Dstermination of Current and
Candle Power Curves of the
Alternating Current Arc
Under Impressed Voltage
Of Various Wave Forms
H. E. Brashares
J. R. Armstrong
Brashares, H. E,
Determination of current and
candle power curves of the
DETEEIvIIKATIOH OF CUiSEliTT AITD CixKDLE POWER CimVES
THE ^LTEEHATIl^G CURSEnT ^C
IffiDER ILIPKESSi;!) VOLTAGE 0? V^P.IOUS 7/aVE F0K.1S.
ii THESIS PEESEITED
PRESIDEITT AITD F^iCULTY
^HvIOim IHSTITUTE OF TECmiOLOGY
For the degree
BACHELOR OF SGIEUCE II? 3:iECTRICAL EFr^HTESRIITG
Having completed the prescribed course of study in
ILLINOIS INSTITUTE OF TECHNOLOGY
PAUL V. GALVIN LIBRARY
^ ^ y '"'^-so, .u.e 16. 1905. SforsoeT"
J^^Of^^ (^ (3ul£^t^tA.^ Z^^'.,
Table of cohtmts.
Preface, containing statement of Thesis subject 2
A Brief History of the Electric ixrc 4
The Rotation of the Arc 10
The Photometer 13
The Llagnetic Coupling and Contact llaker 17
The Llotive Power E3
The Curve Tracing Table Z5
The Enown Resistance 28
The Regulation of the Current E8
Calibration of the Standard Lamp 29
The Arc from a Peaked Wave 32
The -arc from a Flat Top V/ave 37
Comparison of the Results Obtained from the Flat
Top and Peaked E. LI. F. Y/aves 42
LIST OF ILLUSTRATIONS,
Machine for Rotating Arc
Photometer Bench Complete
Photometer Bench without Discs
The Llagnetic Coupling
The Contact Maker
The SjoiohronouB Ivlotor
4. The Curve Tracing Table
f Print 1)
The oTDJcot of this thesis is to point out the
relation of the useful light emitted by the alternating
current arc to the wave form of the voltage impressed
The existence of different wave forms of voltage
in alternating current practice is a well knovna and use-
ful fact. ^11 forms varying from the peaked to the
flat or rectangular are found, both in regular and ir-
regular ways. For a given effective value, the flat top
wave gives more nearly a constant val'ae to the current
through a half cycle, while the peaked form gives a rap-
idly varying current which rerches c. high maxirauja. It
is desired that the light given out by the alternating
current arc be as uniform as possible in order that it
may be applicable to commercial problems of lighting.
Since the current producing the arc pa.sses through a
zero value twice per cycle, the wave form which best
adapts itself to the arc light is that one which will
store the greatest amount of energy in the arc to main-
tain its brightness through these periods.
The plan adopted for the study of the effects as
pointed out above is that of the photometric determina-
tion of the light given out by the arc for the instanta-
neous values of the current and the voltage producing it,
This plan necessitates a combination of the well known
contact maker and the flicker photometer. This appara-
tus is described in detail in the following pages, as it
was constructed and operated in this work. In these
descriptions, numerous references are made to lettered
prints of the different parts, by which it is hoped that
they may be more clearly understood.
A BRIEF HISTORY OF THE ELECTRIC -tiRC.
The Electric ^rc v/as probably first discovered
by Volta in 1802 or 1803 v/hile experiment iiig with the
heating power of a current, especially with the sparks
produced on the breaking of a live circuit. It v/as
first brought to general notice by Davy in 1808, during
his experiments with the great battery of the Royal In-
The arc is the most intense artificial illuninant
and the chief comuercial source of very powerful light.
The slow breaking of an electric circuit, carrying con-
siderable current at a medium voltage, does not inter-
rupt the flow of the current even after the conductor is
parted. The current jumps across the break however Vi/ith
the evolution of great heat and a vivid light. If this
separation is at the ends of two carbon rods, the light
is enormously brilliant, and by proper mechanism can be
maintained fairly constant. This passage of current
from one carbon tip to the other is accompanied by the
production of great heat, and the tips of the carbon
rods will grow white hot, thus serving as a so\u-ce of
light. In the ordinary direct current arc lamp, the
upper carbon is the positive pole of the circuit and is
fed slowly downwurd, bo as to keep the arc wiiform as the
cartoB. is consulted.
