1 PDF

The Alternating-Current Commutator Motor
BY B. G. LAMME
Chief Engineer, Westinghouse Elec. and Mfg. C o . , East Pittsburgh, Pa.
The alternating-current commutator motor in its various forms has been developed principally for the purpose of
obtaining adjustable or variable speeds, the usual induction type motor, both in the single-phase and polyphase forms,
being inherently constant speed. In general, control of speed with the alternating-current induction motor implies
variable or adjustable frequency in some form, and this, in turn, involves a commutator of some type. In consequence,
the problem of commutating alternating current usually goes hand in hand with that of speed control.
The simplest method, according to the author, for viewing both the a-c. and the d-c. commutating problems, is to
consider primarily the actual e.m.fs. short-circuited by the brushes, and the resistance in the short-circuited paths,—or
in other words, the problem of commutation is largely one of the permissible amount of short-circuited current. In the
d-c. machine the e.m.f. short-circuited by the brushes is that generated by rotation of the short-circuited armature coils
in the armature and external field fluxes. In the same way, in the a-c. commutator motor, there is an e.m.f. due to the
armature flux or field, as in the d-c. machine, and in addition, there are other e.m.fs. due to the primary or field fluxes.
These latter may be classified as primary rotational and transformer e.m.fs. One limitation in the design of a-c. com-
mutator motors, in general, is that these rotational and transformer e.m.fs. are often larger in degree than the e.m.f. due to
the armature flux. However, the author contends that the commutation problem in a-c. motors is the same as for d. c.
when all e.?n.fs. are taken into account.
The major part of the paper covers the consideration of the different e.m.fs. which should be taken into account in
the various types of a-c. commutator motors, and it is shown in a general way that the e.m.fs. involved in speed control
also appear in the commutation problem. In the latter part of the paper certain general conditions of commutation and
brush operation are treated and some figures are given for comparison of a-c. and d-c. commutating limits.
T HE alternating-current commutator motor in some

