INDUCTOR MACHINES

281

longitudinally to the shaft, and thus magnetic stray flux tends
to pass along the shaft, closing through bearings and supports,
and causing heating of bearings. Therefore, in the types 136
and 137, magnetic barrier coils have been used where needed,
that is, coils concentric to the shaft, that is, parallel to the field
coil, and outside of the inductor, that is, between inductor and
bearings, energized in opposite direction to the field coils. These
coils then act as counter-magnetizing coils in keeping magnetic
flux out of the machine bearings.

The type, Fig. 139, is especially adapted for moderate fre-
quencies, a few hundreds to thousands of cycles. A modifica-
tion of it, adopted as converter, is used to a considerable extent:
the inductor, I, is supplied with a bipolar winding connected to a
commutator, and the machine therefore is a bipolar eommutating
machine in addition to a high-frequency inductor alternator
(16-polar in Fig. 139). It thus may be operated as converter,
receiving power by direct-current supply, as direct-current motor,
and producing high-frequency alternating power in the inductor
pole-face winding.

161. If the inductor alternator, Fig. 139, instead of with direct
current, is excited with low-frequency alternating current, that

FIG. 140.—Voltage wave of inductor alternator with single-phase excitation.
is, an alternating current passed through the field coil, F, of a
frequency low compared with that generated by the machine as
inductor alternator, then the high-frequency current generated
by the machine as inductor alternator is not of constant ampli-
tude, but of a periodically varying amplitude, as shown in Fig.
140. For instance, with 60-cycle excitation, a 64-polar in-
ductor (that is, inductor with 32 teeth), and'a speed of, 1800
revolutions, we get a frequency of approximately 1000 cycles,
and a voltage and current wave about as shown in Fig. 140.
The power required for excitation obviously is small compared
with the power which the machine can generate. Suppose,
therefore, that the 'high-frequency voltage of Fig. 140 were
rectified. It would then give a voltage and current, pulsating