The main consximption of energy appears to be at
the tip of this positive carbon which is by far the most
brilliant part of the arc. Tlie carbon fairly boils away
into vapor, producing a slight hollow in the center of
the carbon, commonly Icnown as the crater. The carbon
of this pole is consujued about twice as fast as that of
the negative pole. The positive carbon resembles in
shape a blxmt pointed pencil, while the negative carbon
becomes more nearly like a sharp pointed pencil.
With the introduction of the alternator into the
commercial world, the alternating current was at once
applied to arc lighting. QQie alternating current arc
has grown in the lighting field until it is most v/idely
used in the United States as well as on the Continent.
In this country, it is used on both the constant poten-
tial and constant current systems.
It is very evident that there is neither a posi-
tive nor a negative carbon, each one being positive and
negative alternately and changing from one to the other
between 3,000 and 14,000 times per minute, according to
the frequency of the current in the circuit. Under these
conditions, no marked crater is formed on either carbon
and the two are consumed at about an equal rate. ^s a
natural result of this intermittent coi:dition, the heat
whioh is stored up in the carbons is depended upon to
produce sufficient ligiit while the current is passing
through the zero value, to maintain the effect constant.
This intermittent supply and the lack of a localized
crater causes a lower average temperature in this arc
than in the similar direct current arc. The tempera-
ture of the alternating arc varies hetween 2,800 and
4,000 degrees 0.
Careful tests have been made with the alternat-
ing arc with regard to the freq.uency and it has been
found that it operates best at from 40 to 50 cycles
per second. whien operated belov; 40 cycles, the light
flickers to a troublesome extent, v/hile above 60 cycles
it becomes very noisy. Enclosed arcs, or very soft
carbons, e.nd heavy currents in open arcs, decrease these
above defects to some extent.
Exhaustive tests have been carried on in the
various engineering laboratories of the electric compan-
ies to determine the best carbon for use in the alternat-
ing arc. It is found that a specially fine soft carbon
is the best adapted for open arc work. These carbons
not only muffle the vibration to a remarkable degree but
also when used at high current densities give off a large
quantity of vapor in the arc, which adds stability to it.
The result of using soft volatile carbons is a q.uiet
steady brilliant arc of the most excellent illuminating
powers. V/here this class of ce.rbon is either too costly
or impossible to obtain, the harder carbons must be used
and especially with enclosed arcs.
The alternating current arc produces a singing
sound \fhich is partially due to the vibrations produced
in the mechanism and partially to the pulsations im-
pressed directly on the air by the oscillatory action
in the arc itself. The former can be luodified to a
considerable degree by proper design of the mechanism
and care in the manufacture of it. The latter however
is hard to suppress. With all of the precautions possi-
ble and V7ith the best carbons possible, the arc still
sings ii: a tone pitched about a semi-tone below bass G
with very ujipleasant harmonics.
Tests to determine the difference of efficiency
between direct and alternating current arcs have given
a collection of varied results which are hard go formu-
late into a definite rule. Ki'. Louis Bell drav/s the
conclusion that for xmit energy actually applied to the
arc, the direct current arc v/ill give about 25^o more
light than the alternating current arc. "Tien operated
on a constant potential lighting system (comparing en-
closed i,-rcs!, the suarplus voltage is taken up in a re-
active coil which wastes very little energy, while a
direct current arc light has a dead resistance which
wastes considerable, so that the two arcs are on about
even terms. The following table will c-ive some idea of
the relative efficiencies of the different arcs.
Kind of ^rc. 7/atts per S.C.P.
Direct current open 1»0
Direct current shaded 1«3
i-:.lternating current open 1»7
Alternating current shaded 2«a
Direct current enclosed 2»4
Direct current enclosed 2*9
Direct current enclosed 3»0
Alternating current enclosed 2»6
Alternating current enclosed 3.3
Alternating current enclosed 3.6
The sizes of carbons which are on the market vary
from 5/l6 to 3/4 inch in diameter. The current em-
ployed varies from 6 to 10 amperes according to condi-
tions of locality and the size of carbon used. The volt-
age also varies under different conditions.
l^s. -.yrfcon, I,Ir. Stein and other prominent photo-
metricians have carried out extensive tests with differ-
ent carbons varying sizes, current densities and voltages
under similar conditions to determine the best adapted
and the most economical combinations for these conditions.
