel circuits
will be changed if the rotor runs off center, as may happen
with a worm bearing, thus causing overheating. On account
of the small air gap usually found in alternating-current
motors, there is small probability of this happening because
worn bearings would cause the rotor to rub the stator before

FIG. 273. Motor winding before and after changing from 440 volts to 220

volts.

it would bring about overheating. Unequal distribution
of current between parallel windings due to a change of imped-
ance may be avoided by connecting the coil groups which are
diametrically opposite Of course, this cannot be done with
two parallel windings if one-half the number of poles is an
odd number.

Multiple Connection Diagram for A.-C. Motor Windings.
Maurice S. Clement of Youngstown, Ohio, makes use of the
diagram shown in Fig. 274 and claims that it saves a great deal
of time in re-connecting alternating-current motors (Electrical
Record, February, 1919). This multiple connection winding
diagram takes the place of four diagrams, since it can be used
to show connections for single star, single delta, parallel star

METHODS USED TO SOLVE SPECIAL TROUBLES 437

and parallel delta windings for alternating-current motors.

In this instance a four-pole, three-phase diagram is shown, but

the same plan can be used with different polarity and phasing.

The following are the formulas for the different connections :

SINGLE STAR

7, 9, 11 = Line

8 & 10 & 12 = Star

1 & 4 2 & 5 3&6in Series

PARALLEL STAR

4 & 7 = Line
6 & 9 = Line
2 & 11 = Line
1&3&5 8&10&12= Stars.

SINGLE DELTA

7 & 8 = Line

9 & 10 = Line

11 & 12 = Line

1&4 2&5 3&6in Series

PARALLEL DELTA

7& 4 & 8 & 5= Line
2 & 11 & 3 & 12 = Line
1&10&9& 6= Line

When connecting a stator it is always well to look forward
to the fact that some day the connections may be changed to
suit other conditions, and that if the long connection is used
it will be much easier to change. By long connection is meant
this. If the reader will trace out the first phase of a single
star on the sketch, he will
find it to be 1 to 7 to 10 to
4; whereas, it is sometimes
connected as follows: 1 to
4 to 7 to 10. By study-
ing the sketch closely it will
be seen that if the long
connection method is used,
the connection can be
changed to any of the other
three much easier than by
the latter method, as half
the connections do not have
to be opened. Although
the short connection is per-
haps a wire saver by a couple of inches, the long connection
is easier to understand and makes a much neater job.

Brush and Slip-ring Sparking Traced, to Absence of
Rotor Balancing Weights. Manufacturing companies take
particular care that the rotors of alternating-current machines
and the armatures of direct-current machines are perfectly

FIG. 274. Diagram that can be used
for connecting terminals of pole-phase-
groups for single and parallel star and
delta connections.

438 ARMATURE WINDING AND MOTOR REPAIR

balanced before they are shipped either .in machines or as
extras. Some go so far as to balance rotors before and after
installing the coils and the higher speed rotors of the slip-ring
type may be balanced before and after installing the slip rings.
Most of the balancing weight is applied by forcing lead into
pockets provided in the core end plates for that purpose.
That the absence of these weights may set up vibrations that
cause roughening of the slip rings and ultimate sparking, is
proven by the following experience:

In one shop (Electrical Record, August, 1918) a slip-ring motor
had given much brush and slip-ring trouble. Turning of the
rings and changing of the rings, brushes and holders, had given
no permanent relief. A factory inspector at once diagnosed
the cause of the trouble as vibration which no one had sus-
pected because it was in one of those shops where everything
that moves vibrates. He tested the rotor in a lathe and found
that the shaft was bent. The shaft was then straightened
and it was thought that all trouble was at an end; but not so.
The machine ran better, but there was considerable vibration
still there.

On questioning the operator, the statement was obtained
that the motor had run all right up to the time that it had
passed through a fire about a year previously. The fire had
melted out the bearings and had destroyed everything in the
nature of insulation; also it had distorted the shaft and melted
the balance weights although no one had considered the pos-
sibility of the last-named conditions at the time of re-winding
che machine and installing new bearings. The rotor was sent
back to the factory with a statement of its history and when
it was returned and re-installed the motor, in the words of the
owner, "ran like a new machine. " Probably it was a new
machine,

Overheating of an Induction Motor Traced to a Variation of
Frequency. Provided that an induction motor is not already
overloaded, it is not considered abusive to operate the motor
at less than 10 per cent, over voltage or at 10 per cent, under
frequency. Manufacturers are careful to specify, however,
that the motor should not be expected to stand without mate-
rial increase in heating, the continuous application of 10 per

METHODS USED TO SOLVE SPECIAL TROUBLES 439

cent, over voltage and 10 per cent, under frequency, because
both of these variations are in the direction that tends to in-
crease the heating. If the ammeter of an induction motor
circuit be observed while the motor is being started, it will
be seen that when the motor reaches nearly normal speed, the
current drops suddenly to a value far below what existed imme-
diately preceding the sudden drop. It is conceivable that
certain voltage-frequency-load conditions might exist whereby
the motor would be operating just on the high side of the crit-
ical point indicated. Under this condition the current would
be abnormally large.

In one case a mill operator complained of the heating of one
of his larger induction motors. He admitted that the motor
was overloaded, but the overload was constant and this did
not explain why the heating was so much greater during some
periods than during others. When taking the motor speed
during a period of maximum load, the inspector was much
surprised to find that the speed was five per cent, above syn-
chronism. The mill operator stated that over speeds had never
been noticed, but he felt certain that there were times when
the speed was considerably under normal. Investigation of the
engine room equipment, disclosed that at times the steam pres-
sure became so low that it was necessary for the engineer to
operate the overload valve of the turbine in order to maintain
its speed, hence frequency. The turbine generator was excited
from an engine-driven exciter, the speed of which was not
materially affected by the variations of steam pressure. As
the frequency of the service applied to the motors depended
on the speed of the turbine, which speed depended on steam
conditions, while the voltage ,of the service applied to the
motors depended on the turbine voltage which in turn de-
pended not only on the turbine speed but on the turbine excita-
tion which was independent of steam pressure variations, the
extreme conditions of high voltage and low frequency, which
may have obtained at times, can better be imagined than
estimated. As the engine room conditions could not imme-
diately be relieved, the duty of the abused motor was lightened
by removing a few of the connected machines that constituted
its load.

440 ARMATURE WINDING AND MOTOR REPAIR

Relief for a Hot Bearing. In case of a hot bearing, feed
plenty of heavy oil, loosen the nuts on the bearing cap and
slacken the belt if one is used. If no relief is shown, take off
the load and run the machine slowly until the shaft is cool
so that the bearing will not " freeze. " If the bearing is of
babbitt examine it to see that the oil grooves are still intact.
If they are and the surface of the bearing has not been injured,
renew the oil supply and start the machine again. Watch
the oil rings to see that they are revolving properly and carry-
ing plenty of oil to the shaft. A new machine or one in which
the bearings have been renewed should be run at slow speed
for an hour or more before the load is applied in order to see
that the bearings are properly adjusted and worked in.

Static Sparks from Belts (Instruction Book, Westinghouse
Electic & Mfg. Co.). It sometimes occurs on belted machines,
especially in dry weather, that charges of static electricity
of considerable potential on the belt cause discharges to the
ground. If the frame of the machine is not grounded, these
charges may jump to the armature or field winding and then
to the ground, puncturing the insulation. The belt and frame
may be discharged by placing close to the belt, at a point near
the machine pulley, a number of sharp metal points like a
comb, which are carefully grounded. If the field frame is
grounded, there should be no danger to the insulation.

Ratings of A.-C. Generators. In the case of a single-phase
generator, the rating in kva. is equal to the product of amperes
and volts.

For a two-phase generator, the rating in kva. is equal to
twice the output of one phase when the load is balanced.

For a three-phase generator the total rating in kva. is equal
to the output of one phase multiplied by 1.732. That is, the
readings or amperes and volts for one phase times 1.732 is
equal to the kva. rating of the machine, when it carries a
balanced load.

Alternating-current Motor Phase-rotation (Instruction Book,
Westinghouse Electric & Mfg. Co.). In order that the
alternating-current wiring connections between the motor
and its supply circuit may be correctly made to obtain a given
direction of rotation, it is necessary to know the phase-rota-

METHODS USED TO SOLVE SPECIAL TROUBLES 441

tion of the motor and the supply circuit. By " phase-rotation
of the motor" is meant the order in which each phase reaches
its maximum voltage of one polarity. When the machines
are arranged for clockwise rotation, looking at the commutator
end, the phase-rotation is given by the order of terminal
notation. Using letters to designate the terminals of West-
inghouse machines, the order for three-phase machines is
B-C-A and for two-phase Bi-Ai B 2 -A 2 .

The sequence of phase-rotation of the supply system can be
found by tracing the wiring back to the generating station or
else by the use of a phase-rotation indicator. When there are
flexible cable leads on the alternating-current side, the simplest
method of determining the proper order in which to connect the
motor leads is to connect the leads in any convenient order and
start the motor. If the direction of rotation is opposite to
that desired, reverse any two leads in the case of a three-phase
motor or interchange the two leads of either phase, in the case
of a two-phase motor.

End Bells or Heads. An end bell which is cracked can
easily be welded, which process makes it just as good as new.
If a bearing has been worn out or burned out, it must be re-
lined. This work must be done by a machinist who appre-
ciates the importance of a good bearing.

Brushes and Brush Holders. When a brush holder is
damaged to any great extent, it is advisable to order a new
set from the factory as it would not pay to make a set in a
repair shop. Carbon brushes are easily made, however, and
dimensions should be exact so as to insure their working easily
in the holder.

The Rotor. Under certain conditions, a rotor will become
so heated as to cause the solder to melt from its bars. When
this happens the affected parts will rattle when the machine
is run; resoldering the bars will remedy this. If the bars
become damaged or bent, they should be removed, straight-
ened, the slots reinsulated and rotor reassembled.

The Stator. The troubles occurring in an alternating-cur-
rent stator may be said to be very similar to those of an arma-
ture and most of the tests described can be used. For details
of checking up errors in connections see Chapter XI, page 288.

442 ARMATURE WINDING AND MOTOR REPAIR

Sizes of Fuses for A.-C. Motors. The National Electrical
Code does not specify the size of wire which should be used to
connect up any given motor, nor does it give the sizes of
starting and running fuses to be used. A committee of the

SQUIRREL-CAGE THREE-PHASE INDUCTION MOTORS EQUIPPED WITH
AUTO-STARTERS

Average
horsepower

Full-load amp.

Starting fuse amp.

Running fuse amp.

220 volts

550 volts

220 volts

550 volts

220 volts

550 volts

0.5
1.0

1.8
3.5
6.5
9.5
15.4
22.4
29.0
42.5
55.0
68.0
80.0
94.0
105.0
130.0
155.0
192.0
252.0
368.0
484.0
595.0
710

0.7

1.3
2.6
3.8
6.2
9.0
11.8
17.4
22.5
27.0
32.0
37.0
42.0
52.0
62.0
77.0
101.0
148.0
195.0
240.0
285.0

5
10

20
30
40
60
70
85
110
140
160
190
210
260
310
390
500
730
920
1200
1420

5
10
10
15
25
30
40
55
65
70
75
85
110
125
160
200
300
390
480
570

5
10
15
20
25
35
45
60
75
90
110
115
145
170
210
280
410
530
650
780

5
5

5
5
10
15
15
20
25
30
35
40
45
60
70
85
110
160
215
265
315

2.0
3

5.0
7.5
10.0
15.0
20.0
25

30.0
35

40.0
50.0

60 0... .

75.0

100.0
150.0
200

250.0
300

Western Association of Electrical Inspectors has compiled the
accompanying table based on the code rules which give this in-
formation for single-phase, two-phase and three-phase motors.
Excerpts from the report and the recommendations for three-
phase motors follow:

No consideration was given to limiting the voltage drop at
the motor. Motors without a starting device and those
operating at less than 600 r.p.m. require, in the majority of
cases, one size larger wire or cable than the table calls for.

METHODS USED TO SOLVE SPECIAL TROUBLES 443

The wires in the table are calculated for two and one-half
times the full load current of motors up to 30-amp. rating
and for twice the full-load current of larger motors.

Wire or cable sizes for other types of continuous-duty
induction motors should be based on the following multiples
of the full-load current: Squirrel-cage motors up to 7.5 hp.
without starters, three times full-load current; squirrel-cage
motors with star-delta starting switch, one and a half times
full-load current; wound-rotor motors with resistance in rotor,
one and a tenth times full-load current, and single-phase
repulsion motors up to 15 hp., twice, and single-phase motors
with split-phase starting, three times full-load current.

CHAPTER XVII

MACHINE EQUIPMENT AND TOOLS NEEDED IN A
REPAIR SHOP

The work that comes to the average electrical repair shop
varies from the re-winding a fan motor to the overhauling and
re- winding of large station- generators. In the latter case the
work must, in the majority of cases, be done at the location of
the machine and calls for more portable tools than machine
equipment. To meet the requirements of work between these
limits in size of machine there is needed a fairly well equipped
machine shop. Most motor repair and armature winding
work can be properly handled if the following equipment is
available :

Lathe.

Coil- winding lathe heads.

Shaper, or undercutting attachment for lathe.

Drill press.

Band saw.

Emery wheel.

Portable hand crane.

Chain block.

Welding outfit.

Vices.

Coil-pulling machine.

Coil-taping machine.

Banding machine or tension device for lathe.

In case only one lathe is available this should be large enough
to accommodate a good sized armature and have a spread
between head and tail post of about 60 inches. The shaper
and band saw are rather special machines for a repair shop
but very useful. The former serves as a rapid and accurate
method for undercutting the mica of a commutator. The

444

REPAIR SHOP EQUIPMENT AND TOOLS

445

latter saves a great deal of time when many coil forms are to
be cut for several coil winders.

In the electrical repair shop of a large industrial plant where
a number of armatures are re-wound and repaired the following
equipment is provided:

1 12-inch speed lathe.

1 24-inch armature-banding lathe.

1 24-inch commutator grooving lathe.

1 Commutator turning device.

1 Double head emery wheel grinder.

1 3-inch bench vice.

1 5-inch bench vice.

2 6-inch bench vices.

2 Field-winding machines.

2 Armature coil-taping machines.

1 One-ton electric hoist and carriage.

FIG. 275. Armature winders' hand tools. (Numbers refer to their names and
uses, page 446.)

Armature Winder's Tools. The experienced armature
winder usually has a variety of tools either designed according
to his own ideas or convenience for a particular job or for use
in connection with armatures of different sizes. In Figs. 275
and 276 a varied assortment of hand tools and drifts are shown.
The names of these tools and their uses are given in the
following tabulations.

446 ARMATURE WINDING AND MOTOR REPAIR

FIG. 276. Drifts used by armature winders. (Numbers refer to types and

uses, page 447.)

ARMATURE WINDER'S HAND TOOLS (Fio. 275)

Number

in
illustration

Name of tool

Uses in winding armatures

1
2

3
4

5 and 6

11 and 12

Coil hook. . .
Medium file.

Knife edge file

16-oz. machinist's hammer

No. 1 and No. 2 rawhide mallets ,

12-oz. machinist hammer

Tinsmith's hammer

Cold chisel

Coil lifter and shaper'.