form is doubtless the earliest of ail alternating-
current motors. Probably everyone of the old-
With clearer and better conceptions of the commuta-
tion problem, it began to be seen, b y a few, that
alternating-current commutation was quite closely
timers who dabbled in alternating currents tried his related to that of direct-current; in fact, if all the condi-
hand, sooner or later, at this type of machine. Some tions were included, it was evidently the same problem
even went so far as to put such motors on the market. except in degree.
Series-wound, shunt-wound, repulsion types, induction However, in one vital feature, the commutation in
types, as they were formerly named, all had their trial, a-c. machines has been assumed to be fundamentally
either experimentally or commercially, and all the different from the d-c. In the former, the condition in
early motors withdrew for one common reason ; namely, many cases is largely that of short-circuiting part of the
sparking at the commutator. It must be kept in mind armature winding, by the brushes, under a magnetic
that all the earlier frequencies were quite high and it field, or when enclosing an alternating magnetic flux.
was recognized, early in the game, that this was a most Superficially, this problem appears to be quite different
serious handicap to successful commutation. One from that of simply reversing the current in those
principal reason why the Niagara Falls Power Commis- armature turns lying midway between poles, as in the
sion in 1892 chose 2000 alternations per minute (16 / 2
s usual d-c. machines. This short-circuiting under a
cycles per second) as its preferred frequency, was that magnetic field (or enclosing an alternating flux) was
commutator-type a-c. motors would thus be feasible supposed to introduce new and most intricate phenom-
for general power work. ena, and, apparently, the action was such that good,
In consequence of the commutation difficulties with or even passable, commutation was not even theoretic-
all the early a-c. commutator motors, and their conse- ally possible.
quent withdrawal from the commercial field, the belief As the theories and practises in d-c. commutation
became most firmly established that alternating developed, new ideas began to grow and new viewpoints
currents, in general, could not be commutated success- were obtained, which had a direct bearing on the cor-
fully and that it was hopeless to attempt further work responding a-c. problem. It was perceived, perhaps
on the problem. dimly at first, that the d-c. problem could be considered
This fallacy was the direct outcome of lack of knowl- simply as a case of the brushes short-circuiting part of
edge of the general problem of commutation itself. the armature winding in a magnetic field, such field
The alternating current simply introduced additional being set up by the armature winding as a whole, or by
conditions, all of which made the whole subject more a combination of armature and field fluxes. From
complex and difficult to understand. In fact, in this such a viewpoint the problems of a-c. and d-c. commu-
early period, in most cases, the methods of treating the tation began to approach each other, and it was per-
direct-current commutation problem were more or less ceived that it was largely a matter of the values of the
incomprehensible and there was little or no physical e. m. f. short-circuited, rather than the nature of the
conception of what was really taking place. current itself.
Paper to be presented before the Schenectady Section of the Others may have anticipated the writer in this
Α. I. Ε. E., April 2, 1920. method of looking at the commutation problem, but
249
250 LAMME: ALTERNATING-CURRENT COMMUTATOR MOTOR Journal Α. I. Ε. E.
possibly he was the first to go fully into the analysis of from the armature m. m. f. as in d-c. machines. In
this method and to show its equivalence to former consequence, on the d-c. conditions are superposed
methods. He found that this way of viewing the
1
additional ones, which often are more difficult than
problem was of very great help in the consideration of those corresponding to the d-c. However, the whole
a-c. commutating machines, both polyphase and single- problem turns on the total short-circuited e. m. f. and
phase. It put the short-circuiting and commutation the means for limiting the circulating or local current.
problems on a comprehensible basis and at once elimi- So much for commutation. Now as to the reasons
nated from the treatment a large number of inter- for the use of commutator motors, when induction
mediate phenomena. motors, both polyphase and single-phase, have been
To express this method in its simplest form,—if a available. It has been well known, almost since the
given e. m. f. at the commutator is short-circuited, induction motor first appeared, that this type of ma-
with a given brush, similar commutating conditions chine has inherently constant speed characteristics.
are obtained, whether the current is alternating or With the growth of the art and increase in the field
direct. But all the e. m. fs. must be taken into for electric motors, there has come a demand for vari-
account, and here has been the principal source of able and adjustable speed alternating-current motors.
trouble, both in a-c. and d - c , but more in the former The true induction motor cannot in itself give the
than in the latter. variable speed characteristics of the series d-c. motor,
In accordance with this method of treating the prob- nor the adjustable speed characteristics of the shunt
lem, the function of commutating poles, or their equiv- machine. True, variations in speed with the poly-
alent, is simply to reduce the short-circuited e. m. f. phase induction motor are obtainable by means of
under the brush to a value within the resistance limits resistance in the secondary circuit, but such speed
of the brush, as fixed principally by the brush contact regulation is directly comparable with the use of arma-
and by any additional resistance in the winding which ture resistance in d-c. shunt motors and this character-
may be inserted for commutation. With the d-c. istic is not a desirable one except for certain very
machine, this has been quite easy, as both the e. m. fs., limited uses.
due to the armature flux and due to the commutating The problem of true speed control in the induction
pole, are proportional to the speed, and therefore, if motor is complicated b y the fact that with change in
properly set for one speed, the conditions hold for all speed, both the secondary voltage and the secondary
speeds. But in the a-c. motor some components of frequency are directly related to the speed changes.
the e. m. f. short-circuited by the brush are not a It has been known for many years that, by the addition
function of the speed, while others are, and thus the of suitable voltages in the secondary circuit of the
problem of a correct commutating field for a wide range ordinary polyphase induction motor, the speed could
of speed becomes a very complex one. The difficulty is be controlled over any desired range, without rheo-
inherently one of design, rather than one of theory or static loss, but for each change of speed the regulating
principle. voltage supplied to the secondary winding requires
Narrowed down to its simplest form, in the d-c. ma- a different frequency, and the difficulty in the past
chine, the commutation is dependent upon the e. m. f. (as well as at present) has been to obtain the required
short-circuited b y the brush, and the permissible short special frequency for each speed. It is true that a
circuit or local current. The permissible work current very limited number of speed changes is possible by
enters directly as one of the elements which determines means of pole changing and by cascade connection of
the short-circuited e. m. f. The amount of short- motors, but these give no true flexibility in speed ad-
circuit current permissible is dependent largely upon justment. Up to the present, the only really flexible
the characteristics of the brush and brush contacts. means for obtaining variable frequency for such motors
The amount of short-circuit current which can flow is has been through some form of a-c. commutating device,
dependent upon the e.m. f. short-circuited, and the such as a multi-bar commutator connected to the
conditions of the short-circuited path. Due to the secondary or the primary winding; or a commutator-
characteristics of commercial carbon brushes, there are type frequency changer, a synchronous converter, or
fairly definite limits to the e. m. f., which can be short- a polyphase commutator-type motor, structurally in-
circuited while keeping the short-circuit current within dependent, but connected to the primary or secondary
the limitations of the brush itself as fixed by burning, circuits of the induction motor. It may be noted that
disintegration, etc. 2
the synchronous converter, transforming from alterna-
In the a-c. machine quite similar limits hold, except ting to direct current, is here included among the fre-
that the short-circuited e. m. f. is partly, or in most quency-changer devices. This should be considered
cases largely, due to fluxes other than those resulting as a frequency changer when used in this manner,
converting from direct current (zero frequency) to any
1. A Theory of Commutation and its Application to Interpole
a - c frequency depending upon the adjustments. This
Machines, T R A N S . , Α . I . Ε . E . , 1 9 1 2 .
implies direct-current circuits in addition to the alter-
2 . See Physical Limitations in D-C. Machines—TRANS.,
Α . I. Ε . E . , Sept., 1 9 1 5 .
nating-current supply circuits. However, many of the
M a r c h 1920 LAMME: ALTERNATING-CURRENT COMMUTATOR MOTOR 251
other devices proposed imply transformation from the along axis Â will generate a transformer e. m. f. b. In
primary or supply frequency to some other frequency the winding under pole A, therefore, there will be two
or range of frequencies, or vice versa. The commutator component e. m. fs. a and 6, which are in opposition to
thus enters as a frequency-changing, rather than a each other, a being in phase with the flux in phase A,
current-rectifying device. and b being 90 deg. out of phase with the flux B, which
in turn is 90 deg. out of phase with flux A. Thus a
T H E POLYPHASE COMMUTATOR T Y P E INDUCTION and b are 180 deg. part, neglecting minor effects.
MOTOR In the same way, under pole B, e. m. fs. c and d are
Although the polyphase motor without commutator approximately 180 deg. apart.
is a well understood device, yet there are various ways Obviously, the rotational e. m. fs. a and c are de-
of treating its characteristics with regard to speed pendent upon the speed of the secondary and therefore
variations, etc. Some of these methods serve better vary from zero, at standstill, to values equal to b and
for illustration than others. The following treatment, d when the rotational frequency (poles times revolu-
which has been used extensively by Eichberg, Mc- tions) is equal to the line frequency; for with equal
Allister and others, probably gives the best visual fluxes, equal frequencies, and other conditions, e. m. fs.
conception of what takes place in such machines, in set up by rotation are equal to the e. m. fs. set up by
regard to the component voltages upon which com- so-called transformer action. Therefore, at synchron-
mutation and speed regulation depend. The writer ous speed of the secondary, the. rotational and trans-
has accordingly adopted this for purposes of easy former e. m. fs. in the secondary windings practically
illustration, in connection with the several types of balance each other, in a well-designed machine.
apparatus which are considered. Obviously, in the external secondary circuits the
e. m. fs. will vary from maximum at standstill to zero
at synchronism, considered simply from the standpoint
of the component e. m. fs. a, b, c and d, as shown in
Fig. 1.
It is evident that the rotational e. m. fs. a and c
correspond to the counter e. m. fs. in an ordinary motor,
while the transformer e. m. fs., b and rf, correspond to
the impressed e. m. fs. Thus by comparison, the
rotational or counter e. m. fs., a and c are zero at stand-
still and rise in value until they are practically equal to
the impressed e. m. f., as in ordinary types of electric
A motors.
FIG. 1 Now as to the frequency of such secondary circuits,—
at standstill obviously this must be the same as the
Assume as the first and simplest case, a two-phase primary or supply frequency; and as the secondary
induction motor, in which the primary is illustrated winding is rotated in the direction of the primary
as a two-pole construction with two sets of poles at rotating magnetic field, the secondary frequency will
right angles to each other, as shown in Fig. 1, and with decrease, reaching zero at synchronism with the pri-
a secondary of a ring type with four taps at A A é and
X mary field. Thus the secondary voltage and frequency
ΒιΒι for two-phase connection. The e. m. f. relations both are dependent upon the speed of rotation. This,
in the secondary winding may be explained as follows: of course, is all an old story, but is brought in here
Considering the rotational e. m. fs. first, the alter- simply as part of the problem of commutation and
nating flux from pole or phase A through the secondary speed variation.
will generate an e. m. f. a, as indicated in Fig. 1. If the secondary circuits are connected to the second-
Also the alternating flux from pole or phase Β will ary windings by means of permanent connections, then
generate in the winding a rotational e. m. f. c, as obviously, the points of connection will rotate with the
shown. These e. m. fs. are dependent upon the secondary core. If, however, the secondary winding
primary flux and the speed of rotation and are in- is provided with a commutator and the secondary
dependent of the primary frequency. Due to the fact circuits are connected to the secondary winding by
that the flux and poles A and Β are at 90 deg. relation means of brushes on such commutator, then the con-
to each other, these two rotational e. m. fs. a and nection of the secondary circuits to the secondary
c are at right angles to each other. winding becomes a variable or movable one, depending
Let us now consider the transformer action of the upon the motion of the brushes relative to the commuta-
primary fluxes acting on the secondary. The rotor tor. If the brushes are in a fixed position on the com-
winding along the axis of pole or phase A will generate mutator, as shown in Fig. 2, then the secondary
an e. m. f. d 90 deg. out of phase with the flux of phase connections are fixed in space, so to speak, instead of
A. This is indicated in Fig. 1. Also the winding rotating with the secondary winding. In such an
252 LAMME: ALTERNATING-CURRENT COMMUTATOR MOTOR Journal Á . I. Å . E .
arrangement, the frequency of the external secondary increases, for the rotational e. m. f. per turn is rising
circuits at any speed is the same as with the usual and more nearly approaches in value the opposing
induction motor at standstill. In other words, it is transformer e. m. f. per turn. The e. m. f. short-
always the same as that of the supply circuit, although circuited by the brush decreases accordingly, reaching
the e. m. fs. in the secondary circuits will vary with practically zero value at synchronous speed in a well-
the speed, as in the ordinary induction motor. Con- designed machine, and again rising in value above
sequently, this arrangement differs from the usual synchronism. This is the resultant e. m. f. due to the
induction motor, in having full line frequency, with primary flux. If this resultant were the only e. m. f.
varying voltage, in the external secondary circuits, to be considered, the problem would be fairly simple in
whereas the usual induction motor has varying fre- its explanation. However, there are other magnetic
quency and varying voltage in such circuits. The fluxes due to the rotor magnetomotive forces, as in
interesting feature in this arrangement is that we now d-c. machines, which also set up e. m. fs. and the
can supply e. m. fs., at line frequency, to the secondary inclusion of these complicates the explanation. These
circuits for speed control. From an inspection of Fig. latter, however, are usually small compared with the
2, it is obvious that for stable conditions, e. m. fs. a and resultant short-circuited e. m. f. due to the primary
b must balance if no other conditions are introduced ; field, and in most cases are largely masked by the latter,
that is the motor will tend to run up to synchronous except near synchronism. Ordinarily the conditions
speed. However, if external e. m. fs. respectively in which tend to give a low short-circuited e. m. f., due to
phase with a and c are added, then these e. m. fs., plus the primary field, also tend to reduce these lesser
a and c can balance b and d at some other than syn- e. m. fs., so that in general one need consider only the
chronous speed, and thus stability is obtained at other larger e. m. f. due to the primary field. By reducing
than synchronous conditions. It is thus evident that, this to a value within the limits of resistance commuta-
tion, and proportioning the armature so that it would
be very good for commutating a corresponding direct
current, one usually need consider nothing else, for
practical purposes. Obviously, in the motor just
described the commutating conditions will be easiest
near synchronism and will become increasingly difficult
with departure from synchronous speed. In practise,
the permissible limit of short-circuited volts for opera-
tive commutation may be reached anywhere between
synchronism and standstill, depending upon the design.
From the preceding description, it is evident, in the
general problem of speed control, that we must take
into consideration the frequency of the current handled
FUÎ. 2
by the circuits connected to the brushes on the commu-
by means of the commutator, we are able to introduce tator. While this may have little bearing on the gen-
into the secondary circuits suitable e. m. fs. at line eral question of commutation, it is one of the important
frequency to accomplish this result, and this is one of features in the general scheme of speed variation or
the methods which has been used in the past for speed control of the induction motor. Therefore where there
control of polyphase motors. The A. E. G. Co., of are two (or more) sets of circuits connected to one or
Berlin, has built such motors, in which the supply cir- both motor windings, it may be advisable at this point
cuits for the secondary voltage regulation are obtained to show in a simple way the relation of the frequencies
from special windings on the primary, with a number in these circuits. Such circuits may be connected to
of taps to give different voltages. In other arrange- the windings, either by permanent taps or by variable
ments transformers have been used for converting taps through the medium of brushes on one or more
from the line to the desired secondary voltages, with a commutators. One set of these circuits may always
number of taps for varying such voltages. be considered as a primary or supply system and the
The above covers broadly this general method of others as secondary circuits, whether they are on the
speed control. Let us look now into the problem of primary or the secondary core. A simple rule can be
commutation in such machines, as determined by the given for the relation of these frequencies in polyphase
e. m. fs. short-circuited by the brushes. In the first motors,—namely, the frequency of any external secondary
place, at standstill the volts per conductor or turn are circuit relative to the primary or supply circuit is deter-
practically the same as in the primary winding, and the mined by the relative angular movement of its points of
e. m. f. spanned by the brush is dependent upon the contact with respect to those of the primary circuit. If
volts per turn and the number of turns short-circuited. these contacts have a fixed angular relation to each
In the second place, as the secondary speeds up, its other, then the frequency of the secondary circuits is
resultant volts per turn decrease directly as the speed always the same as that of the primary circuits. If
these are moving with respect to each other, the In this arrangement the commutating brushes are
frequency of the secondary circuits will vary from that fixed in position while the winding, commutator and
of the primary according to the relative movements. primary supply taps revolve. Here the primary field
For example, in the usual induction motor, with or flux travels at a definite speed with respect to the
stationary primary and wound secondary with collector rotor core and windings, the primary current being
rings, at standstill the stator and rotor taps on the supplied through the collector rings and taps to the
windings have a fixed relation to each other and the winding. Just as in a stationary primary, the pri-
frequency of the secondary current is therefore the mary counter e. m. f. and the effective e. m. f. per turn,
same as that of the primary. As the secondary core are constant, as they depend upon the rotation of the
and windings increase in speed, the rotor taps move primary field or flux with respect to the primary core
with respect to the primary and the frequency of the and windings. Therefore the short-circuited e. m. f.
secondary external circuits is thus varied. under the brushes on the primary winding is constant
In the same way, the motor with a commutator on and corresponds to that portion of the primary e. m. f.
the secondary and with fixed brushes, will, according to spanned by the brushes. Herein lies one reason for the
the above rule, give full primary frequency on the use of the double primary winding. With a very high
secondary external circuits at all speeds, simply because applied e. m. f., the number of primary turns would be
the secondary taps, or points of contact, hold a fixed high and the turns per coil or per commutator bar
relation to the primary taps. Any rotation of the would be unduly large unless an enormous number of
secondary brushes will give a varying frequency in the commutator bars is used ; but the addition of the low
secondary external circuits, depending upon the speed voltage winding allows this element to be proportioned
of rotation of the brushes. to suit the short-circuited conditions and the number
There are various modifications of the polyphase of commutator bars, independently of the primary
commutator-type motor, several of which involve winding proper.
special commutating conditions, and therefore should This arrangement short-circuits the full value of
be described briefly. One of these is known as the primary e. m. f. per turn at all speeds, but the brushes
Schräge motor. This is a polyphase induction motor do not handle current of the primary frequency, but of
with a rotating primary with polyphase taps and col- a frequency depending upon the speed, for the primary
lector rings and also with a commutator and polyphase taps and the brushes on the commutator are moving
brushes. In its simplest form the commutator and the with respect to each other, except at standstill. The
collector ring taps are connected to a common winding frequency delivered at the commutator therefore
of the closed coil direct-current type. In the more varies from full-line frequency at standstill to zero
usual form there is a second low voltage winding on the frequency at synchronous speed. However, the volt-
rotor core which is connected to the commutator, the age delivered from the winding to the brushes is
primary winding proper being of higher voltage and not practically constant at all speeds. The result therefore
connected to the commutator in any way. By this is just the opposite of that described in connection with
latter arrangement, the primary supply e. m. f. and the the polyphase motor with a commutator on its second-
low voltage required on the com- (-Rotating Primary
ary, for in that case the frequnecy delivered at the
mutating winding become inde- commutator was constant and the voltage varied with
pendent of each other and each the speed, whereas in this latter case, the frequency
can be proportioned for its own varies with the speed and the voltage is constant with
best requirements. This may be a given brush setting. As the frequency and voltage
illustrated, in its arrangement of of the stator winding of this motor both vary with the
circuits by Fig. 3. Here only a speed, as in the usual induction motor, it is thus possible
single primary winding is shown, in this motor to obtain speed regulation by connecting
for simplicity, with both the col- Stationary Secondary'
the circuits from the primary commutator brushes
lector rings and the commutator FIG. 3 directly to the secondary winding, provided voltage
connected to it. AAA represent three sets of fixed adjustment is obtainable. This, of course, can be
brushes, while Â Â Â represent movable ones. accomplished by means of transformers and in various
According to the foregoing method of determining other ways, but in the Schräge motor, as built, this is
the frequency etc., between brushes A and B, there is obtained by the two sets of brushes on the commutator,
a fixed e. m. f., but a frequency depending upon the as shown in Fig. 3, which can be displaced with respect
speed of rotation. Connecting the circuit from A Â to each other, so that the voltage between them can
to a phase group on the stator core, the speed will thus be varied up or down by suitable setting of the
balance at such point that the e. m. f. and frequency of brush arms. In this way the armature winding and the
the brushes will equal the stator e. m. f. and frequency. commutator are used as the equivalent of a transformer
By rocking Â to some new position, a different voltage is with variable voltage taps, so that both frequency and
obtained between brushes A and Â and thus a new speed voltage are thus under control.
is required for balancing the e. m. fs. and frequencies. As far as commutation is concerned, this scheme is
254 LAMME : ALTERNATING-CURRENT COMMUTATOR MOTOR Journal Á . I. Å . E.
at a disadvantage compared with the previously speed will vary directly with that of the primary supply
described arrangement with the commutator on the brushes.
secondary winding in which the short-circuited e. m. f. Looking into the conditions of commutation in this
varies with the slip from synchronism. With this type of machine, it is at once noted that, as the brushes
latter scheme, therefore, commutation is to a consider- short-circuit part of the primary winding which gener-
able extent independent of speed, simply because the ates the counter e. m. f. of the machine, the short-
short-circuit conditions are equally difficult at prac- circuited e. m. f. will be proportional to the counter
tically all speeds. If such a machine can be made to e. m. f. and therefore will be of practically constant
operate sufficiently well at any one speed, it should value. Here the conditions and limitations of, com-
then be suitable for operation over a fairly wide range mutation are fixed almost entirely by the primary
of speed. e. m. f. short-circuited, and practically no other e. m. f.
It is apparent from the description of this Schräge conditions need be considered. If five volts, for
motor that the rotor becomes simply a changer of instance, could be taken care of by the brushes, then
frequency, but not inherently a transformer of voltage. they would stand an average of five volts on the com-
A similar arrangement was devised many years ago by mutator at any speed within reason. As far as commu-
the writer purely as a frequency changer, and it is being tation is concerned, this motor is subject to the same
used at present in connection with speed control of limitations as the Schräge motor with a single primary
large induction motors. The rotor of this machine, as winding.
shown in Fig. 4, consists of a primary winding on the The above covers briefly the polyphase induction
rotor core connected to both a commutator and a set of motor with commutator. Polyphase series and shunt
collector rings. In operation the core is rotated inside types of commutator motors might be considered as
a laminated ring, which may or may not carry any next in order, but they are so closely related to various
windings. At synchronism with the a-c. supply or single-phase types, that their treatment, according to
primary system, the commutator delivers direct current the foregoing method of illustration, should properly
at a voltage proportional to that^of the alternating follow the single-phase. Therefore, they will be taken
up later.
SINGLE-PHASE COMMUTATOR MOTORS