The mean spherical candle power, the horizontal candle
power and the distribution of alternating arcs have been
carefully tested on both constant potential and constant
current circuits by Prof. Schaler Katthews of Purdue
University, under the direction of the national Electric
Light AsBOCiation. Extensive experiments have been made
to determine the time resistance of the arc iinder differ-
ent conditions of current and voltage. This, and the
determination whether there exists a "back E. M, F. in the
aro are two prohlems which have never heen satisfactorily
solved. The effect of the different wave forms on the
light giving (jualities of the aro has received, seemingly,
little attention hy the experimenters. Since this
subject of wave form has "been given positions of consider-
able weight in other commercial lines of electricity, it
should receive soirie consideration.
In the above discussion, it has only been desired
to give a brief history of the electric arc and the treat-
ment it has received in fche hands of experimenters whose
conclusions have decided the practice at the present time.
For a detailed and extensive history of the arc, the reader
is referred to specially prepared works on the subject,
chief of which is that of Ivlrs. Ayrton, "The Electric ^rc."
The Rotation of th.e Arc.
One of the peculiar actions of the alternating
current arc, v*hich has not been most satisfactorily under-
stood, is that it wanders continually around the tips of
the carbons maintaining it. To counteract this action
and obtain a condition which more accurately represents
aoi average effect of the arc, a special machine is used.
Vhit nit chine, es shov.r ir Print 1, was designed by Prof.
Snow of the Armour Institute of Technologj;^, and built in
the Institute shops under the direction of Mr. Sommers.
At the beginning of this work, it was altered in some re-
spects, chief of which was the support of the lower car-
bon. This alteration was m6.de to eliminate the shadows
produced by the supports as much as possible. The main
purpose of the machine is to rotate the arc a.^ a high
speed and thus eliminate the effects of its wandering.
The machine consists of a suitable v/rought iron
frame, a, v/hich supports the cast iron ?/heel, b, at its
central point, c. This ?;heel is graduated on its back
side in degrees from zero to ninety in each direction
from d, which represents the zero or vertical position
of the arc. The wheel, b, carries the white metal frame,
e, which is secured rigidly to it. The frame, e, supports
the meohanism for adjusting and rotating the carbone.
The upper- carhon is supported on the I'od-, f',' wjiich
passes through the sleeve, g, through6ut its entire
length and is -secured to it by a feather. key which permits
of easy movement up or down, by the thumb .screw, k, v/hile
the key insures rotation with the sleeve. Upon sleeve,
g, and insulated from it is the second sleeve, i, which
Bupi'Orts the lower carbon. The brush, j, serves to con-
nect the lower carbon with the binding post, o, which
is connected to the line. The other terminal of the
line is connected to the binding post, n, v/hich makes
connections with the upper carbon through the frame, e,
and the sleeve, g. Keyed to sleeve, g, is the pulley,
m,' which is driven by a belt running over two small idler
pulleys in the back of the frame to the shaft of the smc:.ll
Crocker-lVheeler motor, q.. The shaft of this motor is
in line with the horizontal axis of the wheel, b, v/hich
permits of three movements of b about its axis without
disturbing the rotation of the motor armature. The en-
tire part supported uy the sleeve, g, may be moved through
a Blight distance along its axis by the thiimb screw, 1,
which allows of close adjustment. The r;lobe, h, is
supported by the cover plate, s, by means of the copper
springs which permit it to be easily removed. ^ white
metal frame, t, is also supported by the frame, e.
for the purpose of holding an outer globe. It may be
raised or lowered by the thumb screw, v, and adjusted as
desired. The motor is connected to the mains through a
starting box uud rheostat by which its speed can be gov-
erned as desired. Thus the arc is rotated and at the same
time easily adjusted as to length and position.