Metal drifts.

13 and
14
16
17
18
19
20
21
22
23
24
25
26
27
28
29

Two sizes of wedge drivers .

Undercutting tool

Coil scraper

Diagonal cutters

Half-round duck bill pliers.

Round-nose pliers

Small screw driver

Shoemaker's knife

Scissors

Tinsmith's shears

6-in. parallel pliers

8-in. side cutting pliers

Flat duck bill pliers

Long metal coil drift

Monkey wrench

Large screw driver

To lift coils out of slots.

For cleaning commutator necks before
soldering.

For trimming edges of mica segments.

For use on metal surfaces.

For drifting coils into slots.

For use on metal surfaces.

For peening slots of commutator necks.

For cutting coil leads to commutator.

For lifting coils out of slots and shaping
when rewinding. !*

For driving down coil terminals in com-
mutator necks.

For driving wedges into slots.

For undercutting mica on commutators.

For scraping cotton insulation off wires.

Special pliers for rewinding work.

For general use.

For triming and cutting slot insulation.

For cutting tape, and cotton duck.

For cutting tin clips for banding wires.

For gripping wires in winding coils, etc.

For cutting heavy wire coil terminals.

For shaping coils.

For lifting tight coils out of slots.

For tightening commutator, etc.

For general use.

REPAIR SHOP EQUIPMENT AND TOOLS
DRIFTS USED BY ARMATURE WINDERS (Fie. 276)

447

Number

in
illustration

Name of drift

Particular uses

3, 4 and 5

7
8

9, 10, 11
and 12

13, 14, 15,
18, 19, 20,
21 and 22

16
17 and 23

Coil sliders or guides

Different sizes of fiber slot drifts.

Fiber coil shapers .

Fiber coil drift tapered.
Small metal coil lifter. .
Center punch

Lead lifters.

Metal drifts.

Ordinary dividers.
Tee slot drifts...

For sliding top sides of coils into slots.
For driving coils into slots. They vary

in width and length according to size

of slot.
The curved surfaces are used to shape

ends of coils.

For lifting tight coils in small machines.
For lifting tight coils out of slots.
For marking location and pitch of coil

when stripping armature.

For driving soldered ends out of com-
mutator necks when stripping arma-
ture.

For seating coil terminals in commutator
necks.

For laying out fiber collars.
For partly closed slots to drive
down in slots.

oils

Device for Shaping Insulating Cells of Armature Slots.

For shaping the fish paper in making cells for armature and
stator slots the cell shaper shown in Fig. 277 is very useful.
It consists of two pieces of wood hinged together so that they
will make a neat 90-deg. fold. The permanency of the correct-
fold maker is insured by means of a
metal strip attached to the wood slot.
The cell shaper is used by inserting
a piece of fish paper in the opening
between the two blocks of wood, which
is the length of the slot plus twice the
height and whose width is the width
of the slots. The metal straight-edge,
which is adjustable by means of wing nuts, allows the paper
to be folded so as to be made the height of the slot.

Tool for Cutting Cell Lining at Top of Slot. For cutting the
projecting insulation of an armature slot after the coils have
been assembled, the tool shown in Fig. 279 is recommended by
Maurice S. Clement (Electrical World, Oct. 12, 1918). This

CELL SHAPR

FIG. 277. Device for
shaping insulating cells of
armature slots.

448 ARMATURE WINDING AND MOTOR REPAIR

S j*
fl "fl

,,-. <D
P 3

REPAIR SHOP EQUIPMENT AND TOOLS

449

riLEHANDLE

tool as well as those of Figs. 280 and 282 and the insulating cell
shaping device shown in Fig. 278 are home-made designs he has
devised. It consists of a piece of forged steel 14 inches long
by % inch wide by %Q inch thick with a set of beveled knife-
edges at one end and a file handle at the
other.

Special Winding Tools. In Fig. 280
three convenient tools are shown that can
bs easily made by any armature winder.
From left to right in the illustration they FlG 279 Tool for
are a coil-taping needle, coil raiser, and cutting insulation that
wire scraper. The coil-taping needle can projec
be made from one foot of No. 14 banding wire shaped so that
it can be used for taping coils in closed-slot stators. After
the user is accustomed to this device high speed may be
attained.

The coil raiser consists of a piece of steel, 16 inches by 1 by
mc h w ^h a four-inch one-sided taper on one end for strip-

CELL CUTTER

*tt

v'/SEnd

~jj

)

Plan

CO/

i
r V

LRAISEK

L \

Section

MEDLE" ^MRESCRAPE*

FIG. 280. Convenient designs of coil taping needle, coil raiser and wire

scraper.

ping open-slot armatures and stators. This also can be used
to good advantage in removing grounded coils from a newly
wound armature or in raising coils sufficiently to allow for
insulating weak spots in the coils, the main object in this case

450 ARMATURE WINDING AND MOTOR REPAIR

REPAIR SHOP EQUIPMENT AND TOOLS

451

being to lift out a tight-fitting coil without damaging the
insulation.

The wire scraper is very simply made and very economical,
because it eliminates the use of a knife, whose life is short on
account of the rough treatment accorded it. This device is
made of spring metal, 12 inches by % inch by Jfe mcn - The
knife-edges can be sharpened by means of a file and the tool
used indefinitely. A section the shape of a rectangle is cut
from the metal at the handle end, greatly increasing the spring
effect of the device.

For driving fiber wedges between the top of a coil and the
lamination overhanging closed-slots, a wedge drift, made of
a piece of tool steel, eight inches by five inches by ^2 inch,
over which is fitted a loose-fitting steel sleeve, J^g inch thick,
is very convenient. This is used by inserting the fiber wedge
about y inch into the slot; then, with the drift pulled back into
the sleeve, the sleeve is fitted over the wedge, which is driven
into the proper place, the sleeve holding the wedge in position.

Repair Tools that Can be Made from Old Hack-saw Blades.
Seven tools that have been made from old hack-saw blades

FIG. 282. Seven armature repair tools made from old hack-saw blades that
are useful around a motor repair bench.

and found useful in the re-winding of motors are shown in
Fig. 282. A small wooden case to stow away these tools can
be made as follows: Cut out two pieces of any suitable hard
wood eight inches long, one and one-half inches wide and
]4, inch thick. Then on the flat side of one of these pieces
cut several grooves, large enough to allow a hack-saw blade
to fit in it snugly. The two pieces should then be nailed

452

ARMATURE WINDING AND MOTOR REPAIR

together so that the grooves will be between both pieces.
When a third piece of wood, eight inches long, one inch wide
and y inch thick is nailed on as a bottom piece, the
case is ready for the tools.

While each of the tools can be used for several different
operations, a few of the most important uses are as follows:
Tools one, two and five (from the right) are used mostly in
taping coils. Tools one and two are used to best advantage in
digging stray drops of solder out of the winding of a stator
after the connections have been soldered. Tools four and six,
which have both edges of the V sharpened to a knife-edge,
are very well adapted to cutting insulation and scraping wires
or leads. Number seven is used to drive wedges into tight
slots. By setting the saw-teeth on the wedge and pounding
on the beveled end with a hammer, the teeth are made to
grip the wedge and drive it to its proper place.

Special Coil-winding Device. For winding coils of practi-
cally any shape the special device shown in Fig. 283 has been

Ji Hole off Center

J^St'd. Thread
and Washer

FIG. 283. Framework for use in winding different shapes of armature coils.

devised by Frank Huskinson (Electrical World, July 27, 1918).
It consists of an iron framework held together with four bolts
shown at B. By loosening the nuts of these bolts the two
outside members can be adjusted for any width of coil needed.
To give the coil the proper shape, disks such as marked C are
clamped on the piece D and the latter inserted in the slots
marked A in the frame. The view at E shows how these
disks appear when the bolt D is in its proper place in the frame.
When forming the coils the wire is wound around six of the
bolts or pegs and between the disks mounted on them.

REPAIR SHOP EQUIPMENT AND TOOLS

453

FIG. 284. Convenient bench device for holding the stator of a small motor
when repairing windings.

FIG. 285. At the left a vertical stand is shown for use in mounting the
commutator on small armatures. At the right, a large armature being wound
with pushed-through coils.

464 ARMATURE WINDING AND MOTOR REPAIR

Steadying Brace for Repairing Small Motors. The device
shown in Fig. 284 has been devised by Maurice S. Clement
(Electrical Record, December, 1918) for use in repairing fan
motors or other small motors when it is necessary to remove
the winding of the stator from the frame. To remove these
windings from the frame the laminations must be removed
with the winding. When once removed it is very difficult
to work on it as it has a tendency to roll away all the time.

This brace is constructed as follows: A block is cut as
shown and bolted to the bench. One end of a length of
ribbon copper wire about }{$ inch thick and ^ inch wide is
then screwed to the end of the block and passed over the wind-
ing which sets in the cut out portion of the block. The other
end of the copper ribbon is held down by means of a wing
nut placed between the bolts. A leather strap is arranged
from one side of the block, over the copper ribbon and fastened
to a buckle on the other side. This is a tension strap and takes
the slack out of the copper ribbon.

Tension Block For Use When Banding Armatures. A tool
is shown in Fig. 286 which has been devised by Maurice S.

W* Ifl*

maatm | "' *W*-W*| ,

-* w^~^ ;

lii^BRC

IBREi^rCWWE

FIG. 286. Tension block for use when winding armatures.

Clement (Electrical World, Oct. 12, 1918) for banding an arma-
ture in a small repair shop where a banding lathe is not avail-
able. By use of the device the armature can be banded on an
ordinary winding stand which is securely fastened to the floor
or work bench. When this device is used about one foot of
stout line with a hook attached to one end is made fast to the
ring on the tension block and hooked to an eye-bolt which is
set in the floor for that purpose. The spool of banding wire
is placed on a small stand beside the eye-bolt and the wire is
passed between the two blocks at the rear end through the
hole in the first wire guide over the tension curve and through

REPAIR SHOP EQUIPMENT AND TOOLS

455

the second wire guide hole and then to the armature. The
tension can be regulated by the wing nut placed at the forward
upper end of the block. By screwing down the wing nut
both sides of the block are brought nearer together, thus

Double ply leather

%"thick i!o"

V *

7 Brass rivets *

5^4

Iron triangle
FIG. 287. Armature sling made up of double-ply leather belting.

narrowing the tension curve over which the wire must pass.
This increases the tightness of the band when a pipe wrench
can be used to revolve the armature.

FIG. 288. Method of using a rope sling in handling armatures with a crane.

Armature Sling. For handling heavy armatures in repair
shops a sling is sometimes used instead of lifting by means of a
rope attached to the ends of the shaft as a bale handle for the

456 ARMATURE WINDING AND MOTOR REPAIR

crane hook. The sling prevents the possibility of springing
the shaft. A sling which can be used for handling small
armatures is shown in Fig. 287. It is constructed of a piece
of double ply belting about % inch thick, 10 inches wide and

5% feet long, provided with triangles
of steel bar for the hook of a shop
crane.

Pinion Puller. A device which
can be made up as shown in Fig.
289 is convenient for removing
pinion gears from motor shafts with-
out injury. By making the head
sufficiently long and providing holes
at suitable distances, pinions of
different sizes can be removed by
adjusting the end pieces.

Coil-winding Machines. In Fig.

290 a device is shown which does not require the use of forms
when winding armature coils. It consists of an ad j ustable metal
frame with jaws which may be clamped in various positions so
as to make a form over which the wire can be wound into the

FIG. 289. Convenient de-
vice for removing pinions
from a motor shaft.

FIG. 290. Segur armature coil winder for winding hair-pin loop, obtuse loop
and rectangular loop coils. (Electrical Manufacturers Equipment Company.)

With this device loops from. 3 inches to 36 inches can be wound. By means of the
spreader shown in Fig. 291 the coil shapes shown in Fig. 292 can be formed from these
loops.

shape of the coil desired. Jaws are provided for forming hair-
pin loop coils, that is, coils with parallel sides and curved ends.

REPAIR SHOP EQUIPMENT AND TOOLS 457

Another set of jaws provided with the machine has four points
and gives four curves to the coil, the jaws being set so as to
form a rectangular coil with two angles obtuse and two angles
acute. These same jaws may be set in another way so as to
give a triangular or square loop. With these attachments,
coils of various shapes can be wound, some of the shapes being
suitable for use directly in various types of motors and gen-
erators and other coils requiring further operations on a
machine called a spreader. The purpose of this machine
(Fig. 291) is to form the coil to the throw required when it is

FIG. 291. Coil spreader for shaping the coils shown in Fig 292 as wound
on the machine shown Fig. 290. (Electrical Manufacturers Equipment Com-
pany.)

inserted in the armature or stator. The spreader takes the
coil as it is wound upon the winding machine and opens it
and twists it at both ends, so that it conforms to the shape
desired. Some of the forms of coils as wound on these ma-
chines and as finally shaped on the spreader are shown in the
accompanying illustrations, Fig. 292.

Coil-taping Machines. Field and armature coils for me-
dium and larga-sized motor and generators are wound from
bare copper wire and are insulated after the winding operation
has been completed. The material usually employed for
winding coils is treated or untreated tape which is wound on
the complete coil by hand using rolls of treated or untreated
tape or by the use of a taping machine. The taping machine
consists essentially of a rotating circular element which has an
opening, by means of which the coil is placed inside of the ring

458 ARMATURE WINDING AND MOTOR REPAIR

(e)

FIG. 292. Coils that can be shaped on the spreader shown in Fig. 291.
(1) Common loop, 3 wires wide with 5 turns that can be spread as shown in (2)-
(3) ao-coilgroup 3 wires wide, wound 5 wires vertical and spiral shown f ormed in (4)-
(5) a rectangular loop 3 wires wide, 2 turns of 6 wires that can be shaped to make a
3-coil bundle. (7) Obtuse loop, 5 turns, 3 wires wide that can be formed as shown in
(8) and spread to make an Eickemeyer coil like (9). (10) Ribbon copper coil formed into
a hair-pin loop by bar bender (Fig. 293) and shaped in coil spreader.

FIG. 293. Top view of a Segur bar bender for shaping armature coils made

of copper strip from 0.05 to 0.125 inch thick by any width up to 1H inch.

(Electrical Manufacturers Equipment Company.)

REPAIR SHOP EQUIPMENT AND TOOLS

459

(Fig. 294). Tape is then attached by hand and the rotating
element which carries the spool of tape winds it symmetrically
about the bundle of wires forming the coil.

FIG. 294. Segur armature coil-taping machine.
(Electrical Manufacturers Equipment Company.)

FIG. 295. Portable commutator slotting machine that will handle arma-
tures up to 18 inch in diameter. (Electric Service Supplies Company.)