These may be of various types or forms, such
as series, repulsion, doubly-fed and induction types.
In these, in general, the commutating conditions
are more difficult to analyze and to explain than
in the polyphase induction motor, just as the phe-
FIG. 4
nomena of single-phase induction motors, in general,
e. m. f. impressed. With change in speed, either up or are usually more difficult to grasp than those of the
down from synchronism, the brushes deliver current at polyphase. Also just as the assumption of two
a frequency varying with the slip from synchronous component fields of opposite rotation can replace the
speed, but the e. m. f. delivered is constant with con- single-phase field in considering some of the characteris-
stant supply e. m. f., regardless of the slip. This is tics of single-phase motors, so in commutation of
therefore a variable-frequency, constant-voltage ma- single-phase machines this assumption may be of
chine and the conditions of commutation are fixed assistance in some instances, in explaining the actions,
principally by the primary flux or field, practically although the writer has not used the method in this
regardless of speed. By the use of several sets of paper.
brushes, properly spaced on the commutator, polyphase Single-phase motors are structurally of two types,
currents are taken from the commutator. namely, of the polar and of the distributed type fields.
A third form of polyphase commutating motor was In practise these are equivalent, but the distributed
proposed in Europe some years ago, in which a sta- type, with its overlapping windings and fluxes, intro-
tionary primary winding of the closed-coil type is duces such complexity, in any general analysis, that
connected to a commutator, and the current is admitted for purposes of illustration, a polar arrangement should
by brushes which are rotatable on this commutator. be shown. Consequently, in the illustrative diagrams,
It is at once evident that this is simply an induction which are given later in connection with single-phase
motor primary in which the points of current supply can motors, polar arrangements are given. It might be
be rotated around the core. In consequence, the speed well, therefore, to show the equivalence of the two
of rotation of the primary field, or flux, corresponds to arrangements insofar as principles are concerned.
the line frequency, plus or minus the speed of rotation Equivalence of Polar and Distributed Windings.
of the brushes. The secondary core and winding, of Taking the simplest type of distributed winding,—
usual polyphase type, at full speed rotates at practically namely, the closed-coil, ring type, other types and
the speed of the primary field; and accordingly its arrangements can be developed from it quite easily.
Assuming such a ring-type closed coil, as shown in of the field winding equal to double these angles, be-
Fig. 5, then with taps at A and Â there will be a come magnetizing, and the remaining parts become
maximum m. m. f. at A and Â tapering off to zero mid- compensating. This relation is just as definite, and in
way between. Thus there will be magnetic poles with fact is the same thing, as using two separate windings
maximum value at A and Â and zero at midway points. for exciting and compensating, connected in series, and
The same holds true for a closed coil armature. The with armature brushes at a midway point in the dis-
maximum m. m. fs. of the armature windings corres- tributed exciting winding. With the two distinct
pond to the points of current connection, as shown in winding arrangements, the relation of exciting to com-
Fig. 6. pensating strength can be changed without chang-
Let us consider such an armature in a ring field of ing the brush angle, which is not the case with the single
this type. Fig. 6 illustrates this. With the brushes
FIG. 5
FIG. 7
on the armature at right angles to the field taps, and
corresponding to the points of zero field flux, the winding. Or stating the matter in another way, the
arrangement is equivalent to the usual d-c. motor field strength, and thus the speed, may be changed
with two poles and with the armature brushes on without shifting the brushes. However, in principle,
neutral. Here all the field winding is magnetizing. the arrangements are the same.
However, let us next shift the armature brushes The disadvantage in directly representing the polar
around to another angle than 90 deg. to the field axis. conditions with distributed windings, in the various
Assuming a 45 deg. angle, the arrangement is as in single-phase motors illustrated in the following figures,
Fig. 7. Then all the field winding between A and C, is that the overlapping poles and fluxes are confusing
representing double the angle between the armature to almost everyone except the designer. In conse-
and field axes, becomes magnetizing or exciting field quence, the writer has shown the fluxes as separate
winding and sets up a magnetic field at right angles to
FIG. 6
FIG. 8
the armature axis or brushes; while the part of the
field winding between Â and C represents a m. m. f. magnetizing and compensating or transformer com-
acting parallel to the armature axis, but in opposite ponents in order to illustrate more easily the principles
direction to the armature m. m. f. This part of the of operation and of commutation.
field winding thus automatically becomes a compensating Single-Phase Series Motors. Of all the single-phase
winding. motors, presumably the straight series type is one of
Taking again a much smaller angle between the field- the simplest to explain from the standpoint of commuta-
winding and armature-winding axes, as in Fig. 8, here tion, as well as otherwise. This motor is, in effect,
again the field winding between A and C becomes just like a d-c. series motor of similar general construc-
magnetizing, while that between C and Â becomes tion and the performance is much the same. However,
compensating. Thus by setting the armature brushes it differs from the d-c. series motor in one essential point ;
at various angles to the field-winding axis, components namely, there are two e. m. fs. set up in the armature
winding, although not along the same axis, these e. m. fs. which could be short-circuited successfully by the
being the rotational and the transformer e. m. fs. due brushes, and his early experimental designs were all
to the main field flux. This is illustrated in Fig. 9. along lines of obtaining a sufficiently low transformer
The rotational e. m. f. is along the axis of the brushes, e. m. f. Low frequency, of course, was one of the
while the transformer e. m. f. is along the field axis, and recognized factors in accomplishing this. Mechanical
at right angles to the rotational, and therefore has no limitations in the number of commutator bars, width
direct influence on the latter. However, the brushes of brush, etc, and the electrical limits of volts per bar,
short-circuit part of the transformer e. m. f., as will be at once dictated that the counter e. m. f. of such a
considered further. machine must be relatively low. There was much
In its commutation the series motor embodies the criticism at that time regarding the adoption of such
equivalent of the usual d-c. problems, plus the addi- low voltage for a-c. series motors,—about half that of
tional one due to the transformer e. m. f. in the arma- d-c. railway practise,—but it may be noted that no one
ture windings. Thus there are two short-circuited in the succeeding twenty years has been able to increase
e. m. fs. to take into account, one being due to rotation greatly over the armature voltages which were first
in the armature field or flux, and the other to trans- used, with motors of corresponding speed and capacity.
former action. These are 90 deg. out of phase with Here appeared to be a fundamental limitation, partly
each other. This, together with the fact that one of of a mechanical and partly of an electrical nature.
these e. m. fs. is dependent upon armature speed while In the study of the short-circuited transformer
the other is independent of it, makes any commutating e. m. f., it appeared at that time that somewhat higher
pole problem quite difficult, for this type of motor is short-circuited e. m. f.s. under the brush would be
used principally where speed variation over a wide allowable, by including a considerable amount of
range is one of the requisites. However, it is the resistance in the short-circuited paths, in order to
transformer component which is the great handicap in reduce the short-circuit current. In consequence, the
use of resistance, or preventive, leads, between the
commutator and windings, was proposed and tried with
the result that materially higher short-circuited e. m. fs.
proved to be practicable, and thus correspondingly
higher main flux densities were permissible. The
writer has never carried this construction to its extreme
limits, as no very good opportunity offered itself. The
use of resistance leads for such purpose was not new or
novel, but the writer believes the results he obtained
FIG. 9 were much more satisfactory and conclusive than in
former trials, largely because the leads were applied
a-c. commutation in general, for it is not possible to with due regard to the electrical limitations of the
counterbalance completely such an e. m. f. under all short-circuited e. m. f. and the values of local and work
conditions, by any corrective e. m. f. depending upon current, etc.
rotation, as is done in d-c. machines; for at zero speed In the early days the writer also gave serious consid-
this transformer e. m. f. still exists at full value, whereas eration to the problem of commutating poles on series
no rotational e. m. f. is possible under this condition. a-c, single-phase, motors and even carried this to the
Therefore, in this, as in all a-c. commutator motors, the extent of trying it out on a 300-h.p. motor. This was
standstill condition represents a limit where compensa- not for the purpose of eliminating the resistance leads,
tion, or the commutating pole in any form, is ineffective. as he did not believe at that time that this was practic-
Under this condition, however, only the transformer able, on account of starting conditions, but it was for
e. m. f. is short-circuited by the brushes, all rotational reducing the short-circuited e. m. f. under the brushes,
e. m. fs. being absent. while in rotation, by introducing a counter-acting
As this transformer e. m. f. in the short-circuited e. m. f. Unfortunately, these tests were carried out
coils is proportional to the main field flux, it might be entirely with series excited commutating poles. There
assumed that there could be a considerable gain by was a very considerable improvement at high speeds,
reducing the field flux during starting and at lower but at moderately low speeds no improvement was
speeds. However, this cannot be carried very far, for apparent. Of course, it was recognized, even at that
the torque of the motor also depends upon the main time, that the transformer e. m. f. under the brush was
field flux, and thus any reduction in this flux means out of phase with the armature current and it was
correspondingly increased armature current,—so that doubtless recognized that the phase relation of the
the reduction in the short-circuit or local current is commutating pole flux was not just what it should be.
accompanied by an increase in the work current. However, the general results, as tested, held so little
In his early work on single-phase commutator motors, promise, except at quite high speeds, that it was felt
the writer recognized the condition of a limiting e. m. f. that it was not worth while to follow this matter up
March 1 9 2 0 LAMME: ALTERNATING-CURRENT COMMUTATOR MOTOR 257
further. At that time the proportioning of commutat- Assume first a series motor with its armature fed
ing poles and commutating-pole windings was not very from the secondary of a transformer, the primary of
well understood, even on direct machines, so it is not which is in series with the motor field. Also assume for
surprising that anyone should be misled on the a-c. simplicity a 1 : 1 ratio of transformation. Fig. 10
machine. illustrates this arrangement. 1
More recent work on commutating poles in such With a given field flux, the armature winding will
motors has shown that properly proportioned poles and generate along the brush axis a rotational e. m. f.
windings with suitable phase shift of the current will directly proportional to the speed, and counter to that
materially improve the commutation over a relatively of the secondary of the transformer, to which its
wide range of speed beginning at fairly low speeds. brushes are connected. This, therefore, fixes the value
Experience also shows that, by such means, series of the secondary e. m. f. of the transformer and,
motors without preventive le^ds can be made to give therefore, its primary e. m. f. Accordingly, with a
results comparable with motors with such leads, but 1 : 1 ratio the transformer primary e. m. f. is practically
without commutating poles. Very considerable devel- equal to that of the armature at all times; and the
opment has been carried on recently along such lines. main field, with this primary e. m. f. in series with it,
These good results are due, largely, to better knowledge should act just like the field and armature in series in
of the most desirable proportions, etc This will be the straight series motor. Assuming a given main field
treated more fully, later. flux, then the armature rotational or counter e. m. f.,
No mention has been made of compensating windings and consequently the e. m. f. of the transformer, will
on series motors. Such windings were developed very vary up and down with the armature speed. Therefore
early, but principally for the purpose of improving the
power factor by neutralizing the armature m. m. f.
Apparently this did effect some slight improvement in
the commutation, but it was possibly a mere question
of opinion. Both series-connected and shunt-con-
nected compensating windings were tried, the latter
operating fairly well but seemingly representing no nut
physical improvement over the series arrangement.
Over-compensation with the winding series connected,
appeared to represent little gain. In fact, the com-
pensating winding short-circuited on itself seemed to
be just about as effective as any other arrangement, in
most of the tests.
Repulsion Motor. Another early development was
the so-called repulsion motor, now known as a trans- FIG. 10
former type. We will retain the older term, as it is
better known. At first, this was believed to be a the magnetic flux of the transformer will thus vary with
distinct type of single-phase motor, although its the armature speed.
characteristics were recognized as being very similar to The primary current of the transformer necessarily
those of the series machine. In an Institute discussion is in phase with the main field current and the secondary
of repulsion motors, in January, 1904, the writer showed current from the transformer will be in phase with the
that this type of motor was simply a series type, in primary, excepting the necessary magnetizing com-
which the current in the armature was generated by ponent. The current supplied to the armature will,
transformer action, instead of being fed in directly at therefore, be practically in phase with the rotational
the brushes. In other words, it is a transformer type e. m. f., just as in the straight series motor, excepting
series motor. Possibly there was no novelty in this the small magnetizing current for the transformer,
form of treatment, but it enabled one to analyze this which is 90 deg. out of phase with the primary current,
type of motor from the series standpoint. This method as is also the flux in the transformer. The secondary
of treatment also allows the commutation problem to e. m. f. of the transformer must be just enough out of
be considered in a fairly simple manner. 180 deg. phase with the armature rotational e. m. f. to
A brief review of this method may be given here, not allow the necessary transformer magnetizing current
simply for the purpose of showing the relation of the to flow.
repulsion to the series motor, but to bring out some of Such an arrangement, in general, can replace the
the commutating conditions not included in the straight series motor. It is different from the series
former discussion, and also to show the close resem- motor, in that the armature can be proportioned for
blance to some of the conditions in the polyphase its own best voltage and current conditions, regardless
induction motor, as described and illustrated in con- of line conditions. It also differs in the fact that the
nection with Figs. 1 and 2. armature furnishes exciting current for the transformer.
It should be noted also that minor effects due to mag- synchronous speed, which, as already described under
netic leakages, etc., are neglected in the above treat- Polyphase Motors, gives zero frequency and e. m. f. in
ment. the rotating armature coils.
In the repulsion type motor, instead of a separate Under the condition of equal fluxes the commutation
transformer with its own primary and secondary wind- is best, as in the polyphase motor with a commutator
ings, the magnetic circuit of the motor itself is used as a type armature, fpr the coils short-circuited by the
transformer core, and the armature winding is used as brushes have in them a transformer e. m. f. generated
the secondary winding of the transformer. Thus the by the flux of the main poles, and a practically equal
secondary voltage supplied to the motor, and the rota- and opposite rotational e. m. f. generated by the trans-
tional e. m. f., are both generated in the same winding. former poles. Due to these two principal fluxes being
90 deg. out of phase, as already explained, the two
Main Field
t e. m. fs. balance each other in the short-circuited coils;
>
>
which is simply another way of stating the above con-
dition of zero e. m. f. at synchronous speed.
At standstill, the transformer e. m. f. in the short-
circuited coils, due to the main flux, exists just as in the
Transf. series motor, but there is no rotational e. m. f. from the
Field
transformer poles to neutralize it. This, therefore, is
the same as in the series motor, but between standstill
and synchronous speed, the resultant short-circuited
e. m. f. falls with increase in speed. The commutation
thus improves with speed, much more than with the
series motor.
FIG. 11 It has been shown in connection with Figs. 10 and 11,
that for a fixed main field the transformer field is low
Fig. 11 illustrates such an arrangement. Here the at low speed, and increases with the speed until at
transformer magnetic circuit is across, or at right synchronism, it is equivalent to the main field,—that
angles to, the main magnetic circuit. is, it varies directly with the speed. Let us see the
In operation, the counter e. m. f. a is generated along effect on commutation above synchronism. At all
the brush axis by rotation under the main field. Also, speeds the short-circuited e. m. f. due to the transfor-
due to the transformer flux, an e. m. f. b is generated in mer action of the main poles, is constant with constant
the winding along the same axis. If the main current flux, but the rotational e. m. f. in the same short-
in the transformer winding should generate the trans-
former flux, then such flux would be in phase with the
primary current, and the transformer e. m. f. b would
be at 90 deg. to the rotational e. m. f. a. But this
represents an unbalanced condition with the brushes
short-circuited on themselves, as in Fig. 11, and heavy
currents will tend to flow, correcting the unbalanced
voltage conditions. The result of this is a magnetizing
current for the transformer magnetic circuit which
gives a resultant flux in the transformer pole lagging 90
deg. from that in the main poles. This 90 deg. lag in
flux thus throws the transformer e. m. f. b practically Curve of Short-Circurted Rotational
180 deg. out of phase with a; that is, directly in opposi- Emf. du$ to Transformer Field-
tion to it, as described in connection with Fig. 10.