In order to measure the light given out by the
arc as operated by the machine described above, it must
be compared with the light of a stuiidarcl lamp, the candle
power of which is known under certain conditions of im-
pressed voltage. To do this, use is made of the physical
principle that the intersity of a light at any point varies
inversely as the square of the distance from the source.
The comparison is made between the standard light and the.t
of the one to be measured, by placing a screen between them
which is illuminated by both. The screen is then moved in
the line connecting the two lights Ujitil the ill--aninations
on both sides are eq.ual. The distance from the screen
to each light is noted, and thus the candle power of the
unknown light can be calculated from the relation that the
candle powers of the two lights are to each other as the
squares of their distances from the screen.
In common incandescent light photometry, the con-
stancy of the light and the accuracy of reproduction are
the primary considerations, but in arc light photometry
the color of the light is of first importance. The criter-
ion for the accurate arc liglit phot one try is that compar-
isona be made only between equally lighted and similarly
colored fields. The impossibility of obtaining any
standard giving the same colored liglit or any where near
the same intensity as the arc compels the adoption of other
methods, to be used in connection with the ordinary photom-
etric scherae. The incandescent lamp is the one standard
v/hose quality of light may be made to approach closely to
that of the arc and that can be easily and accurately
standardized. This necessitates the balancing of a yellow
light against that of the arc which varies in color from
a yellowish V7hite through a white to a bluish white. To
accomplish this comparison and also obtain instantaneous
values, if possible, of the light emitted by the arc as
the current passes through a complete cycle, a form of
flicker photometer is used in this work.
The flicker photometer as here used is shown com-
plete in Print 2. The steel frame, d, rests upon the table,
c. This fra,me serves as a runway for the two carriages, e
and i, the former carrying a Lummer-Brodhun screen, f, and
the latter carrying the standard lamp, b. The arc light
is placed at the opposite end of the bench, at a. This is
the machine as described above. The discs, g and h, are
secured to the shaft, 1, which runs in the bearings, m, n
and 0. These discs are better shov/n by Print 4. They
are made of aluminum, and are 20 inches in diameter by 3/52
of an inch thick. Two of these discs are mounted on the
oast iron hub, d, and secured "by the screws, e. TJae slits,
a, "b and c, are spaced 120 degrees apart and may be adjust-
ed in size by moving the two disos I'elative to each other.
The maximuia opening is as shown.
Referring again to Print 2, these discs are so
placed that when one of the three slits of h is directly
between the screen, f, and the standard light, b, one
slit of disc, g, is 60 degrees from this position. That
is, at this position of h, the line joining the arc and
the screen would cut disc, g, midway between tvo slits.
Thus, as the shaft, 1, is rotated, light is admitted to
the screen, f , alternately by the standard light and the
arc. The purpose of this is, as suggested above, to
illuminate the effects of the two different colors of
light, and thus to obtain a balance on the screen. Since
these slits are small, the light passing through them and
being measured repx'esents practically the instantaneous
values, as the current of the arc passed through a cycle.
Thus, one position of the discs gives one point on the
instantaneous candle power curve. Therefore, it requires
only to alter the position of the discs relative to the
armature of the machine ftirnishing the power to the arc,
to obtain the candle power curve for a complete cycle.
The Llagnetic Coupling and Contact l.tlcer.
i*s a means of altering the relative positions
of the machine armature and the shaft carrying the flicker
disos, the magnetic coupling shown in Prints 5 and 5 was
designed and constructed by the authora. This instrument
not only serves as a coupling "but also as a contact maker.
