Introduction to

ELECTRICAL MACHINES
Other Titles in Electrical and Electronic Engineering

G. B. Clayton: Experiments with Operational Amplifiers

G. B. Clayton: Linear Integrated Circuit Applications
J. C. Cluley: Electronic Equipment Reliability
R. F. W. Coates: Modern Communication Systems
C. W. Davidson: Transmission Lines for Communications
W. Gosling: A First Course in Applied Electronics
B. A. Gregory: An Introduction to Electrical Instrumentation
Paul A. Lynn: An Introduction to the Analysis and Processing of Signals
A. G. Martin and F. W. Stephenson: Linear Microelectronic Systems
R. G. Meadows: Electrical Communications-Theory, worked
examples and problems
J. E. Parton and S. J. T. Owen: Applied Electromagnetics
A. Potton: An Introduction to Digital Logic
J. T. Wallmark and L. G. Carlstedt: Field-effect Transistors in Integrated
Circuits
G. Williams: An Introduction to Electrical Circuit Theory
Introduction to
ELECTRICA L
MACHINES

A. R. Daniels
University of Bath

M
MACMILLAN
© A. R. Daniels 1976

All rights reserved. No reproduction, copy or transmission

of this publication may be made without written permission.
No paragraph of this publication may be reproduced, copied
or transmitted save with written permission or in accordance
with the provisions of the Copyright Act 1956 (as amended).
Any person who does any unauthorised act in relation to
this publication may be liable to criminal prosecution and
civil claims for damages.

First published 1976

Reprinted 1978, 1979, 1982, 1985

Published by
Higher and Further Education Division
MACMILLAN PUBLISHERS LTD
Houndmills, Basingstoke, Hampshire RG21 2XS
and London
Companies and representatives
throughout the world
ISBN 978-0-333-19627-4 ISBN 978-1-349-15689-4 (eBook)
DOI I 0.1007/978-1-349-15689-4
Contents

Preface Vlll

1 The Basis of Operation of Electromagnetic Machines

1.1 Basic elements 1
1.2 Production of a rotating m.m.f. 4
1.3 The general machine 6
1.4 Particular cases of the general machine 8
1.5 Circuit conventions 15
1.6 Normalization and per-unit values 18

2 Armature Winding Arrangements 20

2.1 Polyphase winding arrangements 21
2.2 D.C. machine armature windings 25
Tutorial problems 29

3 Process of Energy Conversion 30

3.1 Electromechanical energy conversion 30
3.2 Generation of e.m.f. 38
3.3 Winding factors for a.c. machine windings 39
3.4 Production of torque 40
Tutorial problems 42

4 The Transformer 43
4.1 The ideal single-phase transformer 43
4.2 The practical single-phase transformer 45
4.3 Derivation of equivalent circuits 48
4.4 Determination of transformer parameters 50
4.5 Performance calculations 53
4.6 Voltage regulation 55
4.7 Efficiency 56
4.8 Parallel operation 58
4.9 The waveform of no-load current 61
VI CONTENTS

4.10 3-Phase transformer connections 66

4.11 Parallel operation of 3-phase transformers 70
4.12 The auto-transformer 70
Tutorial problems 71

5 Polyphase Induction Motors 74

5.1 Derivation of equivalent circuits 74
5.2 Power balance equations 77
5.3 Torque calculations 78
5.4 Output power calculations 81
5.5 Determination of circuit parameters 84
5.6 Starting methods for induction motors 87
5.7 Speed control of induction motors 90
5.8 Dynamic operation of induction motors 97
Reference 98
Tutorial problems 99

6 Single-phase Motors 101

6.1 The single-phase induction motor 101
6.2 Locus diagrams for single-phase motors 106
6.3 The shaded-pole motor 108
6.4 The series commutator motor 109
6.5 The repulsion motor 110
Reference Ill
Tutorial problems Ill

7 The Polyphase Synchronous Machine 113

7.1 Basis of operation of cylindrical rotor machines 113
7.2 Equivalent circuits for cylindrical rotor machines 116
7.3 Open-circuit and short-circuit characteristics 118
7.4 Basis of operation of salient pole machines 120
7.5 Determination of synchronous reactances 122
7.6 Voltage regulation of generators 123
7. 7 Paralleling of polyphase synchronous machines 124
7.8 Cylindrical rotor machines on infinite busbars 125
7.9 Salient pole machines on infinite busbars 129
7.I 0 Parallel operation of generators 130
7.II Generator operating charts 131
7.12 Starting of synchronous motors 134
References 135
Tutorial problems 135
 CONTENTS Vll

8 The D.C. Machine 137

8.1 Armature reaction 137
8.2 The commutation process 139
8.3 Generator performance characteristics 141
8.4 The separately excited generator 142
8.5 The shunt generator 143
8.6 Performance characteristics of motors 145
8.7 The shunt motor 146
8.8 The series motor 147
8.9 The compound motor 151
8.10 Speed control of motors 151
8.11 The Ward-Leonard method of speed control 152
8.12 Motor starters 153
8.13 Losses and efficiency 154
8.14 Dynamics of separately excited motors 156
Tutorial problems 160

9 Semiconductor Control of Motors 164

9.1 The thyristor as a power device 164
9.2 Thyristor firing mechanisms 165
9.3 Gate firing 167
9.4 Series and parallel operation 169
9.5 Overcurrent protection 171
9.6 Natural commutation 172
9.7 Forced commutation 176
9.8 Speed control of d.c. motors 177
9.9 Speed control of a.c. motors 181
References 184

Index 185
Preface
The attention given to the study of the performance of electrical machines
has, in recent years, been reduced in many higher-education courses.
Developments have taken place which, in the early stages, emphasize the
similarities, rather than the differences, between the various machine types.
In preparing this text, a basic approach to the formation of operating equa-
tions has been made using the concept of a so-called 'general machine'
whose operation covers all the well-established different machine types. In
this manner, the concepts of synchronous and asynchronous operation
and the action of both a.c. and d.c. commutator motors are introduced at
the earliest possible stage. Since the action of all electrical machines depends
on energy storage in a coupling field, a simplified treatment of electro-
mechanical energy conversion has been included.
Once the common principles governing the operation of all forms of
electrical machine have been established, each different machine type is
dealt with separately in a conventional manner. Modem teaching courses
take place at a rapid pace and the salient facts in the basic analysis of each
machine type are considered first in each chapter. In this way, more advanced
topics in certain chapters can be omitted without disrupting the continuity
of the text. The first section of the book (chapters l-3) considers general
principles and also deals with winding arrangements. The performance of
the transformer and the various machine types is considered in the next
section (chapters 4--8) and the final section (chapter 9) deals with the charac-
teristics and application of the power thyristor when used to control the
speed of both d.c. and a.c. motors.
Throughout the text, distinction between phasor and scalor quantities is
made by the use of bold type to designate phasors. In the particular case of
phasor diagrams which, in any case, always depict phasors, bold type is not
used. A simple circuit convention for a source of energy (generator) and for a
sink of energy (motor) is used throughout the text. This particular conven-
tion is only one of many possible conventions and its only requirement is
that it should be correctly and continuously applied.
I would like to thank the Council of the University of Bath for permission
to use past examination papers. Any dedication made must be to my past
students, all of whom survived after treatment with the material contained
in this textbook.
Vlll
1 . The basis of operation of
electromagnetic machines

1.1. Basic elements

All electrical machines are forms of energy converter and the vast majority
are electrical-mechanical, arranged to convert from electrical to mechanical
energy (motors or sinks) or from mechanical to electrical energy (generators
or sources). There will always be some form of coupling field between the
electrical and mechanical systems and, in most cases, this takes the form of a
magnetic field. The structure on which the field circuit is located may be
either stationary or rotating, depending on the particular form of machine.
In many cases the field will be electromagnetic and field coils carrying the
field current will be wound on a magnetic structure. The iron forming this
structure will be laminated in order to reduce the field iron losses if the field
current is alternating or contains an alternating component. Two different
forms of construction are used for the field circuit. Firstly the so-called
'salient-pole' arrangement can be used, in which field coils are concentrated
and wound around protruding poles. This form of construction is only used
for machines with direct-current (d.c.) field supplies. The second form of
construction is that in which the field coils are distributed in slots cut into a
cylindrical magnetic structure, and this arrangement is commonly used on
certain forms of alternating current (a.c.) generators. It is important to note
at this stage that the field structure can, in general, be physically situated on
either the stationary member (stator) or rotating member (rotor) on the
machine.
The action of all forms of electromagnetic machine is based on two funda-
mentallaws. Firstly Faraday's law, which states that, when there is relative
motion between the magnetic field and a coil situated in that field, there will
be an e.m.f. induced in that coil, whose magnitude is equal to the time rate
of change of flux linkage between the field and the coil. The second law is
based on an observation by Maxwell which states that, if a current-carrying
conductor is placed in a magnetic field, a force will be exerted and relative
motion will take place between the structure carrying the conductor and that
carrying the field. There will be two electrical members of an electrical
machine: the field and the so-called armature in which work is done. In
2 THE BASIS OF OPERATION OF ELECTROMAGNETIC MACHINES

general, the electrical circuit representing the armature of the machine will
consist of coils distributed in slots cut into a cylindrical magnetic structure.
The armature currents will always be alternating and, in these circumstances,
the iron forming the magnetic circuit of the armature will be laminated in
order to reduce the associated iron losses. When there is current in the
armature coils, an m.m.f., known as the m.m.f. of armature reaction, will
be set up. There will be a magnetic flux associated with this m.m.f., which
will combine with the field flux to produce the resultant total flux in the
machine. If it is assumed that the two component fluxes, those of field flux
and armature reaction flux, exist separately, then the speed of rotation of
these two component fluxes in space must be the same if a steady electro-
magnetic torque is to be produced.
In general, either the field or the armature can physically rotate and, for
example, when the armature of the machine is on the rotor, the machine is
known as a rotating armature machine. It is important to note that the basic
laws stated previously explicitly govern the operation of all forms of electro-
magnetic machinery. The form of different rotating machines (a.c. or d.c.,
shunt or series, etc.) is governed by external constraints such as the form of
electrical supply connected to the field circuit or the mechanical method of
connection to the rotating member.
Mechanically the main components of an electrical machine consist of a
stationary element, known as the stator, and a rotating element, known as
the rotor, with an air-gap between them. Special mechanical arrangements
must be provided when electrical connections are to be made to the rotating
member. Such connections are normally made through carbon brushes
bearing on either a so-called slip-ring or a commutator, rotating with the
rotor. A slip-ring is a continuous ring, usually made of brass, and there will
only be one electrical connection to each slip-ring. A commutator is, in
effect, an elegant mechanical switch and consists of hand-drawn copper seg-
ments separated and insulated from each other by mica. A simplified
schematic diagram showing a typical commutator together with the winding
connections to the commutator is given in Fig. 1.1 in both radial and
developed form. The winding considered in Fig. 1.1 is known as a double-
layer winding and consists, basically, of a number of interconnected coils.
Each coil has two sides and in a double-layer winding there are two separate
coil sides in each slot cut into the cylindrical magnetic structure. The par-
ticular rotor shown is cylindrical and has 16 slots carrying a total of 16 coils.
The pitch of the coil in this case is 180 mechanical degrees ("m) so that a coil
is formed between the 'top' of slot I and the 'bottom' of slot 9 and so on.
This arrangement is more readily understood by inspection of the developed
winding diagram of Fig. l.l(b). This type of winding is known as a dis-
tributed, double-layer, lap connected winding and a useful schematic repre-
sentation of this arrangement is given in Fig. l.l(c).
Many forms of electrical machine operate from a 3-phase a.c. supply and
 THE BASIC ELEMENTS 3

the armature winding of such a machine will consist of a 3-phase distributed

winding for which the three separate phase windings (a, b, and c) are wound
in an identical manner. In, for example, the particular case of the 4-pole
machine, 360° mechanical must correspond to 720° electrical since for a
full-pitch coil, one pole-pitch must always equall80°e. Then, in general, for
a machine with p poles:

9
Fig. 1.1. Simplified form of commutator winding:
(a) Radial diagram showing front coil connections.

'BACK' CONNECTIONS

Fig.1.1(b) Developed diagram of complete winding.

4 THE BASIS OF OPERATION OF ELECTROMAGNETIC MACHINES

or ~ electrical radians = I mechanical radian. (1.1)

When 3-phase voltages are applied to a 3-phase winding, an m.m.f. will be

set up and the distribution of the total m.m.f. around the periphery of the
stator is non-sinusoidal but the error involved in assuming a sinusoidal space
distribution of m.m.f. will, in the basic analysis, be small.

Fig.l.l(c) Schematic representation of commutator winding.

1.2. Production of a rotating m.m.f.

In the case when the supply to the machine is a 3-phase balanced a. c. supply
of frequency fs Hz, such that the angular velocity w. = 2nf., it is possible to
produce an m.m.f. which rotates, in synchronism with the supply frequency,
relative to the structure on which the coils are situated. This structure may
be either the stator or rotor of the machine and the speed of rotation of this
rotating m.m.f. in space will then be the algebraic sum of the speed of the
m.m.f. relative to the structure and the mechanical speed of the structure
itself. The stator of the machine can be assumed to be stationary in space and,
in general, the rotor can be assumed to be rotating at some mechanical speed
w rad/s.
Consider a simple 2-pole, 3-phase winding arrangement of which each
phase is assumed to produce a sinusoidal space distribution of m.m.f. The
peak value of the three phase components of m.m.f. will each pulsate
sinusoidally with time between positive and negative maximum values as
the a.c. current in each phase varies sinusoidally with time. The total m.m.f.
produced will be the algebraic sum of the three separate phase values at any
given instant in time. The distribution of the m.m.f. of each phase and that of
the total m.m.f. are shown in Figs. 1.2(a) and (b) for two different instants in
time corresponding to positive maximum values of two of the phase cur-
rents, and this time interval is l20°e. It can be seen from Fig. 1.2 that, during
 PRODUCTION OF A ROTATING m.m.f. 5

A
ic
Instant in time
ib

..

Angle(}

(a)

A
ib

Ia lc

-
Instant in time

.......
' .......
'

...........

(b)

Fig. 1.2. Production of a rotating m.m.f.:

(a) Spatial m.m.f. distribution when the phase 'a ' current is a maximum.
(b) Spatial m.m.f. distribution when the phase 'b ' current is a maximum.
6 THE BASIS OF OPERATION OF ELECTROMAGNETIC MACHINES

this time interval, the total waveform is sinusoidal and has rotated through
120°e of the periphery of the structure carrying the 3-phase winding, i.e. the
wave of total m.m.f. rotates in synchronism with the time variations in cur-
rent. It is also apparent from Fig. 1.2 that the peak amplitude of the total
m.m.f. wave remains constant at 1·5 times that of the phase value. This effect
results in the production of a total m.m.f. rotating in synchronism with the
supply frequency and, in general, for a 3-phase machine with p poles operat-
ing from a 3-phase supply of frequency /Hz, the speed of rotation of the total
m.m.f. is known as the synchronous speed, w. and is given by:
2nf
w. = p/2 rad/s (1.2)

120/
or N. = - - r.p.m. (1.3)
p
The same result can be obtained analytically. For any point p, in Fig. 1.2 at
angle() from the origin, the total m.m.f. F is:
F = Fa sin() + Fb sin(() - 120°) + Fe sin(() - 240°).
But Fa = Fm sin w.t, Fb = Fm sin (w.t - 120°), Fe = Fm sin (w.t - 240°).
Then, on substitution and solution:
3
F = 2 Fm COS(() - W 8 t). (1.4)

Equation (1.4) shows that the total m.m.f. has a constant amplitude, is a
sinusoidal function of the angle() and rotates in synchronism with the supply
frequency.
The corresponding flux density B is given by

B = Jl.oF (1.5)
g
where g is the 'length' of the air-gap with the reluctance of the magnetic
circuit neglected.

1.3. The general machine

It has previously been stated that the principles on which all electro-
magnetic machinery operate are the same and that different machine types
are obtained by the use of different forms of supply and different mechanical
elements. In these circumstances it is useful to introduce the concept of a
so-called 'general machine' and to show that the different machine types are
all particular cases of this general arrangement.
Consider a machine with a 3-phase, 2-pole winding on the stator and a
2-pole lap-connected winding on the rotor connected to both a commutator
 THE GENERAL MACHINE 7

carrying a 3-phase arrangement of brushes, and to three slip-rings with

tappings spaced at l20°e connected to the three separate slip-rings. The
general arrangement is given in Fig. 1.3. When a 3-phase voltage system is
applied to the stator winding, an m.m.f. rotating at synchronous speed ws
with respect to the stator is set up. An m.m.f. rotating at some speed w2 with

)
Fig. 1.3. Arrangement of the basic general machine.

respect to the rotor will be induced in the rotor winding and the rotor itself
will rotate at some mechanical speed w. Then the speed of the rotor m.m.f.
in space is w ± w2 and the speed of the stator m.m.f. in space is W 8 • In order
that a steady torque is produced, the two component m.m.f.s must rotate at
the same speed in space, so that:
(1.6)
Since the winding on the rotor is connected to both a commutator and slip-
rings, it is reasonable to assume that alternating 3-phase voltages will appear
at both. The frequency of the voltage at the slip-rings will be assumed to be
/ 3 Hz such that w 3 = 2nf3 and that at the commutator will be assumed to be
/ 4 Hz such that w 4 = 2nf4 . The basic difference between the action of slip-
rings and that of a commutator is illustrated in Fig. 1.4. The conducting coil
connected to the slip-ring is always connected to the brush, regardless of the
speed w and position of the rotor. Thus the frequency of the voltage at the
slip-ring brush must be equal to that of the rotor m.m.f. relative to the rotor,
so that w 3= w 2 • However, in the case of the commutator, the conducting
coil is only conducting current while it is physically under the commutator
brush, i.e. while it is stationary with respect to the commutator brush. It
follows that the frequency of the voltage at the commutator brush must be
that of the rotor m.m.f. in space, so that w 4 = w ± w 2 •
8 THE BASIS OF OPERATION OF ELECTROMAGNETIC MACHINES

~ instant 1
~ \)

instant 1

~ instant 2
~ instant 2
Slip-ring action Commutator action

Fig. 1.4. Commutator and slip-ring frequencies.

Fig. 1.5. Final arrangement of the general machine.

The arrangement of the so-called general machine can now be redrawn in

the manner shown in Fig. 1.5 in which all the required frequencies have been
written in terms of the mechanical speed of rotation, w, of the rotor and the
angular speed, w., of the m.m.f. set up when polyphase a.c. of frequency
fHz is applied to the stator. Each of the well-known electromagnetic rotating
machines can then be obtained as a particular case of this general machine.

1.4. Particular cases of the general machine

(a) The synchronous machine
The polyphase synchronous machine operates with direct current applied
to the field winding, assumed to be on the rotor and supplied through two
 PARTICULAR CASES OF THE GENERAL MACHINE 9

slip-rings, and polyphase a.c. supplied to the armature, assumed to be on the

stator. This mode of operation is illustrated in Fig. 1.6 and the rotor m.m.f.,
which is obtained from a d.c. source, is stationary with respect to the rotor
structure. Thus under steady-state conditions 0 = w. ± w and a synchron-
ous machine will only operate under steady-state conditions at synchronous

)o=w. + w

Fig.l.6. The general machine as a 3-phase synchronous machine.

speed, as its name implies. It follows that, when the load on a synchronous
machine is to be changed, the machine cannot operate at synchronous speed
during the load transition. Thus in the case of a generator supplying power
to a high capacity, fixed frequency system (infinite busbar) an increase in
the electrical output power of the generator is brought about by increas-
ing the power supplied by the prime mover. The speed of the rotor will
then momentarily increase and the axis of the field m.m.f. will advance on
the axis of the armature m.m.f. by some angle 0. This results in an increase
in electrical output power and the angle () continues to increase until the
output power is equal to the nett input power. At this point, the machine
will lock into synchronism and continue to rotate at synchronous speed until
a further load change takes place. The angle () is known as the load angle {J of
a synchronous machine and there will be a sinusoidal relationship between
power and load angle (Pr:x sin J) for a machine with a cylindrical stator and
rotor construction.
The same general argument can be applied to the operation of the syn-
chronous motor, except that an increase in mechanical load will decrease the
speed of the rotor during the transition period, so that the axis of the field
m.m.f. falls behind that of the stator m.m.f. by the required value of load
angle. When a synchronous motor is stationary, it is not rotating at syn-
chronous speed and cannot, therefore, produce a steady torque. It follows
immediately that a synchronous motor is not self-starting.
When the load on a synchronous machine changes, the load angle must
change from one steady value to another and, during this change, oscillations
10 THE BASIS OF OPERATION OF ELECTROMAGNETIC MACHINES

in the load angle must occur. Such oscillations must be controlled and it is
common practice to fit an additional short-circuited winding, known as a
damper winding, on the field structure, in order to damp out these oscilla-
tions. When the machine is operating under steady-state conditions, at
synchronous speed, there will be zero rate of change of flux linkage with this
damper winding and, therefore, no voltage induced in it. Under these circum-
stances, this winding will have no effect on the steady-state performance of
the synchronous machine.

(b) The direct current machine

Such a machine operates with direct current applied to the field winding
which is generally situated on the salient-pole stator of the machine and
direct current applied to a commutator connected to the armature winding
situated on the cylindrical rotor. This mode of operation is illustrated in
Fig. I. 7 and the operating conditions satisfy those of the general machine.
Thus the d.c. machine will operate under steady-state conditions, whatever
the rotor speed, w, and the d.c. motor will produce starting torque. The
armature current is alternating and, in this particular case, the action of the
commutator is to frequency change the armature current from a frequency
governed by the mechanical speed of rotation to zero frequency at the
commutator brushes.

~0

)o
Fig.J.7. The general machine

Since both the armature and field circuits carry direct current, they can be
connected together in either series or parallel. In the particular case when
the armature and field circuits are connected in parallel, the machine is
known as a shunt machine and the field coils will be wound with a large
number of turns and will carry a relatively small current. When these circuits
are connected in series, the machine is known as a series machine and the
field winding which now carries the full armature current will be wound with
a small number of turns. If the machine has both a series and a shunt field
winding, it is known as a compound machine.
 PARTICULAR CASES OF THE GENERAL MACHINE II

(c) The induction machine

The polyphase induction motor operates with polyphase a.c. applied to
the primary winding, usually on the stator, and carries a short-circuited
secondary winding on the rotor. As its name suggests, the induction machine
has e.m.f. induced in the short-circuited secondary winding by virtue of the
primary rotating m.m.f. and, in many cases, no external electrical connection
need to be made to the secondary circuit. Such a machine is then singly excited
and is of the simplest form of construction with the winding on the rotor
consisting of a series of copper or aluminium bars short-circuited together
at each end and carried in a slotted magnetic structure. This so-called
'squirrel-cage' induction motor is probably the most widely used electro-
magnetic machine.
The mode of operation of the polyphase induction motor is illustrated in
Fig. 1.8 and, once again, the conditions governing its operation are a par-
ticular case of the general machine. It is convenient to introduce the concept
of fractional slip S given by:
s= w.- w (1.7)
w.
at the earliest possible stage. The frequency / 2 Hz of the secondary currents
is given by:

w2
/2 = 2n so that / 2 = Sf (1.8)

wherefHz is the supply frequency. At standstill, w = 0 so that S = I and

= f The machine then acts as a simple transformer with an air-gap and a
/ 2
short-circuited secondary winding. Thus a steady starting torque is produced
and the polyphase induction motor is self-starting. At synchronous speed

Fig. 1.8. The general machine as a 3-phase induction motor.

12 THE BASIS OF OPERATION OF ELECTROMAGNETIC MACHINES

w = w., S = 0 and/2 = 0. Since the rotor carries a polyphase short-circuited

winding, a stationary m.m.f. cannot be produced and it must be concluded
that at synchronous speed the value of the secondary m.m.f. will be zero.
Thus no torque will be produced at synchronous speed and, since any practi-
cal machine must develop torque to overcome friction and windage losses,
the practical polyphase induction motor cannot run at synchronous speed.
The running-light or no-load speed will normally be of the order of 99·5%
of synchronous speed corresponding to a slip of 0·005 and the full-load slip
will be of the order of0·05. Thus the polyphase induction motor is, effectively,
a constant speed machine.
An induction machine can be made to generate if it is driven mechanically
at above synchronous speed, so that w > w. and the slip is negative. If the
machine is driven mechanically in the opposite direction to its primary
rotating m.m.f. then the slip S becomes greater than unity and the machine
acts as a brake. If, for example, the machine is operating normally as a
loaded motor when two of the supply lines to the stator are reversed, the
direction of the rotating m.m.f. will reverse. The rotor will then be rotating
in the opposite direction to that of the rotating m.m.f., the machine will act
as a brake and the speed will rapidly become zero. If, when zero speed has
been reached, the supply is removed from the machine, this reversal of two
supply lines is a useful method of rapidly stopping the motor and is generally
known as plug-braking. If, however, the supply is not removed at zero speed,
the machine will reverse its direction of rotation. The general idealized form
of the torque-speed curve for a polyphase induction motor between rotor
speed limits of - w. :;:; w :;:; 2w., corresponding to a range of slips of
2 :;:; S :;:; -1 is given in Fig. 1.9(a).
It is important to note at this stage that the previous discussion applies
only to machines operating from a polyphase supply. In the particular case
of the single-phase machine, the primary m.m.f. cannot be rotating and pul-
sates in phase with the variations in the single-phase primary current. Any
pulsating m.m.f. can be resolved into two rotating m.m.f.s of equal magni-
tude, rotating in synchronism with the supply frequency but rotating in
opposite directions. Then the slip, SP, of the machine with respect to the
m.m.f. rotating in the same direction as the rotor is given by the usual formula
of (1.7) as:
s = w.- w (1.9)
p w.
and is known as the forward or positive-sequence slip. However, the slip,
Sn, of the machine with respect to the m.m.f. rotating in the opposite direction
to the rotor is given by:
-w- w
sn = •w =2-Sp (1.10)
s
 PARTICULAR CASES OF THE GENERAL MACHINE 13

Torque

c
2

Speed

Fig. 1.9. Torque-speed curl!es for: (II) Polyphase induction motor. (h) Single-phase motor.

and is known as the backward or negative-sequence slip. If these two com-

ponents of m.m.f. are assumed to exist separately, the frequency and mag-
nitude of the component e.m.f.s induced in the rotor by their presence will,
in general, be different since SP i= Sn. Then the machine will, in general,
produce a steady total torque as the algebraic sum of the component torques.
At standstill, however, w = 0 and SP = Sn = l. Thus the component torques
are equal and no starting torque is produced. It follows that the single-phase
induction motor is not self-starting but that it will continue to rotate once
started. Most such machines are, in practice, forms of asymmetrical 2-phase
motors operated from a single-phase source. The form of the torque-speed
curve for the single-phase motor can readily be obtained from that of the
3-phase machine. For any positive slip S = OA in Fig. 1.9(a), the positive-
sequence torque is AB. Then, the negative slip 2 - S = OC such that
OC = OA, and the negative-sequence torque is CD. The resultant torque
is AB - CD and this process can be repeated for a range of values of slip
(1 ~ S ~ 0) to give the form of the single-phase induction motor torque-
speed curve of Fig. l.9(b).