Commutator-slotting and Grinding Machines. In Fig. 295
is shown a commutator-slotting outfit. This machine consists
of a frame in which the commutator is held either before or

460 ARMATURE WINDING AND MOTOR REPAIR

after the commutator has been placed on the armature, and
a rotating disk saw mounted on slides which are parallel to the
direction of rotation of the disk. The rotating saw is driven
by belts or other suitable means of transmission from the
source of power. The guide or frame upon which the saw is
mounted is adjustable according to the diameter of the com-

Fia. 296. Commutator slotting maching that can be attached to a standard
engine lathe. (Electric Service Supplies Company.)

mutator to be worked on and after the correct height and depth
of the cut have been ascertained, the frame is clamped firmly
in position. The operator then starts the saw and moves it
forward on its guide by means of a lever, so that it comes into
contact with the commutator, thus cutting a slot. The arma-
ture is moved through a space corresponding to the width of a
single bar, and the saw is again brought forward cutting

REPAIR StiOP EQUIPMENT AND TOOLS

461

another slot, and so the operation continues until the commu-
tator has been slotted all around.

The face of the commutator must be smooth and perfectly
circular so as to secure sparkless operation and the minimum
wear of brushes. This result is obtained by grinding the com-
mutator after it has been assembled or slotted. The com-

FIQ. 297. Combination armature banding and tension machine. (Electric
Service Supplies Company.)

mutator grinder operates by bringing a stationary grinding
tool or abrasive material into contact with the rotating com-
mutator which is mounted in its final position on the armature,
in this case the entire armature being mounted in a lathe
and rotated. Another method is to apply a rotating grinding
wheel to the armature. The commutator grinding or truing

462

ARMATURE WINDING AND MOTOR REPAIR

machine is essentially a lathe and operates in the same manner
as any other machine for grinding a circular surface.

Armature-banding Machine. This machine usually con-
sists of a suitable stand with bearings upon which the arma-
ture shaft rests, and some means for rotating the armature by
means of a belt, chain or gear drive from a line shaft or from
an individual motor. A band wire tension device similar to the
tension device employed for winding coils is incorporated in the
machine or is built as a separate machine, which may be at-
tached to the armature stand. With an armature in position,
the band wire is attached at several points throughout the length

FIG. 298. A machine equipped so that coil-winding, banding and grinding
operations may be performed on it.

of the armature, one at each end and one or more at equal dis-
tances between the ends, and the machine is started. The ro-
tation of the armature causes the wire to be wrapped about the
armature, the tension device and guide laying the wires evenly,
to a width which is determined by the setting of the guide.

Combination Machines. A combination coil- winding,
banding and grinding machine is shown in Fig. 298. This
machine combines into one, the following tools: (a) a band-
ing machine with a self-contained tension carriage for the
band wire, designed to handle large or small railway, locomo-
tive or stationary armatures; (6) a commutator slotting ma-
chine with independent motor drive; (c) a commutator grinding
machine with independent motor drive; (d) a commutator

REPAIR SHOP EQUIPMENT AND TOOLS 463

cutting or turning attachment; (e) a field and armature coil
plate mounted on the main spindle suitable for all classes of
heavy form coil winding. This machine closely resembles a
large power lathe but is provided with the special attachments
for performing the work above outlined. After an armature
has had the coils placed in the slot and the leads soldered to the
commutator, it may be placed in this machine and the work
completed without removing the armature from the machine.
The banding attachment is capable of heavy duty, the rotation
being under control of the operator. When the machine
is stopped it automatically locks and prevents slack in the
band wire due to backing motion of the armature. Uniform
tension is secured by a tension device mounted on the feed
carriage and traveling with it. The feed carriage is moved
along by means of a rack and pinion and pilot wheel and is
adjustable vertically in or out.

The slotting attachment is supported by a bracket clamped
to the tail-stock spindle and is removed by loosening two cap
screws. The commutator truing or grinding attachment
consists of a traveling grinding wheel supported from the tail
stock and motor driven. The casting which supports the two
rods carrying the grinding wheel and the independent motor
are adjustable along two other steel rods projecting backward
from the tail stock and bringing the grinding wheel parallel
to the face of the commutator. These rods are adjustable in
or out to suit the length or location of the commutator, and
the grinding wheel is moved along by means of the screw and
hand wheel. Cutting is done in both directions of travel of
the wheel. Power is applied either through belt and pulley
from a countershaft or by an individual motor drive which
may readily be installed. The control is by means of a clutch
which is operated by a treadle running the full length of the
machine. The machine is provided with two changes of
speed, a low speed for banding and coil winding and a high
speed for commutator cutting, grinding and truing. The
head stock is provided with a coil-winding plate to which may
be attached forms for winding coils of various sizes and shapes,
upon the same principle as the independent coil-winding
machines, designed to do this work only.

APPENDIX

DATA AND REFERENCE TABLES

How to Remember the Wire Table. The wire table for B. & S.
gauge copper wire has simple relations, such that if a few constants
are remembered the whole table can be constructed mentally with approxi-
mate accuracy.

A wire which is three sizes larger than another wire has half the resist-
ance, twice the weight and twice the area. A wire which is ten sizes
larger than another wire has one-tenth the resistance, ten times the
weight and ten times the area.

No. 10 Wire is 0.10 inch in diameter (more precisely 0.102); it has an
area of 10,000 circular mils (more precisely 10,380); it has a resistance of
1 ohm per thousand feet at 20 C. (68F.), and weighs 32 pounds (more
precisely 31.4 pounds) per thousand feet.

The weight of one thousand feet of No. 5 wire is 100 pounds.

The relative values of resistance (for decreasing sizes) and of weight
and area (for increasing sizes) for consecutive sizes are: 0.50, 0.60, 0.80,
1.00, 1.25, 1.60, 2.00.

The relative values of the diameters of alternate sizes of wire are:
0.50, 0.63, 0.80 1.00, 1.25, 1.60, 2.00.

The "mil," whose value is one-thousandth (0.001) of an inch, is the
practical basis for determining the diameters and thereby the areas of
all wires used as electric conductors. The diameter being given, the area
is obtained by the well-known rule, "the area of a circle, in circular units,
is equal to the square of its diameter," hence, the square of the diameter
of a wire expressed in mils equals the area of its cross-section. D 2 = A,
which area is expressed in circular mils or CM; hence, D 2 = CM.

Circular Mils. Conductors of large size are usually specified in cir-
cular mils as 500,000 circular mils, 750,000 circular mils.

To find resistance, drop one cypher from the number of circular mils;
the result is the number of feet per ohm.

To find weight, drop four cyphers from the number of circular mils
and multiply by the weight of No. 10 Wire.

Copper for Various Systems of Distribution. When the power trans-
mitted, distance, line loss and voltage of lamps is constant and all wires
30 465

466 ARMATURE WINDING AND MOTOR REPAIR

of each system are of the same size, the following is the relationship of
the copper required.

Copper
System. Required

2 Wire, single-phase or direct current 1 . 000

3 Wire, single-phase or direct current . 375

4 Wire, single-phase or direct current 0. 222

4 Wire, two-phase 1 . 000

4 Wire, three-phase with neutral . 333

3 Wire, three-phase Delta ...' 0. 75

Classification of Wire Gauges. Wire gauges are known under a variety
of names so that it is important to know the difference when conditions
require that the values of one gauge shall be converted into those of an-
other. The following is a classification of gauges by names and uses:

Brown & Sharpe (B.& S.) = American Wire Gauge (A. W. G.).

New British Standard (N. B. S.) = British, Imperial, English Legal
Standard and Standard Wire Gauge, and is variously abbreviated by
S. W. G. and I. W. G.

Birmingham Gauge (B. W. G.) = Stubs, Old English Standard and
Iron Wire Gauge.

Roebling = Washburn Moen, American Steel and Wire Go's. Iron
Wire Gauge.

London = Old English (not Old English Standard).

As a further complication:

Birmingham or Stubs' Iron Wire Gauge is not the same as Stubs'
Steel Wire Gauge.

Uses of Various Gauges. B. & S. Gauge. All forms of round wire used
for electrical conductors. Sheet copper, brass and German silver.

U. S. S. Gauge Sheet iron and steel. Legalized by act of Congress,
March 3, 1893.

B. W. Gauge. Galvanized iron wire. Norway iron wire.

American Screw Co.'s Wire Gauge Numbered sizes of machine and
wood screws, particularly up to No. 14 (0.2421 inch).

Stubs' Steel Wire Gauge Drill rod.

Roebling & Trenton Iron and steel wire. Telephone and tele-
graph wire.

N. B. S. Hard drawn copper. Telephone and telegraph wire.
London Gauge Brass wire.

APPENDIX
DIFFERENCE BETWEEN WIRE GAUGES

467

Gauge

No.

Brown &
Sharpe's

Old English or
London

Stubs' or
Birmingham

New British
Standard

0000

0.460

0.454

0.400

000

0.40964

0.425

0.372

. 36480

0.380

0.348

0.32495

0.340

0.324

0.28930

0.300

0.25763

0.284

0.276

0.22942

0.259

0.252

0.20431

0.238

0.232

0.18194

0.220

0.212

0.16202

0.203

0.192

0.14428

0.180

0.176

0.12849

0.165

0.160

0.11443

0.148

0.144

0.10189

0.134

0.128

0.09074

0.120

0.116

0.08081

0.109

0.104

0.07196

0.095

0.092

0.06408

0.083

0.080

0.05706

0.072

0.05082

0.065

0.064

0.04525

0.058

0.056

0.04030

0.049

0.048

0.03589

0.040

0.042

0.040

0.03196

0.035

0.036

0.02846

0.0315

0.032

0.025347

0.0295

0.028

0.022571

0.027

0.025

0.024

0.0201

0.025

0.022

0.0179

0.023

0.020

0.01594

0.0205

0.018

0.014195

0.01875

0.016

0.0164

0.012641

0.0165

0.014

0.0148

0.011257

0.0155

0.013

0.0136

0.010025

0.01375

0.012

0.0124

0.008928

0.01225

0.010

0.0116

0.00795

0.01125

0.009

0.0108

0.00708

0.01025

0.008

0.010

0.0063

0.0095

0.007

0.0092

0.00561

0.009

0.005

0.0084

0.005

0.0075

0.004

0.0076

0.00445

0.0065

0.003965

0.00575

0.003531

0.005

0.003144

0.0045

468

ARMATURE WINDING AND MOTOR REPAIR

EQUIVALENTS OF WIRE SIZES (B. & S. GAUGE)

0000 = 2

No. 0=4

No. 3 = 8

No. 6 = 16

No. 9 = 32

No. 12 = 64

No. 15

000 = 2

No. 1=4

No. 4 = 8

No. 7 = 16

No. 10 = 32

No. 13 = 64

No. 16

00 = 2

No. 2=4

No. 5 = 8

No. 8 = 16

No. 11 = 32

No. 14 = 64

No. 17

= 2

No. 3 = 4

No. 6 = 8

No. 9 = 16

No. 12 = 32

No. 15 = 64

No. 18

1 = 2

No. 4=4

No. 7 = 8

No. 10 = 16

No. 13 = 32

No. 16 = 64

No. 19

2 = 2

No. 5 = 4

No. 8 = 8

No. 11 = 16

No. 14 = 32

No. 17 = 64

No. 20

3 = 2

No. 6 = 4

No. 9 = 8

No. 12 = 16

No. 15 = 32

No. 18 = 64

No. 21

4=2

No. 7=4

No. 10 = 8

No. 13 = 16

No. 16 = 32

No. 19 = 64

No. 22

5 = 2

No. 8 = 4

No. 11 = 8

No. 14 = 16

No. 17 = 32

No. 20 = 64

No. 23

6 = 2

No. 9=4

No. 12 = 8

No. 15 = 16

No. 18 = 32

No. 21 = 64

No. 24

7 = 2

No. 10 = 4

No. 13 = 8

No. 16 = 13

No. 19 = 32

No. 22 = 64

No. 25

8 = 2

No. 11 = 4

No. 14 = 8

No. 17 = 16

No. 20 = 32

No. 23 = 64

No. 26

9 = 2

No. 12 = 4

No. 15 = 8

No. 18 = 13

No. 21 = 32

No. 24 = 64

No. 27

10 = 2

No. 13 = 4

No. 16 = 8

No. 19 = 13

No. 22 = 32

No. 25 = 64

No. 28

11 = 2

No. 14 = 4

No. 17 = 8

No. 20 = 16

No. 23 = 32

.No. 26 = 64

No. 29

12 = 2

No. 15 = 4

No. 18 = 8

No. 21 = 16

No. 24 = 32

No. 27 = 64

No. 30

13 = 2

No. 16 = 4

No. 19 = 8

No. 22 = 16

No. 25 = 32

No. 28

It = 2

No. 17 = 4

No. 20 = 8

No. 23 = 16

No. 26 = 33

No. 29

15 = 2

No. 18 = 4

No. 21 = 8

No. 24 = 16

No. 27 = 32

No. 30

16 = 2

No. 19 = 4

No. 22 - 8

No. 25 = 16

No. 28

17 = 2

No. 20 = 4

No. 23 = 8

No. 26 = 16

No. 29

18 = 2

No. 21 = 4

No. 24 = 8

No. 27 = 16

No. 30

19 = 2

No. 22 = 4

No. 25 = 8

No. 28

20 = 2

No. 23 = 4

No. 26 = 8

No. 29

21 = 2

No. 24 = 4

No. 27 = 8

No. 30

General Wiring Formula For Alternating- and Direct-current Circuits.
The following general formula may be used to determine the size of
copper conductors, volts loss in lines, current per conductor, and of cop-
per per circuit or any system of electrical distribution.

Area of conductor, circular mils =

Volts loss in lines

Current in main conductors =

Pounds copper

D X W X C
P X E*

PXEXB
100

W XT
E

D* X W X C X A
P X #X1,000,000

W = Total watts delivered.

D = Distance of transmission (one way) in feet.

P = Loss in line in per cent, of power delivered, that is of W.

E = Voltage between main conductors at receiving or consumer's

end of circuit.
For continuous current. C = 2160, T = 1, B = 1, and A = 6.04

APPENDIX

469

System

Value
of A

Value of C

Value cf T

Per cent., power factor

100

Single-phase
Two-phase (4 wire)
Three -phase (3 wire) . .

6.04
12.08
9.06

2160
1080
1030

2400
1200
1200

2660
1330
1330

3000
1500
1500

3380
1690
1690

1.00
0.50
0.58

1.05
0.53
0.61

1.11
0.55
0.64

0.17
0.59
0.68

1.25
0.62
0.72

APPLICATION OF THE FORMULA

The value of C for any particular power factor is obtained by dividing
2160, the value for continuous current, by the square of that power fac-
tor for single-phase, and by twice the square of that power factor for three-
wire, three-phase, or four-wire, two-phase.

The value of B depends on the size of wire, frequency and power fac-
tor. It is equal to 1 for continuous current, and for alternating current
with 100 per cent, power factor. For sizes of wire given in the tables of
wiring constants (pages 470 and 471), and other power factors and cycles
the values of B are given.

The figures given are for wires 18 inches apart and are sufficiently
accurate for all practical purposes provided the displacement in phase
between current and emf . at the receiving end is not very much greater
than that at the generator; in other words, provided that the reactance
of the line is not excessive, or the line loss unusually high. For example,
the constants should not be applied at 125 cycles if the largest conductors
are used and the loss 20 per eent. or more of the power delivered. At
lower frequencies, however, the constants are reasonably correct even
under such extreme conditions. They represent about the true values
at 10 per cent, line loss, are close enough at all losses less than 10 per
cent., and often, at least for frequencies up to 40 cycles, close enough for
eren much larger losses. When the canductors of a circuit are nearer
each other than 18 inches, the volts loss will be less than given by the
formula, and if close together, as with multiple conductor cable, the loss
will be only that due to resistance.