Thus there are generated in the armature winding two
l· i—é u~
FIG. 12
practically equal and opposite e. m. fs. However, the
fluxes in the two fields are not equal or equivalent except circuited coils, due to the flux from the transformer
under certain conditions. The e. m. f. a depends upon poles, increases rapidly with speed, both on account
the main poleflux,and upon the speed, which is variable of the increase in speed and the increase in the trans-
while e. m. f. b depends upon the transformer pole flux, former flux, which also varies with the speed. Conse-
and the frequency of the main system which is con- quently, the rotational e. m. f., due to the transformer
stant. When the rotational frequency (revolutions field, increases as the square of the speed.
times poles) becomes equal to the line frequency, the These e. m. f. conditions may be illustrated in Fig.
two sets of field fluxes become equal, or equivalent, in 12. Here it may be seen that the resultant e. m. f.
generating e. m. f., and therefore, the motor is similar under the brush falls slowly, slightly above standstill,
to a polyphase motor with its armature rotating at but quite rapidly near synchronism. Above synchron-
ism it increases rapidly. At 41 per cent above syn- current capacity is concerned, as the exciting current is
chronous speed, it is equal to the standstill value, and relatively small, compared with the normal work
at 100 per cent above synchronism, it is three times as current carried by the other set of brushes. The
great as at standstill. Of course, this is simply an short-circuiting conditions, however, may be the same
illustrative condition, as the actual values do not as for the main brush set. It may be mentioned,
follow rigidly the above conditions, the phase relations before leaving this type, that in practise the exciting
of the two e. m. fs. not being at exactly 180 deg. But current is usually supplied through a small transformer
this method of showing the commutating conditions the primary of which is connected in series with the
gives a good conception of why the repulsion motor transformer-pole winding.
sparks badly at much above synchronous speed. Doubly-Fed Motor. The above treatment of the
repulsion motor leads directly to another later type,
which has been used considerably; namely, the doubly-
fed motor, as it is usually called. On first inspection
this appears to be a combination of the series and the
repulsion types, in that the armature is connected in
series with the main field winding, and that it also has
a transformerfieldlike the repulsion motor. However,
the transformer field differs in the fact that it is
connected across a section of the main supply trans-
former, and thus, in reality, is in shunt relation to the
armature and main field windings. This allows
control of the transformer e. m. f. in the armature
winding, and also of the rotational e. m. f. due to the
FIG. 13
transformer flux. The general arrangement is illus-
trated in Fig. 14. Here it is seen that the main field and
Armature Excited Repulsion Motor. A modification the armature windings are in series, as in the series
of the above type is the Winter-Eichberg motor built motor, the transformer winding being across part of
by the A. E. G. Co. of Germany. This is illustrated in the main transformer. In consequence of this shunt
its simplest form in Fig. 13. The only material dif- connection of the transformer, its magnetizing current
ference between this form and the repulsion motor is supplied from the transformer, instead of the arma-
shown in Fig. 8 is that the stator exciting windings are ture winding, as in the true repulsion motor, and the
omitted and excitation is supplied to the armature magnetizing current and flux lag 90 deg. The trans-
instead, by a set of brushes midway between the usual former e. m. f. b generated in the armature winding is
brushes. at 180 deg. to e. m. f. a, just as in the repulsion motor,
To generate the main field flux, it takes a given but b now can have any value, regardless of a, depend-
m. m. f. whether the windings are placed on the main
poles or on the armature. However, when placed on
Main Field
the poles the main flux sets up a counter e. m. f. in the
field coils at 90 deg. to the current, thus materially
affecting the power factor. When placed on the
armature, there is a corresponding m. m. f. set up, with
the armature at standstill, but, as shown before, at
synchronous speed the armature e. m. f. is practically
zero; that is,—the exciting m. m. f. can be supplied at
practically zero volts, thus greatly improving the power
factor. Such improvement lessens, however, with
departure from synchronous speed.
Another way of looking at this is that the main flux
through the armature winding sets up a 90 deg. dis- FIG. 14
placed e. m. f., just as in a stator exciting winding, but
the transformer field sets up, by rotation, an opposing ing upon the'flux in the transformer pole, which is
e.fm. f., so that the resultant e. m. f. varies from full dependent in turn, upon the main transformer e. m. f.
reactive e. m. f. at standstill to practically zero at supplied to the transformer pole winding. Thus the
synchronous speed. principal condition of this arrangement is the non-
This motor has practically the same commutation dependence of b on a, although they are in practical
problems as the ordinary repulsion motor except that opposition. Their difference appears as an e. m. f.
it has twice as many brush sets. The exciting set, between the brushes. By adjusting the e. m. f. sup-
however, does not present much of a problem as far as plied to the transformer winding, to make the e. m. f.
b equal to a, the two brushes could be connected higher speeds and the main field is restored to normal
together, as in the repulsion motor. value by the time any considerable speed is attained.
When it comes to commutation, this motor differs The Commutating Pole Series Motor. The next step
materially from the repulsion motor. It is evident that from the doubly-fed motor is to the series motor with
if the transformer pole flux is adjustable in value, then commutating poles with phase displacement. In fact
the e. m. f. d due to transformer action of the main the doubly-fed motor with its adjustable transformer
field can be balanced by the rotational e. m. f. c at other pole strength may be defined as one form of commuta-
than synchronous speed. Consequently, the resultant ting pole series motor. The commutating pole series
short-circuited e. m. f. under the brush can have its motor, proper, is simply a series a-c. motor, with main
minimum value at any desired speed, within the limita- and compensating windings in series with the armature
tion of this type of motor. This can be illustrated as in and with small commutating poles in addition, as
Fig. 15. shown in Fig. 16. The commutating poles may
generate a small opposing e. m. f. b in the armature
é winding, but on account of the small size of the poles
themselves, this effect is almost negligible, compared
with the doubly-fed motor, and it is opposed by a
corresponding e. m. f. in the compensating winding.
However, it is intense enough, in its flux, to set up the
necessary opposing e. m. f. in the short-circuited
armature coils. By suitably exciting it in the 90 deg.
relation, as in the doubly-fed motor, it can do what the
latter motor does, and in addition, can be over-excited
below synchronous speed, without the disadvantages of
the transformer pole in the doubly-fed type, as its total
flux is quite small, in comparison. Thus it can give
FIG. 1 5 considerable correcting effect at quite low speeds,
which is a material advantage. Furthermore, by
Assume, as in the repulsion motor conditions in Fig. exciting at some other than the 90 deg. relation, it can
12, that the main field is of constant strength. Then also be made to assist in neutralizing the so-called
the transformer e. m. f. in the coils short-circuited by reactive voltage of the armature (corresponding to d-c).
the brush, due to the main pole flux, is of constant For this latter the commutating pole flux should be
value, as shown in I in Fig. 15. However, the e. m. f.
in the short-circuit coil due to rotatiqn in the trans-
former field, is proportional to the speed, instead of the
speed squared, for the transformer flux does not vary
with the speed, as in the repulsion motor. This is
shown as curve 2 in Fig. 15. By weakening the trans-
former field, the rotational e. m. f. could be changed
to 3 or 4. Thus the point of complete neutralization
of the two short-circuited e. m. fs. can be shifted to
various speeds, and the commutation is thus under
better control than in the corresponding repulsion
motor. The principal advantage taken of this in is
running above synchronous Speed, where the usual
repulsion motor is bad, as shown in Fig. 12. The FIG. 1 6
method could be used below synchronous speed also,