(d) The frequency changer

The general arrangement of the rotating frequency changer is given in
Fig. 1.10 and the stator of such a machine carries no winding. The machine
must be driven at speed w and polyphase voltages are applied to the rotor
slip-rings such that a rotor m.m.f., rotating at speed w. with respect to the
rotor, is set up. Then, under the conditions governing the operation of the
14 THE BASIS OF OPERATION OF ELECTROMAGNETIC MACHINES

general machine, the polyphase voltage appearing at the commutator

brushes will have an angular velocity of w. ± w and frequency changing
action takes place. The magnitude of the commutator voltage can easily
be controlled by variation of the voltage applied to the slip-rings. If the
phase angle of the commutator voltage relative to that at the slip-rings is to
be controlled, then the position of the commutation brushes must be changed.

)w, + w
Fig. 1.10. The general machine as a 3-phasefrequency changer.

It has previously been noted that the polyphase induction motor is,
effectively, a constant speed machine. Speed control of such a machine can
be obtained in various ways and, in the particular case of an induction motor
with a rotor winding brought out through slip-rings, a method very popular
in the past has used a rotating frequency changer. This method, known as
the Leblanc system, is illustrated in Fig. 1.11 where the auxiliary drive for the
frequency changer is the induction motor under control. Since both machines
rotate at the same speed w, it follows that the rotor circuit of the induction
motor can be electrically connected to the commutator output in the manner
shown in Fig. 1.11. If the magnitude of the voltage at the commutator is less
than that at the motor slip-rings, energy will be transferred from the second-
ary of the induction motor, through the frequency changer and back into

)w. + w

)w. +w
Fig. 1.11. Use offrequency changer for induction motor speed control.
 CIRCUIT CONVENTIONS 15

the supply. Under these conditions, the motor will slow down and speed
control below the synchronous speed can be obtained. When, however, the
voltage at the commutator brushes exceeds that at the motor slip-rings,
energy will be transferred from the supply, through the frequency changer
and into the rotor circuit of the induction motor. This provides extra power
for the induction motor and it can then be made to run above its synchronous
speed. This method of control is a particular case of the use of injection of
power into the rotor circuit of a slip-ring induction motor and the general
method is capable of producing variation of the speed of the motor between
wide limits, both above and below synchronous speed. It does, however,
have two major disadvantages. Firstly, electrical connections must be made
to the rotor circuit of the induction motor under control and the major
advantage of this motor, that of no connection to the rotating parts, is lost.
Secondly it involves the use of an auxliliary but expensive commutator
machine. In many cases, however, the main motor and the commutator
machine can be combined into one frame and two particular forms of poly-
phase a.c. commutator motor, the Schrage motor and the doubly fed motor,
are still in relatively wide use.

1.5. Circuit conventions

A form of circuit convention to be used in any text should be established at
an early stage. Various different conventions are in current use and the one
used in this context is only one of many available. This circuit convention will
be used continuously and consistently throughout the main body of this text.
The form of connection diagram representing a sink of electrical energy,
i.e. an electrical motor, to be used throughout this text, is illustrated in
Fig. 1.12, in which the armature current i in the direction specified will have
a component in phase with the voltage rise e across the armature. The
generated e.m.f. e is specified by Faraday's law in the general form:
dA
e =- (1.11)
dt

Fig. 1.12. Schematic representation of a sink of electrical energy.

16 THE BASIS OF OPERATION OF ELECTROMAGNETIC MACHINES

where A. represents the flux linkages. The e.m.f. e acts in the direction shown
so that, in general:
v = e + iz. (1.12)
The torque T. produced by the interaction of field and armature currents in
the machine will result in the angular velocity w and will be opposed by the
load torque TL on the shaft. Under steady-state conditions T. = TL, but
under dynamic conditions when the speed is changing:
dw
T. = Jdt + TL (1.13)

where J is the polar moment of inertia of the rotating parts and dw/dt is the
angular acceleration.
In the case of a source of electrical energy, i.e. an electrical generator,
electrical output energy will be obtained from mechanical input energy. The
simplified form of connection representing a generator, to be used through-
out this text, is given in Fig. 1.13, in which the armature current i in the
direction specified will have a component in phase with the voltage rise e
across the armature. The generated e.m.f. e is, again, specified by (1.11) such
that, for a source:
v = e- iz. (1.14)

Fig. 1.13. Schematic representation of a source of electrical energy.

The torque T. applied to the shaft of the generator will be opposed by the
machine torque E. and, under dynamic conditions:
dw
T. = J dt + T.. (1.15)

In general most electromagnetic machines are reversible in that a motor

can be made to generate or vice versa. However, mechanical drives and
loads are generally irreversible and, in many cases, care must be taken to
ensure that, should this electrical energy reversal occur, damage to the corres-
ponding mechanical system does not take place.
 CIRCUIT CONVENTIONS 17

Analysis of the performance of many electromagnetic machines involves

the concept of mutually coupled circuits, and the so-called 'dot' convention
to be used throughout this text to define the direction of induced voltage in a
mutually coupled circuit is illustrated in Fig. 1.14. The voltage e2 induced
in circuit 2 of Fig. 1.14 by mutual coupling is given in magnitude bye 2 = L 12
(ditfdt) and acts in the direction shown, i.e. a current 'entering' the dot in
one winding will always induce a voltage in the second winding which is
positive towards the dot in the second winding and vice versa. In many
cases, the mutual inductance L 12 is given the symbol M and in a reciprocal
circuit, L 12 = L 21 = M.

• •

r.:
Fig. 1.14. The convention for mutually coupled circuits.

A practical form of a mutually coupled circuit is shown in Fig. 1.15 and is

the equivalent of a sink (circuit 1) and a source (circuit 2). Then for circuit
resistance R and self-inductance L with a mutual inductance L 12 , the
voltage-current relationships can be written in the form:

(sink) (1.16)

(source) (1.17)

where e 1

Fig. 1.15. A practical form of mutual coupling circuit.

18 THE BASIS OF OPERATION OF ELECTROMAGNETIC MACHINES

In order to avoid confusion between instantaneous and steady (mean or

root mean square) values, all instantaneous values will be in lower-case
symbols (v, i, e, etc.) and steady values in capitals (V, I, E, etc.). There is an
equal possibility of confusion between scalar and phasor quantities and all
phasor quantities will be in bold type. Then, for the example V = I(R + jX),
the use of bold type signifies that V and I are phasor quantities having both
magnitude and direction. The previous example can be rewritten in terms of
the complex impedance Z = R + jX, so that V = IZ. Each of the quantities
V, I and Z can be specified in terms of magnitude and direction such that:

V = VLcx, I= ILP, Z = ZLO = R +jX.

Then V = IZ and ex = p + e.
It must be emphasized that the conventions listed are arbitrary. Other
conventions can be used and the only requirement is that they be continu-
ously and correctly applied throughout the text.

1.6 Normalization and per-unit values

In many cases, considerable simplification in calculations can be made by
the use of the technique of normalization, such that the actual quantity is
expressed as a fraction of some arbitrary value of that quantity.
One widely used particular form of normalization is the use of per-unit
(p.u.) values. Most electrical equipment will be specified in terms of its rated
volt-amperes SN and its rated voltage V N· The rated current IN is then
given by IN = SN/VNon a per-phase basis. Then, for an operating voltage V,
current I and power P, per-unit voltage Vpu = V/VN, per-unit current
lpu = 1/IN and per-unit power Ppu = P/SN.
Since equipment will not have a rated impedance the term base impedance
Zb must be introduced such that:

VN VN2
Zb=-=-·
IN SN

Then, for a circuit of impedance z = R + jX:

z R X
Zpu = z' b
Rpu = z'
b
X
pu
=-·
zb
The base admittance Yb is given by Yb = 1/Zb, so that, for a circuit of
admittance Y = G - jB:
y G B
Ypu Y,' Gpu Y,' B =-·
=
b
=
b
pu Yb
 NORMALIZATION AND PER UNIT VALUES 19

In some cases, the impedance will be specified as a percentage value defined

as the impedance voltage drop for rated current, expressed as a percentage
of rated voltage. Then

percentage impedance = IN z x 100 = Zpu x 100

VN
so that conversion from percentage to per-unit values can be made directly.
2. Armature winding arrangements

It has been noted in chapter 1 that machine windings can be classified as

either field windings or armature windings. The forms taken by the field
winding have been discussed in chapter 1, and the present chapter will be
limited to forms of armature winding for both a.c. and d. c. machines. Such
windings will generally be formed by wire wound coils of one or more turns
placed in slots arranged to form either single-layer or double-layer windings.

/
'' / ''
'v /
/
/

---Coils ---Interconnections

(a) Lap (b) Wave

Fig. 2.1. Methods of interconnection of armature coils:

(a) Lap co1111ection. (b) Wave connection.

A single-layer winding is one in which one side of a coil occupies the whole
of one slot whereas a double-layer winding is one in which there are two
separate coil sides in any one slot. Most polyphase a.c. windings and armature
windings for d.c. machines will be double-layer windings with multi-tum
coils. A double-layer winding will, in general, have a lower leakage reactance
and will produce a better waveform than the corresponding single-layer
winding. The separate coils of a winding can be interconnected in several
different manners and the two most general methods of interconnection, lap
and wave connection, are illustrated in Fig. 2.1.
 POLYPHASE WINDING ARRANGEMENTS 21

2.1. Polyphase winding arrangements

The armature winding of an a.c. machine can be considered in terms of
several separate sections or' phase belts'. A 3-phase double-layer winding can
be produced with either three phase belts per pole pair (120° spread), or six
phase belts per pole pair (60° spread). A 2-phase machine will normally have

1 3 4 5 6 7 8 9 10 11 12
A -C C -B -B

A A -C -C B B -A -A C C -B -B

I A
30° F(-B)KD(-C)
..... , .,.., ..,. ~I
Coil pitch C I
t B
180° E(-A)
Fig. 2.2. Forms ofpolyphase windings with full-pitch coils:
(a) 3-phase, 60° spread winding.

a double-layer winding with four phase belts per pole pair (90° spread). The
more usual arrangement is to use 60° spread for 3-phase windings and 90°
spread for 2-phase windings, but it should be noted that a 3-phase, 60° spread
winding is effectively a 6-phase winding and a 2-phase, 90° spread winding is
effectively a 4-phase winding. A developed sectional diagram for a 12 slot,

1 2 3 4 5 6 7 8 9 10 11
A A A A B B B B c c c

-B -B -C -C -C -C -A -A -A -A -B -B

,-(\
Fig. 2.2(b) 3-phase, 12fr spread winding. C B

2-pole arrangement with a coil pitch of 6 slots (180° = slots 1-7) connected
in each of the above forms is shown in Fig. 2.2 and it is apparent that the
3-phase, 60° spread form of connection illustrated in Fig. 2.2(a) will produce
a sequence of ACBACB in the phase belts of the winding for a phase sequence
22 ARMATURE WINDING ARRANGEMENTS

ABC. All the coils shown in Fig. 2.2 are full pitch coils in that the coil pitch
is equal to the pole pitch. The pitch of a coil in a polyphase winding will often
be less than the pole pitch and such an arrangement is known as a short-pitch
or chorded winding. The effect of using chorded coils is to reduce the length

J __ r:.
A A A B B -A -A -A -B -B -B

I. Coil pitch
180°
D(-B) 1
tQ-A)
Fig. 2.2(c) 2-phase, 9ff' spread winding.

of the end connection and to reduce the magnitude of certain harmonics in

the waveform of the generated e.m.f. It is standard practice to use chorded
windings in most a.c. machines and the coil pitch will generally be between
2/3 pole pitch and full pole pitch. The developed diagram corresponding to

2 3 4 6 7 9 10 11 12
A A A A A A B B B B B B

-B -B -A -A -A -A -A -A

J
-B -B -B -B

I. Coil pitch
180°

Fig. 2.2 (d) 2-phase, JB(f' spread winding.

a 12 slot, 2-pole, 3-phase 60° spread arrangement with a coil pitch of 5/6 pole
pitch (150°e = slots 1-6) is illustrated in Fig. 2.3.
Simplified 'clock' diagrams illustrating the arrangement of the phase belts
for one layer of a 3-phase, double-layer 60° spread, 2-pole, 4-pole, and 6-pole
winding are shown in Fig. 2.4, and it can be seen that for a phase-sequence of
ABC, the origin sequence of the points X, Y, and Z displaced by 120°m in
space is ABC for the 2-pole case, ACB for the 4-pole case, and AAA for the
6-pole case. It immediately follows that if entry points displaced by 120om
are made to the phases of a 3-phase machine with a double-layer winding
 POLYPHASE WINDING ARRANGEMENTS 23
with 6K + 2 poles, where K is an integer (i.e., 2, 8, 14, 20, poles) they must be
made in the order of the phase sequence, ABC, and that for a machine with
6K + 4 poles (i.e., 4, 10, 16, 22, poles), they must be made in the order of
reversed phase sequence, ACB. For a machine with 6Kpoles (i.e., 6, 12, 18,

1 2 3 4 5 6 7 8 9 10 11 12
A A -C -C B B -A -A c C -B -B

A -C -C B B -A -A c c -B -B A

I~ Coil pitch
J
5/6 180° = 150°
c,Z~
Fig. 2.3. Form of a 3-phase, 60° spread winding with 5/6 pitch coils.

poles) it is never possible to make entry points displaced by l20°m. In many

practical cases, the points of entry to the three phase windings will be as close
together as possible and are shown as the arrowed lines in Fig. 2.4. In all
cases, the arrangement of the other layer of the double-layer winding will be
identical to that shown in Fig. 2.4, displaced by an angle corresponding to the

(a) (b) (c)

Fig. 2.4. Clock diagrams for a 3-phase, 60° spread winding:
(a) 2-pole arrangement.
(b) 4-pole arrangement.
(c) 6-pole arrangement.

coil pitch and the arrangement of both layers of the winding for the 4-pole
case, shown in Fig. 2.4, when the coil pitch is 5/6 pole pitch (75om) is shown in
Fig. 2.5.
Consider the particular case of a 3-phase machine with a 36 slot stator
having a double-layer winding with a coil pitch of 90om (9 slots). If this
winding is to be arranged in a 3-phase, 4-pole, lap connected form, there will
24 ARMATURE WINDING ARRANGEMENTS

be 36/(3 x 4) = 3 coils/pole/phase (c.p.p. = q). Such an arrangement for

which the c.p.p. is an integer is known as an integral slot winding. Thus each
phase belt of 60°e will occupy 3 slots of the stator. The phase and polarity of

Fig. 2.5. Clock diagram for both layers of a 3-phase, 600 spread, double layer, 4-po/e winding with
5/6 pitch coils.

all 72 coil sides forming this winding is illustrated in Fig. 2.6, in which it is
assumed that the first phase, A, belt occupies slots 1, 2, and 3. It is important
to note that, when the coil side in the bottom of slot 1 is assumed to have a
'positive' polarity it follows that the coil side in the top of slot 10 must have a
'negative' polarity, since a coil is formed between these two coil sides. This
process must be repeated for corresponding coil sides in all other slots. It is
apparent that the four phase belts of phase A must be connected in series,

Fig. 2.6. Phase and polarity of coil sides for a 3-phase, 600 spread, double-layer, 4-po/e winding in
36 slots.
 POLYPHASE WINDING ARRANGEMENTS 25

parallel, or series/parallel to form the phase A winding and the method of

interconnection will depend on the operating voltage of the machine. It is,
however, standard practice to connect pairs of phase belts in any one phase
in series opposition, so that the three coils formed between the bottoms of
slots 1, 2, and 3, and the tops of slots 10, 11, and 12, respectively, in Fig. 2.6,
will be connected in series opposition with the three coils formed between the
bottoms of slots 10, 11, and 12, and the tops of slots 19, 20, and 21 respectively
to form two series connected phase belts. The other two coil groups in phase
A will be connected in an identical manner and it is then possible to connect
the two separate groups of phase belts in phase A in either series or parallel.

2.2. D.c. machine armature windings

These windings are always of the closed type of double-layer lap or wave
form in the forms illustrated in Fig. 2.1 connected to a commutator. The
simplex lap winding will have as many parallel paths as there are poles, while
for the simplex wave winding, there will always be two parallel paths. A
multiplex winding is specified by the number of parallel paths in the winding
relative to that for the corresponding simplex winding.
The arrangement of the interconnections between coils and the method of
connection to the commutator can be specified in terms of the back pitch Yb•
and the front pitchyr of the coil and the commutator pitchyc. The back pitch
of any d.c. armature winding will generally be as nearly as possible equal to
the pole pitch, so that for a P-pole winding in Q slots the back pitch will be
the integer equal to or less than Q/ P. In all cases, the commutator pitch will
equal the algebraic sum of the front pitch and the back pitch of the coil.

Fig. 2.7. Simplified form of lap connected winding for a d.c. machine.
26 ARMATURE WINDING ARRANGEMENTS

The commutator pitch for a simplex lap winding will always be one
commutator segment and the commutator pitch for a multiplex lap winding
of degree m will be m commutator segments. Simplified arrangements of a
lap winding are illustrated in Fig. 2. 7.
The commutator pitch Yc for a simplex wave winding is governed by the
equation

y = s -+ 1 (2.1)
c pole pairs
where Sis the number of commutator segments and a simplex wave winding
can only be obtained if Yc is an integer. In some cases where Yc is not an
integer, a so-called 'dummy coil' is introduced into the winding. This coil
will normally be on open circuit and its function will be to maintain mechani-
cal symmetry of the armature. Simplified arrangements of a wave winding are
illustrated in Fig. 2.8.

~------Yc------~

Fig. 2.8. Simplified form of wave connected winding for a d.c. machine.

Consider a 20-slot armature carrying a 4-pole, double-layer, simplex lap

connected winding. The pole pitch is 90°m and the back pitch of the coil will
be the integer less than or equal to 20/4, i.e., 5 slots. The commutator pitch
will be chosen as + 1 so that the winding is progressive, and it follows that the
front pitch of the coil is given by Yr = -4. The winding diagram in radial
form is shown in Fig. 2.9 and it should be noted that there are four parallel
paths through the armature. The brushes on the commutator will be situated
as nearly as possible on the centre of the poles so that the coil undergoing
commutation is situated in the interpolar gap. As a particular armature coil
rotates, it must pass from the influence of one pole (say, a north pole) to the
influence of an adjacent pole (a south pole), and during this process there
 D.C. MACHINE ARMATURE WINDINGS 27

must be a reversal of current through the coil with an associated short-

circuited position. It follows that this reversal of current should take place
when the coil is situated in the interpolar gap.
It has so far been assumed that the machine is geometrically and magnetic-
ally similar so that the generated voltages in each of the parallel paths of a lap
connected winding will be equal. In practice, these voltages will not have

Fig. 2.9. A double-layer, simplex lap connected, 4-po/e d.c. armature winding.

exactly the same value and circulating currents through the armature and
brushes will exist. In these circumstances, most lap connected machines will
be fitted with 'equalizers', in the form of copper straps with a large cross-
sectional area, which join together points on the winding exactly 360°e
apart. This process of equalizing requires that the number of slots per pole
pair is an integer and the major advantage of equalization is that excessive
brush currents with associated commutation difficulties can be avoided. A
lap connected winding is said to be 100 per cent equalized when all points on
the winding exactly 360°e apart are joined together and, for a simplex
28 ARMATURE WINDING ARRANGEMENTS

winding, the number of equalizer bars is given by the slots per pole pair. For
the case shown in Fig. 2.9, ten equalizer connections would be required, each
ofwhich would join two coils together.
As an example of a simplex wave connected winding, consider a 17-slot
armature with a double-layer winding connected for 4-pole operation. The
commutator pitch y c is given by (17 ± 1)/2 = 8 or 9 and will be taken as 9.
The back pitch of the coil will be the integer less than 17/4, i.e., 4 slots so that
the front pitch of the coil will be 5 slots. The complete arrangement of this
winding in radial form is shown in Fig. 2.1 0.

Fig. 2.10. A double-layer, simplex wave connected, 4-po/e d.c. armature winding.

In many cases, there will be more than two coil sides in any one slot so that
the number of commutator segments will be greater than the number of
armature slots. This has the major advantage that, for a given output voltage,
the voltage between adjacent segments decreases as the number of segments
increases with a consequent reduced risk of flashover between segments.
The main difference between simplex lap and wave connected armature
 D.C. MACHINE ARMATURE WINDINGS 29
winding is the number of parallel paths through the armature. The voltage
developed across each of the parallel paths in a lap connected winding must
equal the system voltage, although the current in each of the parallel paths
will be considerably less than the total armature current. The wave connected
winding has only two parallel paths through the armature so that the current
in each path will be one-half of the total armature current. In practice, the
wave connected winding would generally be used in cases where the total
armature current was less than 500 A and the lap connected winding fitted
with the necessary equalizers, in applications where armature currents
greater than 500 A are necessary.

Tutorial Problems
1. A 72-slot stator is to be wound for 3-phase operation. If the coil pitch is 6 slots,
derive a suitable form for the arrangement as (a) an 8-pole winding, (b) a 10-pole
winding and draw a clock diagram to show the position of the phase belts in each
layer of the winding in each case.

2. A 36-slot stator carries a double-layer winding with a coil pitch of 13 slots and is to
be wound with a 3-phase, 8-pole winding. Derive a suitable winding arrangement
and find its fundamental pitch and distribution factors.
(Answer: 0·985; 0·975)

3. A 48-slot stator carries two identical separate single-layer windings, each with a
coil pitch of 1-12 = 11 slots. Show how such windings can be connected to give
the m.m.f. distribution associated with a 4-pole, 3-phase, double-layer, 60° spread
winding in 48 slots with (a) full pitch coils, (b) 5/6 pitch coils.

4. Draw a developed diagram for a 6-pole, double-layer d.c. armature winding in 28

slots for lap connection. Draw the brushes and the necessary paralleling connec-
tions. The winding should be progressive.
3. Process of energy conversion

The conditions governing the generation of e.m.f. and the production of

torque in a heteropolar electrical machine, in which the coupling field is
electromagnetic, require that there shall be a time rate of change of flux
linkage between the armature conductors and the field flux.
In the case of a d.c. machine, the flux is stationary and the coil moving;
for a normal synchronous machine the flux is rotating and the coil stationary,
while for the polyphase induction motor, the coil moves with respect to a
constant rotating flux. For the ideal case, the field set up in a d.c. machine
will be constant over the pole, while that of an a.c. machine will be sinu-
soidally distributed in space. In practice, however, the presence of space
harmonics in the field flux wave and the armature m.m.f. wave of an a.c.
machine can be of considerable importance but no account of harmonic
effects will be taken in this simple treatment.

3.1. Electromechanical energy conversion

The process of electromechanical energy conversion must always satisfy
the conservation of energy, so that in the case of a sink of electrical energy,
i.e., a motor, it follows that:
Electrical input energy = Mechanical output energy + Energy loss
+ Increase in stored energy. (3.1)
The energy losses associated with this form of energy conversion are the
energy loss because of the resistances of the circuits, the energy loss associated
with the coupling field and the friction and windage loss always associated
with motion. In most cases, the coupling field will be a magnetic field with its
associated small iron losses and this text will be limited to the treatment of
devices of this form. Then the nett electrical input w. can be equated to the
increase in energy stored in the magnetic field We and the mechanical output
w
energy m so that, in incremental form:

(3.2)

The incremental mechanical energy for a virtual displacement dx in time dt

 ELECTROMECHANICAL ENERGY CONVERSION 31

when the force is f can be written:

dWm =Jdx (3.3)
and it follows from (3.2) and (3.3) that:
f= oWe_ oWr. (3.4)
ox ox
The induced e.m.f. e can be written, by Faraday's law, as:
dA.
e=- (3.5)
dt
where A. represents the flux linkages. Then the nett incremental electrical
energy in time dt can be written:

dWe = eidt = (~~}dt =idA..

Then != o(idA.) _ awr.
(3.6)
OX OX
If the displacement occurs with the flux linkages A. held constant it follows
from (3.6) that:

1_ -(oWr) . (3.7)
- OX .l.const

The general form of the relationship between the flux linkages A. and the
current i is that of the normal magnetization curve shown in Fig. 3.1 for

w;
Co-energy

Current i
Fig. 3.1. The magnetization curve.
32 PROCESS OF ENERGY CONVERSION

which the energy stored in the magnetic field is the area shown and is given by:

We= f: idA. (3.8)

The area between the x-axis and the curve must also have the dimensions
of energy and is known as the co-energy We' such that:

We' = I~ Adi (3.9)

and We+ We'= Ai,

so that We= Ai- We'.
Then, from (3.2) and (3.3)
dW. = d(Ai- We') + fdx.
Now dW. =idA

so that f= -A~+ oWe'. (3.10)

ox ox
If the displacement occurs at constant current it follows from (3.10) that

! -(oWe')
- -- . (3.11)
OX icons!

The foregoing analysis has been concerned with force f and linear dis-
placement x. In the case of a rotational transducer, torque T and angular
displacement(} must be introduced and (3.7) and (3.11) can be written:

T = -(oWe) = (oWe') . (3.12)

o(J ). const o(J i const

For the particular case of a linear magnetic circuit, the stored energy is
equal to the co-energy and for a singly excited system is given by:

We = We' = !Ai = t~ (3.13)

where F = Ni is the m.m.f. and ci> is the flux.