The value of T depends on the system and power factor. It is equal
to 1 for continuous current and for single-phase current of 100 per cent,
power factor.

The value of A and the weights of the wires in the table are based on
0.00000302 pound as the weight of a foot of copper wire of 1 circular mil
area.

In using the above formula and constants, it should be particularly
observed that P stands for the per cent, loss in the line of the delivered

470 ARMATURE WINDING AND MOTOR REPAIR

power, not for the per cent, loss in the line of the power at the generator;
and that E is the potential at the end of the line and not at the generator.

When the power factor cannot be more accurately determined, it may
be assumed to be as follows for any alternating system operating under
average conditions: Incandescent lighting and synchronous motors,
95 per cent. ; lighting and induction motors together, 85 per cent.; induc-
tion motors alone, 80 per cent.

In continuous current, three-wire systems, the neutral wire for feeders
should be made of ^ the section obtained by the formula for either of the
outside wires. In both continuous and alternating-current systems, the
neutral conductor for secondary mains and house wiring should be taken
as large as the other conductors.

The three wires of a three-phase circuit and the four wires of a two-
phase circuit should be made all the same size, and each conductor should
be of the cross-section given by the first formula.

GENERAL WIRING DATA FOR FORMULA FOR DIRECT- AND ALTERNATING-
CURRENT CIRCUITS 25 AND 60 CYCLES

Size
of wire
B. & S.

Area
wire
cir.
mils

Wt. Ibs.
bare
wire
per
1000 ft.

Resist-
ance
ohms
per
1000 ft.

at 2or i

Value of B for formula (page 468)

Size
of
wire
B. &

25 cycles

40 cycles

Per cent, power factor

at 4\j iv.

0000

211,600

640.73

0.04879

1.23

.29

1.33

.34

1.52

1.53

.61

1.67

0000

000

167,805

508.12

0.06154

1.18

.22

1.24

.24

.40

1.41

.48

1.51

000

133,079

402.97

0.07758

1.14

.16

1.16

.16

.25

1.32

.35

1.37

105,560

319.00

0.09775

1.10

.11

1.10

.09

.19

1.24

.26

1.26

83,694

253.43

0.1234

1.07

.07

1.05

.03

.14

1.17

.18

1.17

66,373

200.98

0.1556

.05

.01

1.02

.11

1.12

.12

1.10

52,633

159.38

0.1962

.03

.02

.07

1.08

.07

1.05

41,742

126.40

0.2473

.02

.05

1.06

.03

33,102

100.23

0.3120

.03

1.01

26,250

79.49

0.3934

.02

20,816

63.03

0.4959

.01

16,509

49.99

0.6250

13,090

39.60

0.7886

10,382

31.40

0.9940

APPENDIX

471

GENERAL WIRING DATA FOR FORMULA FOR DIRECT- AND ALTERNATING-
CURRENT CIRCUITS 60 AND 125 CYCLES

Size

of wire
B. & S.

Area
wire
cir.
mils.

Wt. Ib.
bare
wire

ioo er ft.

Resist-
ance
ohms

1000 ft.
at 20C.

Value of B for formula (page 468)

Size
of
wire
B. &

60 cycles

125 cycles

Per cent., power factor

0000

211,600

640.73

0.04879

.62

.84

1.99

2.09

2.35

2.86

3.24

3.49

0000

000

167,805

508.12

0.06154

.49

.66

1.77

1.95

2.08

2.48

2.77

2.94

000

133,079

402 . 97

0.07758

.34

.52

1.60

.66

.86

2.18

2.40

2.57

105,560

319.00

0.09775

.31

.40

1.46

.49

.71

.96

2.13

2.25

83,694

253.43

0. 1234

.24

.30

1.34

.36

.56

1.75

1.88

1.97

66,373

200.98

0. 1556

.18

.23

1.25

.26

.45

.60

1.70

1.77

52,633

159.38

0. 1962

.14

.17

1.18

.17

.35

.46

1.53

1.57

41,742

126.40

0.2473

.11

.12

1.11

.10

.27

.35

.40

1.43

33,102

100.23

0.3120

.08

1.06

.04

.21

.27

.30

1.31

26,250

79.49

0.3934

.05

.04

1.02

.16

.20

.21

1.21

20,816

63.03

0.4958

.03

.02

.12

.14

1.13

16,509

49.99

0.6250

.02

.09

.10

.09

1.07

13,090

39.60

0.7886

.06

.04

1.02

10,382

31.40

0.9940

.04

.03

Amperes in Alternating-current Circuits. The following tables give
the amperes per lead wire per kilowatt for single-phase and three-phase
balanced loads. The single-phase table can be used for two-phase
balanced loads by using a current value corresponding to twice the stated
potential of the circuit or by dividing the current value at the potential
of the circuit by two. That is, each wire of a two-phase circuit carries
one-half of the current indicated at the load specified. These tables
show the value of the current at power factors varying from unity to
70 per cent. The power of any circuit in kilowatts can, therefore, be
computed by dividing the reading of the ammeter by the tabulated value
corresponding to the measured power factor and voltage of the circuit.
These values are correct only for a balanced load (and there is generally
a slight unbalancing of the loads on the phases), but the table is useful in
computing the sizes of wire required for transmission purposes.

This table was derived from the following formulas :

For single-phase circuits: Amperes per wire = watts * (volts X
power factor).

For three-phase circuits: Amperes per wire = total watts -r (volts
between wires X power factor X A/S)-

For two-phase circuits: Amperes per wire = total watts -i- (volts
between wires of one phase X power factor X 2).

In making the computations the number of watts was assumed as
1000, and the amperes were computed for various values of emf. to a

472

ARMATURE WINDING AND MOTOR REPAIR

sufficient number of decimal places to insure accuracy. The tables were
then extended by multiplication and division. If desired, these tables
can be further extended to cover voltages outside of their limits by using
the tabular values corresponding to potentials of one-tenth (or 10 times)
the desired potential, care being used to shift the decimal point in the
proper direction.

The values for intermediate power factors can be approximated from
the tables. For lower power factors, the value of the current for unity
power factor can be divided by actual power factor of the circuit or
multiplied by the reciprocal of this power factor.

SINGLE-PHASE CIRCUITS AMPERES FOR ONE KILOWATT AT DIFFERENT
POWER FACTORS

Power factor in per cent,

Volts

100

110

9.0909

9 . 5693

10. 1010

10.6952

11.3636

12.1211

12.9870

220

4.5455

4.7847

5.0505

5.3476

5.6819

6.0606

6.4936

440

2.2727

2 . 3923

2.5252

2.6738

2.8409

3 . 0303

3 . 2467

550

1.8182

1.9139

2.0202

2.1390

2.2728

2.4242

2.5974

1100

0.9091

0.9569

1.0101

0.0695

1 . 1364

1.2121

1 . 2987

2200

0.4545

0.4785

0.5050

0.5348

0.5682

0.6061

0.6494

3300

0.3030

0.3190

0.3367

0.3565

0.3788

0.4040

0.4329

6600

0.1515

0.1595

0.1684

0.1783

0.1894

0.2020

0.2165

11000

0.0909

0.0957

0.1010

0.1070

0.1136

0.1212

0.1299

THREE-PHASE CIRCUITS AMPERES PER WIRE FOR ONE KILOWATT AT
DIFFERENT POWER FACTORS

Power factor per cent.

Volts

100

110

5.2486

5.5249

5.8319

6.1749

6.5608

6.9982

7.4980

220

2.6243

2 . 7624

2.9159

3 . 0874

3.2804

3.4992

3 . 7490

225

2.5660

2.7010

2.8511

3.0188

3.2075

3.4213

3.6657

440

1.'3122

1.3812

1.4579

1 . 5437

1 . 6402

1 . 7495

1 . 8745

550

1.0497

1 . 1050

1 . 1664

1 . 2350

1.3121

1.3996

1.4996

1100

0.5249

0.5525

0.5832

0.6175

0.6561

0.6998

0.7498

2200

0.2624

0.2762

0.2916

0.3087

0.3280

0.3499

0.3749

3300

0.1749

0.1842

0.1944

0.2058

0.2187

0.2333

0.2499

6600

0.0875

0.0921

0.0972

0.1029

0.1093

0.1167

0.1249

11000

0.0525

0.0552

0.0583

C.0617

0.0656

0.0700

0.0750

APPENDIX

473

MINIMUM SIZE WIRE FOR MOTOR SERVICES WHEN CONCEALED OR
PARTLY CONCEALED WIRES ARE USED

HP.

Size wire B. & S. gauge

110 volts

220 volts

550 volts

000

0000

....

000

....

0000

....

000

100

...- v ;

0000

VALUES OF FIELD CURRENT IN DIRECT-CURRENT GENERATORS

It has been found that a fair average for the field amperes of different
sizes of generators is as follows :

Kw....

100

Per cent

3 5

2 75

The field current (expressed as a percentage of full-load current on
lines) is determined with all of the resistance cut out, that is, with the
rheostat on the first notch.

474 ARMATURE WINDING AND MOTOR REPAIR

ui %

ON 8J TA1

-ON

.2 a

JO '0 N

edray

V* V* V* V* V* V* \* \* \* V* \* V* V* V* V* V* V* v*
W\ \ WN WN N ^ "X *X WX X "X *X WN \ \ rt\ \ cK

<* <N O

.2=8

sdray

c.t

\^|i \* v* \c

\ M\ w\ \

sdray

ui '

jo -ON

APPENDIX

475

hes

-HOOOOC5

OOMOOM OOOOOfN O?OIN(N

*t>O-*O5 (Nt^<NOOO -*)IOOJ

II-IT-I ININCOPO-^ "501^00

OOOOOOCO-* 00ilNOO

<N 00 CD O CO 00 CO <-* CN

coec^iocD t^Oii-ieo

00 <N Tjn Tjl O 00 00 O5 CO 00 t>^ O5 CO

-Ci-HlNCC ^M<CiCOOO O5-tCOCO

a3nT?3 -g y -g

uorjBjnsui

fOOOOCOOO
CO <** CO OS W

W CO O5 CO l^ CO OS CO Hi

rH *-ii-ie^<NC3 COrtfUS

'M-f-^OOM

00 O (N C>

SS 55 S

(MCOOOIN

i-i (NCOiO(N O5OCOO

-O<Ni-ieO <N<N
<N CO *' CO 00 O <N

Ja o

nd a number equal to or greater than the given hp.
mber of the smallest wire permissible. For other e
for 0.75, and 0.78 for 0.70 per cent, efficiency.

conditions of wiri
, will be found th
85; 0.89 for 0.80;

nder
the he
by 1.06

476

ARMATURE WINDING AND MOTOR REPAIR

aximum hp. allowed on wires according to Na
ities. P. F., 0.85; Efficiency, 0.90.

Table showing
and carrying ca

O -H t- 00 <$< N. OS

3 S S 2 ?3i <N co

SS^ggg

a3ntj3 -g 35,

ing
ties

2-
og

uoi^jnsut

^H ^H 1 (N

*H rt (N (N CO

APPENDIX

477

rHOO

,-H rH ^H r-l (N <N CC

T-iOO

'-iM>COO(NOi-c
_ _ ^ ^H IN (N CO

tfl

c 2 <B 5
S3 *<

I* "a *3

0> ,0 W Q.I3

ft* *

1^1*

2 Q *& i

UI4

2 >> 5

fl a,

^^^^^
*" "^s *"

ic o oq S e

|I|>1

5 !!i

> ^ * -a *"

T3 ^ "O

te-2. - J

o > ^ >. **

*" - 3 -

llfll

478

ARMATURE WINDING AND MOTOR REPAIR

General Motor Data. The following table gives the horsepower, vol-
tage and speed of standard motors at various frequencies for two- and
three-phase operation:

VOLTAGE, HORSEPOWER AND SPEEDS OF STANDARD MOTORS

Windings

Cycles

Volts

H P .

Rpm.

2- and 3-phase

110, 220, 440, 550

1, 2,3, 5, 7.5, 10, 15

1800

2- and 3-phase

110, 220, 440, 550

0.75, 1.5, 2, 3, 5, 7.5

1200

3-phase

220, 550

1,1.5, 2, 3, 5,

1200

3-phase

220, 550

1,2,3

800

3-phase

110, 220, 440

1, 2, 3, 5

1500

3-phase

110, 220, 440

1, 2, 3, 5

750

STANDARD POTENTIALS FOR WHICH SKELETON FRAME MOTORS ARE

WOUND

Style of winding

Cycles

Volts

Under 50 hp.

50 hp.
and above

3- and 2-phase

60
40
25

220
220
220

440-2200
550
440

3-phase

Standard motors will develop considerably more torque than that given
at the rated speed and voltage, so that there is ample margin to carry full
load or temporary overloads under ordinary variations of voltage.

The maximum output of an induction motor varies with the square
of the voltage at the motor's terminal, but motors will give their rated
output even with a drop of 10 per cent, in the voltage, as their maximum
output is greatly in excess of the rated value. At the lower potential
the efficiency and power factor will be increased at light loads; the full
load values, however, are usually somewhat lower.

Transformer Rating for Alternating-current Motors. For the larger
motors the capacity of the transformers in kilowatts should equal the out-
put of the motor in hp. Small motors should be supplied with a somewhat
larger transformer capacity, especially if, as is desirable, they are expected
to run most of the time near full load, or even at slight overload. Trans-
formers of less capacity than those given in the following table should
not be used even when a motor is to be run at only partial load.

For the operation of industrial motors, from three-phase systems, three
single-phase units or one three-phase unit are recommended, although,
if desired, two single-phase transformers may be used. The use of
the three-phase transformer greatly reduces the space required and makes

APPENDIX

479

Phase B

Taps

FIG. 299. Connections for a 2-phase auto-starter for a 4-wire circuit.

Generator

Phase A Phase B

Taps

FIG. 300. Connections for a 2-phase auto-starter for a 3-wire circuit.

Tape

FIG. 301. Connections for a 3-phase auto-starter.

480 ARMATURE WINDING AND MOTOR REPAIR

the wiring very simple, while the only advantage gained in using three
single-phase transformers rather than a three-phase transformer is that
in the case of one transformer burning out, the other two may be used to
operate the motor at reduced load.

RATINGS OF TRANSFORMERS REQUIRED FOR INDUCTION MOTORS

Size of motor
hp.

Kilowatts per transformer

Two single-phase
transformers

Three single-phase
transformers

One three-phase
transformer

0.6

1.5

1.0

2.0

1.5

3.0

2.0

5.0

4.0

3.0

7.5

'10

5.0

4.0

10.0

7.5

5.0

15.0

10.0

7.5

20.0

15.0

10.0

30.0

25.0

15.0

50.0

40.0

25.0

75.0

100

50.0

30.0

100.0

RATING OF TRANSFORMERS FOR THREE- AND TWO-PHASE INDUCTION
MOTORS ON VARIOUS CIRCUITS

Single-phase transformer voltages

jjeiiverea
voltage of
circuit

110-volt motor %

220-volt motor

Primary

Secondary

Primary

Secondary

1100

2200

1100

2200

122

1100

2200

244
244

SIZE OF WIRES FOR SINGLE-PHASE MOTORS

220 volts

Jlp.