Print 5 shows the coupling mechanism while Print 6 shows
the arrangement of the contact maker. Referring to Print
5, a is an annular "brass gear having 144 teeth. The driver,
h, is of "brass and carries the two electro-magnets, d and
f, (f and d are symmetrical with respect to the axis of
rotation of b). These two magnets are identical in con-
struction, and are of the horseshoe type with the coil
wound on one leg. The coil was wound on a wrought iron
core 3/8 of an inch in diameter and fitted into the outer
casing of iron pipe of V/s inch inside diameter and 1 inch
outside diameter. A cap was riveted to the bottom of the
core and also to the bottom of the casing, thus completing
the horseshoe. The armatures of the magnets are of
wrought iron, and are hinged at g to the clip h, support-
ing the magnet. An arm projects from the armature through
the slot, m, in the driver, b, and again is pivoted to a
bar sliding the full length of b, in a slot l/4 inch deep
and 9/I6 inch wide. The ends of this bar are constructed
to fit into the spaces between the teeth of the gear, a,
and are so £ rranged that when the finger on one end occu-
pies the space between two teeth, that on the opposite
end is directly over the point of a tooth. The length
of this sliding bar is such that in moving from one
position to the other, the two fingers are never dis-
engaged from a tooth of gear, a, at tlie same time, but the
finger on one end enters the space before that on the other
end disengages the tooth. When the current is passed
through the coil of one magnet, the armature is dravm to-
ward it, thus causing the sliding bar to engage a tooth
of the gear. It is held in this position as long as the
current remains in the coil. How if the current be changed
to the other magnet coil, its armature drav/s the sliding
bar in the opposite direction, and since the finger on
this end is directly over a tooth, the friction of the
parts driven and attached to gear, a, is depended upon to
cause it to drop back slightly and the finger engages a
tooth which has moved the gear, a, backward relative to
the driver, b, through a distance equal to one-half the
distance between two teeth, i.e., one-half the circumfer-
ential pitch of the gear teeth. Since there are 144 teeth,
this distance corresponds to 1-J- mechanical degrees. Thus,
by shifting the current from one coil to the other of the
magnets, the space represented by any number of teeth may
be moved through as desired.
One terminal of each magnet coil is connected to
a slip ring carried on the hub of the driver, "b, and in-
sulated from it. The copper brush, n, slips on this ring
and thus serves as the common connection to the outside
circuit. The remaining terminal of each magnet is connect-
ed to a separate slip ring, one to o and the other to p.
Copper brushes again supply a comieotion to the outside
circuit . A two point switoli serves to control the oper-
ation of the magnets, and a current of Z^ amperes is suf-
ficient for successful operation.
The gear, a, carries on its circumference the
hai'd ruhber ring, f, (Prints 5 and 6) which is secured to
it by screws as shown in Print 6. Lt the point, q, in this
hard rubber ring, a narrow brass strip is inserted, which
is connected by hidden connections to slip ring, r, on the
hub of the gear, a, A copper brush, the terminal of
which is shown at s, makes sliding contact with this ring,
while a second brush, t, slips upon the circumference of
the hard rubber ring. At the instant the brush, t, comes
in contact with point, q., the circuit is closed between
the binding posts, u and v, if no break exists in the out-
side circuit connected to these points. The brush, t,
remains stationary, but the point, q., changes its posi-
tion relatively when the magnetic coupling is operated.
Thus, instantaneous values of the current and voltage can
be obtained as with an ordinary contact maker.
Print 2 shows the connection of the coupling to
the disc shaft of the photometer. The gear, a, (Print 6)
is connected to the shaft, 1, by the universal joint, p.
For operation, the relative positions of the contact point
and the disc slits are so adjusted that at the instant of
contact, the slit on the disc, g, is in the line of the
arc and screen. This is done by placing the brush, t.
(Print 6) in the correct position where it remains fixed.
Tlie coupling is mounted upon the wooden pedestal,
y. Print 7 shows its connection to the rotary converter
used to drive the apparatus. This connection is made by
the steel coupling, a, which is fixed to the shaft of
the driver of the coupling, while on the generator end it
has a pin which fits into the keyway of the armature shaft,
and is free to move along the axis of this shaft. This
does not interfere with the end play of the rotary arma-
The Motive Pov/er.
The power required to rotate the flicker photom-
eter discs as described above, is furnished by a 5 E.^.
wood rotary converter, shown in Print 7 at b. This ro-
tary has a capacity of 36 amperes at 140 volts at the
direct current end, and 85 volts at the alternating current
end. It is a six pole machine and its armature is wound
for one, two, three or six phases. In this work, the
rotary is operated as a single phase synchronous motor by
the machine furnishing the power to the arc. The synchro-
nizing is arranged for by the lamps, c, in series with one
side of the line controlled by switch, d. The direct
current side of the rotary is connected through the start-
ing box, e, and has an auxiliary rheostat, and single pole
switch, h, in the armature circuit, jx rheostat, f, is also
used in the rotary field circuit. By using a rotary in
this position, a convenient method for synchronizing is
afforded, "by getting synohronous speed as a direct current
The Curve Tracing Tatle.