The reluctance S of the magnetic circuit is given by F = SCI> so that:

(3.14)

The inductance L is given by A = Li so that:

 ELECfROMECHANICAL ENERGY CONVERSION 33

Then from (3.7)

j-
-
-(oWe)
OX ).const
(3.15)

and from (3.11)

1= (ow;)
OX iconst
= t;2 dL_
dx
(3.16)

For the case of a rotational transducer (3.15) and (3.16) become:

T= -~2dS (3.17)
dO

(3.18)

In the case ofthe simple doubly excited system shown in Fig. 3.2, the total
flux linkage A. 1 with circuit 1 will be given by:
A.l = Luil + Ll2i2
where L 11 is the self-inductance of circuit 1 and L 12 is the mutual inductance
between the two circuits.
In a similar manner, the total flux linkage A. 2 with circuit 2 is given by:
A.2 = L12i1 + L22i2
where L 22 is the self-inductance of circuit 2.

-
i1
"' ........
(' ~
(( i2

_ i
~ 1-1
v 1

-·1
~ ~ ~
~ ~ ~ 1
I'
....... ......
/

Fig. 3.2. A doubly excited electromechanical transducer.

34 PROCESS OF ENERGY CONVERSION

The total energy Wr stored in the coupling field is given by:

Wr = !Lui/ + iL22i2 2 + L12i1i2.
The torque T can be written in the form:

T =
8Wr).
(----a8 =
1. 2
21 1
dL 11
~
1. 2
+ 2z2 dO+
dL22 .. dL 12
l1l2 "de" (3.19)
1 const

The voltage-current relationships for the two circuits can be written as:
. dA.l
Vt = zlrl + dt

and (3.20)

In general, the inductances L 11 , L 22 , and L 12 will be functions of the

angular position 0 of the rotor and the currents will be time functions so that:
dA.l d [ . .
dt = dt L 11 (0)1 1(t) + L 12(0)1 2 (t)]
dA. 1 di 1 . dL 11 dO di 2 . dL 12 dO
or dt = Lu dt + 11 de dt + L 12 dt + 12 ~ dt. (3.21)

In a similar manner:
dA. 2 _ L di 1 . dL 12 dO L di 2 . dL 22 dO
df - 12 dt + It de dt + 22 dt + 12 "de dt. (3.22)

When the values of dA.tfdt and dA. 2/dt from (3.21) and (3.22) are substituted
in (3.20) it follows that:

Vt
.
= [ ltrt di 1 ] +
+ L 11 dt [(· dL
It~
11 dL 12 ) dO]
+ l2. de dt + [ L12 dt
di2 ] (3.23)

and

. r2 + L 22 dt
v2 = [ z2 di2 ] + [(· dL 12
It de + . dL
12
22
~) dO]dt + [ L12 dt
dit] · (3.24)

The first terms on the right-hand side of (3.23) and (3.24) represent the
so-called' voltage of self-impedance', the second terms represent the so-called
'speed voltage', and the third terms represent the so-called 'transformer
voltage'.
In many cases, the self-inductances Ltt and L 22 will not depend on the
angular position of the rotor so that (3.19), (3.23), and (3.24) reduce to:

(3.25)
 ELECfROMECHANICAL ENERGY CONVERSION 35

(3.26)

(3.27)

EXAMPLE 3.1. Many single-phase synchronous machines will be made in the

form illustrated in Fig. 3.3, and such a machine is known as a single-phase

Fig. 3.3. The single-phase reluctance motor.

reluctance motor. The air-gap of the machine is not constant and the reluct-
ance S will be assumed to vary periodically with the angular position 0 of
the rotor between a value Sq when 0 = 90° or 270° and a value Sd when
0 = oo or 180°. The reluctance can be written as:
s -_ sq +2 sd -
sq - sd
2 cos 20 .

The torque Tis given from (3.17) as

- 21lm2 ds __ 1 2
T = "'I"'J! dO -~ (Sq - Sd) sin 20.

If the flux <I> is assumed to be given by <I> = <I>m cos w.t, it follows that
T = --!<l>m2 (cos 2 w.t)(Sq - Sd) sin 20.

But cos 2 w.t = 1{1 + cos 2w.t)

36 PROCESS OF ENERGY CONVERSION

so that T = -~m 2 (Sq - Sd) (sin 2(} + sin 2(} cos 2w.t)
or T = -~m 2 (Sq - Sd) [sin 20 + t sin 2(0 + w.t) + t sin 2(0- w.t)].
If(} = wt - <5, such that <5 is the displacement of the rotor when t = 0, the
torque equation can be written:

T = -~m 2 (Sq- Sd) [sin 2(wt - D) + t sin (2(w + w.)t + 2<5)

+ t sin (2(w - w.)t - 2<5)].
The average value TA of this torque will be zero unless w = w., and then:

TA = <D; 2
(Sq - Sd) sin 2<5. (3.28)

Thus the reluctance motor will only produce steady torque at synchronous
speed and this therefore is a synchronous machine. The angle <5 is the load
angle of the machine and the torque (or power}-load angle characteristic of
the machine will be sinusoidal with maximum torque at <5 = 45°. Such a
torque is known as reluctance torque and is produced by the geometric
asymmetry of the rotor.
EXAMPLE 3.2. The dynamometer type of instrument illustrated in Fig. 3.4
consists of a fixed coil 1, and a moving coil 2, suspended on a pivot acting
against a spring. The self-inductances of the two coils are independent of the
angular displacement (} between them and the mutual inductance L 12
between the coils will be a maximum when (} = oo and zero when (} = 90° so
that:
L12 = M cos 0.

In the linear case the torque is given by:

T =
8We')
(aB i const =
(8----ae
Wr);. const

But Wr = fL 11 i 1 2 + fL 22 i/ + i1i2 M cos(}

then T= -i 1 i2 Msin0.
When the instrument is used as an a.c. ammeter

so that
. - Im 2 M sin(} )
T = - Im M sm (} cos w.t =
2 2
2 (1 + cos 2w.t .
The average value Ta of this torque per cycle of the alternating current will
be given by:
 ELECTROCHEMICAL ENERGY CONVERSION 37

Coil Coil
1 2

Fig. 3.4. The dynamometer instrument.

Thus the deflection of the moving coil will be proportional to (current) 2

for small displacements.
When the instrument is used as an a.c. wattmeter, the system voltage
V1 = V m cos w.t will be applied to the moving coil of resistance R 2 and
negligible inductance and the system current i = Im cos (w.t - ¢) will be
applied to the fixed coil.

Then i1 = Im cos (w.t - c/> ), i2 = ~:cos w.t.

Thus T = V ;m (cos w.t) cos (w.t - cf>)M sin()

= -Vmlm . e[cos cf> - cos (2w.t - 'I'

lR M sm A-)] .

The average value T w of the torque is given by:

Tw = J2J2
Vm Im M .
cos cf> R sm ()

so that the deflection is proportional to power.

38 PROCESS OF ENERGY CONVERSION

3.2. Generation of e.m.f.

In the basic analysis, the winding will be assumed to be concentrated with
full-pitch coils.
Consider the case of a sinusoidally distributed flux, such that the flux
density in the air-gap varies sinusoidally around its periphery. When an
armature coil of N turns is displaced by an angle tpO from the position of
maximum flux density, the flux linkage A. with the coil as a function of time is
given by
A. = NcJ>m sin w.t (3.29)
where cJ>m is the maximum value of flux and w.t = tpe. Then

(3.30)

where the frequency,fhertz, of the generated e.m.f. is given by

2n 1 = w =
" s
~ (p(})·
dt 2
(3.31)

The first term in (3.20) is known as a speed voltage and the term involving
dcJ>.Jdt, which is only present when the maximum value of flux density is a
function of time, is known as a transformer voltage.
Then the speed voltage in (3.30) can be written
e = 2nfNcJ>m cos 2nft (3.32)
and its r.m.s. value E is given by

E = 2nj;tm = 4·44/NcJ>m. (3.33)

For the case of a d.c. machine with a flux per pole cJ>, the total flux cut by
one conductor per revolution will be cJ>p. If the speed of rotation is N r.p.m.
the e.m.f. generated in a single conductor will be given by
cJ>pN
e=6o·
For an armature with z conductors and a parallel paths, the total generated
e.m.f. E is given by
E = f!.. cJ>Nz_ (3.34)
a 60
The angular velocity w rad/s is given by w = 2nN/60 so that
E = pz cJ>w (3.35)
2na
or E = KcJ>w. (3.36)
 WINDING FACTORS FOR A.C. MACHINE WINDINGS 39

3.3. Winding factors for a.c. machine windings

In practice, the coils of an armature winding will not be concentrated but
distributed in space so that the induced voltages in each coil will not be in
phase but will be displaced from each other by the slot angle IX,. The effect of
coil distribution is taken into account by introducing the distribution factor
kd defined as the ratio of the actual resultant e.m.f. E, to the arithmetic sum
of the individual coil e.m.f. E. For a winding with q coils/pole/phase, the
distribution factor is given by
E,
kd = -· (3.37)
qE
The arrangement for the particular case of q = 3 is illustrated in Fig. 3.5,
and it follows from Fig. 3.5 that

E, = 2Xsm
. qiX, d 2 . IX,
2 an E = XsmT
. qiX,
sm-
2
Then kd = - - - · (3.38)
. IX,
qsm 2

where IX, is measured in electrical units. When the winding is a 3-phase, 60°
spread double-layer winding it follows that q~X, = 60ae = i-n. Then
. n
sm-
6 0·5
kd = --- (3.39)
. n . n
q sm 6q qsm-
6q

.~
E

~\
\ I
\\ \x I
\ \ I
X \ ~. 1X
\'4 q~, /

\\7
\~/
Fig. 3.5. Fundamental distribution factor. Fig. 3.6. Fundamental pitch factor.
40 PROCESS OF ENERGY CONVERSION

The expression for the induced voltage given in (3.33) can then be written
(3.40)
When the coil pitch is less than a pole pitch (180oe), it follows that the flux
linking the coil and hence the induced voltage will be less than that for a
full-pitch coil. For the conditions shown in Fig. 3.6 the coil pitch A. is less than
the pole pitch and this effect can be accounted for by introducing the pitch
factor kP such that

. A.
k = SID-·
That is, p 2 (3.41)

The winding factor kw is given by

kw = kdkp (3.42)
and the general form of the induced voltage is
E = 4·44kw<J)mNf (3.43)

3.4. Production of torque

The instantaneous value of the gross output power p 9 from an electrical
motor can be written
(3.44)
where e is the generated e.m.f. and ia the armature current.
For the case of an a.c. motor with a sinusoidally distributed flux, it follows
from (3.32) that
e = (w.Na<J)m cos w.t)kw
where N 3 is the effective number of armature turns in series and kw is the
winding factor.
The armature current ia can be written
i3 = Im. sin (w.t - A.)
where 90 - A. is the phase angle between e and i 3 • Then
P9 = w.(Nakw<J)m)/ma cos w.t sin (w.t - A.). (3.45)
Then the electromagnetic torque produced by the machine is given by
pg pPg
t =- = --
e W 2 W8
 PRODUCTION OF TORQUE 41

so that te = ~ kw(Nalma)(Cl>m) cos wst sin (wst - A). (3.46)

The maximum value of the flux, <l>m, can be written in terms of the maxi-
mum value of the field m.m.f., F 1 , as

G) =
<I>m = B m -dl
2d/ j.l 0 2dl
J.l 0 Hp- = -gp - F1

where d is the mean diameter at the air-gap, I is the length of core, and g is
the air-gap length.
Equation (3.46) then becomes
p J.lo2dl .
/e = - - - kwF1F2 COS W.f SID (W 5 / - A)
2 gp

J.lodl kw -F1F2
=g [ . A) - . AJ. (3.47)
/e 2- SID (2W 5/ - SID

The average value Te is given by

_ _ J1. 0 dlk F1F2 . ,
Te- g (3.48)
w 2 SID A.

The field flux density Bm is given by Bm = J.loF1/g, so that

Te = - ~ kwBmF2 sin A (3.49)

In the case of a d.c. machine the generated e.m.f. E is given from (3. 35) as

E = pz <l>w.
2na

Then (3.50)

It follows that the torque Te is given by

T = p 9 = pz <1>1 (3.51)
e w 2na a

or (3.52)
where K = pz/2na.

The general analysis given here has assumed that the rotating armature
m.m.f. is sinusoidally distributed in space. In practice, the presence of m.m.f.
harmonics must be taken into account.
42 PROCESS OF ENERGY CONVERSION

Tutorial Problems
1. A 6-pole, double-layer, d.c. armature winding in 28 slots has 5 turns per coil. If
the field flux is 0·03 Wb per pole and the speed of the rotor is 1000 r.p.m., find the
value of the induced e.m.f. when the winding is (a) lap connected, (b) wave con-
nected.
(Answer: 140 V; 420 V)

2. A 3-phase, 4-pole, star-connected synchronous machine, driven at 1500 r.p.m. has

a 48-slot stator carrying a 60° spread double-layer winding with 83·3% pitch coils.
There are 10 conductors in each slot. If the maximum value of the sinusoidally
distributed field flux is 0·088 Wb calculate the fundamental distribution and pitch
factors and the phase and line r.m.s. values of the generated e.m.f.
(Answer: 0·958; 0·966; 1450 V; 2510 V)

3. A certain electromechanical device has a magnetism characteristic given approxi-

mately by:
kF
<II= --FWb, k = (b +ex)
a+
where x is the displacement. By considering both energy and co-energy in turn,
show that the force expressions derived from either of these two methods of
approach are identical.
Also find the mechanical work done if xis permitted to change slowly from zero
by an amount b/c with exciting m.m.f. value F.

(Answer: c(F- a log. a : F} bF- a log. a : F)

4. The stator and rotor of a machine each carry a single winding with the rotor short-
circuited. The stator winding has a self inductance of 3 H, the corresponding value
for the rotor is 2 H and the mutual inductance between the windings is given by
2·83 cos () H where () is the angle between the winding axes. All resistances can be
neglected. Derive an expression for the torque as a function of() when the machine
is stationary when the stator current is = 1·41 sin 100 nt. If the rotor is held with
() = 45°e and then released, find the corresponding value of 0.
(Answer: 2 sin 20(1 - cos 200t))

5. A 4-pole, d.c. series motor has a lap connected, double-layer armature winding
with a total of 400 conductors. Calculate the gross torque developed for a flux per
pole of 0·025 Wb and an armature current of 40 A.
(Answer: 63·6 N.m)

6. A 60-slot stator is to be wound with a 3-phase, I 0-pole, double-layer, 60° spread

winding with (a) full-pitch coils, (b) 5/6 pitch coils. Plot the total m.m.f. distribution
in space for the instants in time when the phase A current is a maximum and when
the phase A current is zero.
4. The transformer

The action of a transformer is a particular case of the principle of mutual

inductance, and a transformer consists essentially of two windings, the
primary and secondary on a common magnetic core. A transformer will be
of either core-type or shell-type construction, and typical forms of construc-
tion are shown in Figs. 4.1 and 4.2. It can be seen from Figs. 4.1 and 4.2 that

A High Voltage Winding B Low Voltage Winding

B Low Voltage Winding A High Voltage Winding

Fig. 4.1. Core-type construction. Fig. 4.2. Shell-type construction.

for a core-type construction the primary and secondary windings are wound
as a pair of concentric coils on each limb, whereas for a shell-type construc-
tion the primary and secondary windings form interleaved layers on a single
limb. In all cases, the core will be of laminated construction in order to
reduce iron losses to a minimum.

4.1. The ideal single-phase transformer

When developing the theory of a single-phase transformer it is first
convenient to consider the operation of the so-called 'ideal transformer', for
which all the flux links all the turns on both windings; winding resistance and
44 THE TRANSFORMER

iron losses can be neglected and negligible magnetizing current is required to

set up the flux.
An elementary diagram for an ideal transformer is shown in Fig. 4.3.
When an alternating voltage V 1 volts at frequency f Hz is applied to the
primary windings, a mutual flux «D weber is set up which links both windings,
and primary and secondary e.m.f.s E 1 and E 2 are set up such that

(4.1)

where N 1 and N 2 are, respectively, the primary and secondary turns.