Full-load current-amp.

Size of wire B. & S. gauge

APPENDIX

481

SIZE OF WIRES OF DIRECT-CURRENT MOTORS

Horsepower

220 Volts

Full load current,
amp.

Size of wire, mains,
B. & S. gauge

fize of wire, branches,
B. & 8. gauge

1.0

2.0

3.0

4.0

5.0

7.5

10.0

12.5

15.0

17.5

4 ,

20.0

25.0

30.0

113

2/0

35.0

131

2/0

3/0

40.0

150

2/0

4/0

45.0

169

3/0

4/0

50.0

188

4/0

250,000 C.M.

55.0

206

4/0

300,000 C.M.

60.0

225

4/0

300,000 C.M.

65.0

244

250,000 C.M.

350,000 C.M.

70.0

263

300,000 C.M.

400,000 C.M.

75.0

281

300,000 C.M.

500,000 C.M.

80.0

300

350,000 C.M.

500,000 C.M.

85.0

319

400,000 C.M.

500,000 C.M.

90.0

338

500,000 C.M.

600,000 C.M.

95.0

356

500,000 C.M.

600,000 C.M.

100.0

375

500,000 C.M.

700,000 C.M.

125.0

463

700,000 C.M.

900,000 C.M.

150.0

563

800,000 C.M.

1,100,000 C.M.

200.0

750

1,300,000 C.M.

1,700,000 C.M.

250.0

938

1,700,000 C.M.

2-900,000 C.M.

300.0

1,125

2-800,000 C.M.

2-1,100,000 C.M.

Column headed " Size of wire, branches " gives size of wire for branches
and for mains supplying one motor and is based on 50 per cent, over-
load.

Column headed "Size of wire, mains" gives size of wire to be used for
mains, but in no case must the size of these mains be less than that re-
quired for the 50 per cent, overload on the largest motor such mains
supply.

The question of drop is not taken into consideration in these tables.

482 ARMATURE WINDING AND MOTOR REPAIR

aT - M

- 8 a

TtH ^ TiHT H^^ C s q0o0o0cOOTt ,^ cv . )(N ^ |000

g "?

TtHTjHT^T^TtHT^^fNfNOQOOOCOCOO^TH^COC^

O10B8

lit

N . l ocooc5 22g ,og 3 .. 32 o S o E:

3 5) *

Sja

O Q

O pg CO

<N CM CO rt< TfH O O

ffl-

(N CO

ire main
. gauge

(N CO CO

'" V

Oj"fl

r2 2 a

X) T _22^coco^io^bIoi2c5^ioc

OOOOOiCOiOO^OOOOOOOOOOO

APPENDIX

483

000 .00. 00

d 9 O, Q d

o o

1-Hr-lOOOOOO
CO ^^

odd

OOOOOOOOQQOO

d d d d d d

p o o o o

CO ^ Tt< Tj< Tt<

Ifl

^i tO I s *

OOOOOOOOOOOO

oooooidotocJooc5

t>.t>.QOOOO5OiO<MOOOO

i I i I rH C<l O< CO

484 ARMATURE WINDING AND MOTOR REPAIR
WIRING DATA FOR DIRECT-CURRENT MOTORS, 115 VOLTS

Horsepower

Approx.
full
load
current,
amp.

Size of
fuses, amp.

Size of
fused switches,
amp.

Size of wire,
B. & S. gauge

Size of conduit,
inches

2.5

y 2

4.4

30 i

8.4

17.0

23.6

38.7

57.6

100

75.1

100

113

150

200

151

175

200

000

191

225

400

0000

226

275

400

300000

2>i

264

325

400

400000

300

375

400

500000

375

*450

500

600000

405

*500

500

700000

480

*600

600

900000

520

*650

700

1000000

555

*700

700

1100000

*When fuse sizes exceed 600 amperes, circuit breakers of approved
type should be substituted.

APPENDIX 485

WIRING DATA FOR DIRECT-CURRENT MOTORS, 230 VOLTS

Horsepower

Approx.
full
load
current,
amp.

Size of
fuse, amp.

Size of fused
switches, amp.

Size of wire,
B. & S. gauge

Size of conduit,
inches

1.3

2.3

4.2

8.6

11.8

19.0

28.6

37.6

55.0

100

73.8

100

125

200

113

150

200

130

175

200

000

150

175

200

000

185

225

400

0000

200

250

400

250000

219

275

400

300000

238

300

400

350000

260

325

400

400000

274

350

400

450000

288

375

400

500000

308

400

500000

328

*425

*500

500000

100

368

*450

*500

600000

*When fuse sizes exceed 600 amperes, circuit breakers of approved
type should be substituted.

486

ARMATURE WINDING AND MOTOR REPAIR

WIRING DATA FOR DIRECT-CURRENT MOTORS, 550 VOLTS

Horse-
power

Approx.
full-load
current-amp.

Size of
fuses, amp.

Size of fused
switches, amp.

Size of wire,
B. & S. gauge

Size of
conduit,
inches

0.52

0.94

1.80

3.50

y 2

4.93

8.00

12.00

15.60

22.80

30.70

40.00

47.00

55.00

100

62.00

100

78.00

100

84.00

100

92.00

125

200

98.00

125

200

108 . 00

150

200

114.00

150

200

120.00

150

200

129.00

150

200

138.00

175

200

000

100

152.00

200

0000

WIRING DATA FOR INDUCTION MOTORS SINGLE-PHASE 110 VOLTS,
ALL FREQUENCIES, STANDARD SPEEDS

Horse-
power

Approx.

full current-
amperes

Size wire
B. & S.
gauge

Size of
switches
in amperes

Size of
starting
fuses,
amperes

Size of
running
fuses,
amperes

Size of
conduit,
inches

16.4

24.0

33.6

100

43.6

100

54.0

100

80.0

200

100

106.0

200

125

APPENDIX

487

WIRING DATA FOB INDUCTION MOTORS SINGLE-PHASE 220 VOLTS,
ALL FREQUENCIES, STANDARD SPEEDS

Horse-
power

Approx.

full current-
amperes

Size of
wire,
B. & S.
gauge

Size of
switches
in amperes

Size of
starting
fuses,
amperes

Size of
running
fuses,
amperes

Size of
conduit,
inches

8.2

12.0

16.8

21.8

27.0

40.0

100

53.0

100

WIRING DATA FOR INDUCTION MOTORS THREE-PHASE 110 VOLTS,
ALL FREQUENCIES, STANDARD SPEEDS

Horse-
power

Approx.

full current-
amperes

Size of
wire,
B. & S.
gauge

Size of
switches in
amperes

Size of
starting
fuses,
amperes

Size of
running
fuses,
amperes

Size of
conduit,
inches

' 10

100

200

125

200

150

112

101

400

225

150

125

000

400

250

187

150

0,000

400

325

225

175

300.000

400

262

210

400,000

500

*500

315

246

450,000

500

*500

369

* When fuse sizes exceed 600 amperes, circuit breakers of approved
type should be substituted.

488

ARMATURE WINDING AND MOTOR REPAIR

WIRING DATA FOR INDUCTION MOTORS THREE-PHASE 220 VOLTS,
ALL FREQUENCIES, STANDARD SPEEDS

Horsepower

Approx.
full
current-
amperes

Size of
wire,
B. &. S.
gauge

Size of
switches
in
amperes

Size of
starting
fuses,
amperes

Size of
running
fuses,
amperes

Size of
conduit,
inches

1.0

2.0

/ t

3.0

6.0

9.0

13.3

19.5

25.5

100

37.5

100

50.5

200

130

62 '.5

200

150

75.0

200

100

87.5

200

125

105.0

400

225

150

123.0

000

400

250

175

186.0

300,000

400

275

100

243.0

450,000

500

*550

350

150

362.0

800,000

800

*850

*550

200

480.0

1,200,000

1,000

*1,150

*725

* When fuse sizes exceed 600 amperes, circuit breakers of approved
type should be substituted.

APPENDIX

489

WIRING DATA FOR INDUCTION MOTORS THREE-PHASE 440 VOLTB,
ALL FREQUENCIES, STANDARD SPEEDS

Horse-
power

Appro*,
full current-
amperes

Size of
wire,
B. & S.
gauge

Size of
switches
in amperes

Size of
starting
fuses,
amperes

Size of
running
fuses,
amperes

Size of
conduit,
inches

1.5

3.0

y 2

4.5

i.7

9.8

12.8

18.8

25.2

100

31.3

100

37.5

100

43.8

100

i'K

52.5

200

135

tot

61.5

200

175

93.0

400

225

140

100

121.5

000

400

275

175

150

181.0

300,000

400

275

200

240.0

450,000

400

*550

350

* When fuse sizes exceed 600 amperes, circuit breakers of approved
type should be substituted.

Circuit-breakers for Overload Protection of Motors. There are two
overload conditions that may injure an electric motor. (1) A continu-
ous load beyond the overload rating of the motor and (2) by an exces-
sive momentary overload Overloads of the second class can obviously
be much larger than those of the first class without exceeding safe limits.
For example, a continuous overload of 50 per cent, may injure a motor,
while the safe momentary overload may be three or four times the full
load. If a motor is subjected to overloads of both these classes, ade-
quate overload protection may be difficult with either a circuit-breaker
or fuses. Both devices must be set for maximum allowable current,
usually considerably in excess of the continuous safe carrying capacity
of the motor, and a continuous overload current inside this limit may
work injury. Such conditions are unusual, however, and either fuses
or a properly designed circuit-breaker will ordinarily afford all necessary
protection.

For Induction Motors. In case the squirrel-cage induction motors is
protected by a circuit-breaker, the motor should be so connected as to be
without overload protection when the starter handle is in the starting
position; in addition, the circuit-breaker should be provided with a time

490 ARMATURE WINDING AND MOTOR REPAIR

element device to prevent opening when the starter handle is thrown to
the running position. Without this time element the circuit-breaker
when connected as just described, must be calibrated for two or three
times the rated full-load current of the motor. If the circuit-breaker
must be effective in the starting position it must be provided with a time
element and be calibrated for from three to five times the rated full-load
current; with this calibration no protection will be afforded from contin-
uous overloads which might work injury.

For Motors Carrying a Uniform Load. That is, motors driving line
shafts, fans, machine tools, or any machines which give them a fairly
uniform load, carrying maximum continuous rated output at least part
of the time and possibly full-rated overload for short periods. In general,
the overload protection for such motors should have a normal calibration
equal to 125 per cent, of normal full-load motor current. If no standard
circuit-breaker is listed for this rating, the next larger listed size should
be used. A circuit-breaker selected in this manner will have a maximum
calibration at least twice full-load current. Where overloads greater
than 200 per cent, of normal current are regularly experienced for periods
not exceeding two seconds, a time element device should be added to the
circuit-breaker. For heavy overloads frequently recurring and continu-
ing for periods of five seconds or longer, the circuit-breaker should have
a maximum calibration of at least 10 per cent, in excess of the maximum
current.

For Motors Started and Stopped Frequently. Motors on cranes, ele-
vators, pumps, air compressors, or other apparatus requiring frequent
starting and stopping, should have circuit-breakers with a calibration
of at least 25 per cent, in excess of the maximum current actually required
by the motor during any continuous period of more than five seconds.
A time element device should be supplied to care for greater overloads
lasting less than five seconds.

Belting. Rubber belts should always be kept free from grease or
animal oils. If they slip, moisten the inside of the belt with boiled
linseed oil. Some fine chalk, sprinkled on over the oil, will help the belt.

Length of Belts. Add the diameter of the two pulleys together, multi-
ply by three and one-seventh, divide the product by two, add to the
quotient the distance between the center of the shafts, and the product
will be the required length.

ORDINARILY ACCEPTED EQUIVALENTS OF BELTING

2-ply Rubber Light single leather.
3-piy Rubber Medium single leather.
4-ply Rubber Heavy single leather.
5-ply Rubber Light double leather.
6-ply Rubber Medium double leather.
7-ply Rubber Heavy double leather.
8-ply Rubber Triple leather.

APPENDIX

491

The thickness of rubber belting is usually figured at J-f 6 inch per ply.

Horsepower Transmitted by Belting. One-inch single belt moving at
a velocity of 1000 ft. per minute equals one horsepower. One-inch
double belt moving 700 ft. per minute equals one horsepower. The
horsepower of any belt equals its velocity in feet per minute, multiplied
by its width and divided by 1000 for single, and by 700 for double belts.

The following table is based on single belts running on pulleys of equal
diameter, revolving at 100 rpm. The power transmitted at other speeds
is in direct proportion. For double belts, multiply by ten-sevenths.

HORSEPOWER TRANSMITTED BY BELTING

Diameter of
pulley in
inches

Width of belt in inches

.44

.65

.87

1.09

1.31

.51

.76

1.01

1.27

1.53

.58

.87

1.16

1.45

1.75

.65

.98

1.31

1.64

1.97

.73

1.09

1.45

1.81

2.18

.80

1.20

1.60

2.00

2.40

.87'!. 31

1.75

2.18

2.62

.95

1.42

1.89

2.36

2.83

1.02

1.52

2.02

2.53

3.05

1.09

1.64

2.19

2.73

3.29

1.16

1.74

2.32

2.91

3.48

1.24

1.85

2.47

3.09

3.70

1.31

1.96

2.62

3.27

3.92

1.39

2.07

2.76

3.45

4.14

1.45

2.18

2.91

3.64

4.36

1.60

2.40

3.20

4.00

4.80

3.50

4.40

5.20

7.0

8.7

10.5

12.2

14.0

16.0

17.0

19.0

3.80

4.70

5.70

7.6

9.5

11.3

13.2

15. 1

4. 10

5. 10

6. 10

8.1

10.2

12.2

14.3

16.3

4.40

5.40

6.60

8.7

10.9

13. 1

15.3

17.4

19.0

22.0

24.0

4.70

5.80

7.00

9.3

11.6

14.0

16.3

18.6

4.90

6.20

7.40

9.9

12.4

14.8

17.3

19.8

5.20

6.50

7.80

10.5

13.1

15.7

18.3

20.9

24.0

26.0

29.0

5.50

6.90

8.30

11.0

13.8

16.6

19.3

22.1

25.0

28.0

30.0

5.80

7.30

8.70

11.6

14.6

17.5

20.4

23.3

26.0

29.0

32.0

6.10

7.60

9.20

12.2

15.3

18.2

21.4

24.3

28.0

31.0

34.0

6.40

8.00

9.60

12.8

16.0

19.2

22.4

25.6

29.0

32.0

35.0

6.70

8.40

10.00

13.4

16.8

20.1

23.4

26.8

7.00

8.80

10.40

14.4

17.4

21.0

24.4

28.0

31.0

35.0

38.0

7.20

9.00

10.90

14.6

18.2

21.8

25.4

29.0

33.0

36.0

40-0

7.80

9.80

11.80

15.6

19.6

23.6

26.4

31.2

35.0

39.0

43.0

....