but the flux from the transformer pole would have to more nearly in phase with the main flux. Conse-
be larger than that in the main poles, which, in general, quently by exciting the small commutating poles by a
means a larger machine. As a matter of simplicity in combination of series and shunt connections, an average
control, such motors are sometimes operated as repul- best condition is obtained; but this changes with the
sion motors below synchronism, and as doubly-fed at speed, and must be adjustable for best results. One
the higher speeds, where weaker transformer fields relatively simple and effective means for obtaining the
are used. desired phase-shift, in the commutating poles, is by
In practise, in order to reduce the short-circuiting means of a series winding on these poles, this winding
at start and low speeds, it is not unusual to weaken the being shunted by a suitable resistance. This arrange-
main field a small amount, possibly as much as 25 per ment, patented by Latour, has been used to a consider-
cent to 30 per cent. This simply helps the low speed able extent in Europe, especially by the Oerlikon
conditions, but does not have any material effect at the Company.
The foregoing descriptions cover only the general ously this represents quite good commutating condi-
arrangements of single-phase series type motors, with tions, as in the repulsion motor at synchronism.
regard to commutation. Many variations have been Like the doubly-fed motor, balanced voltage condi-
tried out from time to time, but it is beyond the scope tions can be obtained at the repulsion motor brushes,
of this paper to go into all such methods. Latour and at other than synchronous speeds, by the introduction
Alexanderson have been pioneers in the doubly-fed of an e. m. f. at the repulsion motor brushes. Assume,
type and they have devised most excellent schemes and for example, that an e. m. f. of proper phase relation
methods, both for this and other types, which, in with a value of 2 0 per cent of b is introduced, thus mak-
principle, are embodied more or less in the preceding ing b equal to 1 2 0 per cent of a. To balance this, the
descriptions. speed should rise, or the main field should strengthen,
Single Phase Commutator Type Induction Motors. or the transformer flux should weaken, or part or all
These are of various types resembling, in many ways, of these may modify in value. Let us assume, as a
the polyphase commutator and the repulsion types rough approximation, that the speed rises 1 0 per cent,
already described. However, an analysis of one the main field is strengthened 5 per cent and the
characteristic type, from the commutation standpoint transformer flux weakened 5 per cent. Then e. m. f.
will be sufficient. The so-called repulsion induction a becomes 1 1 5 per cent, due to the speed and the
type will be considered, as it bears a very close resem- stronger main field, while b with its 2 0 per cent addition
blance in many ways to the repulsion and the doubly- is reduced to 1 1 5 per cent by the weakening of the
fed motors. An illustrative arrangement of such a transformer field, thus balancing these two voltages.
Then d due to the transformer action of the 5 per cent
stronger main pole, becomes 1 0 5 per cent and c due to
1 1 0 per cent speed and 9 5 per cent transformer pole
flux, becomes 1 0 5 per cent also, thus giving balanced
voltages between the rotational and the transformer
actions. Here we have unequal voltages along the two
axes. As far as commutation is concerned, there is
practically a balance in the short-circuited e. m. fs.
under the A set of brushes. However, it is obvious
from the above, that under the Β set the two e. m. fs.
do not balance and the unbalance increases with
departure from synchronous speed.
POLYPHASE SERIES AND SHUNT MOTORS