(
,---------...
<II
\
I I
I I
--~(~.~ (4-~~-t---
I ~

v1
( I l I
!P t'IP
~~~~~~ t.~~~~P___
I
\ , ________ __, I
I

Primary Secondary
N 1 turns N 2 turns
Fig. 4.3. Elementary diagram for ideal transformer.

It can be seen from (4.1) that the e.m.f.s E 1 and E 2 are in phase and dis-
placed from CJ) by 90°. The primary applied voltage V 1 is then equal to the
primary e.m.f. E 1 , and the secondary terminal voltage V 2 is equal to and in
phase with the secondary e.m.f. E 2 .
If the mutual flux is assumed to vary sinusoidally with the supply frequency
f Hz, the equation for the e.m.f. E 1 can be written

Em, = N 1 :t Cl>m sin 2nft (4.2)

where Em, is the peak value of the primary e.m.f. and Cl>m is the peak value of
the mutual flux, then
Em, = N12nfci>m COS 2nft.
Thus the r.m.s. value E 1 of the primary e.m.f. will be given by

E1 = -::n1 2nfN1ci>m (4.3)

or
 THE PRACTICAL SINGLE-PHASE TRANSFORMER 45

In a similar manner, the r.m.s. value E 2 of the secondary e.m.f. will be given
by
(4.4)
When a load resulting in a secondary current 12 is connected to the trans-
former, an m.m.f. N 212 is set up. If the primary applied voltage is maintained
constant, the primary e.m.f. and hence the resultant flux <D must be constant.
Thus the secondary m.m.f. N 212 must be balanced by a primary m.m.f. N 1/ 21
set up by a primary current 121 , such that N 212 = N 1/ 21 •

Fig. 4.4. Phasor diagram of ideal transformer.

The primary current 121 is then the primary reflection of the secondary
current 12 and is known as the load component of the primary current or as
the secondary current referred to the primary. A phasor diagram representing
the operation of an ideal transformer can then be drawn as shown in Fig. 4.4.

4.2. The practical single-phase transformer

Consider now a practical transformer with the secondary on open-circuit.
A magnetizing current lm will be required to set up the mutual flux fD and
this current will, under linear conditions, be in phase with the flux «D. A
practical transformer will obviously have iron losses in the core and a core-
loss component of current lc will be required to provide the power associated
with these losses. This current will be in phase with the primary e.m.f. The
no-load current 10 of the transformer will then be the phasor sum of 1m and
lc. A phasor diagram representing this condition is shown in Fig. 4.5.
It has previously been assumed that the windings of the transformer have
no resistance or leakage inductance. Since the windings consist of copper
conductors it immediately follows that both the primary and secondary will
have winding resistance. The total flux linking the separate windings can, for
analytical purposes, be divided into two components, the mutual flux which
links both windings and the leakage flux linking one winding only. The
relevant flux paths are illustrated in Fig. 4.6. The leakage paths are mainly
through air and the effect is to introduce leakage inductances L 1 and £ 2 into
46 THE TRANSFORMER

the primary and secondary, respectively, given in each case by the leakage
flux linkage per unit current. Each leakage inductance will result in a leakage
reactance.

Fig. 4.5. Phasor diagram with secondary on open-circuit.

/'"
Leakage Leakage

Fig. 4.6. Main and leakage flux paths.

Consider now the practical transformer on load. The total primary current
11 must now meet two requirements. It must supply the no-load current 10 of
the transformer and supply a current 121 to counteract the demagnetizing
effect of the secondary load current 12 . The voltage V 1 applied to the primary
winding can then be equated to the phasor sum of the voltage drops in the
primary resistance R 1 and the primary leakage reactance X 1 and the primary
e.m.f. E 1 • The no-load current 10 can be resolved into two components, the
magnetizing current Im in phase with the mutual flux and hence in quadrature
with the e.m.f. and the core-loss current Ic in phase with the e.m.f. In an
equivalent circuit, the effect of the no-load current can be simulated by means
of a shunt branch of a non-inductive conductance gin parallel with an induc-
tive susceptance b. Thus an equivalent circuit representing the primary of a
practical transformer on load is shown in Fig. 4. 7. In the parallel combination
shown, the power E 1 2g accounts for the iron loss and the current Im = E 1 b
 THE PRACTICAL SINGLE-PHASE TRANSFORMER 47

is the magnetizing current. Both g and b are usually determined at rated

primary voltage and frequency and are then assumed to remain constant for
the small departures from rated conditions associated with normal operation.

+ +

Fig. 4.7. Equivalent circuit for primary with secondary on load.

Ideal
transformer
Fig. 4.8. Practical transformer on load: (a) Equivalent circuit.

Secondary Primary

Fig. 4.8(b) Phasor dillgram.

48 THE TRANSFORMER

The transformation ratio is taken into account by introducing an ideal

transformer into the equivalent circuit such that E 1/N1 = E 2 /N2 • The
secondary e.m.f. E 2 can then be equated to the phasor sum of the secondary
terminal voltage V 2 and the voltage drops in the secondary resistance R 2 and
leakage reactance X 2 • An equivalent circuit representing a practical trans-
former on load is shown in Fig. 4.8(a) and the corresponding phasor diagram
in Fig. 4.8(b). It can be seen from Fig. 4.8(a) that the equivalent circuit for a
practical transformer is that of an ideal transformer plus external im-
pedances.

4.3. Derivation of equivalent circuits

It is, in general, not convenient to base calculations on a circuit which
includes an ideal transformer, and the process of removing the ideal trans-
former is known as 'referring' all quantities to one side (either the primary
or secondary) of the transformer.
Consider the secondary voltage equation,
E2 = V2 + lz(R2 + jX2 ). (4.5)
N2 E1
Since E 2 = E 1 - = - · and 12 = NI 21 ,
N1 N
equation (4.5) can be written in the form

~= V2 + N1 21 (R 2 + jX2), or E 1 = NV 2 + 121 (N2 R 2 + j~X2 ).

Thus the equivalent circuit of Fig. 4.8(a) can be redrawn as shown in Fig. 4.9
in which R 21 = N 2 R 2 and X21 = N 2 X 2 are, respectively, the secondary
resistance and leakage reactance referred to the primary, and V 21 = NV 2 is
the secondary terminal voltage referred to the primary. The circuit of Fig. 4.9

Ideal
transformer

Fig. 4.9. Equir~lllent circuit with referred r~alues.

 DERIVATION OF EQUIVALENT CIRCUITS 49
is usually drawn with the ideal transformer omitted as shown in Fig. 4.10,
and this circuit is usually known as the exact equivalent circuit. Various
approximations to the exact equivalent circuit can be made and the usual

Fig. 4.10. Exact equivalent circuit.

approximate equivalent circuit for small transformers is shown in Fig. 4.ll(a).

This circuit differs from the exact equivalent circuit in that the small voltage
drop produced by the no-load current in the primary impedance has been
neglected. For a medium-size power transformer the no-load current can
usually be neglected, and for large power transformers the winding resistance
can also be neglected. The forms of equivalent circuit for such transformers
is shown in Figs. 4.1l(b) and (c).
The complete operating performance of any single-phase transformer can
then be evaluated on the basis of the appropriate equivalent circuit. It is
usual to base calculations on one of the circuits of Fig. 4.11, which have the
advantage that the parameters can be determined by two simple tests, whereas
determination of the component resistances and leakage reactances for use
in an exact equivalent circuit can be difficult.
1 R X
~

(a) (c)
Fig. 4.11. Forms of approximate equivalent circuits:
(a) Small transformer.
(b) Power transformer.
(c) Large power transformer.
50 THE TRANSFORMER

4.4. Determination of transformer parameters

The parameters on the equivalent circuit can, of course, be obtained from
design considerations. There are, however, two simple tests, the open-
circuit and short-circuit tests, which can be carried out to determine the circuit
parameters.

(a) Open-circuit test

The high-voltage winding is on open-circuit and the low-voltage winding
connected to a variable voltage supply, at normal frequency as shown in
Fig. 4.12, in which the primary is assumed to be the low-voltage winding.

Fig. 4.11. Connections for open-circuit test.

The input current and power and the voltage across the open-circuited wind-
ing are measured for a range of applied voltages up to 125% of the rated
voltage. Then, at rated voltage,
Iron loss Pi = Pg, i.e., g = PJP.
Open-circuit admittance y = 1/V.
Magnetizing susceptance b = J(y 2 - g 2 ).
In these tests the small copper loss in the primary resistance has been
neglected.
If this test is performed at two different frequencies/a andfb and the input
power plotted to a base of input voltage divided by frequency, Vjf, as shown
in Fig. 4.13, the iron losses can be separated into the component hysteresis
and eddy current losses. At a given flux density, hysteresis loss is proportional
to frequency and eddy current loss is proportional to the square offrequency.
Now fixed flux density corresponds to a fixed voltage-frequency ratio and for
the ordinate LMN in Fig. 4.13,
Pa = Afa + Bf.. 2 at frequency fa,
where Pa and Pb are the total iron losses at frequencies/a and/b respectively
and A and B are constants.
 DETERMINATION OF TRANSFORMER PARAMETERS 51

The constants A and B can now be determined and the component iron
losses at any frequency for this fixed value of flux density can be determined.
By considering several different values of flux density, the component iron
losses at any value of flux density can be found.

§:l':l P.- - - - - - - -
.9
1::

·=3
0

Voltage/frequency
Fig. 4.13. Separation of iron losses.

(b) Short-circuit test

The low-voltage winding is short-circuited through an ammeter and the
high-voltage winding is connected to a variable voltage supply at normal
frequency as shown in Fig. 4.14 in which it is assumed that the primary is the
high-voltage winding. The input current and power and the voltage across the
high-voltage winding are measured for a range of short-circuit currents up to
125 per cent of the rated current. The input current is made up of the reflection
of the short-circuited current and the no-load current and the input power is
made up of copper loss, iron loss, and dielectric loss.
Copper loss Pc = P R and Total resistance R = PJP..
Short-circuit impedance Z = V/ I.
Total leakage reactance X = .j(Z2 - R 2 ).
Thus the total resistance and total leakage reactance referred to the high-
voltage winding can be found.

Fig. 4.14. Connections for short-circuit test.

52 THE TRANSFORMER

EXAMPLE 4.1. The results of the open-circuit and short-circuit tests of a

230/100 V 10 k VA single-phase transformer are as follows:
Primary Secondary
Open-circuit test Open-circuit 100 V, 6 A, 154 W
Short-circuit test 18 V, 43·5 A, 240 W Short-circuit
Determine the parameters of the approximate equivalent circuit of Fig.
4.ll(a) and express these quantities in per-unit values.

Solution: From short-circuit test

p 240
R = R1 + R21 = p = 43 _52 = 0·127 Q

v 18_ = 0·4I4n.
z =I= 43 5

Then x = xl + X21 = ~(Z2 - R 2) = 0·394 n.

Since the open-circuit test was performed with the primary on open-circuit,
the results of this test must be referred to the primary if parameters applicable
to an equivalent circuit referred to the primary are to be obtained. Then
results of open-circuit test become
230 100
100 X 100 = 230 V, 6 X 230 = 2·61 A, 154 w.

p 154
Thus g = y2 = 2302 = 0·00291 mho

I 2·61
y = - = - = 0·01136mho
v 230 '
Then the approximate equivalent circuit can be drawn. Now,
z V2 v 230 2
Zpu = zb and Zb = I: = (V~), = 104 = 5·28.

X 0·394
Then Rpu = !!:.._ = 0 "127 = 0·024 Xpu = zb = 5 _28 = 0·0784.
zb 5·28 '
I, 1
Similarly, Yb = - = - = 0·189.
v, zb
Then

= _f!__ = 0·00291 = 0·0 154 b = !!_ = 0·01094 = 0·058

Bpu Yb 0·189 ' pu Yb 0·189 •
 PERFORMANCE CALCULATIONS 53

4.5. Performance calculations

It should be noted at this stage that, in practice, the primary will be
supplied at rated voltage and a particular load current at a particular power
factor will be supplied from the secondary. If it is assumed that the parameters
of the equivalent circuit are known, calculations based on the approximate
equivalent circuit of Fig. 4.11 (a) can become complicated. A phasor diagram
corresponding to the circuit of Fig. 4.11(a) is shown in Fig. 4.15 in which the
quantities V t> R, X, g, b, I 21 , and c/J are assumed to be known. It should be
noted, at this stage, that Fig. 4.15 is drawn for a lagging power-factor and

Fig. 4.15. Phasor diagram for approximate equivalent circuit.

under these conditions the power angle c/J is assumed to be negative. Then by a
simple application of Pythagoras's theorem

That is,
V/ = ~1 + 2/21 V21 (R cos c/J - X sin c/J) + l~ 1 (R 2 + XZ).
Now R = Z cos (} and X = Z sin (}
where Z 2 = R 2+ X 2 , tan (} = X/ R. Thus
Vi = ~1 + 2121ZV21 (cos(} cos c/J - sin(} sin c/J) + /~ 1 Z2 ,
or V1 2 = ~1 + 2/21zv21 cos((}+ c/J) + /~1Z 2 (4.6)
where c/J is negative for lagging power factors.
This is a quadratic equation which can be solved for V21 . The phasor
diagram can then be drawn and the input current I 1 and power-factor cos c/J 1
found.
Some simplification can be made when the angle between V 1 and V21 is
small and (4.6) can be modified to the approximate volt drop equation.
Referring to Fig. 4.15, OC = V1 and, if il( is small, OC = OE. Now
54 THE TRANSFORMER

OE = OA + AF + FE. But OA = V2 ~> AF = 121 R cos(-</>),

FE = 121 X sin (-</>).Then
V1 = V21 + / 21 (R cos</> - X sin</>),
or V21 = V1 - 121 (R cos</> - X sin</>).
Thus V21 = vl - I21z cos (O + </>) (4.7)
where </> is negative for lagging power-factors.
Both (4.6) and (4.7) can be rewritten directly in terms of per-unit (p.u.)
quantities. In particular, if the transformer is operating on full load
(/21 pu = I) with rated primary voltage (V21 pu = 1), (4.6) can be written as

Then
2V21 pu = -2Zpu cos (0 + </>) ± [4z;u cos 2 (0 + </>)- 4(z;u- l)]t,
that is,
V2tpu = -Zpu cos (0 + </>) ± [I - z;u sin 2 (0 + </>)]t.
The only practical solution is
V21 pu = [I - Z~u sin 2 (0 + </>)]t - Zpu cos (0 + </>). (4.8)
In a similar manner, (4.7) can be written
V21 pu = 1 - Zpu cos (0 + </>). (4.9)
Comparison of (4.8) and (4.9) shows that the approximate volt drop equation
is exact when z;u sin 2 (0 + </>) = 0. This can only be true in practice when
0 + </> = 0 or 180°.

EXAMPLE 4.2.A 10 kV A, 2000/400 V, single-phase transformer has resistances

and leakage reactances as follows:
R 1 = 5·2 n; X 1 = 12·5 n; R2 = 0·2 n; X 2 = 0·5 n.
Determine the value of the secondary terminal voltage when the transformer
is operating with rated primary voltage with the secondary current at its
rated value with power-factor 0·8 lag. The no-load current can be neglected.

Solution: The parameters of the equivalent circuit referred to the primary are

R = R1 + N 2R 2 = 5·2 + 52 x 0·2 = 10·2 Q

x = X1 + ~X2 = 12·5 + 52 x 0·5 = 25 n.
 VOLTAGE REGULATION 55

Then
z = .j(R2 + _x2) = .j(I0·2 2 + 25 2 ) = 21 n,
1 X 1 25 o
()=tan- -=tan- - = 67·8
R 10·2 .
Now
vr vr 2000 z 27
s:
2 2
zb = /r = = 10 103 = 400 n,
X Zpu = zb = 400 = 0·0675.

Given V1 pu = 1 and /pu = 1, cos <P = 0·8lag, and <P = -36·8°, (4.8) can
be written as
v21 pu = [1 - 0·0675 2 sin 2 (67·8 - 36·8)Jt - 0·0675 cos (67·8 - 36·8)
= 0·9415.
Secondary voltage v2 = 0·9415 X 400 = 376·6 v.
If the approximate voltage drop equation (4.9) is used,
V21 = 1 - 0·0675 cos (67·8 - 36·8) = 0·9421 v.
Secondary voltage v2 = 0·9421 X 400 = 376·8 v.
It can immediately be seen that, in this case, the error involved in using the
approximate equation is negligible.

4.6. Voltage regulation

The voltage regulation is defined as the rise in the secondary terminal
voltage when full load is removed with constant primary applied voltage.
Then, from (4.9)
p.u. regulation= 1 - V21 pu = 1 + Zpu cos(()+ </J)
- [1 - z;u sin2 (() + </J)]t (4.10)
and, from the approximate volt drop, equation (4.9),
p.u.regulation = 1- V21 pu = Zpucos(() +4J). (4.11)
Conditions for maximum and zero regulation can be obtained directly
from (4.10) and (4.11 ).
From (4.10), zero regulation occurs when
0 = 1 + Zpu cos (8 + </J) - [1 - Z~u sin 2 (8 + </J)]t,
i.e.,

or z;u + 2Zpu cos (8 + </J) = 0.

56 THE TRANSFORMER

Hence Zpu = 0 or cos (0 + c/J) = - ZPU.

2 (4.12)

In a similar manner, from the approximate equation (4.11), the regulation

will be zero when Zpu cos (0 + c/J) = 0, i.e.,
Zpu = 0 or cos (0 + c/J) = 0. (4.13)
The first condition (Zpu = 0) in (4.12) and (4.13) will not be considered
further. Comparison of the second condition in (4.12) and (4.13) shows that,
if Zpu is relatively large, considerable error can be involved in using the
approximate voltage drop equation for the calculation of the condition for
zero regulation.
Maximum regulation will always occur at large values of the angle cjJ and
under these conditions the approximate volt drop equation can be used.
Then, from (4.11) directly, maximum regulation will occur when
cos(O + c/J) = 1. (4.14)
Thus maximum regulation will occur when the lagging power angle cjJ is
numerically equal to the impedance angle 0 of the transformer.

4.7. Efficiency
Since the transformer is a static piece of apparatus, the losses are limited
to:
(1) Copper loss in the resistance of the windings which is variable with load
current.
(2) Iron loss in the core made up of component hysteresis and eddy current
losses which are usually considered to be constant.
(3) Stray loss produced by stray flux producing eddy current losses in the
conductors.
(4) Dielectric loss in the insulating material which is appreciable only in the
particular case of high-voltage transformers.
In general, efficiency can be written in terms of any two of input power,
output power, and losses. That is,
.
Effictency Output Output
= I
nput Output + Losses
For the approximate equivalent circuit of Fig. 4.ll(a)

.
Effictency v21pu/21pucoscjJ (415)
p.u. = -r2 2 · ·
V21pu/21pucoscjJ + L2lpuRpu + Vlpugpu
When (4.15) is differentiated with respect to the load current / 21 pu and the
resulting equation equated to zero, it follows that maximum efficiency occurs
when
(4.16)
 EFFICIENCY 57

i.e., when the variable loss equals the fixed loss. Then, with the primary
supplied at rated voltage,
.
Maxtmum effi ctency
. p. u. v21 pu/21 pu cos <P
= -=----'-'---"""--,---,--- (4.17)
V21Pul21Pu COS </J +
2gpu

The maximum efficiency for any power-factor occurs at the same load,
and the highest possible efficiency occurs at unity power-factor.
A transformer which is to operate continuously on full load will be de-
signed to have maximum efficiency on full load. In the case of transformers
used for distribution purposes which are to operate for long periods on light
load, the point of maximum efficiency would be arranged to be between one-
half and three-quarters of full load.
If it is required to measure the efficiency of a transformer on load, it is
normally inconvenient to dissipate the rated output of the transformer in

Single-phase
supply 1

Fig. 4.16. Back-to-hack test.

load banks, and a method of testing known as 'phantom load' testing can be
used. For the particular case of the transformer, this method is known as the
Back-to-Back or Sumpner Test and requires two identical transformers. The
circuit used is shown in Fig. 4.16 in which the primary windings are connected
in parallel and supplied at normal voltage and frequency, and the secondary
windings are connected in series opposition and supplied through a variable
58 THE TRANSFORMER

voltage regulator. Then, as the voltage applied to the secondary winding is

increased, the phantom load on the transformer increases. Under these
conditions, the wattmeter W1 in the primary circuit will record the total iron
loss and the wattmeter W 2 in the secondary circuit will record the total
copper loss. If facilities for temperature measurement are included, a heat
run can also be performed and the final temperature rise of the transformer
on load can be obtained.

4.8. Parallel operation

Transformers are said to be connected in parallel when their primary
windings are connected to a common voltage supply and their secondary
windings are connected to a common load. If the two transformers are to
operate satisfactorily in parallel, the turns ratio of the two transformers
should be the same and only points at the same potential should be joined
together.

Fig. 4.17. Equivalent circuit for parallel operation.

An equivalent circuit, with all quantities referred to the secondary,

representing the parallel operation of two transformers, A and B, is shown in
Fig. 4.17. Then, by inspection,
v= EA - IAZA = EB - IBZB (4.18)
and lA + IB = IL. (4.19)
Consider the case of parallel operation of two transformers A and B with
equal open-circuit voltages so that, in equation (4.18), EA = E8 .
Then, from (4.18)
IAZA = IBZB.
Ifl8 is eliminated between (4.18) and (4.19) it follows that
ZB
(4.20)
lA = IL ZA + ZB = IL I + ZAfZB
 PARALLEL OPERATION 59

In a similar manner
ZA
Is= IL = IL---- (4.21)
ZA + Z8 1 + Z 8 /ZA
If the load is specified in terms of k VA and not current, (4.20) and (4.21)
can be written

(4.22)

and (4.23)

where ST = VILis the total load kVA, SA is the kVA of transformer A, and
S 8 is the kV A of transformer B.
It is often convenient to specify the percentage or per-unit values of resist-
ance and leakage reactance for a transformer, and in these circumstances
(4.20) to (4.23) do not, in general, apply directly.
It has previously been shown that per-unit impedance Zpu is given by
Zpu = Z/Zt,, where Zb is a base impedance. Then

Z = Zpuzb = Zpu- =
Vr zpu-'
V/
/r Sr

where Vr is the rated voltage and Sr is the rated kVA. Then

ZA ZApu V:A Srs (4.24)

Zs = ZBpu v;B srA 0

It has previously been assumed that the open-circuit voltages are equal
and under these conditions (4.24) reduces to

ZA
- - -srs
= -ZApu -· (4.25)
Zs ZBpu srA
It is immediately apparent from (4.25) that, if the transformers are to share
the load in proportion to their ratings, their per-unit impedances must have
the same magnitude and that, if the transformers are to operate at the same
power-factor, their per-unit impedances must have the same phase angle.
When the load is specified in kVA some difficulty can arise in the calcula-
tion of the terminal voltage and the technique applicable to the solution of this
problem is best illustrated by means of a numerical example.
EXAMPLE4.3. A 500 kVA, 33/3·3 kV single-phasetransformerwitharesistance
voltage drop of 1·5% and a reactance voltage drop of 6% is connected in
parallel with a 1000 kVA, 33/3·3 kV single-phase transformer with a resist-
ance voltage drop of 1% and a reactance voltage drop of 6·2%. Find the kVA
60 THE TRANSFORMER

loading and operating power-factor of each transformer when the total load
is 1200 kVA at power-factor 0·8 lagging. If the transformer primaries are
connected to constant frequency, 33 kV infinite busbars, calculate the load
terminal voltage.

Solution: Transformer A, 33/3·3 kV, 500 kVA;

Rpu = 0·015, xpu = 0·06. :. ZApu = 0·06185/76°.

Transformer B, 33/3·3 kV, 1000 kVA;

Rpu = 0·01, xpu = 0·062. :. ZBpu = 0·06201/80·8°.

Now from (4.22),

From (4.25),
ZA _ ZApu SBr _ 0·06185M 1000
ZB - ZBpu SAr - 0·062/80·8° 500'
= 1·995\4·8° = 1·988- j0·167.

Hence, 1 + ~; = 2·988 - j0·167 = 2·993\3·2°.

1200\36·8
Then, from (4.22)
0

SA = 2·993\3-2" = 402\33·6 .

Then SA= 403 kVA, cos ¢A= cos 33·6 = 0·833lag.

Now from (4.23) -

S B- ST
1 + ZB/ZA
then ZB = 1·502 + j0·0422 = 1·503~
ZA

and = ST = 1200'\36-8 = 798\38·4.

SB I + ZJZA 1·503\Rl
Then SB = 798 kVA, cos ¢B = cos 38·4° = 0·784.
If the transformers are supplied at rated primary voltage the approximate
voltage drop equation can be written as
1 = v21pu + lpuzpu cos((}-¢).
Then, for transformer A
1 = Vpu +. /Apu X 0·06185 COS (76 - 33·6°).
This equation contains two unknowns, but if both sides are multiplied by
 THE WAVEFORM OF NO-LOAD CURRENT 61

V pu the equation becomes

vpu = v;u + VpJApu X 0·06185 cos 32·4°.
Now
SA 403
vpu/Apu = SAr = 500 = 0·804.

Hence v;u - vpu + 0·804 X 0·06185 cos 32·4° = 0. This is a quadratic in

V pu and can be solved to give V pu = 0·996 or 0·004. The value of V pu = 0·004

is obviously absurd and the load voltage VL is given by
VL = 3300 X 0·996 = 3287 v.

4.9. The waveform of no-load current

It has previously been assumed that when the mutual flux and hence the
e.m.f.s vary sinusoidally, the no-load current of the transformer will also vary
sinusoidally. It is, however, common practice in transformer design to use
high core flux densities which will, in general, lead to saturation of the core
and a non-linear magnetization curve for the transformer. It immediately
follows that if the flux is assumed to vary sinusoidally, the magnetizing
current cannot vary sinusoidally. If, at this stage hysteresis loss is neglected,
the waveform of the magnetizing current can be obtained if the magnetization
curve for the transformer core is known and the method is best illustrated by
means of a numerical example.
Consider a 50 Hz single-phase transformer with a 200 V primary winding
with 260 turns. The effective cross-sectional area of the core is 2·5 x 10- 3 m 2
and the mean length is 0·65 m. The magnetization curve is given in Fig. 4.18.
Now from the e.m.f. equation for a transformer

E = 4·44./N<I>m

and, if the primary leakage impedance is neglected on no-load,

200 = 4-44 X 50 X 260 X Bm X 2·5 X 10- 3 so that Bm = 1-4 Wbjm 2 •

If the flux is assumed to vary sinusoidally it follows that the instantaneous

value b of the flux density is given by
b = Bm sin 2nft = 1·4 sin 314·2t = 1·4 sin 8.

Then the corresponding instantaneous values in the magnetizing current

can be obtained by the method shown in Table 4.1. For a particular instant
in time, the instantaneous flux density can be found and the corresponding
value of magnetizing force obtained from the magnetization curve plotted in
Fig. 4.18. Thus the magnetizing current i 0 can be obtained from the value of h.
62 THE TRANSFORMER

TABLE 4.1

Time Flux density Magnetizing force Magnetizing current

(0°e) (b= 1·4 sin 0) (h) (io = Hl/N)

0 0 0 0
15 0·366 80 0·2
30 0·7 120 0·3
45 0·99 200 0·5
60 1·21 500 1·4
75 1·35 1072 2·68
90 1-4 1452 3·63

1·6

0·4

0·2

oL---~~--~r---~----~----~-----r---
500 1000 1500 2000 2500 3000
Magnetizing force h (AT/m)

Fig. 4.18. Magnetization curve.

 THE WAVEFORM OF NO-LOAD CURRENT 63

Since the waveform is symmetrical about (} = 90°, values of current need

only be obtained for one-quarter of a cycle. The waveform of flux density and
magnetizing current are shown over one-half cycle in Fig. 4.19 and it is
immediately apparent that the waveform of magnetizing current is far from
sinusoidal.
If the hysteresis loop for the transformer is known, it is possible to de-
termine a more accurate form for the waveform of the no-load current, since
the component of the no-load current required to set up the flux and to supply
the hysteresis loss can be found in a manner similar to that used to determine
the magnetizing current from the magnetization curve. Then Table 4.2 can

2{) 4

1·5 3 $
~
._e
"'E
~
~ 1·0
.c
·;;;
1=:
<U
"0
>< 0·5
&:
0
30 60 90 120 150 180
Angle in degrees electrical
Fig. 4.19. Waveforms offlux density and magnetizing current.

TABLE 4.2

Time Flux density Magnetizing force Current

(O"e) (b = 1·4 sin 8) (h) (i)

0 0 170 0·425
15 0·366 230 0·575
30 0·7 290 0·725
45 0·99 400 1·0
60 1·21 672 1·68
75 1·35 llOO 2·75
90 1·4 1500 3·75
105 1·35 852 2·13
120 1·21 332 0·83
135 0·99 104 0·26
150 0·7 -40 -0·1
165 0·366 -104 -0·26
180 0 -172 -0·43
64 THE TRANSFORMER

be drawn up in exactly the same manner as Table 4.1 for the hysteresis loop
given in Fig. 4.20.
Values of current must be obtained for a range of() from 0 to 180° and the
current obtained by this method is the sum of the magnetizing current im and
the hysteresis loss component ih of the core loss component of current. This
waveform is shown in Fig. 4.21, and if a sinusoidal component of current to
take account of the eddy current component ie of the core loss component of
current is added to this waveform, the no-load current i 0 can be drawn. The

1·6

1500 1000 500 0 500 1000 1500

Magnetizing force h (AT/m)

1·6

Fig. 4.20. Hysteresis loop.

 THE WAVEFORM OF NO-LOAD CURRENT 65

waveform of the no-load current i 0 obtained by this method is also shown in

Fig. 4.21. It can then be said that, to a close degree of accuracy, the no-load
current can be subdivided into a sinusoidal waveform of current which
provides for the core loss in the transformer, and a non-sinusoidal waveform
of magnetizing current. Thus the waveform of the no-load current can be