8.80

10.80

13.10

17.4

21.8

26.2

30.6

34.8

39.0

44.0

48.0

9.60

12.00

14 40

19. 2

OQ Q

oo a

OQ A

4. O

AO n

rq f)

10.40

13.00

15.60

21.0

^^ . u
26.2

^O . o

31.4

Oo . O

36.6

Oo . 4

41.8

"0 . U

47.0

4O . U

52.0

)O U

58.0

11.40

14.20

17.00

22.6

ofi 4

04 A

4^ 4

e t f\

C'7 f\

AO A

....

12.20

15.20

19.40

24.4

^O . rt

30.6

o4 U

36.4

oy . o
42.8

fcO . 4

48.6

51 * U

55.0

Of .U

61.0

\Z I/

67.0

492 ARMATURE WINDING AND MOTOR REPAIR

Belting Rules. The power transmitted by a belt depends upon its
width and its thickness and the speed at which it travels. Hence it is
customary (as mentioned under "Horsepower Transmitted by Belting")
to express the transmitting capacity as the speed in icet per minute
required by a belt one inch wide to transmit one horsepower. For single
or light double belt, this expression is usually given a value of about
700, and for heavy double belt approximately 450. Thus a double belt
four inches wide, running at 2250 ft. per minute will transmit 20 hp.

Roughly> belt speeds should not exceed one mile per minute ; this speed is
given when the diameter of either pulley in inches multiplied by its rpm.
equals 20,000 (D X rpm. = 20,000).

Minimum diameter of pulleys for long life of heavy belts :

For double belts ............................... 12 in.

For double belts, extra flexible ................... 10 in.

For double 3-ply belts .......................... 18 in.

hp. 126,500
Ordinarily, wd = CL- X r~

where w width of belt, d = diameter of pulley (both in inches), and
P = belt stress per inch width. Safe values for P are as follows :

Single belts, above 4000 ft. per min., 33 lb.; below 4000, 46 lb.; max.
stress never over 50 lb. Double belts, 40 to 70 lb. ; 3-ply belts, 70 to 95 lb.

With standard construction and normal belt stress, the stress on the
shaft and bearings is within safe limits as long as d exceeds w. To find
length L (length) of belt when the pulley diameters D, d and the distance
A between centers are known, use the formula
L = 1.62 (D + d) + 2A

To determine distance between pulley centers use the formula

A = K (D - d}

where the value of the constant K ranges from 3 to 4, 3.5 being a good
average value when the diameter D is from 4 to 6 times that of d.

The following precautions should be observed :

a. The formula does not apply when pulleys are of nearly the same
diameter; in this case the distance should be great enough to allow some
belt sag.

6. As a general rule the distance A should not exceed 25 ft. in any case.

Rules for Pulley Sizes.---The following formula can be used to calculate
sizes of pulleys for motor drives :

, DXN DXN

d = n = - j

n d

D = Diameter of driver.
d = Diameter of driven pulley.
N = Revolutions per minute of driver.
n = Revolutions per minute of driven pulley.

Speed of Pulleys. When the diameter of the driven pulley is given,
to find its number of revolutions proceed as follows: Multiply the dia-

APPENDIX 493

meter of the driver by the number of its revolutions, and divide the pro-
duct by the diameter of the driven. The quotient will be the number of
revolutions of the driven pulley.

When the diameter and revolutions of driver pulley are given, to find
the diameter of the driven pulley that will make any given number of
revolutions in the same time, proceed as follows : Multiply the diameter
of the driver pulley by its number of revolutions, and divide the product
by the number of revolutions of the driven pulley. The quotient will
be the diameter of the driven pulley.

To ascertain the size of the driver pulley proceed as follows : Multiply
the diameter of the driven pulley by the number of revolutions you wish
to make, and divide the product by the revolutions of the driver pulley.
The quotient will be the diameter of the driven pulley.

Chain Drives. Chains are now made to transmit power for any pur-
pose in sizes varying from ^ hp. at 3000 rpm. to 3000 hp. at 50 rpm.
By the use of a chain instead of a belt or gears, power may be transmitted
with a positive speed ratio on short centers quietly and with a very high
efficiency. Chains should be lubricated with a heavy paste grease con-
taining no solid matter such as automobile transmission grease.

Approximate Weight of Solid Pinions and Armed Sprockets. When
T is the number of teeth; F the face in inches; C a constant in pounds per
inch in face per tooth. Then T X F X C = weight of armed sprocket.
Add 25 per cent, if the sprocket is split and add 50 per cent, if spring
sprocket and split. For solid pinions, weight = C X T 2 X (F + 1).

Points to Consider when Calculating Size of Chain Morse Chain
Co. When the number of teeth equals T and the exact outside diameter
is D, then T should be less than 20 when D equals the pitch diameter.
When T is more than 20 teeth, D is equal to the pitch diameter plus twice
the addendum.

The following points should be considered in this connection:

(1) Use sprockets having an odd number of teeth whenever possible.

(2) When specially authorized, a larger number of teeth than shown
may be cut in large sprocket.

(3) Thickness of sprocket rim, including teeth, should be at least 1.2
times the chain pitch.

(4) The number of grooves in the sprocket, their width and distance
apart, varies according to pitch and width of chain.

(5) The width of the sprocket should be ^ to K inch greater on small
drives, and K to ^ inch greater on large drives than nominal width of
the chain.

(6) An even number of links in the chain and an odd number of teeth
in the wheels are desirable.

(7) Horizontal drives preferred with slack on top strand, but for
short drives without center adjustment slack thould be on the bottom
strand.

(8) Adjustable wheel centers desirable for horizontal drives and neces-
sary for vertical drives.

494'

ARMATURE WINDING AND MOTOR REPAIR

J"^

^-t

o o

i (

11 O

i 1

(*Ti

C^l

l^- CO

h ^

t 1

1 1

"* CO

o o

I 1

CM <M

^^ CO

T t

i-H

c-.

T 1

' *

CO
1 1

1 1

-N
iH

i I

O CO

xao

T 1

1 1

i-H

'"I

' '

i-K

i-H

O 10
tX) CO

i o

t-l

. ^

+3
3

la 03

f-i rt 43
g -

03
>

> 'S

sl -

'g -H

O Cd

228

"T3 ^"^

* *Z*

CD 1^ 5

OQ
, '

.a"

of teeth j Small sprocket
Desirable number of teeth

sprockets
Maximum number of teeth i

ets. (See note 3.)
Desirable number of teeth

sprockets
To find pitch diameter of wh

ply number of teeth by (
Addendum For outside dii

CO
i-H

-2

3 *2

II
1^

Maximum rpm

Tension per 1 ^
- u -jxu Small sprock
inch width >
, . ,, Small sprocke
chain, Ibs. }

Radial clearance beyond

-5

'5
cr

Constants for solid pinions .
Constants for armed sprock

APPENDIX

495

(9) Avoid vertical drives.

(10) Allow a side clearance for chain (parallel to axis of sprockets and
measured from nominal width of chain) equal to the pitch.

(11) Maximum linear velocity for commercial service, 1200 to 1600
feet per minute.

HORSEPOWER TRANSMITTED BY STEEL SHAFTING

Diam-
eter of
shaft in
inches

Revolutions per minute

100

125

150

175

200

225

250

300

350

400

1.2

1.4

1.1

2.1

2.4

2.6

3.1

3.6

4.3

5.0

1 Me

2.4

3.1

3.7

4.3

4.5

5.5

6.1

7.3

8.5

9 7

1 Me

4.3

5.3

6.4

7.4

8.5

9.5

10.5

12.7

14.8

16 9

I 1 He

6.7

8.4

10.1

11.7

13.4

15.1

16.7

20.1

23.4

26.8

1'Mft

10.0

12.5

15.0

17.5

20.0

22.5

25.0

30.0

35.0

40.0

2 Me

14.3

17.8

21.4

24.9

28.5

32.1

35.6

42.7

49.8

57.0

2 ?6

19.5

24.4

29.3

34.1

39.0

44.1

48.7

68.5

68.2

78.0

2 1 Me

26.0

32.5

39.0

43.5

52.0

58.5

65.0

78.0

87.0

104.0

2'Me

33.8

42.2

50.6

59.1

67.5

75.9

34.4

101.3

118.2

135.0

3 Me

43.0

53.6

64.4

75.1

85.8

96.6

107.3

128.7

150.3

171.6

3 Y\o

53.6

67.0

79.4

93.8

107.2

120.1

134.0

158.8

187.6

214.4

3i He

65.9

82.4

97.9

115.4

121.8

148.3

164.8

195.7

230.7

243 . t>

3i Me

80.0

100.0

120.0

140.0

160.0

180.0

200.0

240.0

280.0

320.0

4 #6

113.9

142.4

170.8

199.3

227.8

256.2

284.7

34J.7

398.6

455.6

4'Me

153.3

195.3

234.4

273.4

312.5

351.5

390.6

468.7

646.8

625.0

HORSEPOWER TRANSMITTED BY SINGLE ROPES
(Working strain = 200e? 2 where d is diameter of rope)

Diameter of rope in inches

minute

1.0

1.25

1.50

1.75

2.0

1000

1.24

2.25

3.57

5.59

8.02

10.85

14.20

2000

2.70

3.84

6.84

10.68

15.39

20.93

27.36

2500

3.30

4.71

8.38

13.10

18.86

25.66

33.54

3000

3.83

5.46

9.80

15.39

21.87

29.74

38.88

3500

4.30

6.23

10.09

17.33

24.94

34.03

44.35

4000

4.74

6.83

12.15

18.98

27.33

37.17

48.59

4500

5.01

7.24

12.89

20.15

29.00

39.45

51.57

5000

5.20

7.47

13.29

20.76

29.89

40.65

53.15

5500

5.29

7.60

13.53

21.14

30.43

41.39

54.11

6000

5.08

7.32

13.10

20.36

29,32

39.77

52.12

6500

4.74

6.83

12.13

19.00

27.34

37.21

48.63

7000

4.12

5.93

10.54

16.47

23.72

32.26

42.18

Smallest sheave, diameter,

inches

26.00

30.00

42.00

54.00

60.00

72.00

84.00

Allowable weight, tension

carriage, Ib . .

80.00

110.00

200.00

300.00

450.00

600 . 00

750.00

NOTE. The horsepower decreases when the velocity is above 5500 feet on account
of centrifugal force.

496

ARMATURE WINDING AND MOTOR REPAIR

s i

I :

I 1 5 5 B
S So o o

APPENDIX

497

I I

r . ^

Cu ^
"^ T3

-d -e

.&^

ii -S

03 .S

S.-fl'**^!

1 "2 5 15 -a

| PH PH g -g
^ -S S ^ S "S 2

2 S S g S a -5

fPpfl

^ fl} W fl} .

^3 Jsl j s

o ^! > t 3 *> ? ^

!> ?^ 10 r~> IO V^ w

10 A3 ^ 42 ^

i-H iH i-H (M C<l H

""^ F ^ "T^ *^ T3 P C '^
.^ .^ .21 .zi .zi ,s ,ii

> >

^^3333333

^ .- J

ill^i

Slols

^3 43 ^3 ^5
5 -2 5

ffi S ffi ffi
Ill's

^ ^-

o o o c^

a s s a

.2 .2 .25 .S

p p p p

f-3
If

Q.3

S "3

|5 1 1

3^21

& ^ 6 o

498 ARMATURE WINDING AND MOTOR REPAIR

Some Handy Rules

Diameter of a circle X 3.1416 = Circumference.

Radius of a circle X 6.283185 = Circumference.

Square of the radius of a circle X 3.1416 = Area.

Square of the diameter of a circle X 0.7854 = Area.

Square of the circumference of a circle X 0.07958 = Area.

Half the circumference of a circle X by half its diameter = Area.

Circumference of a circle X 0.159155 = Radius.

Sqare root of the area of a circle X 0.56419 = Radius.

Circumference of a circle X 0.31831 = Diameter.

Square root a the area of a circle X 1.12838 = Diameter.

Diameter of a circle X 0.86 = Side of inscribed equilateral triangle

Diameter of a circle X 0.7071 = Side of an inscribed square.

Circumference of a circle X 0.225 = Side of an inscribed square.

Circumference of a circle X 0.282 = Side of an equal square.

Diameter of a circle X 0.8862 = Side of an equal square.

Base of a triangle X by ^ the altitude = Area.

Multiplying both diameters and .7854 together = Area of an ellipse

Surface of a sphere X by % of its diameter = Solidity.

Circumference of a sphere X by its diameter = Surface,

Square of the diameter of a sphere X 3.1416 = Surface.

Square of the circumference of a sphere X 0.3183 = Surface.

Cube of the diameter of a sphere X 0.5236 = Solidity.

Cube of the radius of a sphere X 4.1888 = Solidity.

Cube of the circumference of a sphere X 0.016887 = Solidity.

Square root of the surface of a sphere X 0.56419 = Diameter.

Square root of the surface of a sphere X 1.772454 = Circumference.

Cube root of the solidity of a sphere X 1.2407 = Diameter.

Cube root of the solidity of a sphere X 3.8978 = Circumference.

Radius of a sphere X 1.1547 = Side of inscribed cube.

Square root of (K of the square of) the diameter of a sphere = Side
of inscribed cube.

Area of its base X by ^ of its altitude = Solidity of a cone or pyramid,
whether round, square or triangular.

Area of one of its sides X 6 = the surface of a cube.

Altitude of trapezoid X % the sum of it parallel sides = Area.