FIG. 17
a. Series Motor. The polyphase series motor differs
motor is shown in Fig. 1 7 . It has been shown in from the polyphase induction motor with commutator,
connection with Fig. 1 1 and the repulsion type motor, in the fact that the counter e. m. f., opposing the pri-
that at synchronous speed of the armature, with respect mary impressed e. m. f., is not generated in the primary
to the line frequency, there is zero frequency in the winding, but is generated directly by the armature, as
armature winding. The e. m. fs. a and b are balanced in a d-c. series motor. It resembles the single-phase
and so are e. m. fs. c and d, due to transformer action series motor, in that the. total current in the primary
of the main flux and the rotational e. m. f. from the winding becomes magnetizing and the e. m. f. generated
transformer flux. Consequently, under this condition, in the field and armature windings must sum up
another pair of brushes, short-circuited on itself, could vectorially to the line impressed e. m. f. Assuming as
be placed at right angles to the repulsion motor brushes, the simplest case a two-phase series motor, this may be
or in fact, could be placed anywhere. Let us assume considered roughly as two single-phase motors, struc-
them at the right-angle position. turally at right angles, and with a common armature.
In the repulsion motor, at other than synchronous This is illustrated in Fig. 1 8 .
speed, the main field and the transformer fluxes are In the circuits between brushes Αι A two e. m. fs.
h
unequal, or not equivalent, as brought out fully in the are generated, a due to rotation in field A and b due to
former description. Under such conditions e. m. fs. transformer action of field B.
c and d will be unequal and the addition of the second As in the polyphase motor, illustrated in Figs. 1 and
set of brushes, shown in Fig. 14, would result in a con- 2, a can be considered in the nature of an armature
siderable current flow. This will be in such direction counter e. m. f. and b as an impressed e. m. f., in this
and value as will equalize e. m. fs. c and d, or in other case adding to the line e. m. f. However, the addition
words, it equalizes the flux values of the main and trans-of compensating windings on the field structure affect
former poles, so that the motor maintains practically this condition. Such windings are added primarily
synchronous speed over its operative range,—that is, for neutralizing or compensating the armature m. m. f.
it operates very much like an induction motor. Obvi- along the axes Α Αι and ΒιΒ . Diagramatically,
λ λ
around the Β pole or flux there should be compensating windings being shunt (or separately) excited, must
windings in phase with the A phase armature windings, therefore have their exciting e. m. fs. practically 90
but opposing. Practically they would be connected deg. out of phase with their respective armature
in series with the A phase. The effect of the Β flux on rotational or counter e. m. fs., the A poles being
the A phase compensating winding is to generate in it a excited from Β phase, as shown in Fig. 19. With shunt
transformer e. m. f., practically equal and opposite to excitation from constant e. m. f. sources, thefieldfluxes
the transformer e. m. f. b in the armature. This was thus are practically constant, so that such motors tend
referred to briefly in connection with the commutating to maintain constant speed, unlike the series type.
pole series motor, Fig. 16. In consequence, as a The conditions of commutation are so similar to those
resultant,*there is left only the rotational or counter of the series type that further discussion is unnecessary.
e. m. f. a,*as in the series single-phase motor. In the above treatment of the polyphase series and
shunt motors, the e. m. fs. generated by rotation in
Phase A the main and in the transformer fields, and also by
Phase Â
transformer action due to both fields, have been
considered, but there.is no reference to the e. m. f. due
to the armature field corresponding to that in d-c.
commutation. However, as was stated in connection
with polyphase induction motors, this e. m. f. is usually
quite small, compared with the maximum values of
other e. m. fs. short-circuited by the brushes. It is
true that where the rotating and transformer short-
circuited e. m. fs. practically balance each other, as is
the case near synchronism, the e. m. f. due to the
armature field may predominate. However, this is in
the best part of the commutating range.
In general, the low-voltage, one-turn, parallel-type
armature winding and the special proportions used in
FIG. IS
such machines to obtain desirable conditions otherwise,
Phase Β with its poles, armature winding, etc, can all tend toward extremely good commutating charac-
be treated just as phase A ; and thus it is seen that such teristics on the basis of direct-current commutation,—
a motor is quite similar to two superposed single-phase much better than the average commercial d-c machines
compensated motors with their torques at 90 deg.
Thus such a motor will not have the pulsating torque
of the single-phase series motor.
In such a motor the commutating conditions are quite
similar to those of the single-phase motor at standstill
conditions, but are much better as synchronism is
approached, just as in the polyphase induction motor
with commutator. While the transformer e. m. fs. b
and d are opposed by similar e. m. fs. in the respective
compensating windings, so that their resultant effects
between machine terminals are practically zero, yet the
two component e. m. fs. remain in the windings,—so
that in the armature, for instance, a opposes b in the
coils under the Βι Bi brushes and c opposes d under the
A\Ai brushes. As synchronism is approached, these
opposing e.m.fs. in the coils short-circuited by the FIG. 19
brushes more nearly balance each other, just as in the
polyphase induction motor with commutator. The of the same capacity. Moreover, the e. m. f. under
conditions of commutation are therefore similar to the brushes, generated by the armature field, is usually
those in the polyphase commutator motor. out of phase with the other short-circuited e. m. fs., or
b. Shunt Motor. The polyphase shunt type com- their resultant, so that its inclusion would mean but
mutator motor is quite similar to the polyphase series little change in the problem. Therefore, it has been
type, except that the field windings are either shunt or neglected throughout this entire method of treatment.
separately excited, as shown in Fig. 19. The respective It should be noted in connection with the above
armature and compensating circuits (not shown) are series and shunt motors, as well as in all of the preceding
in series and their currents are practically in phase with descriptions, that in splitting up fluxes, m. m. fs., etc.,
their rotational e. m. fs. and field fluxes. The field into certain components, it is only for purposes of
illustration and easy explanation. The actual condi- cuited, regardless of the volts per commutator bar.
tions in the machines themselves may be quite different, From this view-point, as treated, the number of bars
—for example, where there are two or more currents in per brush did not come into the problem. For instance,
a given armature winding, their effects may appear as if a given e. m. f. were short-circuited by the brush,
one resultant m. m. f. of a certain phase relation. The this e. m. f. might be generated in one coil of three
same may hold true for magnetizing, compensating and turns connected to one commutator bar, or in three
other windings, and in place of two or more compon- coils of one turn, each connected to one bar. But a
ents, a single winding may be used in practise. Also little study of the problem from the standpoint of
instead of the main and the transformer fluxes acting losses in the brush contact surfaces, as given below,
independently, at any instant there is simply one shows this is not so, and that the number of bars short-
resultant or combined flux, etc. circuited by a given width of brush may be of consider-
able importance.
At first glance, one would say that the loss due to the
short-circuiting of a given e. m. f. by the brush is simply
a question of the value of the e. m. f. and of the brush
contact resistance, and that the number of bars spanned
is of no consequence as long as the total short-circuit
e. m. f. is the same. However, the following results
show wherein the number of bars may have a very
considerable effect. For example, let us take the case
where an e. m. f. Å is spanned by the brush. If the
brush is the width of one commutator bar, then the
I I J , I I average loss is materially larger than if the brush
spanned two bars of half width with 0.5 Å volts per
FIG. 20 bar. These two cases can be illustrated by Figs. 20 and
21. In Fig. 20, two positions of the brush with respect
In the foregoing treatment of the a-c. commutator to the commutator bars are shown, in one of which the
motors, it should be evident, especially in the case of brush short-circuits no e. m. f., while in the other there
the polyphase motors, that both speed control and the is the maximum short-circuiting of e. m. f. E. The
commutation are dependent upon the relation between loss at the brush contact varies from a minimum of
the transformer and the rotational e. m. fs., the speed
control taking into account the e. m. fs. in the full
winding, whereas the commutation involves only that
part of the e. m. f. short-circuited by the brushes.
This general idea of the relationships of the two charac-
teristics may be expressed in another form,—namely,
speed flexibility in induction motors involves change 1 1 I I1
0.5E
in frequency, which in turn involves the commutation
problem in most of the methods and devices used.
In the foregoing treatment of the a-c. commutator
motor, the e. m. f. short-circuited by the brush has not
been considered in the quantitative sense,—that is, no pfilli,
operative values or limits have been given and the 1 1111 1
subject would certainly not be complete without some 0,5 Å 0.5 Å
data of that nature. The actual limitations of com- FIG. 21