To obtain the instantaneous values of the voltage
impressed upon the arc, the method used consists of balanc-
ing these instantaneous values by the direct current volt-
age of the storage batteries. To obtain the instantaneous
values of the current through the arc, a known resistance
is inserted in series with it and the drop over this re-
sistance balanced, as in obtaining the instantaneous
values of the voltage against the voltage of the storage
batteries. Since this kiiovm resistance is non-inductive,
the current is then easily calculated by Ohn's law, for
each value of the drop obtained.
The connections of the curve tracing table are
shown on Print 8. The leads from the storage batteries
are brought to the binding posts, a, which are connected
to one side of the potential rheostat, c, through the
double pole switch, b. The remaining side of the rheostat
is connected to the middle points of the double throw
switch, t. Across these same points, the terminals of the
voltmeter, d, through the contact switch, e, which pei-mits
the use of either scale of the voltmeter. In series with
the switch, t, are connected the galvanometer, f, the con-
tact key, h, and the contact maker, which is connected
between the binding posts, i. The voltage to be measured
is brought to the binding posts, s, v/hich serve as the
terminals of IE sixteen candle power incandescent lamps
in series. The drop over two of these lamps is used in
balancing. The leads across these lamps are shown at the
"binding posts, k, which are connected to the switch, m.
The connections to the switch, n, and to the galvanometer
circuit are shov/n and are such that with switches, m and n,
as shown, the instantaneous values of the alternating
voltage mfy he balanced against the storage battery volt-
age by means of adjusting the rheostat, o. The balance
is shown by no deflection of galvanometer, f, when the
key, h, is closed. The voltmeter, d, then reads the in-
stantaneous values of the voltage, which multiplied by
ratio of the number of lamps used, gives the total volt-
age at that instant. The binding posts, j, are connect-
ed across the terminals of the known resistance mentioned
above. V/ith switch, 1, in position shown, the drop over
this resistance may be thrown across the galvanometer cir-
cuit by throwing switch, n, to the right, v/hich at the
same time opens the pressure circuit. The balance and
reading are obtained as before.
The voltmeter, q., is used in synchronizing, and
is BO connected that it may be thrown across the genera-
tor or motor terminals by means of the single pole double
throw switch, p. The switch in the position shown places
trie voltmeter across the motor terminals. The voltmeter,
0, is connected to the terminals of electric tachometer
by which the speed at any instant is determined. The
The small two point key, g, controls the magnets of tlie
magnetic coupling. Point 1 is counected tlirough three
storage cells to the common ring of the coupling described
ahove, while points 2 and 3 are connected respectively
to rings 1 and S. The coupling is reversed hy moving
the key from 2 and 2 and vice versa.
The Known Resistance.
The resistance used for measuring the instant-
aneous values of current consists of tv/o twelve inch
carbon rods, one-half inch iu dianeter, connected in series,
This is inserted in series with the arc, and tv/o pressure
leads run from its terminals to the curve tracing table.
The Regulation of the Current.
Ill series \iith tlie arc, there is a carbon rheo-
stat and a low inductive coil. This coil is suspended
by a steel v^ire over a laminated iron core, about which
it may be moved up or down by means of a count erv?eight
to vary its impedence. In this way, the current through
the arc is kept constf-rt.
Tne above described apparatus was carefully test-
ed, each part by itself, and under operating conditions,
before any tests v/ere made on the actions of the arc it-
GALIBKATIOII OF THE ST^TDaRD LAllP.
The light used as a standard is thtit of a 110
volt 50 candle power incandescent lamp. It is designed
for electric headlight and has a spherical globe, while
the filament is wound in the shape of a spiral cone.
This lamp was calibrated on the photometer bench by means
of a 16 candle pov/er standard incandescent lamp, which
was standardized to give 16 candle pov/er at 102 volts.