APPENDIX 499

EQUIVALENT VALUES OF CENTIGRADE AND FAHRENHEIT SCALES

Temperature

Centigrade

Fahrenheit

Centigrade

Fahrenheit

176

185

194

203

100

212

105

221

110

230

115

239

100.4

120

248

104

125

257

107.6

130

266

113

135

275

122

140

284

131

145

293

140

150

302

149

155

311

158

160

320

167

165

329

INDEX

Angle for setting brushes, 333
Acid fumes and gases, troubles due

to, 377
Alternating current machines,

causes and remedies of

troubles in, 386
Alternating-current windings (see

also table of contents for

Chapter II), 26
coils for, 34
coil pitch of, 32
connections for coils of, 43
connecting a chain winding, 48
changing star to delta, 41
checking phase relationship in,

52
double-layer, lap connected,

47
distributed and concentrated,

27
easily remembered rules for,

full and fractional pitches, 37
grouping of coils for, 35
lap and wave, 28
laying out and connecting, 35
phase spread of, 32
polarity of coil groups in, 40
progressive and retrogressive

wave, 41

simple winding diagram for, 38
single and polyphase, 30
spiral or chain, 27
2-phase from 4-phase, 33
3-phase from 6-phase, 33
types of, 27

whole and half-coiled, 30
Armature slots, 23

Armature winder's tools, 444
Auto-starter, overhauling of, 351

Balancing armatures

large d. c. sizes, 87

machine for, 460

small and medium sizes, 150
Banding procedure:

for small d. c. armatures, 65

for large d. c. armatures, 82

for railway armatures, 99

for large rotary armature, 410

for different armatures, 146

precautions, 148

simple scheme, 408

size of wire for bands, 148

tension of bands, 147

tool for applying large bands,
86

use of crane for, 409
Banding wire:

for large armatures, 84

for all sizes of armatures, 146

knock in armature caused by,
395

tension of, 147

tension block, 454
Bar bender for making coils, 458
Bars of commutator :

boring out, 313

copper for, 319

end rings for, 316

excessive wear of, 318

high and low, 302

removal of, 304

repairing of, 305

repair of burned sections, 303

replacing of, 306

tightening, 306

501

502

INDEX

Basket coils for induction motors,

195
Belts, poor joints in cause sparking

of brushes, 394
horsepower transmitted by,

491

pulley sizes for, 492
rules for use of, 492
speed of pulleys for, 492
static sparks from, 440
troubles due to tension of, 391
Bench stand for winding d. c.

armatures, 447
Blackening of commutator:

remedy for, 343

Boring out end of commutator, 313
Bracing windings of turbo-genera-
tors, 228

of large a. c. generators, 223
Break-down test for turbo-genera-
tors, 229

Bridging commutator bars, 258

Brushes, adjusting and correcting

troubles (see also table

of contents for Chapter

XIII), 326

adjustment of brush-holders,

329

angle for setting of, 333
blackening of, 343
causes of rapid wear of, 331
causes and remedies of

troubles, 341
checking setting, 334
chattering of, remedy for, 344
contact drop, 334
current density, 335
effect of defective fields, 339
fitting or grinding-in, 326
friction of, 335
glowing, 335
hardness, 336
heating of, remedy for, 343
honey combing, 335
incorrect thickness of, 340
incorrect spacing, 338
inertia of, 336

Brushes (cont.):

locating causes of troubles, 337

locating electrical neutral for,
332

picking up copper, 343

pressure of, 334

specific resistance, 335

terms used, 334

too low brush pressure, 337

unequal air gaps, 339

wrong characteristics of, 341
Brushes for undercut commutator,

322

Brush-holder, adjustment of reac-
tion type, 406

heating of, 407

making adjustments of, 329
Brushes set incorrectly, 404

wrong setting of, 405
Brush shunts, loosening of, 344
Brush studs, heating of, 403
Burned out starting winding, 429
Burred commutator bars, troubles
resulting from, 402

Calculating wiring circuits, for-
mula for, 468

Centigrade and Fahrenheit tem-
peratures, equivalent
values, 499

Chain winding, connecting of, 48
Chain drives:

calculating size of chain, 493
data for, 494

pinions and sprockets for, 403

Changes in d. c. motors (see also

table of contents for

Chapter X), 237

Changes in speed of d. c. motor,

237

air gaps, 242
brushes, 243
connections to commutator,

243

operating voltage, 237
winding wave to lap, 248

INDEX

503

Changes in duplex wave winding,

252

connections of coils, 244, 253
Changes in induction motor wind-
ings (see heading, proced-
ure when reconnecting in-
duction motors):
factors limiting, 286
practical methods for, 261
reconnections frequently

made, 272
Changing 440 volt motor for 220

volt operation, 435
Changing star to delta connection,

Characteristics of brushes, 341
Chord factor, meaning of, 269
Circuit-breakers for motors, 489
Cleaning motors with compressed

air, 420
troubles due to improper

cleaning, 376
Cleaning slots, solutions for, 140

by filing, 140
Coil throw, 7, 70
Coil and commutator insulation,

test voltage for, 175
Coils:

applying insulation on, 163
connections with dead coils, 98
classification of, 5
for large d. c. armatures, 77
for railway, mill and crane

motors, 90

for a. c. windings, 141
forms for making, 140
inserting in open slots, 69
inserting threaded-in type, 60
insertion in railway armatures,

insertion in large d. c. arma-
tures, 81

insulating for open slots, 67
insulation for different volt-
ages, 164, 165, 167, 168
insulation for d. c. coils, 164,
165, 168

Coils (cont.):

r insulation for induction motor

coils, 164, 167, 170
kinds of insulation for, 153
mica insulation for, 173
pitches of d. c., 7
pitches of a. c., 37
placing on armature, 110
removing from d. c. armature,

repairing damages to, 174
shaping end connections, 70
table for connecting, 110
test voltage for insulation of,

175

thickness of insulation for, 163
types and shapes of, 3
wire, strap and bar wound, 34
Coil raising tool, 449
spreader, 457
taping machine, 457
taping needle, 449
winding device, 452
Commutator:
action of, 1

blackening of, remedy for, 343
bridging bars of, 258
compound for plugging of, 304
connections for lap winding,

102

connections for wave wind-
ing, 104

flat spots on, remedy for, 342
improvised method to turn,

411

insulation of connections to, 76
locating first connection to

for lap winding, 101
locating first connection to
for wave winding, 105
making connections to, 101
number of bars, 24
plugging of, 395
poor soldering, troubles re-
sulting, 396

reconnecting with dead coil,
252

504

INDEX

Commutator (cont.):

reconnecting duplex wave

windings, 252
repairs to, 301

repaired under difficulties, 397
sandpapering, holder for, 400
seasoning and grinding, 148
soldering coil leads to, 144
sparking at brushes of, 341
table for connecting coils to,

111, 246

trouble from burred bars, 402
troubles in, 301
undercutting of, 320
undercutting with electric drill,

400

voltage between bars of, 24
winding pitch for, 6
Commutator, repairs to (see also
table of contents to Chap-
ter XII), 301
baking of, 307
blackening of in spots, 320
boring out end of, 313
brushes for undercut commu-
tator, 322

burn out between bars, 303
causes of excessive wear, 318
causes of troubles in, 301
copper for bars, 319
finishing undercut slots, 322
making micanite end rings,

318

mica used in, 315
micanite insulation for, 317
precautions when tightening,

317

refilling, 311
remedies for high and low

bars, 302

removing bars and mica, 304
removing grounds, 308
repairing burned bar, 305
replacing burned bar, 306
shaping mica end rings, 316
templet for making mica
rings, 316

Commutator, repairs to (cont.) :
temporary cover for, 310
test for oil saturated mica,

320

tightening up bars, 306
tools for undercutting mica.

320

troubles from high mica, 301
turning down surface of, 309
undercutting mica, 320
Compressed air pressure in clean-
ing motors, 420

Concentrated a. c. winding, 27
Chain or spiral a. c. winding,

Conductor or inductor, 3
Connecting coils of d. c. armature

in parallel, 244
Connecting coils of d. c. armature

in series, 253

Connections to commutator (see
table of contents for
Chapter IV).

Contact drops for brushes, 334
Controller, drum type, over haul-
ing of, 355
Cost of making repairs, 346

estimating of, 347
Cutting out coils of d. c. wave

winding, 109

of induction motor winding,
288

Dead coils :

connections of coils with, 98
wave winding with, 108
Device for detecting faults in a. c.

windings, 232

Diagnosis of motor and generator

troubles (see also table

of contents for Chapter

XV), 376

Diagrams for connecting induction

motor windings, 292
combination scheme for
several types, 436

INDEX

505

Diagrams for change of connec-
tions of induction motor
windings, 262

different poles and phases,
295

2-phase motors, 267

3-phase motors, 265
Diamond coils for induction mo-
tors, 198

Direct-current armatures (see also
table, of contents for
Chapter III) :

balancing large sizes, 87

balancing small and medium
sizes, 150

banding of, 65

cleaning slots of, 140

floor stand for, 75

loop winding for, 63

repair shop methods for re-
winding of, 56

six steps in winding of, 72

winding procedure for arma-
tures having partially
closed slots, 60

winding procedure for open
slots, 66

winding procedure for large
armatures, 77

winding procedure for railway
and crane armatures, 90
Direct-current generators :

speeds and poles of, 25

winding procedure for 3-wire
design, 89

winding procedure for rotary,

Direct-current machines, causes
and remedies of troubles
in, 376

Direct-current motors changed for
new operating conditions,
237

adjusting air gap, 242

brush changes for operation
at lower voltage, 243

changes in speed, 237

Direct-current motors (cont.) :
changes in operating voltage,

238

generator used as motor, 241
motor used as generator, 241
operating on one-half or

double voltage, 239
reconnecting for different

voltage, 243

size of wire for coils, 240
speed when reconnecting wave

to lap, 242

Direct-current windings (see also
table of contents for
Chapter I), 1

classification of coils for, 5
equipotential connectors, 21
for different number of poles,

for large armatures, 77
for rotary converters, 88
for 3-wire generators, 89
for railway, mill and crane

motors, 90
formulas for lap, 11
formulas for wave, 17
lap (multiple or parallel), 9
lap and wave for large arma-
tures, 79
loop windings, 63
multiplex lap, 14
multiplex wave, 19
numbering coil sides in slots, 8
parts and terms of, 2
symbols used in formulas for, 8
symmetrical, 20
testing of, 122
types of, 2

wave (series or two-circuit), 16
Dismantling a d. c. armature, pro-
cedure for, 56

Distributed a. c. winding, 27
Drifts used by armature winders,

448

Drying out insulation:
of d. c. machines, 181
synchronous motors, 182

506

INDEX

Drying out insulation of induc-
tion motors, 183
insulation test while, 184
insulation resistance while,

185

Duplex wave winding, reconnec-
tion of, 252

Electrical neutral, location of, 332
Element of d. c. winding, 3
End connections of coils :

appearance in lap and wave

d. c. windings, 10
insulation between, 63
shaping of in d. c. winding, 70
thickness of insulation for, 172
End rings for commutator, 316
End play of armature, 418
Engine type generator, inspection

of, 371
Equalizer rings for d. c. windings,

22
use on multipolar machines,

120

Equipotential connectors or equal-
izers, 21
Estimating cost of repairs, 347

Failure of motor to start, 430, 432
Field coils :

ad jus table shunt for, 416

heating of, 414

insulation of, 176

test for dead coils, 136

test for reversed connections,

136
Field current in d. c. generators,

473
Flat spots on commutator, remedy

for, 342
Floor stand for winding d. c.

armatures, 75

Form for recording winding data
of dismantled armature,
58

Formulas and rules for d. c.

windings :
for calculating wiring circuits,

468

for possible symmetrical wind-
ings, 21

lap (multiple or parallel), 11
multiplex lap, 15
multiplex wave (series-paral-
lel), 19

wave (series or two circuit), 17
Frames for making coils, 140
Frequency change, reconnection
of induction motor for, 276
effect of deer ease on induction-
motor-generator, 433
Friction of brushes, 335
Friction cloth blanket for commu-
tator end connections, 76
Fuse block defective, trouble due

to, 427

Fuses, sizes for a. c. motors, 442
d. c. motors, 484
causes of blowing of, 389
starting a. c. motors, 487

Gears, data for, 496
Generator, engine type, overhaul-
ing of, 371

Generator used as motor, 241
Generators, large a. c. :

bar and connector winding for,

217
coils for partially closed slots,

210

coils for open slots, 212
connecting coils of, 221
diamond coils for, 219
double windings for, 221
glowing of brushes, 335
insulation of coils for, 214
inserting shoved through coils,

215

lap and wave connections for,
214

INDEX

507

Generators, large a. c. (cont.):
bracing heavy windings of, 223
testing windings of, 221

Grinding commutator, 148

Grounds in d. c. windings:
causes of, 130
locating dead grounds, 133
telephone receiver test for,

135
tests for, 131

Grouping coils of induction motor

winding, 289

odd number coils per group, 54
procedure for, 35

Half-coiled and whole-coiled a. c.

windings, 30
Hand tools for armature winders,

445

Heating of brush studs, 403
Heating of motor or generator,

remedies for, 343
caused by poor soldering, 396
caused by variation of fre-
quency, 438
hot stator coils, 391
Holder for sand papering commu-
tator, 400
Honey combing of brushes, 335

remedies for, 343
Hoods for armatures, 99

making of, 145
Hot bearing, relief for, 440

Impregnating compounds (see also

varnishes), 176
Induction motors:

basket coils for, 195

coil insulation for, 163, 167,

170

connecting coils of, 205
diamond coils for, 198
insulation of slots, 195

Induction motors (cont.):
inserting new coil in, 203
open slots, winding of, 201
operated on different voltages

and frequencies, 284
overhauling of, 363
partially closed slots, winding

of, 194

phase wound secondary, 209
procedure when connecting

coils of new winding, 288
reconnecting for changes in

voltage, frequency, phase

and speed, 272
squirrel cage secondary, 20/
testing windings of, 202
2-phase stator, winding of, 202
winding small sizes, 192
winding with basket coils, 196
winding with diamond coils,

199

Inspection and overhauling of:
auto-starters, 351
compound d. c. motor, 358
d. c. motor starters, 349
drum- type controllers, 355
engine type generator, 371
induction motor, 363
single-phase motor, 368
slip-ring motor, 366
Inspection and repair of motor

starters, motors and gen-
erators, 345

cost of making repairs, 346
estimating cost of repairs, 347
Insulation for:

a. c. coils, 164, 167, 170
coils and slots, kinds of, 156
coils used in open slots, 67
coils of railway motors, 92
commutator connections, 76
core of railway armatures, 93
d. c. coils, 164, 165, 168
different voltages, 163, 164,

165, 166

electrical protection, 155, 156
end connections of coils, 63

508 INDEX

Insulation for (cont.) : Insulating materials (cont.) :

field coils, 176 presspahn, 154

formed coils, 163 rawhide fiber, 160

high temperatures, 155 , red rope paper, oiled, shel-

large d. c. armatures, 79 lacked, and varnished,

mechanical protection, 153 161

open slots, 69 shellacked bond paper, 1G1

partially closed slots, 61 test voltage to use, 175

phase coils, 172 varnishes and compounds,

slots, thickness needed, 163 176

Insulation from copper to copper, vulcanized fiber, 154

between coils, 172 Insulation resistance, measuring

Insulating materials: of, 185

treated cloths, 156 \,' . megger test for, 185

cotton tapes and cloth, 156 voltmeter test for, 184
Empire cloth, 156 Interpole motor, checking con-

Kobak cloth, 156 nections of, 412
oiled canvas, 158
oiled cotton drill, 158
oiled muslin, 158

Japan muslin 157 Knock m armature cauged b

Japan duck, 157 band wireSj 3%
varnished cambric, 157
-"./ varnished silk, 157

pressboards, fibres and papers, L

158

asbestos paper, oiled and Lap winding (multiple or parallel) :

varnished, 162 commutator connections for,

drying out, 181 102

express parchment paper, connecting double layer for

express paper, shellacked a. c. machine, 47

and varnished, 161 for d. c. armatures, 9

fish paper, 154 formulas and rules for d. c.

horn fibre, Japanned, oiled, armatures, 11

shellacked and varnished, grouping coils in a. c. machine,

159 35

leatheroids, 160 multiplex, formulas for, 14

manila paper, 155 requirements of, 103

mica, 155 use on d. c. armatures, 115

mica for coils, 173 use in a. c. machines, 116

mica paper and mica cloth, versus wave, 112

155 vs. multiple wave (series-paral-

micanite, 155 lei), 119

micartafolium, 162 wave and lap for a. c. ma-

pressboard, oiled, Japanned, chines, 28

shellacked and varnished, Laying out and connecting a. c.

159 windings, 35

INDEX

509

Lining for slots (see heading of
"insulation" and table of
contents for Chapter
VII).