seen to be a periodic function of time of period 360° electrical. As such it can
be resolved into a family of sinusoidally varying quantities of different
frequencies. An analysis of the waveform shown in Fig. 4.21 gives the result

Fig. 4.2i. Waveform of no-load current.

66 THE TRANSFORMER

No-load current i 0 = 2·05 sin 0 + 0·85 cos 0 - 1·02 sin 30

+ 0·48 sin 50 - 0·2 sin 70 +
Then
Core loss component of no-load current, ic = 0·85 cos 0
Magnetizing current, im = 2·05 sin 0 - 1·024 sin 30
+ 0·48 sin 50 - 0·2 sin 70 +
The third harmonic component of no-load current has an r.m.s. value
which is equal to 41% of the r.m.s. value of the total no-load current and the
fifth harmonic component of no-load current has an r.m.s. value which is
equal to 19% of the r.m.s. value of the total no-load current. Thus in problems
directly associated with no-load conditions, the presence of harmonics is of
considerable importance. It must be noted, however, that the no-load current
should, in practice, never exceed 5% of the rated current for the transformer
and that, under normal operating conditions on load, the sinusoidal require-
ments of the load component of current will predominate the current require-
ments of the transformer and the harmonics in the waveform of no-load
current will be relatively unimportant.
It can then be concluded that if the voltage applied to a transformer is
sinusoidal and the mutual flux set up is assumed to be sinusoidal, the no-load
current must have a large harmonic content. If, on the other hand, it is
assumed that the no-load current of the transformer is sinusoidal, the mutual
flux and primary and secondary e.m.f.s must have an harmonic content.

4.10. 3-Phase transformer connections

There are many ways in which three single-phase transformers can be
connected to form a 3-phase transformer bank and, in these circumstances,
some standard method of terminal marking becomes essential. It is possible
to connect both the primary and secondary windings in star (Y), delta (~ or
D), or zigzag (""(or Z); in all cases the primary terminals will be designated
by capital letters (ABC) and the secondary terminals by small letters (abc).
The successive terminals in each phase will be numbered in sequence (A 1 , A 2 ,
etc.), and when the instantaneous value of the voltage in a primary phase
(say A) is in the direction of A 1 to A 2 , the corresponding secondary phase a
voltage will always be assumed to be in the direction of a 1 to a 2 • The British
Standard Specification (No. 171) for polyphase transformer connections
lists transformers into four main groups according to the phase shift between
corresponding primary and secondary line voltages, and each connection
within a main group will be specified by the form of the primary and second-
ary connections. This phase shift is the angle by which the primary line
voltage leads the corresponding secondary line voltage and is measured in
 3-PHASE TRANSFORMER CONNECTIONS 67

unit increments of 30° in a clockwise direction. Thus the designation Ydll

specifies a star connected primary, a delta connected secondary and an
11 x 30° = 330° angle between corresponding primary and secondary line
terminals with the secondary leading the primary. Table 4.3 summarizes
most of the standard connections and shows, under idealized conditions, the
voltage and current relationships for balanced primary line voltages V with
a primary to secondary turns ratio t. Some of the more important of these
connections will be considered further.

(a) Delta-delta connection (DdO or Dd6)

The primary and secondary phase windings must both be insulated for the
full-line voltages while the phase currents under balanced conditions are
each 57·7% of their line values. This connection is useful for relatively large,
low voltage transformers but suffers from the disadvantage that no neutral
connections are available. One of the main advantages of this connection is
the fact that one phase can be removed without disrupting the operating
system. This condition, known as the open delta or vee connection is capable
of supplying 1/J3 = 57·7% of the load supplied by the full connection.
It has been noted previously that, with sinusoidal applied voltages, the
presence of a third-harmonic component of no-load current can have an
important influence on the operation of 3-phase transformer banks. If the
phase currents in a delta-connected circuit each contain a third harmonic it
follows that since the fundamentals are 120° out of phase, the third harmonics
are in phase. Thus the line currents will contain no third-harmonic component
and these third-harmonic currents will circulate around the delta. Thus, any
transformer with a delta-connected winding will operate with sinusoidal
voltages on both sides with a third-harmonic current circulating within the
delta.

(b) Star-star connection (YyO or Yy6)

This connection can be used for relatively small high voltage transformers
and has the advantage that a neutral connection can be made on both sides
of the transformer. If, however, it is necessary to operate without a primary
neutral conductor, severe imbalance in the secondary line to neutral loads
must be avoided when the bank consists of three single-phase transformers
or a 3-phase shell-type transformer.
When three single-phase transformers connected in star-star operate with
a primary neutral conductor, the neutral, under balanced loading conditions,
will carry only third-harmonic current. Since the third-harmonic component
of the no-load current has a return path through the neutral conductor, the
phase voltages on each side of the transformer will be sinusoidal. If the
primary neutral conductor is removed, third-harmonic currents cannot exist.
Thus the phase voltages on both sides of the transformer will contain a
68 THE TRANSFORMER

TABLE 4.3

GROUP NUMBER
SYMBOL
WINDINGS AND TERMINALS PHASOR DIAGRAMS
PHASE SHIFT Primary Secondary Primary Secondary

~i ;~ ~ ~ ~
11
v~.o
12

lt ~~ ;=±z=z=~ i::s 6
'*'
DdO
o•
13
DzO ~A c
~B b 4 ~
w crt
~ y
o• c
n
4

2 1
~ E~ I ~: ~~
2
V~6
1 o•

lt ~i UJ t;~ \1
22
Dd6
1SOO

~A c~~
~ '(.''
2 3
Dz6 8 b 4 2

u
180' c 4~ 1

rt ti ;~ ~ ~
31
Dy1
-30'

~ ti ~~c;w ~ <f·
32
Vd1
-30'

~~ c:~ ~ ~
33
Vz1
b 1
-30"

~A ~n~ ~ 1--b
4 1 2 1
Dy11 B
2 .

~ ±j ~$~ ~ ~ a!>.
+30" 1
c
42
Vd11
+30"

~~ ~~ :~n
~ 'Y'-
43
Vz11 B
+30" c
 3-PHASE TRANSFORMER CONNECTIONS 69
pronounced third-harmonic component and the neutral point voltage will
oscillate at triple frequency although the line voltages will be sinusoidal.
When a secondary neutral conductor is provided, it would be, in many cases,
earthed and the resulting third-harmonic currents to earth could cause
interference with protection and telephone circuits. In these circumstances,
it is usual to provide such transformers with a third set of windings connected
in delta. Such a winding is known as a tertiary winding and provides a path
for the third-harmonic currents necessary to produce sinusoidal primary and
secondary phase voltages. The tertiary winding can also be used for auxiliary
loading purposes and, under unbalanced loading conditions, helps to balance
the load more evenly between the primary phases. The rating of the tertiary
winding will, of course, depend on its loading but will normally be a minimum
of 30% of the transformer rating in order to protect this winding under short-
circuit conditions.
An alternate method of reducing the effects of third-harmonic fluxes for
the star-star connection is to use the three limb core-type construction. At
any instant in time, the third-harmonic fluxes in the three separate limbs of
the core are in the same direction and their only possible return path is
through high reluctance air. Under these conditions the third-harmonic
fluxes are considerably reduced and their effect on the phase voltages can
normally be neglected. The use of this form of construction also has a very
favourable influence on the performance of the transformer under unbalanced
loading conditions.

(c) Star-zigzag (Yzl and Yzll)

The zigzag connection involves the use of two separate identical secondary
windings on each phase and, for the same output voltage, requires more
copper than the corresponding single winding secondary. However, the
secondary phase voltages are obtained by the addition of two voltages
displaced by 60° and should each of these two voltages contain a third
harmonic, these third harmonics will cancel and not appear in the phase
voltages. Under unbalanced loading conditions each secondary phase current
will be reflected in two primary phases and this effect leads to a much better
performance under these conditions.
The zigzag secondary connection is widely used for three-anode rectifier
circuits in which each anode conducts for one-third of a cycle.

(d) Delta-star (Dyl and Dyll) and star-delta (Ydl and Ydll)
These forms of connection can be used to combine the advantages of both
star and delta connection into one transformer and are widely used in power
distribution systems. The star connection can be used for a stable four-wire
supply and there will be no neutral displacement under unbalanced loading
conditions.
70 THE TRANSFORMER

4.11. Parallel operation of 3-phase transformers

The parallel operation of 3-phase transformers is particularly important
in distribution systems and the following conditions must be satisfied if the
transformers are to operate satisfactorily in parallel:
(a) their voltage ratios must be the same;
(b) the phase shift between corresponding primary and secondary terminals
on each transformer must be the same. It immediately follows that all
transformers in the same main group can be connected in parallel. It is,
however, possible to connect a transformer with a 330° phase shift
(Group 4) in parallel with a transformer with a 30° phase shift (Group 3)
if the phase sequence of the primary and secondary of one transformer
are reversed.
It has been shown previously that, if two single-phase transformers
operating in parallel are to share the total load in proportion to their ratings,
the magnitudes of their per-unit impedances must be the same and that, if
they are to operate at the same power-factor, the phase angle of their
impedances must be the same. The same conclusions can be applied directly
to the parallel operation of polyphase transformers.
Any of the equivalent circuits derived previously for the single-phase
transformer can be applied directly to any one phase of a 3-phase transformer
under balanced loading conditions and it follows that the theory of chapter 4
can be applied on a per-phase basis to the parallel operation of 3-phase
transformers. It is, however, preferable to perform calculations on a per-unit
basis, particularly in cases where the primary and secondary connections are
different.

4.12. The auto-transformer

It has previously been assumed that the transformer is a two-winding device
but the windings of a normal transformer can be joined as shown in Fig.
4.22(a) without affecting the performance. Such an arrangement is known as

I
II I2

v,l tv,
N 1 I(N1;2Nz)

v
Nl :Nz
I
Nz
tv~. :'N,)
(a) (b)
Fig. 4.12. The auto-connection.
 TUTORIAL PROBLEMS 71

an auto-transformer and is normally represented by the arrangement shown

in Fig. 4.22(b ).
The amount of copper used for an auto-transformer will be less than that
for a two winding transformer and the reduction in the effective core length
will result in reduced iron losses. It follows that both the capital and running
cost of an auto arrangement will be less than that of a two-winding winding.
The direct electrical connection between the high- and low-voltage sides is,
however, one disadvantage of the auto-connection.

Tutorial Problems
l. A single-phase core-type transformer is to be designed to have a primary voltage
of 33 kV and a secondary voltage of 6·6 kV. If the maximum flux density per-
missible is 1·2 Wb/m 2 and the number of primary turns is 1250, calculate the
number of secondary turns and the cross-sectional area of the core when the
operating frequency is 50 Hz.
(Answer: 250; 0·0993 m 2 )

2. The parameters of a 200/100 V, 50 Hz, single-phase transformer are

R1 = 0·1 Q; X 1 = 0·4Q; R2 = 0·03Q; X2 = 0·1 Q
g = 1·2 X 10- mho,
4

b = 4 X 10- 4 mho.
The transformer supplies a secondary load current of 20 A at power factor
0·8 lag with rated secondary voltage. Draw a complete phasor diagram to scale
and determine the primary applied voltage and efficiency under these conditions.
(Answer: 216·5 V; 97·1%)

3. A 200/400 V single-phase transformer gave the following test results:

(a) with the /ow-voltage winding short-circuited, measurements taken on the
high-voltage side were: 20 V, 10 A, 100 W.
(b) with the high-voltage winding open-circuited, measurements taken on the
low-voltage side were: 200 V, 1 A, 60 W.
Determine the circuit constants and draw the approximate equivalent circuit as
seen from the low-voltage side for a rating of 4 kVA.
If a load impedance (80 + j35) Q is connected to the high-voltage terminals,
determine the percentage regulation and efficiency when 200 V is applied to the
low-voltage terminals.
(Answer: R = 0·25; X= 0·433; g = 1·5 X 10- 3 ; b = 4·77 X 10- 3 ; 2%; 95·5%)

4. The results of the open-circuit and short-circuit tests on a single-phase transformer

are as follows:
Voltage p.u. Current p.u. Power p.u.
Open-circuit 1·0 0·025 0·01
Short-circuit 0·04 1·0 0·012
Derive the per-unit values of its parameters and find the power-factor at which
the voltage regulation is zero and the corresponding efficiency.
(Answer: R = 0·012; X= 0·0382; g = 0·01; b = 0·0229; 0·947lead, 97·7%)
72 THE TRANSFORMER

5. A 200/100 V, 50 Hz, single-phase transformer has a primary leakage impedance

of 0·15 + jO· 76 n and an actual secondary leakage impedance of 0·04 + j0·19 n.
The transformer supplies a load of impedance 3·92 + j2·62 !l. Calculate the
secondary terminal voltage, the input current and power-factor, the regulation
and the efficiency of the transformer if the primary is supplied at rated voltage.
(Answer: 94·4 V; 10 A; 0·8lag; 5·6%; 97·9%)

6. Show that the voltage regulation of the transformer is given by [ v.c cos (</> - </>.c)]/
Vrated where V.c is the voltage required to circulate rated current on short-circuit,
cos </>sc is the power-factor on short-circuit, and cos </> is the operating power-
factor.

7. Two single-phase transformers A and B with identical turns ratios and ratings
operate in parallel. The per-unit impedance of A is 0·006 + j0·04 and that ofB is
0·009 + j0·05. If A operates on full load at power-factor 0·8 lag, calculate the
per-unit load and operating power-factor of B.
(Answer: 0·801, 0·825lag)

8. A 400 kVA, 33/3·3 kV, single-phase transformer is connected in parallel with a

500 kVA, 33/3·3 kV, single-phase transformer and the combination supplies a
total load of 800 k VA at power-factor 0·8 lagging. The resistances are I% and 2%
and the leakage reactances are 7% and 5% respectively. Calculate the terminal
voltage, the kV A output, the current and the power-factor of each transformer.
(Answer: 30·1 kV; 304 kVA; 498 kVA; 101 A; 166 A; 0·7; 0·852)

9. The hysteresis loop for the iron core of a 50 Hz, single-phase transformer is as
follows:
Flux density b (Wbjm 2 ) 0 0·2 0·4 0·6 0·8 1·0 l-15
110 145 200 240 390 555 800
Magnetizing force h (AT/m) { -110 -80 -30 0 155 370 800
Draw one cycle of the waveform of no-load current for this transformer when the
applied voltage is sinusoidal and one cycle of the waveform of voltage when the
no-load current is sinusoidal.

10. The total iron loss for a 440 V, 50 Hz, transformer is 2500 W. When the applied
voltage is 220 V at 25 Hz the corresponding loss is 850 W. Calculate the eddy
current loss and the hysteresis loss at normal voltage and frequency.
(Answer: 1600 W; 900 W)

II. The total iron loss at constant flux density in a transformer varies with frequency
as follows:
Total iron loss (W) 46 63 78 100 124 175 230
Frequency (Hz) 25 33·3 40 50 60 80 100
Determine the hysteresis and eddy current loss at 50 Hz.
(Answer: 84·5 W; 15·5 W)
12. The input current to a 3-phase, step-down transformer connected to an II kV
supply system is 14 A. Calculate the secondary line voltage and current for (a)
star-star, (b) delta-star, and (c) star-delta connection if the phase turns ratio is 44.
(Answer: 250 V; 616 A; 432 V; 355 A; 144 V; 1070 A)
 TUTORIAL PROBLEMS 73
13. An 11,000/415 V 3-phase transformer with high voltage tappings of ±2t% is
designed to operate on 4 V per tum. Find the number of turns required on each
winding for (a) star-star, (b) delta-star, and (c) delta-zigzag connection.
(Answer: 1630/62; 2818/62; 2818/69)
14. When phasing-out two 3-phase 11,000/400 V delta-star transformers with their
neutrals joined and their h.v. terminals connected to the h.v. supply the following
voltmeter readings were obtained:
c2 - c2 , 230 V.
Deduce how the incoming transformer differs from the original transformer and
explain how the incoming transformer can be made to operate in parallel with the
original transformer.
15. A delta-zigzag, 3-phase transformer is required to operate in parallel with a star-
star, 3-phase transformer with a turns ratioN = 3. Find the required phase turns
ratio for the delta-zigzag transformer.
(Answer: 4·5)
5. Polyphase induction motors

5.1. Derivation of equivalent circuits

It has been shown in chapter 1 that the polyphase induction motor can be
considered as a polyphase transformer with variable frequency in the short-
circuited secondary winding and an equivalent circuit can be derived per
phase of the motor in the same general way as that of a transformer. It is
apparent that each winding will have resistance and leakage inductance. The
leakage reactance of the rotor will, however, be dependent on the frequency
of the rotor current and, since the rotor frequency is Sf, the slip frequency,
where S = (Ns - N)/Ns is the slip andfis the supply frequency, the rotor
leakage reactance is given by sx2 where x2 is the standstill value of the rotor
leakage reactance. Thus an equivalent circuit for the rotor can be drawn as
shown in Fig. 5.1. The stator and rotor will, however, have different turns and

Fig. 5.1. Rotor equivalent circuit.

the actual rotor must be made into an equivalent rotor having the same turns
as the stator. At a corresponding flux and slip the relation between the actual
rotor induced voltage E 2 and the equivalent rotor induced voltage E 2 , is

(5.1)

where a is the ratio, effective stator turns per phase, N 1 , to effective rotor
turns per phase, N 2 .
If the rotors are to be magnetically identical it follows that their m.m.f.s
must be equal and the relation between the actual rotor current / 2 and the
 DERIVATION OF EQUIVALENT CIRCUITS 75

equivalent rotor current / 2 , is

(5.2)

Thus the relation between the slip-frequency leakage impedance Z 2 =

R 2 + jSX2 of the actual rotor and the slip-frequency leakage impedance
Z 2 , = R 2 , + jSX2, of the equivalent rotor is

Z 2, = R 2, + jSX2, = E 2' = a 2 E 2 = a 2(R 2 + jSX2) = a 2Z2. (5.3)

12, 12
The equivalent circuit for a magnetically equivalent rotor is shown in
Fig. 5.2, and the voltages, currents, and impedances in this circuit are now
referred to the stator turns. It will be assumed that the referred rotor constants
are known quantities.

Fig. 5.2. Rotor equivalent circuit referred to stator turns.

Now the resultant flux in the machine can be assumed to link both stator
and rotor and by the laws of electromagnetic induction this flux induces a slip
frequency e.m.f. E 2 in the rotor and a mains frequency e.m.f. E 1 in the stator.
The relative speed of the resultant flux with respect to the rotor is the slip S
times its relative speed with respect to the stator and it follows that, if the
resultant flux is assumed constant,

Then E2 = N2S, or E2Nl - E - SE (5.4)

El Nl N2 - 2, - 1·

It follows that, since, from (5.3) E 2 ,/1 2 , = (R 2 , + jSX2,)

E2, = SE 1 = 12,(R 2, + jSX2)
or E1 =
. (R2,
12 ,
. )
S + JX 2, · (5.5)

The equivalent circuit for each phase of the stator winding of the machine
can be derived in exactly the same way as that for a transformer and is shown
in Fig. 5.3.
76 POLYPHASE INDUCTION MOTORS

It should be noted that the input current 11 can be resolved into two com-
ponents, the load current 12 , and the no-load current 10 • The m.m.f. of the
load current balances the m.m.f. of the rotor current. The equivalent circuits
for the stator and for the rotor when referred to the stator turns can then be
combined as shown in Fig. 5.4.

Fig. 5.3. Stator equivalent circuit.

Frequency Sf
Ideal transformer
1:1 Current
I : S Voltage
Fig. 5.4. Combined equivalent circuit.

It can be seen from (5.5) that the circuit shown in Fig. 5.4 can be reduced to
the usual exact equivalent circuit per-phase of a polyphase induction motor
shown in Fig. 5.5, in which the resistance RdS is an effective resistance

Fig. 5.5. Exact equivalent circuit.

 POWER BALANCE EQUATIONS 77
combining the effects of shaft load and actual rotor resistance. If, as is usual,
it is assumed that the iron losses in the machine are constant, the resistive
branch in the no-load circuit can be removed and the equivalent circuit
redrawn as shown in Fig. 5.6. It should be noted that all quantities used in the
equivalent circuit must be expressed as phase quantities.

Fig. 5.6. Final form of exact equivalent circuit.

5.2. Power balance equations

For an m phase machine with an equivalent circuit of the form shown in
Fig. 5.6,
Input power= mV111 cos c/J 1 ,
where cfJ 1 is the angle between V 1 and / 1 •
Stator copper loss = ml/ R 1 . (5.6)
R
Power transferred to rotor= m/~ 1 ; 1• (5.7)

Rotor copper loss Pc = mftR 21 • (5.8)

1- s
Gross output power P9 = (5.7) - (5.8) = mitR 21 -
8- · (5.9)

Fixed losses are friction + windage + iron losses.

Nett output power = gross output power - fixed losses.

Now 2nN T = P = mit(R )

6() 9 9 2
(1 - s)
-
8- ·

But N = N 5 (1 - S).

m/2 R21
21 s
Then Gross torque T 9 = N.m. (5~10)
2rr
60 Ns
Now 2rrN5 /60 is a constant for a particular machine and it follows that
78 POLYPHASE INDUCTION MOTOR

[(2n/60)N5]T9 , which is equal to the power transferred across the air-gap, is

a measure of the gross torque.
Thus the power transferred across the air-gap is said to be the gross torque
in synchronous watts.

Then Gross torque T 9 = m12,

r2
S
R2
syn. watts. (5.11)

But rotor copper loss Pc = m~,R 2 , watts, and gross output power P9 =
m~,R 2 ,[(1 - S)/S] watts. Then
.
T orque (m ) rotor copper loss
syn. watts = S

gross output power

(5.12)
(1 - S)
The whole performance of an induction motor can be calculated on the
basis of equivalent circuits but some simplifications can be made to the circuit
by the application ofThevenin's theorem and the Maximum Power Transfer
Theorem* if torque slip or power-slip relationships are required.

5.3. Torque calculations

Consider the exact equivalent circuit shown in Fig. 5.7(a). The circuit
viewed from the points 'a' and 'b' can be replaced by a single equivalent
voltage Ve and a single equivalent impedance Ze = Re + jXe such that
jXm
Ve = V 1 - - - - - - - (5.13)
R1 + j(X1 + Xm)
z = (R1 + jXl)jXm .
(5.14)
e R1 + j(X1 + Xm)
In general, it can be assumed that X 1 + Xm is much greater than R 1 and
these equations then reduce to
(5.15)

z = R +"X= (R1 +jXl)Xm. (5.16)

e e 1 e Xl + Xm
* Thevenin's theorem states that any network of circuit elements and voltage
sources as viewed from two terminals can be replaced by a single source in series with
a single impedance. The source voltage is the voltage that would appear across those
terminals when they are open-circuited and the impedance is that viewed from those
terminals when all voltage sources in the original network are short-circuited.
The Maximum Power Transfer Theorem states that maximum power will be
transferred from a source of inductive impedance Z 1 to a sink of inductive impedance
z2 when zl = z2.
 TORQUE CALCULATIONS 79
Thus the exact equivalent circuit can be redrawn as shown in Fig. 5. 7(b). Now
from (5.11),

Torque Te = m~, R;• ·

Thus mvezRzJS (5 17)

Te = (Re + RzJS) 2 + _x2 syn. watts. .

R. X= X.+ X21

(a) (b)
Fig. 5.7. Derivation of Thevenin equivalent circuit:
(a) Usual circuit. (b) Thevenin equivalent.

The general shape of the torque-slip curve for an induction machine is

shown in Fig. 5.8 and the machine will operate as a motor when 1 ::;;?; S > 0,
as a generator when S < 0 and as a brake when S > 1. The machine must be

tTorque

- - - - - Braking ------t--Motoring - t - - - Generating-

-
+oo -.------~~----~~----~~0~----~----a
2 -1
Slip

Fig. 5.8. Torque-slip curve.

80 POLYPHASE INDUCTION MOTORS

driven above synchronous speed for generator operation and must be driven
against the direction of rotation of its m.m.f. for operation as a brake.
Since the torque for synchronous watts is the power dissipated in the
resistor R 2 JS, it follows that the torque will be a maximum when the power
dissipated in R 2 ,1S is a maximum. Then, by the Maximum Power Transfer
Theorem, maximum torque will occur when !R. + JXi = R 2 JS. Thus the
slip S1 for maximum torque is given by
S R2, (5.18)
t = (R.2 + r)t
When this value of slip is substituted in the general equation for torque,
(5.17), the value of maximum torque Tm is given by
T = mV.2J(R.2 + X2)
m [R. + J(R/ + r)] 2 + r
On simplification
mV 2
Tm = 2[R. + J(;/ + r)] (5.19)

It should be noted that from (5.18), the slip S1 at which maximum torque is
available is a direct function of the rotor resistance but that, from (5.19), the
value of maximum torque is independent of the rotor resistance. The typical
torque-slip curves for a motor with variable rotor resistance are shown in
Fig. 5.9 and it can be seen that the speed of an induction motor can be con-
trolled by variation of the rotor resistance. Since the starting torque (S = 1)
and current of the motor also depend on the value of rotor resistance,
variation of rotor resistance can also be used as a starting method. Both of
these techniques can, of course, only be applied to wound-rotor induction
motors.
tTorque

0
-
Slip

Fig. 5.9. Torque-slip curves for variable rotor resistance.

 OUTPUT POWER CALCULATIONS 81

It is sometimes convenient to express the torque in terms of dimensionless

parameters (ref. 1) and a torque ratio t can be defined as the ratio of the torque
Tat any slip S to the maximum torque Tm at slip S,; that is,

t =-=
T 2[Re + J(R/ + ~)]R2JS
.
Tm (Re + RdS) 2 + X2
The slipS, for maximum torque is given by S, = RdJ(R/ +~),so that
R 2, = S,J(Re 2 + X2), or R 2, = S,ReJ{l + (X/Re) 2}.

Then (5.20)

If X/Re is defined as a quality factor Q the torque ratio (5.20) can be written
as

t = s
I + 2; J(1 + Q2) +
(s)2
; (I + Q2) + Q2

or
t = I + J(1 + Q2)
(5.21)
1 S 81 I 2
1+- -+-vO+Q)
2 s, s
If the stator resistance can be neglected, this equation reduces to
2
(5.22)
t s s,
= -----
-+-
s, s
5.4. Output power calculations
The gross output power (P9 ) is given by (5.9) as

P9 = mitR 2 ,
1- s watts
- 8-

1- s
mVe2R2,-S-
or p -
9 - s)2
-r-----~-.-----
(
R+R2,-i- +~
1
watts (5.23)
82 POLYPHASE INDUCTION MOTORS

Again, by the use of the Maximum Power Transfer Theorem, it follows

that the gross output power will be a maximum when + jXI = IR
R 2 , [ (1 - S)/SJ. Thus the slip SP for maximum gross output power is given by

S R2, (5.24)
P = R 2 , + J(R 2 + ~)
When this value of slip is substituted in the general equation for gross
output power, (5.23), maximum gross output power Pgmax is given by
mV 2
P 9 max = 2{R + J(; 2 +~)}watts. (5.25)

It should be noted, at this stage, that the resistance R = R 1 + R 2 , is a

function of rotor resistance and that the value of maximum gross output
power is not independent of the rotor resistance.
Comparison of(5.18) and (5.24) shows that the slip S 1 for maximum torque
is not the same as the slip SP for maximum gross output power.
EXAMPLE 5.1. A 37·3 kW, 3-phase, 4-pole, 50 Hz, induction motor has a full-
load efficiency of 85%. The friction and windage losses are one-third of the
no-load losses and the rotor copper loss equals the iron loss at full load. Find
the full-load speed. Stator resistance can be neglected.

Solution: Now efficiency = 0·85 = nett output/(nett output + losses).

Nett output= 37 300 W.
Losses = copper loss Pc + iron loss Pi + friction loss Pr.
No-load loss= Pi + Pr = 3 Pr. :. Pr =!Pi.
On full load Pc =Pi. Losses= Pc + Pc + !Pc = ~Pc.
37 300
Then 0·85 = 37 X 300 + 2·5Pc' i.e., pc = 2640 W.

Now, from the equivalent circuit,

1- s
Gross output power P9 = mftR 2 , -
8- W.

Rotor copper loss Pc = mftR 2 , W.

1- s
Then pc_S_ = Pg.

Thus 2640 1 ~S= 37 X 300 + 2640 + !{2640)

1
whence, s= 16·65.
 OUTPUT POWER CALCULATIONS 83

Synchronous speed N. = 60f/P = 1500 r.p.m.

Full-load speed N = N.(l - S) = 1500[1 - (1/16·65)] = 1410 r.p.m.