It vas proposed that instead of varying the posi-
tion of the photometer sci-een to obtain a balance, the
candle pov/er of the standard lamp be changed by varying
the voltage impressed upon it. ^ curve plotted betv/een
Candle power and voltage could then be used and the
light given off by the lamp be read off of this curve
directly. To calibrate the lamp in this v/ay, five
different tests were made at different times, between
which was a sufficient period to rest the eyes of the
observer. Sach bench reading recorded is the average
of two different settings of the photometer screen.
Tlie average of the five recorded values was used to
calculate the candle power for that corresponding
voltage. The following curve shows the relation of
the candle power of the lamp and the voltage impressed
From this curve and the data from v/hioh it is
constructed, the ujiifomiity and consistency of the var-
iations are very evident. That is, for a given im-
pressed voltage, the 0c.ndle pov/er of the li.up is con-
stant under normal conditions.
G-o.LIBr;i.TIOII OF THE STxJDaPvD L^^iP.
Standard lamp used #E30, 16 c.p., at 102 volts, position 1.
£ Standard to Screen Centimeters.
12 3 4 5 Average C.P.
75 171.5 170.8 171.1 170.4 171. i3 171.06 S.42
80 150.4 160.2 160.3 160.6 159.5 160.02 5.05
85 142.2 14S.2 142.2 142.0 142.5 142.06 8.80
90 135.1 125.1 125.8 125.8 125.1 125.28 11.50
95 122.1 122.0 122.4 122.7 122.9 122.22 16.90
100 111.7 111.2 110.6 110.8 110.4 110.94 25.25
105 100.8 100.8 1G0.6. 100.4 ICO. 4 100.60 201.80
110 91.5 91.2 81.9 92.2 91.7 91.54 47.90
115 82.2 82.4 82.5 82.2 82.5 82.28 6G.50
Distance from screen to lamp laeing calibrated
equals 250 Centimeters - Distance standard to screen.
THE ARC FROM A PEAEED WixVE.
A peaked pressure wave, because of its applica-
bility to systems in v/hicli transformers must be used,
should be very suitable to arc lighting systems on uhioh
the distance of transmission is comparatively short.
However, it would be expected that on account of the
great variation in the instantaneous values of the volt-
age, the variation of the light given out by an arc oper-
ated on such a wave v/ould be more noticeable than on a
flat top wave.
It was found by operating the arc under test in
this Work from a generator giving a wave form which was
very peaked, that the most satisfactory results were ob-
tained with regard to regulation and steadiness of the
arc. The instantaneous values of the candle power of
the arc at different positions on the wave seem to be
governed to a great extent by the action of the carbons.
ITo distinct v/ave form of candle power was obtained which
followed very closely that of the current producing it.
On the other hand, a candle power curve was obtained
which seemed to be independent of the current curve.
In some instances, the maximum candle pov/er was given
out at the aero values of the cxirrent. This result may
be attributed to some extent to the difficulty in "bal-
anoing two different colors of light on the photometer
screen. This however was reduced greatly by the use
of the rotating discs which have been fully described
The generator used to give this wave form is
a Westinghouse single phase machine with a normal volt-
age of 1000 at 1000 revolutions per minute. For this
work, it was operated at 600 revolutions per minute so
as to obtain the frequency desired and generated a volt-
age of 800. This was stepped down through a 77esting-
house transformer of 18f II.W. to 80 volts, which was
applied directly across the arc and its regulating ap-
paratus. This pressure gave an arc operating under
the normal open arc current of 9.6 amperes and consuming
450 watts which is the nominal rating of a 2000 candle
power arc light. The voltage req.uired to maintain these
conditions was 49. The arc was rotated at about EGO
revolutions per minute throughout the period of taking
the bench readings and the instantaneous values of the
current and voltage. The instantaneous values of the
voltage v/ere taken directly over the arc.
The follov/ing curves show the relation between
the candle powei; current and voltage over the arc. On
the same sheet is given the curve showing the wave form
of the voltage used.
Arc operated at 9.6 amperes, 49 volts, 450 watts, 40 p.p.s.
by Weatingiiouse alternator #2441. normal 1000 volts, 45
amperes, 1000 R.P.K. , 66 p.p.s.
Candle power of standard - 48 at 110 volts.
e over which current drop
- .325 ohms.