Loop winding for small motors, 63
illustrations of, 64

Loose bearing, trouble due to, 423

Machine equipment for repair

shop:

banding machine, 460
bar bender for coils, 458
coil spreader, 457
coil tapping machine, 457
coil winder for lathe, 456
combination machine, 462
equipment needed, 444
slotting and grinding machine,

459

Making new coils, 141
Megger test for insulation resis-
tance, 185
Mica insulation:

built up, 164, 173
for armature coils, 173
Mistakes and faults in induction

motor windings :
device for detecting faults, 232
grounds, 231
improper groups connection,

234

open circuits, 234
reversal of coils or groups, 233
short-circuits, 231
use of wrong number of coils,

234

wrong number of poles, 235
Motor circuits, wire size for, 473
Motor failing to start, 430
Motor reversed direction of rota-
tion at high speed, 412
Motor-starter overhauling of, 349
Motors, safe temperature for, 415
Motors operated on double volt-
age, 419

Noise in three-phase motor, 428
Numbering coil sides in slots, 8

Odd frequencies, winding small

motors for, 191
Open circuits in d. c. windings,

causes of, 127
tests for, 128

telephone receiver test for, 135
transformer testing device,

125

Operations before and after winding
d. c. armatures (see table
of contents for Chapter
VI;. 139

Overhauling and inspection of:
auto-starters, 351
compound d. c. motor, 358
d. c. engine type generator,

371

d. c. motor starters, 349
drum- type controllers, 355
induction motor, 363
single-phase motor, 368
slip-ring motor, 366

Painting windings, 151
Phaoe change, reconnection of
induction motor for, 276
Phase coils, insulation of, 172
placing of, 291
rearranging in induction mo-

*tor winding, 278
when reconnecting 2-phase

to 3-phase, 173

Phase insulation when reconnect-
ing induction motors, 271
Phase rotation, testing of, 420

of a. c. motor, 440
Phase spread of a. c. windings, 32
Pinion puller, 456

510

INDEX

Pitchofd.c. coils:
definition of, 6
front and back, 7
full and fractional, 7
in winding spaces, coil sides

and slots, 7

short-pitch or short cord, 7
Pitch of a. c. coils :

effect of fractional pitch, 38
full and fractional, 37
Plugging commutator, 395

filling compound for, 304
Polarity of a. c. coil groups, 40
Pole-phase groups of induction
motor winding, connect-
ing of, 290
Polyphase windings, connections

for, 43

Potential pitch, 23
Pressure for brushes, 334
Procedure when connecting coils
of induction motor wind-
ing:
connecting pole-phase-groups,

290

constructing connecting dia-
grams, 292
grouping coils, 289
number poles from coil throw,

293

placing phase coils, 291
winding diagrams for different
number poles and phase,
295

Procedure when reconnecting in-
duction motors, 272
changes in voltage only, 275
change in phase only, 276
change in frequency, 278
change in number of poles, 282
change in speed, 273, 282
change in resistance of squir-
rel-cage rotor, 280
constructing connecting dia-
grams, 292
cutting out coils, 288
dead ending coils, 277

Procedure when reconnecting

(cont.) :
double circuit star to single

circuit-delta, 287
effect of high and low voltage

on operation, 283
factors limiting change in

number of poles, 286
grouping coils, 289
phase and voltage change, 278
points to consider, 272
reconnections frequently

made, 272

rearranging phase coils, 278
Scott or T-connection, 2-phase

to 3-phase, 269

testing reconnected motor, 282
winding diagrams for different

poles and phases, 25
Procedure when winding a. c.

machines :
induction motor secondaries,

207

induction motors with par-
tially closed slots, 194
induction motors with open

slots, 201

large a. c. stators, 210
repulsion-start motors, 190
small single-phase motors, 186
small polyphase motors, 192
stator of a. c. turbo-genera-
tors, 223

Procedure when winding d. c.
machines (see also head-
ing direct current arma-
tures) :

banding armatures, 146
cleaning armature slots, 140
connecting coils to commuta-
tor, 101
cutting out coils of wave

winding, 109

insulating coils, 164, 165, 168
large armatures, 77
open slots, 66
partially closed slots, 60

INDEX

511

Procedure when winding (cont.) :

railway, mill and crane arma-
tures, 90

rotary converters, 88

stripping off old winding, 139

3-wire generators, 89
Progressive wave winding, 107

cutting out coils of, 109
Pulleys, rules for size of, 492

speed of, 492

Rating of a. c. generators, 440
Reconnecting d. c. winding:

240 volts lap to 120 volts lap,

243
240 volts wave to 120 volts lap,

248
bridging commutator bars,

258

connecting coils in series, 253
connecting coils in parallel,

244
reconnecting with one dead

coil, 252
reconnecting duplex windings,

252
table for connecting coils to

commutator, 246
table for rewinding coils, 245
Reconnecting induction motors
(see also table of contents
for Chapter XI), 261
changes possible in existing

winding, 272

changes frequently made, 272
change in voltage only, 275
change in phase only, 276
change in frequency, 278
change in number poles, 282
change in speed, 273, 282
change in resistance of squir-
rel cage rotor, 280
cutting out coils, 288
chord factor, meaning of, 269
diagrams for changes in con-
nections, 263

Reconnecting induction motors:
factors limiting change in

number of poles, 286
phase and voltage change, 278
phase insulation, 271
points to consider before recon-
necting, 262

rearranging phase coils, 278
Scott or T-connection, 2-phase

to 3-phase, 269
table of possible reconnec-

tions, 266

testing reconnected motor, 282
with dead ended coils, 277
Reentrant, definition of, 14
Relining split b earings, 151
Remedies for troubles in d. c.

machines, 380
for brush troubles, 341
Removing old coils from d. c.

armatures, 57
Repairing :

auto-starters, 351

coils damaged while winding

armature, 174
compound d. c. motors, 358
cost of, 346

d. c. engine type generator, 371
d. c. motor starters, 349
drum-type controllers, 355
estimating cost of, 347
induction motor, 363
single phase motor, 368
slip ring motor, 366
Repair shop equipment:
banding device, 86
for testing d. c. windings, 138
floor stand for armatures, 75
machines and tools, 444
Repair shop methods for rewinding
d. c. armatures (see also
table of contents for
Chapter III), 56

Repair shop methods for rewinding
a. c. machines (see also
table of contents for
Chapter VIII), 186

512

INDEX

Repairs to commutator:

baking of commutator, 307
blackening of commutator,

in spots, 320
boring out end of, 313
brushes for undercut commu-
tator, 322

burn out between bars, 303
causes of excessive wear, 318
causes of trouble in, 301
copper used for bars, 319
finishing undercut slots, 322
making micanite end rings,

318

mica used in, 315
micanite insulation for, 317
precautions when tightening,

317

refilling of commutator, 311
removing bars and mica, 304
removing grounds, 308
remedies for high and low bars,

302

repairing burned bar, 305
replacing burned bar, 306
shaping mica end rings, 316
templet for making mica rings,

316

temporary cover for, 310
test for oil saturated mica, 320
tightening up bars, 306
tools for undercutting mica,

320

troubles from high mica, 301
turning down surface, 309
undercutting mica, 320
under difficulties, 397
Retrogressive wave winding, 107

cutting out coils of, 109
Reversal of speed of motor while

running, 412

Reversed coils in d. c. winding :
cause of, 131

test for with bar magnet, 131
test for with compass, 132
Ropes, horsepower transmitted by,

495

Rotary converters, winding of, 88

banding of, 410
Rules for a. c. windings:

arrangement of coils, 51

checking* phase relationship,
52

connections for coils of differ-
ent windings, 43

grouping coils, 35

indicating polarity of coil
groups, 40

reconnecting induction mo-
tors, 434
Rules for d. c. windings:

for possible symmetrical wind-
ings, 21

lap (multiple or parallel), 11

multiplex lap, 15

wave (series or two circuit), 17

multiplex wave (series-paral-
lel), 19
Rules, general, 498

Seasoning and grinding commuta-
tor, 148
Shaping insulating cells for slots,

448
Single-phase motors:

connections for windings of,

191

distribution of main and start-
ing windings, 188
inserting skein coils in slots,

187

insulation for slots of, 186
testing windings of, 192
winding of by hand, 190
winding for odd frequencies,

191
winding repulsion start type,

190

winding skein coils in, 186
Single-phase motor, overhauling

of, 368
Single vs. multiple windings, 118

INDEX

513

Size of switches for:
d. c. motors, 484
induction motors, 487
single-phase motors, 486
Shaping mica end rings, 316
Short circuits in d. c. windings:
causes of, 122
tests for, 123
transformer testing device for,

125

telephone receiver test for, 135
Shoved through coils, insertion of,

215
Skein coils for single-phase motors,

186

Slings for armatures, 455
Slip-ring motor, overhauling of, 366
sparking at slip-rings of, 393
Slots:

cleaning and filing, 140
inserting coils in, 61, 69, 81,
94, 187, 196, 199, 202,
203, 215, 227

insulating materials for, 153
insulation at ends of, 172
insulation of partially closed

d. c. type, 61
insulation of open d. c. type,

insulation for large d. c. arma-
tures, 79
insulation for a. c. windings,

186, 195, 214, 226
numbering of coils in, 8
thickness of insulation for, 163
Slotting and grinding machine, 459
Small motors, loop winding for, 63

winding of, 64

Solutions for cleaning slots, 140
Sparking at brushes, remedy, 341
due to poor belt joints, 394
due to absence of balancing

weights, 437
Speed change, reconnecting a. c.

motor for, 273, 282
Speeds of d. c. generators, 25
safe limit of, 25

Spiral or chain a. c. winding, 27
Split bearings, relining of, 151
Squirrel cage rotor, change in

resistance of, 280
Stalling of wound rotor induction

motor, 422

Standard motors on different volt-
ages and frequencies, 284
Static sparks from belts, 440
Stator connections, trouble due to,

421

Steel shafting, horsepower trans-
mitted by, 495

Stripping old d. c. winding, 139
Symmetrical d. c. windings, 20
Synchronous motor troubles, 388
failure to start, 432

Tables for winding d. c. arma-
tures, 110

connecting coils to commu-
tator, 111

changing coil connections, 246
rewinding coils, 245

Tape made from cotton cloth, 181

Temperature, safe for motors,415
CentigradeandFahrenheit, 499

Terms and parts of d. c. winding, 2
armature coil, 4
armature conductor, 3
concentric coils, 5
involute coils, 5
front and back pitch, 7
full and fractional pitch, 7
symbols used, 8
winding element or section, 3
winding pitch or coil pitch, 6

Testing d. c. windings (see table
of contents for Chapter V)
commutator, 136
for short circuits, 123, 135
for open circuits, 125, 128, 135
for grounds, 131, 133, 135
for reversed coil, 131, 132
for reversed or dead field coils,
136

514

INDEX

Testing, equipment for, 138

Testing induction motor windings

for:

device for detecting faults, 232
exploring with a compass, 236
grounds, 231
improper group connection,

234

open circuits, 234
order in making tests, 235
reversal of coils or groups, 233
short-circuits, 231
use of wrong number coils, 234
wrong number of poles, 235

Threaded-in coil, 60

insertion in slots, 61

Three-phase motors, reconnection
for (see reconnection of
indue tion motors), 261

Three-phase motors on single-
phase lines, 425

Throw of coils, 7, 70

Tightening commutator bars, 306

Tools used by armature winder:
band wire tension block, 454
banding machine, 460
bar bender for coils, 458
coil raiser, 449
coil spreader, 457
coil taping needle, 449
coil taping machine, 457
coil winding device, 452
coil winder for lathe, 456
combination machine, 462
drifts, 448

for cutting cell lining, 448
for shaping insulating cells, 448
hand tools, 445
leather sling, 455
made from hack saw blades,

451

pinion puller, 456
rope sling, 455
slotting and grinding machine,

459

steadying brace, 454
wire scraper, 449

Transformer rating for a. c. motors,

478
Troubles in a. c. machines, causes

and remedies:
due to electrical faults, 388
due to mechanical adjust-
ments, 387
fuses blowing, 389
hot stator coils, 391
in windings of, 387
induction motor troubles, 386
rotor windings, 302
sparking at slip rings, 393
stator windings, 392
synchronous motors, 388
tension of belts, 391
testing for grounds, 390
Troubles in d. c. motors and genera-
tors, causes of:
electrical defects, 379
exposure to acid fumes and

gases, 377

lack of inspection and replace-
ments, 378

lack of proper cleaning, 376
operating temperatures too

high, 378

operation in damp places, 377
remedies for, 380
Troubles in commutator, causes of,

307
with brushes, remedies for,

341

with high speed motor, 417
Turbo-generator :

bracing windings of, 228
break-down test for, 229
coils for, 223
connecting windings, 228
inserting coils in, 227
insulation for coils of, 226
forming coils for, 225
testing windings of, 226

Undercutting mica with standard
tools, 320

INDEX

515

Undercutting mica with hand tools,

323
with portable electric drill, 400

Varnishes and impregnating com-
pounds:

baking and air-drying varn-
ishes, 178

characteristics of, 178
classes of, 176

clear and black varnishes, 177
solvent chart for, 178
table of uses, 177

Voltage change, reconnection of
induction motor for, 275
Voltmeter test for insulation re-
sistance, 185

Voltage to use in testing insula-
tion, 175

Wave winding (series or two-cir-
cuit) :

commutator connections for,
104

cutting out coils of, 109

dead coils in, 108

duplex winding, 252

for a. c. machines, 41

for d. c. armatures, 16

formulas and rules for d. c.
armatures, 17

grouping coils in a. c. ma-
chines, 35

lap and wave for a. c. ma-
chines, 28

locating first connection to
commutator for, 105

multiple wave vs. lap, 119

multiplex, formulas for, 19

progressive and retrogressive,
107

reconnecting with dead coil,
252

reconnecting 240 volts to 120
volts, 248

requirements of, 107

Wave winding (cont.):

use in a. c. machines, 118
use on d. c. armatures, 117
versus lap, 112
Wedge driver, 62
Whole-coiled and half-coiled a. c.

windings, 30
Winding d. c. machines:

having partially closed slots, 60
having open slots, 66
large armatures, 77
rotary converters, 88
3-wire generators, 89
railway, mill and crane motors,

90
Winding a. c. machines:

induction motors with open

slots, 201

induction motors with par-
tially closed slots, 194
induction motor secondaries,

207

large a. c. stators, 210
repulsion start designs, 190
single-phase motors, 186
small polyphase types, 192
stator of a. c. turbo-genera-
tors, 223
Winding data needed to duplicate

an old winding, 57
Winding diagrams for a. c. motors:
developed diagram, compared

with circle scheme, 53
for different poles and phases,

295

simple circle diagram, 39
Winding parts and terms, 2
Winding procedure for d. c.
armatures (see heading
"Direct Current Arma-
tures")-

Wire scraper, 449
Wire size for motor services, 473
d. c. motors, 481
motor leads, 474
3-phase motors, 482
Wire table, how to remember, 465
Wire gauges, classification of. 466

UNIVERSITY OF CALIFORNIA LIBRARY
BERKELEY

Return to desk from which borrowed.
This book i

MAY 6 1348

MAR 1 Q 1950

NOV 2 1 1952
APR 1 1953^

MAY 6 1963

LD 21-100m-9,'47(A5702sl6)476

OOUIO

M184967

Ub.

THE UNIVERSITY OF CALIFORNIA UBRARY