EXAMPLE 5.2. A 415 V, 3-phase, 6-pole, 50 Hz, star-connected wound rotor,
induction motor has the following parameters in ohms per phase.
R 1 = 0·04; X 1 = 0·15; R 2 , = 0·05; X2 , = 0·15; Xm = 9·85.
Stator: rotor turns ratio per phase is 1· 5. Determine
(a) the stator current and the gross torque in N.m when the slip is 0·05,
(b) the maximum gross torque, the slip at which it occurs and the gross output
power under these conditions,
(c) the values of external resistance to be inserted to produce 80% of maxi-
mum torque at standstill and state which value would be used.

Solution (a): The input impedance Z is given by

z= 0·04 + ·o·15 + (I + j0·15)j9·85

J I + j10
= 1 + j0·395 = 1·075/21 ° 33'.

Stator current 11 = J33 X

415
1-075/21° 33'
= 222\21 o 33' A.

Input power= J3 V 1
1 1 cos c/J 1 = J3 x 415 x 222cos21o 33' = 148·2kW.
Stator copper loss = 3/1 2 R 1 = 3 x 222 2 x 0·04 = 5·92 kW.
Power transferred to rotor = torque in syn. watts
= (148·2 - 5·92)10 3 = 142·28 X 103 •
60 X 50
Synchronous speed N. = 3 = 1000 r.p.m.

. 142·28 X 103
From (5.10), Torque m N.m = 2n = 1360 N.m.
X 1000
60

Solution (b): Take Thevenin equivalent of stator and magnetizing imped-

ances.
9·85 9·85
R = 0·04- = 0·0394
e 10 ' X 0 = 0·15 X lQ = 0·148.

Then torque will be a maximum when

0·05
10·0394 + j(0·148 + 0·15)1 = -·
s
84 POLYPHASE INDUCTION MOTORS

Slip S 1 for maximum torque is given by

y
From (5.19),

3 X ( 0·985 X
415

Maximum torque Tm = 2[0·0394 + (0·03942 f0·298 2 }~]

= 24,500 syn. watts.
From (5.12),
Gross output power = torque (syn. watts) x (1 - S)
= 24,500 X (1 - 0·133) = 21·2 kW.

Solution (c): Let external resistance referred to stator turns be R. Then,

from (5.17) with S = 1,
. 3(0·985V1 ) 2(0·05 + R)
Startmg torque T. = (0·0394 + 0·05 + R)2 + 0·2982

ButT. = 0·8Tm. Therefore

3(0·985 v1) 2
0"8 2[0·0394 + (0·0394 2 + 0·298 2)-!-] (0·0394 + 0·05 + R) 2 + 0·298 2
On solution, R = 0-481 or 0·191. But turns ratio is 1·5. Actual value of
external rotor resistance = 0·214 or 0·0849. The smaller value of resistance
R = 0·0849 n would be used and this is apparent from the torque-slip
curves shown in Fig. 5.9.

5.5. Determination of circuit parameters

The parameters of the equivalent circuit of a polyphase induction motor
can be obtained from design data but can easily be measured by means of
three simple tests, a running light or open-circuit test, a locked rotor or
short-circuit test, and a measurement of the d.c. resistance per phase of the
stator.

(a) Running light test

The machine is run up to its no-load speed, and the variation of the input
power Poe and the input current /oc for a range of voltages Voc from 125% of
rated voltage at rated frequency to the minimum possible to maintain the
no-load speed constant are recorded.
 DETERMINATION OF CIRCUIT PARAMETERS 85

The no-load losses Pc are usually assumed to be constant and equal to the
value at rated frequency and voltage. Then, at rated voltage,
Pc = Poe - mf;cR 1 (5.26)
where R 1 is the d.c. resistance per phase of the stator winding and m is the
number of phases.
Since the slip Sunder no-load conditions is very small, it is assumed that,
during the running light test, the rotor is on open-circuit and the equivalent
circuit can be drawn as shown in Fig. 5.1 0. Then

(5.27)

Thus the running light test will give the sum of the stator leakage reactance
and the magnetizing reactance and a value for the fixed losses for the machine.
Rl xl

Fig. 5.10. Equivalent circuit for running light test.

(b) Locked rotor test

The machine rotor is locked so that the motor cannot rotate and the
variation of input power Psc• and terminal voltage Vsc for a range of input
currents at rated frequency up to 125% of rated current are recorded. It
should be noted that the value of the input current for a particular applied
voltage can, under locked rotor conditions, be affected by the position of the
rotor. This effect, however, will be, in general, small for squirrel-cage
machines.
The equivalent circuit for the machine under these conditions is exactly
that for a transformer with a short-circuited secondary and is shown in
Fig. 5.11. However, the applied voltage required to produce rated current

Fig. 5.11. Equivalent circuit for short-circuit test.

86 POLYPHASE INDUCTION MOTORS

with the machine at standstill will only be a small proportion of rated voltage
and it is permissible under these conditions to neglect the no-load current.
Then, at rated current,
(5.28)

(5.29)

The sum of the reactances X 1 + X2 , must now be subdivided into its

stator and rotor components and the recommended distribution (ref. 2) is
shown in Table 5.1.
TABLE 5.1
Fraction of X1 + X 21

Class A, normal starting torque and current 0·5 0·5

Class B, normal starting torque, low starting current 0·4 0·6
Class C, high starting torque, low starting current 0·3 0·7
Class D, high starting torque, high running slip 0·5 0·5
Wound rotor 0·5 0·5

Thus values for the stator resistance and leakage reactance, the magnetiz-
ing reactance, the standstill values of rotor resistance and leakage reactance
and a value for the total fixed loss can be obtained directly from the results of
three simple tests.

EXAMPLE 5.3. The results of a locked rotor and running light tests on a 415 V,
30 kW, 3-phase, 50 Hz, delta-connected squirrel-cage induction motor are:
Line voltage Line current Total power
Locked Rotor 130V 77A 6·4kW
Running Light 415 v 22·8 A l·4kW

The stator resistance is 0·48 ohm per phase. Determine the parameters per
phase of the equivalent circuit of the motor.

Solution: The stator resistance R 1 is given as 0·48 ohm per phase. From the
running light test, the total fixed losses Pr are given by (5.26) as

Pr = 1400- 3 x 0·48 x (J;Y = 1150 W.

The input impedance on no-load is given by

415
zoe = 22·8/}3 = 31-4 ohms.
 STARTING METHODS FOR INDUCTION MOTORS 87

The no-load resistance Roc will be given by

1400/3
Roc = (22 .8/)J) 2 = 2·7 ohms.

Then, from (5.27)

Xoc = X 1 + Xm = ,j(31·42 - 2·7 2 )t, i.e., X 1 + Xm = 31·3 ohms.

From the locked rotor test Psc = mP.c(R 1 + R 2 ,). That is,
-6400
3- = J3
( 77 ) 2 (0-48 + R 2 ,), whence R 2 , = 0·60 ohm.

The input impedance on short-circuit is given by

130
Z 2·92 ohms.
>C 77!)3
= -- =

Then from (5.29),

xl + x2, = [2·92 2 - 1·06 2 ]t = 2·72 ohms.
For a standard squirrel-cage motor it can be assumed that X 1 = X 2 , and
X 1 = X 2 , = 1·36 ohm. Thus Xm = 31·3- X 1 = 29·94.

5.6. Starting methods for induction motors

Direct-on starting
The impedance of a polyphase induction motor at standstill is relatively
small and when such a machine is connected directly to a supply system, the
initial current will be high (of the order of six times full-load current) and at a
low power-factor. The size of machine which can be direct-on starting will
depend on the capacity of the available supply system and this method of
starting is only suitable for relatively small (up to 7·5 kW) machines. The
performance of the machine at standstill can be determined from one of the
usual equivalent circuits.

Delta star starting

The machine is designed for delta operation and is connected in star during
the starting period. The impedance between line terminals for star connection
will be three times that for delta connection and for the same line voltage the
line current at standstill for star connection will be reduced to one-third of the
value for delta connection. Since the phase voltage will be reduced by a factor
J3 during starting,it follows that the starting torque will be one-third of
normal.
However, when the machine connection is changed from star to delta, all
three switches must be opened simultaneously. During this period the air-gap
88 POLYPHASE INDUCTION MOTORS

flux and the speed will decrease and, when the switches are reclosed, the
transient current surge can be high. Thus, this method of starting does not
necessarily reduce the peak value" of starting current but should reduce the
time duration of this current.

Auto-transformer starting
The connection for auto-transformer starting is shown in Fig. 5.12 and the
setting of the auto-transformer can be predetermined to limit the starting
current to any desired value. An auto-transformer which reduces the voltage
R

3-Phase 3-Phase
auto-transformer star-connected
motor
S Starting position R Running position
L Transition inductances
Fig. 5.12. Auto-transformer starting.

applied to the motor to x times normal voltage will reduce the starting current
in the supply system and the starting torque of the motor to x 2 times normal
values. This method can be applied to both star- and delta-connected machines
but suffers from the same disadvantage as star-delta starting in that all three
line switches must be opened simultaneously during the change-over from
the starting to the normal running condition. However, the current surge
during switching can be reduced by introducing transition impedances as
shown in Fig. 5.12.

Rotor resistance starting

It is obvious that the starting current can be reduced if the motor impedance
is increased and, in the particular case of the wound rotor machine, this can
be done by introducing additional rotor resistance during the starting period.
This can also be made to increase the starting torque available since the slip at
which maximum torque occurs increases as the rotor resistance increases.
 STARTING METHODS FOR INDUCTION MOTORS 89

Thus, high values of starting torque per ampere of starting current can be
obtained using this method. The rotor resistance will generally be reduced in
steps as the machine runs up to speed.

Multi-circuit starting (ref. 3)

The circuit connections corresponding to this method of starting are shown
in Fig. 5.13 and it can be seen that the motor is connected asymmetrically
A

c B

Switch Sl open for starting (S)

closed for running (R)
Fig. 5.13. Multi-circuit starting.

during the starting period with usually twice as many turns in series in one
phase as in the other two phases. This will, of course, result in unbalanced
currents in the supply system but the advantage of this method is that the
starting torque per ampere of starting current is high and the mean starting
A

c B
Fig. 5.14. Kusa method of starting.
90 POLYPHASE INDUCTION MOTORS

current is low. The main advantage of this method, however, is that during
the transition from the starting connection to the running connection, only
one phase of the machine is open-circuited and the machine will continue to
produce torque as a single-phase motor during the transition period.
A starting method, similar to this starting method, which is known as the
Kusa method of starting, is illustrated in Fig. 5.14, and consists of inserting
a variable impedance in one supply line to the machine during the run-up
period.

5.7. Speed control of induction motors

The simple squirrel-cage induction motor is ideally suited to drives where
an approximately constant speed is required. Where several discreet speeds
or a continuously variable speed is required, modifications can be made to
the induction motor to suit it to such applications. Variable speed can be
produced in one of two ways. First, the synchronous speed of the machine
can be changed by altering the frequency of the supply or the number of poles.
Second, the shape of the torque--speed curve can be changed by varying the
applied voltage or by injecting power into the secondary of the machine. The
various methods by which the speed of the polyphase induction motor can be
controlled will be discussed in this section.

(a) Pole amplitude modulation

A general method of pole changing, known as pole amplitude modulation
(ref. 4), has been recently devised, which enables a single-winding squirrel-
cage induction motor to run equally well at either of two chosen speeds. This
new method is based on the fact that, when an m.m.f. sinusoidally distributed
in space, with a given number of poles, is multiplied by another space-
distributed sinusoid of a different period, the resulting m.m.f. can be con-
sidered as two separate m.m.f.s with different pole numbers. For a particular
instant in time, the m.m.f. F, produced by the current in the phase winding of
a 3-phase machine can be written as
. p()
F =A sm 2 (5.30)

where A is the amplitude of the m.m.f., p is the number of poles and eom is
the mechanical angle in radians around the stator periphery.
When the amplitude A is modulated in space, in such a way that A =
C sin K8, K
where is an integer, (5.30) can be written

F= C sin K8 sin~() = ~[cos(~ - K)e + cos (f + K)e} (5.31)

Thus, from (5.31), the resulting phase m.m.f. can be said to consist of two
 SPEED CONTROL OF INDUCTION MOTORS 91

separate m.m.f.s with different pole numbers, (p + 2K) and (p - 2K). It is,
of course, necessary to eliminate one of the pole numbers produced by
modulation, and the original method consisted of modulating each phase
separately, the individual phase windings being arranged in such a manner
that one of the pole numbers produced by modulation is eliminated from the
total m.m.f. of the winding.
As a particular case, consider the 8-pole arrangement (p = 8) shown in
Fig. 5.15(a) modulated by the rectangular 2-pole wave (K = 1) shown in
Fig. 5.15(b). The process of modulation can be simply performed by reversing

(a)

-+---------------1--------------~r-(b)

Fig. 5.15. Basis ofpole amplitude modulation:

(a) Basic 8-pole m.m.f.
(b) Modulating wave.
(c) Modulated m.m.f.

one-half of the phase winding relative to the other half and the resulting
m.m.f. shown in Fig. 5.15(c) will consist of a 6-pole (p - 2K) and a 10-pole
(p + 2K) m.m.f. If the 3-phase, 8-pole winding is arranged with the phase
entries 120om apart in space, i.e., as if it were a 2-pole winding, the 6-pole
m.m.f. produced by modulation will represent the third harmonic of the
original winding. It has been shown in chapter 3 that the total m.m.f. pro-
duced by a symmetrical 3-phase winding will contain no triple harmonics,
so that the total 6-pole m.m.f. produced by modulation will vanish and the
machine will operate with a 10-pole m.m.f., which can be considered as the
fifth harmonic of a basic 2-pole wave.
Since, in the 10-pole mode, the machine operates on what is essentially an
exaggerated fifth harmonic, the presence of appreciable space-harmonics in
92 POLYPHASE INDUCTION MOTORS

the total m.m.f. could be of considerable importance, and the basic modula-
tion method described in Fig. 5.15 is likely to produce undesirable crawling
effects which can, however, be reduced by chording the winding. A very
great improvement in performance can be brought about by the use of
windings initially of the fractional slot type. In the case of a 36-slot 8/10 pole
machine, the unconventional form of fractional slot winding used is given in
Table 5.2. The coil pitch is 2/3 (pole pitch) for the 8-pole connection and 5/6

TABLE 5.2
Table of number of coils per phase belt for 8/10 pole-changing
POLES

I 2 3 4 5 6 7 8

Phase A 2 4 4 2 2 4 4 2

Phase B 2 2 4 4 2 2 4 4
·-

Phase C 4 2 2 4 4 2 2 4

Total 8 8 10 10 8 8 10 10

(pole pitch) for the 10-pole connection. Modulation is effected by reversal of

one-half of each of the phase windings, and the space distribution of the total
m.m.f. for a particular instant in time, before and after modulation, is shown
in Fig. 5.16. The use of this technique will produce a great improvement in the
performance of the machine with the modulated pole number, with little
reduction in performance for the unmodulated case.
The basic method, for which each phase-winding is considered separately,
precludes pole numbers which are a multiple of three; so that, for example,
while an 8/10 pole arrangement was possible, an 8/6 pole one was not. Later
developments in the technique (refs. 5, 6) have led to the extension of the
available range to include, first, any pole-ratio between 1 and 1·5 in the close
ratio range, and finally all other wide-ratio pole numbers. In all cases only six
external control leads are required and the method of pole-amplitude
modulation has, in effect, turned the standard single-winding squirrel-cage
induction motor into a two-speed machine with performance at each speed
closely approaching single-speed standards, using methods which are simple
and inexpensive.
The squirrel-cage induction motor long ago became the predominant
industrial electrical motor, and it has achieved this because of its many
outstanding merits (cheapness, reliability, good performance, absence of
sliding contacts, etc.) and in spite of its one great defect-a fixed speed. In
many cases, a fixed speed is quite sufficient, in some cases, it is tolerated for
 SPEED CONTROL OF INDUCTION MOTORS 93

simplicity; in other cases, a change in speed is obtained by gears or other

mechanical devices. The method of pole-amplitude modulation has added a
new degree of freedom to the induction motor and the motors so far in
operation have been applied to pumps, fans, grinders, crushers, escalators,
conveyor belts, and machine tools.

C
~ B

(a) Before modulation

A'
C B

(b) After modulation

Fig. 5.16. M.m.f. waveforms for fractional-slot arrangement:

(a) Before modulation. (b) After modulation.

Most textbooks consider a 2/1 pole-changing winding with six terminals,

invented over 60 years ago and manufactured all over the world. This winding
is, in fact, a particular example of pole-amplitude modulation devised without
recognition of the general principle behind its action. Since this one speed-
ratio alone has found many and wide uses, it seems almost certain that a
general method of pole-changing which offers any speed-ratio will be used
even more extensively.
94 POLYPHASE INDUCTION MOTORS
(b) Variation of supply voltage and frequency
The torque developed by the machine at a particular slip is proportional
to the square of the applied voltage and the shape of the torque-speed curve
can be modified by control of the applied voltage.
When the frequency of the voltage applied to the machine is varied, the
synchronous speed is changed and, if the applied voltage is controlled
directly with the frequency to maintain approximately constant air-gap flux
density, torque-speed characteristics of the form shown in Fig. 5.17 are

Fig. 5.17. Torque-slip curves with variable frequency.

obtained. The major disadvantage of this method of speed control is that it

involves the use of a polyphase variable frequency power supply. However,
the discovery of the thyristor (a controlled silicon rectifier), which is capable
of providing large powers at variable frequency from a fixed frequency
source, could do much to overcome this difficulty.

(c) Variation of rotor resistance

In the particular case of a wound-rotor machine, the effective resistance of
the secondary circuit can be controlled by the use of external variable resist-
ances. The form of the torque-speed curve with varying rotor resistance,
derived in section 5.3, is shown in Fig. 5.9, and it can be seen that, for a given
load, the operating speed of the machine varies with the value of the rotor
resistance. However, the rotor copper loss is always high at large slips, so
that this method of speed control can be very inefficient.

(d) Rotor injection using commutator machines

The action of the commutator as a frequency changer has been discussed
in chapter I, and the use of a polyphase commutator machine connected to
the slip-rings of a wound-rotor induction motor, injecting power into the
 SPEED CONTROL OF INDUCTION MOTORS 95

secondary, to produce speed control has been mentioned. Various forms of

commutator machines and their application to the speed and power-factor
control of polyphase a.c. motors will be described in the following sections.
The simplest form of polyphase commutator machine is illustrated in
Fig. 5.18. It consists of a lap-connected rotor winding connected to a commu-
tator with an unwound stator. When 3-phase voltages of frequency f = OJ5 /2n

Fig. 5.18. Basic form ofpolyphase commutator machine.

are applied to the commutator, its action is to frequency change these

voltages to a frequency f 2 = (OJ. ± OJ)/2n, where OJ is the mechanical speed
of rotation. Since the rotor winding is inductive, the equivalent impedance
of the commutator will change in phase and magnitude as the speed OJ is
changed. When the commutator brushes are connected to the slip-rings of the
wound-rotor induction motor under control, the commutator machine
effectively injects voltage into the secondary of the main motor. It is im-
portant to note that this injected voltage will be proportional to the secondary

Fig. 5.19. Polyphase commutator machine with stator winding.

96 POLYPHASE INDUCTION MOTORS

current of the motor under control and this method is only effective when the
main motor is operating on load with a relatively high secondary current.
A considerable improvement to the performance of this arrangement can
be made by the use of a stator winding connected in series with the commuta-
tor as shown in Fig. 5.19. This arrangement will also produce an injected
voltage proportional to the secondary current of the motor under control
and has the advantage that the phase angle of the voltage can be controlled
by the position of the commutator brushes relative to the stator winding. A
second stator winding is now necessary to compensate for the effects of
armature reaction.
When the main stator winding is connected in parallel with the commutator
brushes, an injected voltage independent of the value of the secondary current
of the machine under control is obtained and this arrangement, known as the
Scherbius machine, is shown in Fig. 5.20.

Fig. 5.20. The Scherbius machine.

Fig. 5.21. The frequency changer.

 DYNAMIC OPERATION OF INDUCTION MOTORS 97

The most elegant form of polyphase commutator machine is the Frequency

Changer, described in chapter 1 and illustrated in Fig. 5.21. The injected
voltage will be independent of the secondary current of the machine under
control, its magnitude will be governed by the setting of the regulator and its
phase angle will be controlled by the brush position on the commutator.

5.8. Dynamic operation of induction motors

An induction motor operates under dynamic conditions when the speed
of the motor changes from the steady-state value to another. A change in
speed will result from a change in the voltage applied to the motor or from a
change in load conditions when the machine is operating under stable
conditions. Any electrical transients associated with a change in speed will be
neglected in the following analysis.
The power balance for a polyphase induction motor has been considered
in section 5.2, and, from this analysis it follows that
Torque in synchronous watts = m~ 1 R 2 dS = w.T.
where w. is the synchronous speed and T. N .m is the torque of the m-phase
machine.
The rotor copper loss Pr is given by
Pr = S(w.T.) (5.32)

where S = (w. - w)/w. is the slip when the rotor speed is w.

With the machine on no-load
dw d dS
T. = J dt = J dt w.(l - S) = -w.J di (5.33)

so that from (5.32) and (5.33),

P=-Jw2Sds.
r s dt
Then the energy dissipated in the rotor when the slip changes from S 1 to S 2

J:
is given by

Energy = f Pr dt =
2
-Jw/S dS.

Thus, Energy= J~/ (S 1 2 - S 2 2 ) joules. (5.34)

When an ideal polyphase induction motor accelerates from standstill to

the synchronous speed w., the change in kinetic energy of the rotating mass
is given by: kinetic energy = !Jw/.
98 POLYPHASE INDUCTION MOTORS

The corresponding energy dissipated in the rotor winding is given from

(5.34) as: rotor energy loss = -!Jw/.
For an ideal machine, the input energy is given by
Input energy = Kinetic energy + Rotor energy = Jw. 2 •
Under these conditions the energy efficiency is given by
. Output energy
Energy effictency = I = 50%.
nput energy
Thus the maximum possible efficiency on an induction motor when the speed
is changing is 50%.
When the stator resistance can be neglected, the relationship between the
machine torque and the slip is given in (5.22) as

2
T. = Tm S S (5.35)
-+__!
S1 s
where Tm is the maximum torque and S 1 is the slip for maximum torque.
Then from (5.33) and (5.35)

2Tm =
-=--==- dS or -2Tm dt
-w J-, = - (S S
-st +-1)
dS.
s sl s dt w.J s
-+-
sl s
The timet taken for the slip to change from S 1 to S 2 is given by

2Tm t = - rs2 (§__ + St) dS

w.J Js, sl s
or _ w.J [S/ - S/ s 1 S1J· (5.36)
t - 2Tm 2SI + I og. s2
When the load torque is constant or is a linear function of speed, the
methods of analysis described in this section can be applied directly.

References
1. Fitzgerald and Kingsley, Electric Machinery, McGraw-Hill, 1961.
2. American Standard Test Code for Induction Motors, C.56.20.1954.
3. Alger, P. L., Nature of Polyphase Induction Machines, Wiley, 1951.
4. Rawcliffe, G. H., Burbidge, R. F., and Fong, W., 'Induction motor speed changing
by pole-amplitude modulation', Proc. I.E.E., paper no. 2597 U, Aug. 1958.
5. Fong, W., 'Wide ratio, two-speed, single-winding induction motors', Proc. I.E.E.,
Vol. 112, No.7, July 1965.
6. Rawcliffe, G. H., and Fong, W., 'Two-speed induction motors using fractional
slot techniques', Proc. I.E.E., Vol. 112, No. 10, Oct. 1965.
 TUTORIAL PROBLEMS 99

Tutorial Problems
1. An 8-pole, 200 V, 50 Hz, 3-phase induction motor runs at 726 r.p.m. Calculate the
slip r.p.m. per-unit slip and the frequency of the rotor current. If the machine
develops 5·5 h.p. gross at this speed, calculate the rotor copper loss and torque in
N.m.
(Answer: 24 r.p.m.; 0·032; 1·6 c/s; 136 W; 54 N.m)

2. A 6-pole, 400 V, 50 Hz, 3-phase induction motor has an output of 30 b.h.p. at a

slip of 0·02 p.u. If the torque required for mechanical losses is 20 N.m and the
stator loss is 0·9 kW, calculate the input power and efficiency.
(Answer: 25·9 kW; 0·865)
3. A 415 V, 50 Hz, 3-phase, 4-pole induction motor develops its full load power of
10 h.p. at 1425 r.p.m. and the input is 10 kW at power-factor 0·8lagging. On no-
load the input is 0·5 kW at rated voltage. Calculate the full-load slip, line current,
stator copper loss, rotor copper loss and torque in N.m.
(Answer: 0·05; 17·2 A; 1·62 kW; 0·419 kW; 50·1 N.m)

4. The parameters per phase referred to the primary of a 200 V, 3-phase, 4-pole,
50 Hz star-connected induction motor are as follows:
R1 = 0·11 n; X1 = 0·35 n; R21 = 0·13 n; X21 = 0·35 n; xm = 14 n.
Calculate the percentage error involved when the maximum torque ofthe machine
is determined, neglecting stator impedance.
(Answer: 142%)
5. A 415 V, 50 Hz, 25 h.p., 6-pole, 3-phase, star-connected induction motor is
started with an external resistance of 2 ohm in each supply line. The parameters
of its equivalent circuit are:
R1 = 0·25 Q; X 1 = 0·75 Q; R2 = 0·17 Q; X2 = 0-49 Q; Xm = 32 Q.
If the stator to rotor phase voltage ratio is 415/365, calculate the starting torque
developed and the corresponding operating power-factor for direct-on starting.
(Answer: 44·3 N.m; 0·872)
6. A squirrel-cage induction motor with negligible stator resistance produces a
starting torque of 150% of the full-load torque and a maximum torque of 250%
of the full-load torque. Calculate the slip for maximum torque and the slip for
full-load torque.
(Answer: 0·33; 0·071)
7. The reactance of the stator of a 3-phase induction motor is equal to the referred
value of the rotor reactance at standstill while each resistance is one-fifth of the
reactance. Find the values of the starting torque and the maximum torque if the
full-load torque of 520 N.m occurs at a slip of 0·04. Neglect no-load current.
(Answer: 200 N.m; 928 N.m)
8. A 200 V, 50 Hz, 6-pole, 3-phase, star-connected squirrel-cage induction motor
gave the following test results:
Line Line Total
voltage current power
Open-circuit 200V 7·9A 520W
Short-circuit 133V 48A 5540W
100 POLYPHASE INDUCTION MOTORS

The d.c. resistance per phase of the stator is 0·42 Q and the stator and the referred
rotor reactances are equal. Derive the parameters of the equivalent circuit.
(Answer: R 1 = 0·42 Q; X1 = X21 = 0·692 Q; R 21 = 0·382 Q; Xm = 13-5 Q)
9. A 6-pole, 400 V, 50 Hz, 3-phase induction motor has a star-connected stator and
rotor. The stator impedance is 0·6 + jl·5 Q per phase, the equivalent rotor
impedance at standstill is 0·6 + j2·0 Q per phase and the impedance of the
magnetizing branch is j35 Q. During starting, the rotor is connected to a star-
connected equivalent impedance of 1 + j0·2 Q per phase. Determine the starting
current and torque. If full load occurs at a slip of 0·05, find the normal full-load
current and the b.h.p. allowing a mechanical loss of 300 W.
(Answer: 55·2 A; 171 N.m; 19 A; 4·05 h.p.)
10. A 440 V, 50 Hz, 15 h.p., 6-pole, star-connected wound rotor induction motor has
the following parameters per phase:
R1 = 0·9 Q; R 21 = 0·8 Q; X1 = X21 = 2·0 Q; Xm = 48 Q.
The no-load losses are 494 W. Calculate
(a) the starting torque for direct-on starting.
(b) the maximum torque and corresponding slip.
(c) the maximum output power and corresponding slip.
(d) the values of external referred rotor resistance required to produce two-
thirds of maximum torque at standstill.
(Answer: 74·5 N.m; 187 N.m; 0·198; 14·56 kW, 0·158; 0·59 Q or 10·66 Q)
11. A 3-phase, 4-pole, 50 Hz, wound-rotor induction motor develops its maximum
torque of250% of full-load torque at a slip of0·2 when operating at rated voltage.
Determine the minimum value of per-unit applied voltage at which the machine
will supply full-load torque, and the external referred value of rotor resistance
required to produce maximum starting torque at standstill with the reduced
applied voltage, expressed as a percentage of the referred rotor resistance/phase
of the machine.
(Answer: 63·3%; 400%)
12. Find the ratio of starting to full-load current for a 400 V, 15 h.p., 3-phase
induction motor fitted with a star-delta starter if the full-load power factor is
0·85, the full-load efficiency is 88% and the short-circuit current is 40 A at 200 V.
(Answer: 1·25)
13. An 8-pole, 50 Hz, 3-phase, induction motor has a rotor resistance of0·04 ohm per
phase referred to the primary and the maximum torque occurs at 645 r.p.m.
Determine the ratio of starting torque to maximum torque if (a) direct-on starting
is used, (b) an auto-transformer with a 70% tapping is used for starting. Stator
resistance can be neglected.
(Answer: 0·275; 0·1335)
14. (a) Show that the maximum possible efficiency of polyphase induction motor
during run-up from standstill is 50%.
(b) A 3-phase, 50 Hz, 4-pole induction motor runs on no-load with a slip of 1%
and produces a torque of 40 N.m at a slip of 10%. The machine drives a load of
total moment of inertia 5 kg.m 2 and negligible friction. Determine an expression
for the speed of the machine following the application of a step load of 20 N.m,
if the torque-speed curve of the motor can be assumed to be linear over its
working range.
(Answer: 1425 + 60 e- 0 · 511)
6. Single-phase motors

Many small a.c. motors will be designed to operate from a single-phase

supply and will take the form of single-phase induction motors or a.c.
commutator motors. The single-phase squirrel-cage induction motor is the
most widely used form of machine but suffers from the disadvantage, noted
in chapter 1, that it produces no starting torque. In order to produce starting
torque, a second stator winding will normally be introduced so that, during
the starting period, the machine operates as an asymmetrical2-phase motor
connected to a single-phase supply. In these circumstances, the operation of
the symmetrical polyphase machine on a single-phase system can be used to
illustrate the principles underlying the performance of single-phase induction
motors.

6.1. The single-phase induction motor

It has previously been noted that most single-phase induction motors are,
in fact, 2-phase motors, in which one winding, the so-called auxiliary winding,
is disconnected from the supply when the speed reaches a certain value and
the machine then operates as a true single-phase motor with a single stator
winding, the so-called main winding. The connection diagram for such a
machine is shown in Fig. 6.1, in which the phase converter Y can be resistive,

Auxiliary
winding

Main
winding

v
Fig. 6.1. Single-phase induction motor.
102 SINGLE-PBASE MOTORS

capacitive, or inductive. This machine can be analysed (ref. I) by the method

of 2-phase symmetrical components and the unbalanced 2-phase system
vectors U A and U 8 can be written
UB = sp + Sn (6.1)
UA = ]Sp- ]Sn (6.2)
so that SP = t(UB - ]U A) (6.3)
(6.4)
These equations can only be applied to a symmetrical machine and, in
practice, the number of turns on the auxiliary phase winding will be different
to that on the main phase winding. Ideally, two windings having turns in the
ratio K: I should have impedances in the ratio J<2: 1. Thus, if the admittance
of the main winding is Ym, the ideal admittance of the auxiliary phase is
Yrnl J<2 and any difference between the actual admittance Ya and the ideal
admittance YrnlJ<2 of the auxiliary winding is assumed to be part of the
external phase balance Y. Under these conditions the symmetrical component
equations must be written with reference to the connection given in Fig. 6.1
as
18 = IP +In (6.5)
I
VA= K(JVP- ]Vn), lA = K (JJP - Jln) (6.6)

such that IP = VPYp and In = Vn Yn.

Two inspection equations can be obtained from Fig. 6.1 in the form
v =VB (6.7)
JA (6.8)
v =VA+-·
y
When the values for the sequence quantities from (6.5) and (6.6) are sub-
stituted in (6.7) and (6.8) it can be shown that
VP Yn + K(K- j)Y
(6.9)
V = 2J<2Y + Yp + Yn
Vn YP + K(K + j)Y
(6.10)
V = 2J<2Y + Yp + Yn.
In the particular case of standstill, YP = Yn = Y., so that
vp Y. + K(K- j)Y
(6.11)
v 2(J<2Y + Y.)
vn Y. + K(K + j)Y
(6.12)
v 2(J<2Y + Y.)
and T. = A.(V/ - V/). (6.13)
 THE SINGLE-PHASE INDUCTION MOTOR 103

It should be noted that, since at standstill there will be no coupling between

the two stator phase windings, the equivalent circuit of Fig. 6.1 can be re-
drawn in the manner shown in Fig. 6.2. The phase and line quantities can be
obtained directly from Fig. 6.2 so that
K 2Y
VA=VY.+~Y (6.14)

VY.
Vpc = Y. + KzY (6.15)

(6.16)
YY.
lA = v Y. + ~y (6.17)

IB = VY. (6.18)

IL = Y. :y~y [Y. + Y(l + ~)]. (6.19)

v
Fig. 6.2. Circuit for single-phase induction motor.

Considerable simplification in the analysis can be made by the use of a

normalization technique, by putting

ys Y. l_<E ~,-;; .
y = y~ = y ,IX = y(cos IX - j sm IX). (6.20)

The sequence voltages given by (6.11) and (6.12) can now be written
vp (ycOSIX + ~) -j(ysiniX + K)
(6.21)
V = 2(y cos IX + ~) - 2jy sin IX
Vn (y COS IX + ~) - j(y Sin IX - K)
(6.22)
V 2(y cos IX + K 2 ) - 2jy sin IX
104 SINGLE-PHASE MOTORS

The torque ratio t is given by

A.(Vp 2 - Vn 2 ) Ky sin (J(
t = = --=----=-------,- (6.23)
A.~ y 2 + 2K2 y cos (J( + K4
Equation (6.23) expresses the starting torque ratio t in terms of three
dimensionless parameters y, ()(, and K, so that the variations of the torque
ratio with variation in each of these parameters separately can be fully
investigated. In many cases the form of the external phase converter and the
turns ratio will be specified and only the effects of variation of the parameter
y will be investigated further. From (6.23) the condition for maximum starting
torque is given by

so that (6.24)
is the condition for maximum starting torque tm and
sin
= .
(J(
t (6.25)
m 2K(l + COS ()()
In the particular case of a machine with a main winding standstill im-
pedance angle of 45°, <P = -45°. Then, for a capacitive phase converter,
p = 90° and (6.25) can be written
sin 135° 1·21
t = = --· (6.26)
me 2K(l + COS 135°) K
Similarly, for a resistance phase converter p = oo, so that
sin 45o 0·209
t = = --· (6.27)
mr 2K(l + COS 45°) K
It immediately follows that, for a given K, a phase converter in the form of a
capacitor will produce considerably more starting torque than one in the
form of a resistor.
The imbalance factor U is the ratio of the magnitude of the negative
sequence current to that of the positive sequence current so that, at standstill,

U = In
/P
= Vn
VP
= [y 2

y2
+ 2Ky(K cos
+ 2Ky(K cos
(J( -

(J( +
sin()()
sin()()
+ K 2 (K2 +
+ K 2 (K2 +
l)Jt
1)
(6 . 28 )

and it can be shown that minimum imbalance will occur when

(6.29)
This value of y will also produce maximum starting torque per ampere of
starting current and can be taken as a suitable design criterion for the phase
converter.
 THE SINGLE-PHASE INDUCTION MOTOR 105

The phase and line quantities can also be written in normalized form so
that from (6.14) and (6.15)
VA K2
v = (y +
2 2~y cos IX + K4 )t
(6.30)

y
v
vpc
(6.31)
= (y 2 + 2K2y cos IX + K 4 )t
In a similar manner, from (6.17) and (6.19)
/A
(6.32)
VY. (y 2 + 2~y cos IX + K 4 )t
Jt.
=

~ = [y 2 + 2y(l + K 2 ) cos IX + (1 + K 2) 2
(6.33)
VY. y 2 + 2yK2 cos IX + K 4
Under these conditions, the complete starting performance of a single-
phase induction motor of the form illustrated in Fig. 6.1 can be obtained
directly. The split-phase motor is the particular case for which the phase
converter is in the form of a resistor and the capacitor-start motor is the
particular case for which the phase converter is in the form of a capacitor. It
follows directly from (6.26) and (6.27) that the capacitor-start motor will
produce considerably more starting torque than the split-phase motor. The
capacitor-start-capacitor motor employs a two-value capacitor to produce
(i) good starting performance, using a high value of capacitance and (ii)
practically balanced 2-phase operation under full-load conditions, using a
low value of capacitance.
EXAMPLE 6.1. An induction motor with two stator windings in space quadra-
ture is to be used as (a) a split-phase motor with K = 0·8, (b) a capacitor-start
motor with K = 1·3. The standstill admittance of the main winding is
0·1 f500 mho. If the phase converter is set to produce minimum imbalance
in each case, determine the value of the starting torque ratio, the p. u. phase
converter voltage, and the line current for each machine. Determine the value
of y required to limit the capacitor voltage to the supply voltage for a
capacitor-start motor.

Solution: The condition for minimum imbalance is given in (6.29) as y =

K~(l + K 2 ). Then (a) (b)
Split phase Capacitor start
K 0·8 1·3
(6.29) y = K~(l + K 2) 1·02 2·13
IX=/3-4> 50° 140°

(6.23) 0·41 0·776

106 SINGLE-PHASE MOTORS

(a) (b)
Split phase Capacitor start
vpc y
(6.31) 0·426 1·285
v (y 2 + 2K2y cos + K4}!
[y Jt
IX

~ = 2 + 2y(l + ~) cos IX + (1 - K2) 2

(6.33)
VYS y 2 + 2y~ cos IX + K 4
1·57 1·26
(6.23)
(6 _33 ) Torque per ampere 0·251 0·617

If the phase converter voltage is to be limited to be equal to the supply

voltage for a capacitor-start motor, it follows from (8.55) that

1 = y or 2y cos IX + K2 = 0.
(y 2 + 2K2 y cos IX + K4 )t
-~ -1·3 2
Then y = 2 COS IX= 2 COS 140 = }· 315 ·

6.2. Locus diagrams for single-phase motors

The starting torque produced by a single-phase motor is given by (6.13) as

T. = A(V/ - Vn 2 ) = ~ (I/-!/). (6.34)

s

The phase currents IM and lA can be written as

IM = /M and lA = /A (cos A. +j sin A.).
Now IM = IP + In and KIA = ]lp - Jln so that
21P = /M + KIA sin A.- jKIA cos A.
21n = /M - KIA sin A + jKIA cos A.
Thus I/ - I/ = K(/A/M sin A.). (6.35)
It follows directly from (6.35) that the starting torque is proportional to
the product of the phase currents and the sine of the angle between them. It
can also be seen from Fig. 6.1 that the line current IL is the algebraic sum of
the currents IM and lA.
In the case of a split-phase motor, the reactance X a of the auxiliary phase
is constant while its resistance Ra varies. Thus the locus of the auxiliary phase
current is circular and is of the form shown in Fig. 6.3. Then, for a given
point P on the locus of lAin Fig. 6.3, the distance AP represents lA and the
distance OP represents IL. Now PD = /A sin A., so that, from (6.35), PD is
 LOCUS DIAGRAMS FOR SINGLE-PHASE MOTORS 107

proportional to torque. In a similar manner sin IPOA = PD/OP and is a

measure of torque per ampere of starting current.
Thus, the starting torque will be a maximum when the distance PD in
Fig. 6.3 is a maximum, i.e., when the perpendicular to the line OA produced
from the operating point passes through the centre C of the circle; the
operating point under these conditions is P 1 •

Fig. 6.3. Operating chart for split-phase motor.

In a similar manner, the starting torque per ampere will be a maximum

when the angle POA is a maximum, corresponding to the operating point P 2
in Fig. 6.3 when the line OP 2 is a tangent to the circle.
For the particular operating point Pin Fig. 6.3

lA =
v = a - jb
.
RA + jXA
so that, equating real and imaginary parts,

RA = aXA = V- XAb (6.36)

b a
The required value of RA can then be obtained directly from the known
value of XA and the measured values of a and b.
In the case of a capacitor-start motor, the resistance RA of the auxiliary
phase will be constant while the total reactance varies. Thus the locus of the
auxiliary phase current is again circular and is of the form shown in Fig. 6.4.
The operating point for maximum starting torque is P 1 and that for maximum
torque per ampere is P 2 • A condition for minimum line current, correspond-
ing to the operating point P 3 in Fig. 6.4, can also be introduced in this case.
108 SINGLE-PHASE MOTORS

Fig. 6.4. Operating chart for capacitor-start motor.

For a particular operating point Pin Fig. 6.4

v
lA = =a+ jb
RA - j(Xc - XA)
where X c is the capacitor reactance and X A is the reactance of the auxiliary
phase. Then
bRA V- aR
Xc = XA + ---;;- = XA + b (6.37)

The required value of Xc can then be found from the known values of RA
and XA and the measured values of a and b.

6.3. The shaded-pole motor

The form of construction of the shaded-pole motor is illustrated in Fig.
6.5. The stator is of salient pole construction and carries a single phase
winding. Copper rings are placed around one-half of each pole in the manner
shown in Fig. 6.5, and the effect of the currents induced in these shading
rings is to produce a time lag in the flux under that part of the pole covered
by the shading ring relative to the flux under the other part of the pole. This
produces the effect of a net rotating flux in the air-gap and the machine will
 THE SERIES COMMUTATOR MOTOR 109

produce a small starting torque. This form of single-phase induction motor

has a relatively wide application when only a very low rating is required.

Fig. 6.5. The shaded-pole motor.

6.4. The series commutator motor

The schematic diagram representing the basic form of single-phase series
motor is shown in Fig. 6.6. In practice, both the stator and rotor will be
laminated and the machine can be operated from both a direct and alternating
current system. This motor will have a series characteristic and is used in

Field winding

Fig. 6.6. The series commutator motor.

relatively large sizes for traction purposes. However, low frequency operation
is necessary because of commutation difficulties, and it is also usual to include
a compensating winding connected in series with the armature to compensate
for the effects of the armature m.m.f.
The machine current I is given by I= Im sin w.t, and the flux by Cl> =
Cl>m sin w.t. An expression for the instantaneous torque t. can be obtained
110 SINGLE-PHASE MOTORS

from the torque expression of the d.c. machine given in(3.51) in the form
PZ...,. . 2
te = -2 'Vmlm sm w.t. (6.38)
rca
The average value Te of this torque is given by

Te = ! te = PZ Cl>m ~ or Te = PZ Cl>m I. (6.39)

2 2rca j2 j2 2rca j2
Thus the series motor develops less torque when operating from an a.c.
system than when operating from a d.c. system.
This motor has a wide application in fractional horse-power sizes,
particularly in domestic appliances, and is then known as the Universal
Motor.

6.5. The repulsion motor

Figure 6. 7(a) illustrates the form of the repulsion motor for which the
short-circuited armature winding is identical to that used for a d.c. machine.
The torque produced by the machine is a sine function of the angle oc in
Fig. 6.7(a), and it follows that the speed of the machine is controlled by the
brush position. It can be shown that the arrangement of this machine shown
in Fig. 6.7(a) can be simulated by that shown in Fig. 6.7(b) in which the

(a) (b)

Fig. 6.7. The repulsion motor:

(a) Connection diagram. (b) Simulation of connection diagram.

field winding is resolved along and perpendicular to the axis of the armature
brushes. Undl!r these conditions, the effect of the winding on the brush axis
can be correlated to the effect of the compensating winding on the series
motor so that the machine has essentially series characteristics. Maximum
torque will be obtained when the brush angle oc is approximately equal to 90°e.
In some cases, the action of the repulsion motor is combined with that of
the single-phase induction motor to produce the repulsion-induction motor.
 TUTORIAL PROBLEMS Ill

The machine is started as a repulsion motor with the corresponding high

starting torque and, at some predetermined speed, a centrifugal arrangement
short-circuits the commutator so that the machine then operates as a single-
phase induction motor.

Reference
I. Jha, C. S., and Daniels, A. R., 'The starting of single-phase induction motors
having asymmetrical stator windings in quadrature', Proc. l.E.E., Vol. I 06, Part A,
no. 28, Aug. 1959.

Tutorial Problems

I. The results of the locked rotor test on a 200 V, 4-pole, 50 Hz, star-connected
symmetrical, 3-phase induction motor were as follows: Line voltage 50 V, line
current 20 A; total power 866 W. The stator copper loss is 60% of the total copper
loss at standstill.
The machine is connected to a symmetrical3-phase supply and runs up to speed.
If one line is then disconnected from the supply, calculate the slip at which this
machine will develop maximum torque. The no-load current can be neglected and
an equivalent circuit for a 3-phase machine operating with a single-phase applied
voltage can be assumed.
(Answer: 0·103)

2. A single-phase capacitor motor has the following parameters: Rotor resistance

referred to main 2 n; Rotor reactance referred to main 3 n; Stator resistance main
2 n, auxiliary 8 n; Stator reactance main 3 n; auxiliary 4 n; Magnetizing reactance
referred to main 70 n.
If the turns ratio k = 1·2, calculate the value of capacitance to be used to produce
maximum starting torque and the per-unit value of the corresponding capacitor
voltage.
(Answer: 434 JlF; 0·676)

3. A symmetrical, 2-phase, 50 Hz, induction motor with a standstill impedance of

I 0I 46° ohm per phase has one of its phases provided with tappings. The machine
is to be operated as a capacitor-start, single-phase motor and is required to operate
with the magnitude of the voltage across the capacitor limited to the supply voltage
and the magnitude of the auxiliary phase current equal to its balanced 2-phase
value. Find the required values of capacitance and turns ratio.
(Answer: 221J,lF; 1·44)
4. A symmetrical, 240 V, 50 Hz, 4-pole, 2-phase induction motor is supplied as a
single-phase motor fitted with a phase balancer. If the standstill admittance per
phase of the machine is 0·068160" mho, calculate (a) the minimum possible value
of negative sequence voltage if the admittance Y is in the form of (i) a capacitor,
(ii) a resistor; (b) the value of capacitance to be used to give maximum torque;
(c) the value of capacitance to be used to give maximum torque per ampere of line
current.
(Answer: 0·236; 0·614; 217J,lF; 153J,lF)
112 SINGLE-PHASE MOTORS

5. Show that, for a split-phase, single-phase motor, maximum starting torque is

developed when the angle between the phase currents is equal to one-half of the
standstill phase angle of the main winding.

6. A symmetrical, 2-phase, 50 Hz, induction motor, with a standstill'admittance of

0·1 I 60° mho/phase, is to be operated as a single-phase motor with a phase con-
verter in series with one of its stator phases. The magnitude of the voltage across
the phase converter is to be limited to the supply voltage, and the magnitude of the
current through the phase converter is to be limited to the phase current of the
machine when operating under symmetrical conditions. Find the required com-
ponent values of the external phase balancer and the magnitude of the neutral
current under these conditions expressed as a percentage of the value for balanced
2-phase conditions.
(Answer: 5 Q; 367 ~F; 122·4)

7. An induction motor has stator windings in space quadrature and is supplied with
a single-phase voltage of 200 V at 50 Hz. The standstill impedance of the main
winding is 5·2 + jl0·1 ohms and the standstill impedance of the auxiliary winding
is 19·7 + jl4·2 ohms. If the external capacitance is to be used to start the machine,
sketch the machine stator connections and find, graphically or otherwise, (a) the
value of capacitance to be inserted to give the machine maximum starting torque
per ampere of starting current; (b) the value of capacitance to be inserted to produce
minimum starting current; (c) the value of capacitance to be inserted to give the
maximum starting torque.
(Answer: 128 ~F; 56 ~F; 165 ~F)

8. A single-phase induction motor of the split-phase type is supplied with 240 V

single-phase. The standstill impedance of the main winding is 7·71 + jll·2 ohms
and the leakage reactance of the auxiliary winding is 29·3 ohms. Determine
graphically or otherwise the resistance of the auxiliary winding required to give the
machine (a) the maximum starting torque, (b) the maximum starting torque per
ampere of starting current.
(Answer: 52 Q; 60·2 Q)
7. The polyphase synchronous machine

The polyphase synchronous machine can be classified as either a cylindrical

rotor or a salient pole machine according to its form of construction. Most
turbine-driven synchronous generators are high speed sets with a 2-pole
construction and are highly suited to the cylindrical rotor, rotating field type
of construction. Hydroelectric power generators are generally low speed
multipolar machines of the salient pole type. The synchronous motor can be
of either form of construction, although the majority of motors will be of the
salient pole type.
In the case of the cylindrical rotor machines, it is assumed that the air-gap
between the stator and rotor is constant, so that the reluctance of the mutual
magnetic circuits does not depend on the angular position of the rotor. In the
case of a salient pole machine, however, there are two definite axes of
symmetry and the reluctance will then be a function of the angular position
of the rotor. In these circumstances, the techniques used for the derivation of
equivalent circuits and performance characteristics of the two types of
machine are different.

7.1. Basis of operation of cylindrical rotor machines

The basic principles underlying the operation of polyphase synchronous
machines have been discussed in chapter 1 where the important concept of
load angle has been introduced.
In the basic analysis, it will be assumed that saturation can be neglected
and that the m.m.f. set up by the d.c. excitation is sinusoidally distributed in
space, so that all alternating voltages and currents are sinusoids. The machine
will be considered as a symmetrical polyphase machine and equivalent
circuits and phasor diagrams will be drawn on a 'per-phase' basis.
Consider the operation of a 3-phase, rotating field machine connected to a
constant voltage, 3-phase system with the machine excitation set to produce
an open-circuit voltage equal to the system voltage V. Under these conditions,
the machine is said to be 'floating' on the system and the corresponding
phasor diagram is shown in Fig. 7.1(a) which shows the separate time and
space varying quantities. The m.m.f., F, produced by the d.c. excitation is
equal to the resultant m.m.f., Fr (which sets up the resultant flux <~>r), since
114 THE POLYPHASE SYNCHRONOUS MACHINE

the armature current, 18 , and hence the armature reaction m.m.f., F •• , are
zero. Theintemale.m.f., Er, shown in Fig. 7.I(a) is setup by the flux produced
by the d.c. excitation acting alone and is known as the excitation e.m.f. In the
first stage of analysis, the leakage impedance of the armature will be neglected
so that the resultant e.m.f., E., set up by the resultant flux <D. will equal the
system voltage, V, and will be constant.
When the power applied to the shaft of the machine is increased, with
constant excitation, the rotor speed will increase as the machine enters the
generating mode. During this transient period, the axis of the field m.m.f. is

V = Ec ........... ~
..... \
\
\
\
\
v \

F = F,
Ec = V + I.z Ec = V- I.z
F, = F- F •• F, = F +F.,

(a) (b) (c)

Fig. 7.1. Simplified phasor diagrams:
(a) 'Floating'. (b) Generating. (c) Motoring.

driven ahead, in the direction of rotation, of the axis of the resultant m.m.f.
and this condition will continue until the electrical power output from the
machine balances the increase in applied shaft power. The axis of the field
m.m.f. will then have moved through the necessary load angle <5 9 and the
final steady state condition is shown in Fig. 7.I (b).
If, from the floating condition, mechanical output power is taken from the
shaft, with constant excitation, the rotor speed must decrease as the machine
enters the motoring mode. The axis of the field m.m.f. will then lag behind,
in the direction of rotation, the axis of the resultant m.m.f. and the transient
condition will continue until the electrical input power equals the required
value of load power on the shaft. The axis of the field m.m.f. will then have
moved through the necessary load angle <5m and the final steady state condition
is shown in Fig. 7.l(c).
In both the above cases, it follows from the phasor diagrams of Fig. 7.I (b)
and (c) that
F sin <5 = F•• cos¢. (7.1)
The armature reaction m.m.f., F •., is directly proportional to the armature
current, I., so that, at constant terminal voltage, the quantity F•• cos¢ is a
 BASIS OF OPERATION OF CYLINDRICAL ROTOR MACHINES 115

measure of electrical power. It follows directly from (7.1) that, at constant

excitation, electrical power is proportional to the sine of load angle and the
power-load angle characteristic of the machine is shown in Fig. 7.2. In the
absence of armature resistance, electrical power can be equated to gross
mechanical power.
Generator

Motor
Fig. 7.2. Power-load angle characteristic.

When the machine is operated at constant power, it follows from (7.1) that
F sin J must be constant. The loci ofF and F., with change in excitation are
then as shown in Fig. 7.3. The condition for which the field excitation exceeds

Locus of
F and F., Locus of
F and F.,

(a) (b)
Fig. 7.3. Locus diagram with constflllt power:
(a) Generating. (b) Motoring.
116 THE POLYPHASE SYNCHRONOUS MACHINE

the value required to set up an excitation e.m.f. equal to the system voltage is
known as over-excitation, and the condition for which the field excitation is
less than that required to set up an excitation e.m.f. equal to the system voltage
is known as under-excitation. It immediately follows from Fig. 7.3(a) that an
under-excited generator supplies leading voltamperes reactive (VAr) to the
system and an over-excited generator supplies lagging VArs to the system. In
exactly the same manner, it follows, from Fig. 7.3(b), that an under-excited
motor takes lagging VArs from the system and an over-excited motor takes
leading VArs from the system. Thus an over-excited motor operates at
leading power-factor and acts as a generator of lagging VArs. In these
circumstances, an over-excited motor can be used for power-factor correction
in a power system and if the machine operates on no-load, it is known as a
synchronous condenser.

7.2. Equivalent circuits for cylindrical rotor machines

In practice, the resultant e.m.f., E., produced by the resultant flux, <D., will
not equal the terminal voltage V because of the effects of the armature
resistance Ra, and the armature leakage reactance X 1 . These quantities are
related in the case of a generator by the equation
(7.2)
The resultant e.m.f. will differ from the excitation e.m.f. because of the
effects of armature reaction and, in the unsaturated case, the generator
phasor diagram can be drawn as shown in Fig. 7.4 in which the e.m.f. Ear

I
I
I

Fig. 7.4. Generator phasor diagram.

represents the effect of armature reaction and lags the current Ia by 90o. The
resultant e.m.f. can then be written as
Er = Er + Ear = Er - JlaXar (7.3)
where X.,. is known as the reactance of armature reaction.
EQUIVALENT CIRCUITS FOR CYLINDRICAL ROTOR MACHINES 117

I. x. r. I.

+ +

v v

Fig. 7.5. Generator equivalent circuit and phasor diagram.

I.

Fig. 7.6. Motor eq11ivalent circuit and phasor diagr11m.

118 THE POLYPHASE SYNCHRONOUS MACHINE

When the value of E, from (7.3) is substituted in (7.2) it follows that

Er = V + Ia[Ra + j(X1 + X 3 ,)] or Er = V + Ia(R + jX.)
3 (7.4)
where x. = X1 + Xar is known as the synchronous reactance and z. =
R3 + jX. is known as the synchronous impedance.
Forms of equivalent circuit and corresponding phasor diagrams for a
synchronous generator are shown in Fig. 7.5, and the corresponding equival-
ent circuits and phasor diagrams for a synchronous motor are shown in
Fig. 7.6.
Performance calculations under steady-state conditions will normally be
based on the equivalent circuits of Figs. 7.5 and 7.6, and unsaturated values
for the synchronous reactance can be found from the measured value of