mutation in a-c. as well as d-c. commutating machines,
is fixed principally by the maintenance and life of the zero, to a maximum with the brush midway over two
brushes and commutator. These depend largely upon bars.
the losses due to commutation and the means for pre- In Fig. 21, two positions of the brush are shown, in
venting harmful effects due to such losses. one of which the brush spans just two bars, but short-
circuits only the e. m. f., 0.5 Å, while in the other it is
COMMUTATION LOSSES AND COMMUTATOR WEAR in contact with three bars, and spans a total e. m. f., Å.
It is desirable at this point to consider some of This latter position is that of maximum loss, while the
the losses due to the short-circuited e. m. fs. under first position shows a minimum, but not a zero, loss.
the brushes in a-c. commutating machines. In the As stated before, it would appear, without analysis,
preceding treatment of the rotational and trans- that the average total loss would be practically the
former e. m. fs. short-circuited by the brush, appar- same as in Fig. 21. But calculation shows that the
ently it was a question of total volts short-cir- average total of the first case is 60 per cent greater than
in the second. The general result is indicated in Fig. short-circuited e. m. f. is purposely increased with
3
22. It appears from this loss curve that there is but such resistance in circuit.
little gain, with more than two bars spanned. Also, In all practical a-c. commutator motors the sparking
it usually is thought that the losses would be decreased, and burning action at the brushes appears to be greatest
with the brush spanning less than one bar. This, of at starting, and at the lower speeds during acceleration.
course, is true with a given volts per bar, but is not the This is particularly noticeable under those conditions,
case with a given average voltage short-circuited by the such as railway work, where the torque at start and at
brush, which is the condition where the short-circuit low speed is highest, decreasing with increase in speed
e. m. f. is determined by the primary fields, instead of consequent upon decrease in main field flux. But
the armature fluxes. even with constant torque the sparking appears worse
This question of losses depending upon the number at lower speeds. With those types of a-c. motors with
of bars spanned, may also have a bearing on certain neutralizing e. m. fs. dependent upon speed, the spark-
results noted in d-c. machines. ing and deleterious effects on the commutator and
The losses due to commutation are dependent upon brushes decrease with speed up to fairly high speeds.
the voltages short-circuited and the current values Herein lies one reason why such machines can be made
and distributions at the brush contact and in the leads operative without resistance leads even for quite
frequent acceleration, as experience has shown. It is
because the more severe sparking conditions last for
shorter periods than in the non-commutating-pole
resistance-lead types; otherwise, the brush contact
surface would tend to burn badly and the commutators
would blacken due to such burning action. Herein,
also, is one reason why the commutating-pole type
railway motor, shown in Fig. 16, gives such good brush
and commutating conditions. As stated in the des-
cription of this motor, the correcting effect of the com-
• :
1 2 3
mutating pole can be carried down to quite low speeds
without undesirable effects otherwise.
η
FIG. 22 The real measure of the permissible short-circuited
e. m. fs. must be in the operating characteristics, as they
and coils short-circuited. In ''resistance lead" motors affect the commutator and the brushes. Sparking in
or in small motors with relatively high resistance in the itself is not harmful in either a-c. or d-c. machines,
coils themselves, a considerable portion of the total unless the commutator and brushes show undue
resistance in the short-circuited path may be in such deterioration, or require more than usual attention.
parts, thus relieving the brush contact except where the The so-called "wear" of the commutator copper and
3. T h e relation of average total loss to the a v e r a g e bars the brushes, usually is not true mechanical wear, but is
spanned b y the loss can b e expressed m a t h e m a t i c a l l y as follows : the result of burning at the contact surfaces. All com-
Let e = the e. m . f. per bar. mutating machines are subject to such "wear" in
ç = n u m b e r of c o m m u t a t o r bars spanned. varying degree. Where the burning or wear is at an
Then ne —Å = total e. m . f. spanned b y brush. extremely low rate, the scouring action of the brushes
Let r = resistance of brush c o n t a c t for unit area.
usually will wear down the commutator mica as fast as
W = w i d t h of brush face.
the copper surface burns away, so that the brush keeps
m , , EW / n + 1 \
2 2
in intimate contact with the copper. In those cases
T o t a l loss = —-—( I
12r \· ç / 2
where the rate of burning exceeds the mechanical wear,
the mica will ultimately project above the copper and
ç + 1
2
F o r comparative loss, with ç = 1, then = 2 thus "high mica" will result. As soon as this begins,
?i
even in the slightest degree, the burning action is accentu-
2
n + 12 ated, for the brush makes less intimate contact with

ç =2, — = 1.25 the copper surface, thus further increasing the burning
!
n 2
action. This explains why, in the great majority of

n + 12 cases where commutator burning or undue wear occurs,
ç = 3, — =1.11
ç 2 undercutting the mica makes a very positive improve-
ment,—in many cases apparently entirely eliminating
- ç + 1
2
the commutator wear, due to the better contact
ç = 4, ^ = 1.06
* ç* between the brushes and commutator. This is espe-
cially true in railway motors, both d-c. and a-c. and
modern practise has tended almost exclusively toward
ç= =1.00
ç 2
undercutting the mica in such apparatus. Under-
cutting is particularly beneficial in the former d-c. cient period of scouring action. Starting and accelera-
noncommutating pole types where the short-circuited tion, representing the severest conditions, may be
voltages under the brushes sometimes reached quite frequent in one case and very infrequent in another.
high values. In the same way, in a-c. commutating The whole problem is more or less a question of aver-
motors, where there are often quite high voltages under ages. Unless all the conditions are taken into account,
the brushes, and a relatively high percentage of mica there are many apparent discrepancies.
due to the large number of commutator bars, under- Approximate limits for three operating loads are
cutting of the mica is a practical necessity in all, except, given in the following table:
possibly, quite small machines. Even in well designed
d-c. machines, other than for railway service, where the (a) Load permitting continuous operation without undue deterioration
of the commutator and brushes.
mica represents materially above 20 per cent of the
(b) One-hour loads followed by period of lighter operation.
commutator wearing surface, experience has shown that (c) Momentary or very short period load which would be quite destruc-
undercutting may be needed, even with quite well- tive if continued for any considerable length of time.
proportioned commutating poles. It is obvious, there-

Effective e. m. f. Short-
fore, that the e. m. fs. which can be short-circuited circuited
without harmful effects, will vary greatly, depending Type of apparatus
upon the percentage of mica.