THE E. M. P. lUYK OF THE WESTIKGHOUSE ALTERNATOR #2441
Magnetic Degrees Instantaneous Voltage
THE .iRC FROL: ^ FLLT TOP '.7^VE.
IThe flat top v/ave of E. LI. F. is in modern praooioe
prolDa'bly the one wtLich is most desired for lighting sys-
tems. This results from the faot that it may be easily
adapted to all classes of woi-k and especially for long
distance transmission. ^s compared to the peaked wave,
the light given out from an arc operated on a flat wave
would "be expected to "be steadier and of a more desirable
A generator giving a flat top wave was used l;o
supply the arc under test and the results obtained follow
to some extent in a general way those obtained from the
peaked wave previously discussed. Some difficulty v/as
encountered in obtaining a candle power curve from which
any conclusions could be drawn.
The generaoor used to r-ive the flat top v/ave is
a General Electric three phase machine v;ith a normal volt-
age of 1S6 at ISOO revolutions per minute. For this v;ork,
it was operated at 800 revolutions per minute and a single
phase used. The machine is of the revolving field type
and has six poles. Thus, under the above conditions, the
frequency of 40 cycles per second was obtained. The arc
was operated ae before at the nornal open arc current
of 9.6 amperes and a pressure of 49 volts \ms required to
maintain tlie 2000 candle power rating of 450 watts. The
carbons v/ere rotated at about 200 revolutions per minute
during the period of obtaining the bench readings and in-
stantaneous values of current and voltage.
The following curves show the relation Letv/een the
candle power, current and voltage over the arc. On the
Scixie sheet is given the curve shov/ing the wave form of the
EEE ARC FROM ^ FI^T TOP WAVE.
Aro operated nt 9.6 amperes, 49 volta, 450 watts, 40 p.p.s.
by General Electric alternator #65447. lomal 1£0 volts,
72 amperes, 1200 revolutions per minute, 60 p.p.s.
Candle power of standard - 48 at 110 volts.
Bench Reading cm.
Resistance over whicli current drop was taken - .326 ohms.
Ihe E. M. F. Wave op THE GEI^ERaI. ELECTRIC aITERUaTOR #G5447
Magnetic Degrees Instantaneous Voltage
&■.: j'x, -A
COtilPAKISOII OF THE RESULTS OBT^.IHED FEOIJ THE FLaT TOP -uBD
PEiJaSD E. K. F. iLLYEZ,
The action of the arc liiider tlie flat top and peaked
E. M. F. waves may "be compared "by observing the curves
given on pages 36 and 41. Under both forms, tlie effects
of the arc upon the current curves are very similar. The
single peak anticipated in both oases has been replaced
lay two separate peaks. This is especially noticeable in
the current curve of the peaked S. M. F. , given on page
36. It is only slightly shown on the current curve of
the flat top B. M. F. wave given on page 41. The same
result is observed in the arc voltage curve. It will be
noted that the candle pov/er of the aro tinder the peaked
wave lags behind the current while under the flat top v/ave
the first peak obtained leads the current to a slight ex-
The sudden drop iu the candle pov/er curve on page
41 can not be satisfactorily explained. It is the opinion
however that it is due to the effect of the carbon alone.
Solid carbons were used in both cases, and were practically
of the same hardness.
It is a disappointment to the authors that do little
has "been accompli shed along the escperimental determinations
pertaining to this suhject. It v^as the desire to make
more extensive tests but owing to the fact that almost all
the time given to this work was consumed hy the designing
and making of the greater part of the apparatus, very-
little remained to he given to the experimental wox'k. The
apparatus as it now stands completed requires hut few al-
terations in the mechanical features to insure satisfactory
operation in every respect, ^e recommend that the magnet-
ic coupler he replaced hy one larger, hoth mechanically
and magnetically, as its duty is heavier that was pre-
We desire to thank Prof. G. E. Freeman, Prof. Snow,
Prof. E. H. Freeman, and llr. Somiaers for their many val-
uable suggestions. We also wish to eagoress our appre-
ciation to Kr. Tyler, a class.nate, for his assistance in
our experimental efforts.
H. E. Brashares.
J. R. io-mstron?.