Load (a) Load (b) Load (c)
Experience has shown that the burning of the com-
mutator and brushes is largely dependent upon the
D - C . non-commutating pole traction motor
watts expended at the contact surfaces, as should be (average of 9 sizes) 5 to 7 9 to 11 14 to 17
expected. Many years ago the writer obtained D - C . engine type Generators (non-commu-
tating pole) 5 to 6 7 to 10
valuable results from a long series of tests on the Special d-c. 3rd-brush motor—200 h.p 5 to 6
collector rings of a uni-polar generator, to determine
4
A - C . traction motors—doubly-fed and com-
mutating pole types 5 to 6 12 to 14
the rate of wear due to current and contact conditions. A - C . traction motors—resistance lead, non-
It was found that increased contact drop, as well as commutating pole types 7 . 5 to 9 9 to 1 1 . 5 13 to 15
A - C . industrial motors 5 to 6.5
increased current density, would result in increased Frequency changers 5 to 6 . 5 8 to 10
wear of the rings and brushes, and that the rate of
wear was some function of the actual loss. With a
given current density, the contact drop would vary, The above results were obtained from a compara-
depending upon the formation of a coating or film under tively large number of machines of each class, except
the brushes or on the rings, and if such coating was of in the case of the third-brush machine cited. This
appreciable resistance, the contact drop would increase, special instance is given, for here a very definite e. m. f.
with corresponding increase in the rate of wear. was short-circuited under an active field and good
In consequence of this condition of wear, due to the limiting figures were obtainable.
losses in the contact, the actual current densities in the In the case of the d-c. traction motors, the limits
brush contacts, as well as the e. m. fs. short-circuited, given are on the basis of full-line e. m. f. impressed on
should be taken into account. But as the local or the motor. Under this condition with load (a) the
short-circuited current is a large element in the current commutator would polish quite well, while on load
density in most a-c. commutation, and as this local (b) practically all these motors showed some signs of
current in turn depends upon the e. m. f. short-circuited "smutting" at their commutators and the brushes not
it so happens that the short-circuited e. m. f. in itself infrequently showed evidences of burning. Load (c)
very largely fixes the limiting conditions. Further- for these motors corresponded to 200 per cent of the
more, as it is largely a question of permissible losses, one-hour current rating. However, the assumed con-
the peak value of the short-circuited e. m. f. should dition of full-line voltage on the motor is not a normal
not be used, especially in comparing a-c. and d-c. railway motor operating condition, for, in general, such
limits, but rather some mean value which represents motors handle these excessive currents only during
more nearly the average loss conditions. starting and acceleration,—that is, at reduced speeds
Commutation Limits. It should be evident from with correspondingly reduced e. m. fs. under the
what has already been given that it is quite difficult to brushes.
fix, even approximately, permissible e. m: fs. which It might be asked, here, how it has been possible for
may be short-circuited by the brush, either in a-c. or in the d-c noncommutating pole traction motors to operate
d-c. machines. The grade of brush used, the condition satisfactorily with as high as 10 volts under the brushes,
of the commutator, the duration of the commutating at more or less frequent intervals, and still continue in
period, along with many other conditions, all have a good operating condition. In reply, it may be stated
bearing on the problem. The short-circuited e. m. f. that in traction work, there are, in general, considerable
can be above the harmful point if followed by a suffi- periods of coasting, or light service, in which the scour-
4. Development of a Successful Direct Current, 2000-Kw. Uni- ing action of the brushes on the commutator is suffi-
Polar Generator—TRANS., Á . I. Å . E . , June, 1912. cient to counter-balance the harmful effects of the
heavier periods. Moreover, experience has shown that below the 6.5-volt limit usually made them quite
the effect of burning under the brushes at heavy loads satisfactory.
on these traction motors has been such as to tend to A number of frequency changers have been built for
narrow the effective contact face of the brush, thus operation in the secondary circuits of large inductio ç
automatically lessening the average short-circuited motors for speed control, and fairly definite limits have
e m. f. and the tendency to burn. Also in railway been obtained from these, as indicated in the above
motors which reverse their direction each trip, not table. Some of these machines have shown quite good
uncommonly the brushes tilt slightly in their boxes, commutation at much higher voltages, but after a
depending upon the direction of rotation, so that the considerable period of operation, even without visible
brush contacts show two distinct surfaces, each of half change in the commutation itself, the bruöhes would
the brush width, thus narrowing their effective con- show signs of burning away a part of the contact sur-
tacts, and protecting themselves. face. In such cases, therefore, the real operating
In d-c. generators the extreme conditions of short limits appear to be in the action under the brushes,
circuit were found, in many cases, where the brushes at rather than in the visible commutation. In fact, this
no-load were given the greatest possible forward lead was true in many of the cases which were investigated
in order to take care of the widest range of load without in both d-c. and a-c. machines. The writer has seen
further brush shift. Not infrequently the brushes many machines of both types, which showed sparking at
were shifted into magnetic fields corresponding to five the brushes, but which showed no evidence of serious
to seven volts across the brushes, or even higher, burning under the brushes. Conversely, he has seen
depending upon the amount of sparking which was many cases where the commutation was apparently
considered permissible. Judging from a large number quite good, but in which there were evidences of more
of cases, five to 6 volts short-circuited did not represent or less rapid deterioration. Thus it may be said that
an unduly difficult condition. visible commutation is not always a true measure of
In the a-c. traction motors of the doubly-fed and the permissible short-circuiting action.
commutating-pole types, it has been quite difficult to An interesting feature of the above table is the fairly
obtain any very definite data, for here the short- close agreement between limiting e. m. fs. for the
circuit conditions change with the speed and with the various types of apparatus. While the writer has felt
adjustment of the compensating windings, and there- all along that both a-c. and d-c. commutation were
fore, only average results could be obtained. However, dependent upon approximately the same limits, yet
in the a-c. traction motors of the non-commutating-pole until this paper was undertaken, he had never gone into
type, but with resistance leads, quite definite data was the comparison as fully as covered in the above table.
obtainable. It will be noted from the table that here In former instances, individual cases had been com-
the limiting conditions for continuous operation are pared, but no attempt had been made to analyze any
quite considerably higher than in other types of appara- considerable lines of machines in order to compare the
tus, thus indicating quite clearly that the resistance general results. An interesting fact that developed in
leads, as used, were quite effective in limiting the local this study was that a number of machines, which
currents. On the hour rating, as indicated in column apparently exceeded the above limits, were found, upon
(b), these motors do not show materially better than further investigation, to have proved somewhat
the non-commutating-pole d.c. motors, except that questionable in their operation, after a fairly long period
possibly in some cases the "smutting" was a little less of service.
pronounced. Isolated cases were found, also, which did not fit in
The data in the above table, for a-c. industrial closely with the table, but, as already indicated, no
motors, were compiled from quite a series of small hard and fast rules can be drawn, on account of the large
motors, mostly of variable speed. Here the general number of variables in the problem.
results are also somewhat conflicting, but in practically This paper does not attempt to cover all the charac-
all cases examined, the permissible range of continuous teristics of a-c. commutator motors, but may be con-
operation was within the figures given. A number of sidered simply as a review of certain general conditions
instances were noted where the short-circuit conditions which affect commutation and speed control. Further-
were up to eight or nine volts with apparently fair more, it does not attempt to cover the design of such
commutating conditions, but which eventually showed apparatus. Any wide departure from its present scope
such evidence of deterioration that they were not would lead into the realm of mathematics, with cor-
considered satisfactory. Changing the design of responding reduction in the readability of the article,
such motors to reduce the short-circuit e. m. f. as far as concerns the great majority of engineers.

1 PDF

Uploaded by

Copyright:

Available Formats

1 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 PDF

Uploaded by

Copyright:

Available Formats

The Alternating-Current Commutator Motor

T HE alternating-current commutator motor in some

SINGLE-PHASE COMMUTATOR MOTORS

tion to it, as described in connection with Fig. 10.

method could be used below synchronous speed also,

POLYPHASE SERIES AND SHUNT MOTORS

data of that nature. The actual limitations of com- FIG. 21

n + 12 ated, for the brush makes less intimate contact with

action. This explains why, in the great majority of

proportioned commutating poles. It is obvious, there-

upon the percentage of mica.

You might also like