armature resistance and the results of two simple tests.

7.3. Open-circuit and short-circuit characteristics

The open-circuit characteristic (O.C.C.) is the relationship between the
voltage appearing at the machine terminals on open-circuit and the field
current Ic (or m.m.f. F) when the machine is driven at synchronous speed.
Corresponding values of open-circuit voltage and field current are measured
for a range of voltages up to approximately 130% of rated voltage and plotted
in the manner shown in Fig. 7. 7, which takes the form of a normal magnetiza-
tion curve. The extension of the initial linear part of the O.C.C. is known as
the air-gap line.

/Airgap
1
1 line
I
I
I
I.

oL-------------------
Fieid current

Fig. 7.7. Open-circuit characteristic.

The short-circuit characteristic (S.C.C.) is the relationship between the

armature current / 3 and the field current Ic (or m.m.f. F), when the machine
is driven at synchronous speed with a symmetrical short-circuit applied to its
armature terminals, and corresponding values would normally be measured
for a range of armature currents up to 150% of the rated value. The form of
 OPEN-CIRCUIT AND SHORT-CIRCUIT CHARACTERISTICS 119

the S.C. C. is shown in Fig. 7.8 and will usually be a straight line through the
origin.

Fig. 7.8. Short-circuit characteristic.

A value for the synchronous impedance per phase of the machine can be
obtained directly from these two characteristics which are shown plotted on
the same axes in Fig. 7.9. In this figure, for the excitation OA, the unsaturated
value of synchronous impedance is given by AD/AB and the so-called
'saturated' value by AC/AB.

0 Field current
Fig. 7.9. Combined O.C. and S.C. characteristics.

It is, however, difficult in practice to consider the synchronous impedance

as a variable quantity and for operation near rated voltage, a value for the
synchronous impedance can be obtained by assuming that the machine is
equivalent to an unsaturated machine with an air-gap line passing through
the rated voltage point as shown in Fig. 7.9. Under these conditions, the
synchronous impedance is given by A 1C 1/A 1 B1 •
120 THE POLYPHASE SYNCHRONOUS MACHINE

7 .4. Basis of operation of salient pole machines

It has previously been noted that the air-gap of the salient pole machine
is not uniform and will, therefore, present two definite axes of geometric
symmetry. In these circumstances, it is usual to use the two-reaction theory of
Blondel and to resolve along two axes at right-angles, the so-called direct
axis (d-axis) corresponding to the axis of a field pole and the quadrature axis
(q-axis) corresponding to the axis of the interpolar space in the manner shown
in Fig. 7.10.
Direct axis

Fig. 7.10. Salient pole machine.

Consider the case of an unsaturated salient pole machine. The armature

reaction m.m.f. can be resolved into components along the direct and
quadrature axes and, since the geometric configurations along these two axes
are different, it follows that the component fluxes produced on these axes by
the relevant m.m.f.s will be different. It is apparent that, since the reluctance
on the quadrature axis will be greater than that on the direct axis, the armature
reaction reactance Xaq associated with the quadrature component of armature
current will be less than the armature reaction reactance Xad associated with

Er q-axis

d-axis
Fig. 7.11. Basic phasor diagram for salient pole generator.
 BASIS OF OPERATION OF SALIENT POLE MACHINES 121

the direct component of armature current. It follows directly that the

synchronous reactance Xd on the direct axis will be greater than the syn-
chronous reactance Xq on the quadrature axis, since the armature leakage
reactance X1 will be the same on both axes.

d-axis
Fig. 7.12. Complete generator phasor diagram.

The basic phasor diagram of a salient pole generator is shown in Fig. 7.11,
in which it is assumed that the angle between the excitation e.m.f. Er and the
armature current la is known. The excitation e.m.f. can be written, from
Fig. 7.11, in the form
Er = V + laRa+ jXdld + ]lqXq. (7.5)
Now 18 = (ld + lq) so that
JlaXq = jldXq + ]lqXq. (7.6)
When the value of jlq Xq from (7.6) is substituted in (7.5), Er can be written
Er = V + laRa + JlaXq + JliXd - Xq). (7. 7)

AB = I.(Xd - Xq)
Er = V- jldXd - jlqXq - r.I.
OA = V- I.(r. + jXq)
Fig. 7.13. Complete motor phasor diagram.
122 THE POLYPHASE SYNCHRONOUS MACHINE

In practice, the values of terminal voltage V and armature current Ia will be

known and the corresponding phasor diagram can be drawn by the con-
struction shown in Fig. 7.12 which is based on the form of (7. 7). The phasors
representing V and Ia can be drawn and the quantity V + Ia(Ra + jXq) set up
corresponding to the phasor OA in Fig. 7.12, which fixes the direction of the
quadrature axis as OA. The direct and quadrature axis components ofla can
then be found and the complete phasor diagram can be drawn. The corre-
sponding phasor diagram for a salient pole motor is shown in Fig. 7.13.
Simple equivalent circuits representing the operation of salient pole
machines cannot be drawn unless the direct- and quadrature-axis circuits are
assumed to have separate existences, and performance calculations will
normally be based on the use of phasor diagrams and the known values of
direct- and quadrature-axis synchronous reactance.

7.5. Determination of synchronous reactances

The value of the unsaturated direct-axis synchronous reactance can be
obtained by the method described in section 7.3, and allowance for saturation
made according to the methods of reference 1. In order to obtain a value for
the quadrature-axis synchronous reactance, additional experimental in-
formation is required. One such method is known as the slip test in which
balanced voltages of rated frequency are applied to the armature of the
machine with the field on open-circuit and the rotor is driven at a small slip
with respect to the armature rotating m.m.f. Oscillograms of armature
voltage and current and field voltage are taken and are of the form shown in

I
IDirect I Field
~axis I v~tage

7 :~
Fig. 7.14. Results of slip test.
 VOLTAGE REGULATION OF GENERATORS 123

Fig. 7.14. The direct-axis synchronous impedance is then the maximum value
of the ratio of armature voltage to current and the quadrature-axis syn-
chronous impedance is the minimum value of this ratio. It is normal to obtain
the ratio of these impedances from this test and to use the known value of the
direct-axis value to find the quadrature-axis value. In order to obtain
reasonable accuracy, the value of slip used during this test should be small
and, under these conditions, the machine will tend to synchronize and operate
as a reluctance motor. The machine must therefore be operated at reduced
voltage and the synchronous reactances obtained will be unsaturated values.
An alternative method of obtaining a value for the quadrature-axis
synchronous impedance is the maximum lagging current test, for which the
machine is operated as a synchronous motor on no-load. The excitation is
reduced to zero, then reversed and increased until the stability limit is reached.
It can be shown that, under these conditions, the quadrature-axis synchronous
impedance is the ratio of armature phase voltage to current and this value of
reactance will be a saturated value.

7.6. Voltage regulation of generators

Voltage regulation is defined as the percentage rise in terminal voltage
when the load is removed. That is,
Er- V
Regulation = V x 100%. (7.8)

One of the simplest ways of obtaining voltage regulation is to use the

equivalent circuit and phasor diagram of Fig. 7.5 in conjunction with the
open-circuit and short-circuit characteristics of Fig. 7.9. The value of
synchronous impedance can be obtained directly from the O.C.C. and S.C. C.
using either the air-gap line or making some allowance for saturation as
previously noted in reference 1, and the value of Er for specified load condi-
tions obtained directly. This method is known as the e.m.f. method but
suffers from the disadvantage that, for short-circuit, saturation is low. Hence
the synchronous impedance and regulation will be high and the method is
known as a 'pessimistic' method.
An alternative method of obtaining voltage regulation from the O.C.C.
and S.C. C. is known as the m.m.f. method in which it is assumed that rated
voltage corresponds to the resultant m.m.f. on the O.C.C. Then the resultant
m.m.f. Fr is set up from the O.C.C., and the armature reaction m.m.f. Far
corresponding to the load current is set up from the S.C.C.
The excitation m.m.f. Fis obtained from the relationship,
(7.9)
and the value of Er is obtained on the O.C.C. corresponding to the m.m.f. F.
This method is, however, approximate because it assumes that the leakage
124 THE POLYPHASE SYNCHRONOUS MACHINE

impedance is zero. The value of armature reaction m.m.f. Far is again

obtained from the S.C.C. and will therefore be low. In these circumstances,
the value of excitation e.m.f. and hence regulation will be low and the
method is known as an 'optimistic' method.

7. 7. Paralleling of polyphase synchronous machines

Most synchronous machines will operate in parallel with other synchronous
machines and the process of connecting one machine to another or to an
infinite busbar system is known as synchronizing.
In order that two 3-phase voltage sources can be successfully connected in
parallel, it is necessary that the system voltages are of the same magnitude
and frequency, of the same phase-sequence and that the respective phase
voltages are in phase at the instant of switching. The magnitudes of the system

Al ~ A2 Al CfJ A2

Bl ~ 82 Bl 82

cl ~ c2 cl c2
(a) (b)

Fig. 7.15. Three-lamp synchroscope:

(a) Straight connection. (b) Cross connection.

voltages can be found using a voltmeter and their phase sequences found
using a phase-sequence meter (ref. 2). The conditions governing frequency
and instantaneous phasing can be satisfied using a synchroscope.
The most elementary form of synchroscope consists of three lamps
connected across the paralleling switch in either of the two manners shown in
Fig. 7.15. If the lamps are connected in the manner shown in Fig. 7.15(a), the
instantaneous voltages across all three lamps will be the same and will be
zero when the two systems are in phase. If two of the lamps are cross-
connected in the manner shown in Fig. 7.15(b), the instantaneous voltages
across the three lamps will be different, and the lamps will be illuminated in
cyclic order. The paralleling switch should be closed at the instant when the
straight connected lamp A is out and the other two lamps are at equal
brilliancy. The rate at which the lamps flash depends on the frequency
difference between the two systems. The order in which the lamps flash is
determined by the relative speed between the two systems and this method
shows which of the frequencies is the greater.
 CYLINDRICAL ROTOR MACHINES ON INFINITE BUSBARS 125

In most practical systems, a more elegant indicator in the form of a small

2-pole motor (ref. 3) would be used as a synchroscope.

7.8. Cylindrical rotor machines on infinite busbars

Consider the operation of a 3-phase generator connected to a constant
voltage, constant frequency system. The phasor diagram corresponding to
this condition is shown in Fig. 7.16 in which the system voltage V is taken as

I.

Fig. 7.16. Generator on infinite busbars.

reference. The nett electrical output power Po from the generator is given by
(7.10)
where fli is the 'real part of' and I.* is the conjugate of I •. It follows from
(7.10) that
I = Er - V = Ec[~ - V_
a z. z.~

Thus I *= Ec\1 - V = Er~ - V~. (7.11)

a z.\0 z.
The value ofl. *from (7.11) can be substituted in (7.10) to give

Po= 3fli {
Er~- v~]
z
s
or Po= z3V [Er cos (<5 -
s
(}) - V cos e].
(7.12)
If the machine is operated at constant excitation, Ec will be constant and it
can be seen from (7 .12) that maximum electrical output power P om is obtained
when the load angle t5 is given by

u
~ = (} x.·
=tan - l - (7.13)
R.
126 THE POLYPHASE SYNCHRONOUS MACHINE

Then, from (7.12)

3V
Porn = - (Ec - V COS ()). (7.14)
zs
It follows, from (7.14), that, theoretically, the power available from a
generator increases indefinitely with excitation. In practice, however, there
will be a maximum permissible value of field current and a power limit
imposed by the capacity of the prime mover.
If armature resistance is neglected, Zs [!!. = Xs [2Qa and (7 .12) can be written
as
3VEr sin <5
p = .
0 xs (7.15)

Under these conditions, the electrical output power will be at maximum

when the load angle <5 is 90°.

I.

Fig. 7.17. Motor on infinite busbars.

The corresponding phasor diagram for a motor operating on infinite

bus bars is shown in Fig. 7.17, and the gross mechanical output power P 9
from the motor is given by
P9 = 39fEria*. (7.16)
It follows from Fig. 7.17 that

I -
V - Er V - E/\lJ .
a- zs -- zsl!!_

Then I* = V- Ec[~ = VI!!_- Er~. (7.17)

a Zs\(j Zs
The value ofl.* from (7.17) can be substituted in (7.16) to give

p = 39fE'IJ>[Vl!!_-
g f zEr~J
s

or P9 z
= 3Er [ V cos (<5
s
- ()) - Er cos () . J (7.18)
 CYLINDRICAL ROTOR MACHINES ON INFINITE BUSBARS 127

If the machine is operated at constant excitation, Er will be constant and it

can be seen from (7 .18) that maximum gross output power P gm will occur
when the load angle J is given by

u
~ = (} =tan -1Xs
-· (7.19)
Ra
Then, from (7.18),

Pgm = -3Ec [ V - Ec COS (} ] • (7.20)

zs
When (7.20) is differentiated with respect to Ec it follows that, theoretically,
ultimate mechanical output power is obtained from a synchronous motor
when Ec is given by

Ec = - - - = --·
v vzs (7.21)
2 COS(} 2Ra
In practice, there will be a maximum permissible value of excitation.
When armature resistance is neglected, (7 .18) can be written as
3EcV sin J
(7.22)
p
g
=
xs .

Under these conditions, the gross mechanical output power will be a

maximum when the load angle J is 90°.
It can be seen from (7.15) and (7.22) that the power-load angle characteris-
tic for a cylindrical rotor machine is a sinusoid with its maximum value
depending on the excitation.
Consider a cylindrical rotor machine with negligible armature resistance

Locus of Er
Generator Motor
Fig. 7.18. Operation with fixed power.
128 THE POLYPHASE SYNCHRONOUS MACHINE

connected to infinite busbars and operating with constant power. Since the
nett electrical power Po equals the gross mechanical power P9 , it follows from
(7.15) and (7.22) that

Po = p 9 = 3Ec VX sin J = 3 VI cos 'I'·

,~.. (7.23)
s

Thus Er sin J and /a cos 4J are both constant with constant power and the
loci of Er and 18 for differing values of excitation in both the motoring and
generating cases are shown in Fig. 7.18. The relationship between the arma-
ture current and the field current under these conditions is illustrated in Fig.
7.19, and these curves form the characteristic Vee curves. Lines of constant
power-factor are also shown on Fig. 7.19, and it is apparent that the point of

0·8 0·8

lag lead

Fig. 7.19. Vee curves.

minimum armature current corresponds to the unity power-factor point.

The curves shown in Fig. 7.19 are typical experimental curves and the effect
of armature resistance and saturation is to move the locus of the unity power-
factor line from a theoretical vertical line as shown.
 SALIENT POLE MACHINES ON INFINITE BUSBARS 129

7.9. Salient pole machines on infinite busbars

The phasor diagram corresponding to the operation of a 3-phase, salient
pole, synchronous generator with negligible armature resistance on infinite
busbars is shown in Fig. 7.20. The net electrical output power Po is the sum

Fig. 7.20. Salient pole generator on infinite busbars.

of the products of the in-phase components of voltage and current and can
be written as
P0 = 3((V sin {J)/d + (V cos {J)/q]. (7.24)
Then, from Fig. 7.20,
(7.25)
IqXq = V sin(>. (7.26)
When the value of /d from (7.25) and the value of Iq from (7.26) are sub-
stituted in (7.24), Po can be written
V sin (J V cos (J •
Po= - X [Ec- Vcosb] + --(Vsmb)
d Xq
Ec V sin (J VZ sin (J cos (J VZ sin (J cos (J
----=----
xd xd
+--~=----
xq
P = Ec V sin (J VZ(Xd - Xq) i 2(>
That is, o X + 2XX sn . (7.27)
d d q

The term V:Z(~; ~ Xq) sin 2(> in (7.27) is the reluctance torque introduced
d q

by the effects of the salient poles and is independent of the excitation.

Under the condition of negligible armature resistance, the gross mechanical
output from a salient pole motor will equal the net electrical power from a
generator. The power-load angle characteristic for a salient pole machine is
shown in Fig. 7.21. For a specified operating voltage and fixed direct-axis
synchronous reactance, the salient pole machine will operate at a lower value
of load angle than a cylindrical rotor machine for a given power.
130 THE POLYPHASE SYNCHRONOUS MACHINE

The condition for a maximum power at fixed excitation can be found by

differentiating (7.27) with respect to load angle b. Then
dPo _ ErV cos b V 2 (Xd - Xq) 2~ _
db - X
d
+ X X
d q
cos u - 0

that is, Er cos b + v(x Xq- x)q (2 cos

d 2 b - I) = 0.

''
' \
\
\
\
\
\ 180°
Load angle

Fig. 7.21. Power-load angle characteristic.

On solution, the load angle for maximum power is given by

-Er ± [ E/ + sP(xd .;q xqYT

4 v(xd .;q xq)
cos b = (7.28)

When the salient pole machine is analysed as a cylindrical rotor machine,

by assuming that Xd = Xq, calculated Vee curves are reasonably accurate
but large errors are involved in the values of load angle.

7.10. Parallel operation of generators

Most large power distribution systems consist of a number of generators
in parallel supplying a common bus bar system such that the system capacity
is assumed to be infinite compared with the rating of an individual machine.
In these circumstances, it is usual to assume that the synchronous impedance
of the machine is constant under all loading conditions and, in many cases,
armature resistance will be neglected. The output power from a generator
operating under these conditions will be controlled solely by the power from
the prime mover, and the division of load between the generators can only be
 GENERATOR OPERATING CHARTS 131

changed by control of their respective prime movers. The operating power-

factor of a machine is governed directly by the excitation current.
Consider two identical machines A and B operating in parallel and supply-
ing a common load of fixed impedance. The phasor diagram corresponding
to the condition for which the machines share the load equally and operate at
the same power-factor is shown in Fig. 7.22(a). If the excitation of machine A
is now decreased, the excitation of machine B must be increased in the manner
shown in Fig. 7.22(b), if the terminal voltage is to be maintained constant.
Es

(a) (b)
Fig. 7.22. Generators in parallel.

If the excitations are not controlled in this manner, the terminal voltage must
change and produce a change in the total load. In these circumstances, since
the power from both prime movers has not been changed, the speed of the
machines and hence the supply frequency must change. It is usual, in practice,
to control the throttle setting of the prime mover to obtain constant system
frequency and to ensure that the load is divided between the machines in
proportion to their ratings. Automatic voltage regulators are used in the
excitation circuit to control the operating power-factor of the machine.
The performance of a synchronous generator supplying power to a large
capacity system will normally be obtained from a locus diagram designed to
show the simultaneous variations of the necessary quantities.

7.11. Generator operating charts

Consider the phasor diagram shown in Fig. 7.23 for a cylindrical rotor
machine with negligible armature resistance. For fixed excitation, the length
OC is fixed and the locus of Er will be a circle, centre 0, radius OC. The
machine will be assumed to be connected to constant voltage bus bars, so that
132 THE POLYPHASE SYNCHRONOUS MACHINE

the length OA in Fig. 7.23 is fixed and the points 0 and A are fixed. Then,
from Fig. 7.23,
AM = I.X. cos cjJ and AN = I.x. sin c/J.
Hence the length AM is proportional to output power and the length AN is
proportional to output volt.amp-reactive. A simplified form of operating
C _____ 1 N
I I.x. I
I ¢ I
ML-

0
Fig. 7.23. Phasor diagram for cylindrical rotor generator.

chart for a 3-phase, star-connected, cylindrical rotor machine is shown in

Fig. 7.24, such that
3V
AC = 3 VI. = (I. X.) X and OA = (j3 V) •
2

s x.

Stability
limit
b = 90°

Circles

-Constant
power-factor
I line
Constant
power
line

Fig. 7.24. Simplified operating chart.

 GENERATOR OPERATING CHARTS 133
v2
Now x. = x.pu cvA),
and it follows that
OA = (j3 ~ 2 (VA),= (VA), when J3 V = 1
v, xspu xspu v,
where V, is the rated voltage and (VA), is the rated volt-amperes.
The length OA in Fig. 7.24 is proportional to the excitation e.m.f. and
determines the scale of constant e.m.f. circles. The prime mover will impose a
limit on the available power, and then there will also be a maximum per-

0·6

t
VArs
lead

t
VArs
lag

Fig. 7.25. Practical operating chart.

134 THE POLYPHASE SYNCHRONOUS MACHINE

missible value of field current imposing an excitation limit and a maximum

permissible value of armature current. The theoretical limit of stability occurs
when (J = 90° and it is apparent that there must be a safety margin between
the theoretical limit and that used in practice. The practical stability limit can
be expressed as a fixed value of leading volt.amp-reactive or as a 10%
stability margin on excitation. The form of the practical operating chart for a
cylindrical rotor machine is shown in Fig. 7.25.

7.12. Starting of synchronous motors

It has previously been noted that the synchronous motor does not produce
a steady starting torque and it is therefore necessary to provide a starting
method for this motor.
A small induction motor, a so-called Pony motor, can be direct-coupled
to the main motor and used to run the synchronous motor up to a speed just
below the synchronous speed. If d.c. excitation is then applied, the machine
will pull into step and operate as a synchronous motor.

Fig. 7.26. Synchronous-induction motor.

It is usual to provide all synchronous machines with short-circuited damper

windings on the field circuit, and these windings can be used to produce an
asynchronous torque when the armature is supplied with polyphase voltage.
Thus the machine can be run up to just below synchronous speed with the
field circuit shorted through a resistor.
The most elegant starting method for a synchronous motor is to use a
machine in the form of the so-called Synchronous Induction motor. The
field winding is in the form of a standard 3-phase winding and the machine
is started as a 3-phase wound-rotor induction motor with a variable external
rotor resistance. When the starting resistor has been short-circuited the rotor
is reconnected in the manner shown in Fig. 7.26. Alternatively, the exciter
can be permanently in circuit.
 TUTORIAL PROBLEMS 135

References
1. Kingsley, Charles, Jr., 'Saturated synchronous reactance', Trans. A.L.E.E., Vol.
54, no. 3, pp. 300-305.
2. Russell, R. L., 'The determination of phase rotation of polyphase systems',
Proc. I.E.E., Vol. 102, part 4, no. I, Feb. 1955.
3. Lincoln, P. M., 'Synchronisation and frequency indication', Trans. A.I.E.E.,
no. 18, p. 225.

Tutorial Problems
I. A I 00 h. p., 415 V, 3-phase, star-connected synchronous motor with a synchronous
reactance of 2 ohms per phase and negligible armature resistance has a full load
efficiency of0·92. Calculate the excitation e.m.f., the total power developed and the
operating load angle for full-load at power-factor 0·8lead and power-factor 0·8lag.
(Answer: 81·1 kW; 142 A; 466 V; 29°; 240 V; 73°)

2. A 6·6 kV, 50 Hz, star-connected, 3-phase synchronous generator having a

synchronous reactance of9·5 ohms per phase operates on 6·6 kV infinite busbars
with the field current set to produce an excitation e.m.f. of 1·1 per-unit. Calculate
(i) the maximum value of nett power the machine can deliver to the bus bars, its
operating power-factor and (ii) the total power developed under these conditions.
(Answer: 5·05 MW; 0·74lead; 5·05 MW)

3. Repeat Problem No.2 for a machine with a synchronous impedance of 1·16 +

j9-43 ohms per phase.
(Answer: 4·49 MW; 0·673lead; 4·85 MW)

4. A 3-phase, star-connected turbo-alternator with a synchronous reactance of 10

ohms per phase supplies 220 A at unity power-factor to 11,000 V constant fre-
quency busbars. If the throttle is held constant while the excitation e.m.f. is
increased by 25% determine the new current and power-factor. Neglect armature
resistance.
(Answer: 281 A; 0·806lead)

5. A 6·6 kV, 3-phase, 50 Hz, star-connected synchronous motor has a synchronous

impedance of 12 exp j82o ohms per phase. The field current of the machine is set
to produce a per-unit excitation e.m.f. of 1·3. The machines operate on constant
frequency 6·6 kV infinite busbars. Calculate the operating power-factor and the
line current of the machine when the input power is 800 kW. What is the maxi-
mum value of gross output power the machine can develop with this value of field
current? A graphical, or part-graphical, solution will be accepted.
(Answer: 0·592lead; 125 A; 2·97 MW)

6. A 3-phase, star-connected turbo-alternator with a synchronous reactance of

8 ohms per phase delivers 200 A at unity power-factor when connected to II kV
constant frequency bus bars. If the throttle is held constant while the excitation
e.m.f. is increased by 30% determine the new current and power-factor. The
excitation e.m.f. is now held constant at the new value and the steam supply is
gradually increased. Find the value of output power at which the machine will
break from synchronism and the operating power-factor under these conditions.
Neglect armature resistance.
(Answer: 324A; 0·622lag; 19·8 MW; 0·6lead)
136 THE POLYPHASE SYNCHRONOUS MACHINE

7. A 6·6 kV, 50 Hz, 3-phase, star-connected synchronous motor has an open-circuit

characteristic given by
Field current p.u. 0·80 1·0 1·20 1·40 1·60
Open-circuit voltage p.u. 0·83 1·0 l-13 1·23 1·30
The unsaturated synchronous reactance is 200 Q and resistance can be
neglected.
Find the per-unit field current required for a b.h.p. of 100 at efficiency 0·88
when the operating power factor is 0·9 lead.
(Answer: 1·51)
8. An over-excited 3-phase, 50 Hz, star-connected synchronous motor with a
synchronous reactance of 100 ohms per phase and negligible resistance is to be
used to improve the overall power-factor of a system supplying a factory to unity,
and to provide a constant speed drive with an additional 50 h.p. The existing
factory load is 100 kVA at power-factor 0·6lagging supplied from 6·6 kV, 50 cjs,
3-phase, infinite busbars. Determine the rating, operating power-factor and the
value of excitation e.m.f. for the synchronous motor if, under these conditions,
the per-unit efficiency is 0·8. Describe a suitable method of starting such a motor.
(Answer: 92·5 kVA; 0·5lead; 4540 V)

9. Show that the total reactive volt.amperes Q for a cylindrical rotor machine are
given by
V2 VEe
Q = 3- - -cos i5.
x. x.
Determine the maximum value of Q the machine can deliver with fixed excita-
tion. Resistance can be neglected.

10. A 3-phase, salient pole synchronous motor operates at rated voltage on infinite
busbars. If the per-unit values of the synchronous reactances are Xd = 1·0,
Xq = 0·8 and resistance can be neglected, calculate the percentage of rated power
the machine can develop with (a) zero excitation, (b) 0·8 p.u. excitation. Find the
corresponding load angle and armature current in each case.
(Answer: 0·125; 45°; 1·26; 0·834; 74·65°; 1·26)

II. Construct the phasor diagram for a 20 MVA, 3-phase, star-connected, 50 Hz,
salient pole synchronous generator with Xd = I p.u., Xq = 0·65 p.u., R. = 0·01
p.u. delivering 15 MW at power-factor 0·8lag to an 11 kV, 50 Hz system. Find
the load angle and per-unit excitation e.m.f. under these conditions.
(Answer: 18°; 1·73)
8. The d.c. machine

The basic principles underlying the operation of d.c. machines have been
discussed in chapter 1 where it has been noted that the field winding is
mounted on salient poles on the stator and the armature winding is wound in
slots on a cylindrical rotor. The frequency of the e.m.f. generated in the
armature depends directly on the mechanical speed of rotation, and the action
of the commutator is that of a frequency changer so that the frequency at the
commutator brushes is zero in this particular case. Various arrangements of

i.
+

0
'r

Vr
Lr

r v.

. L dir
Vr = 'r'r + r dt
. di•)
v = e - ( '·'· + L a dt
(a) (b) (c)

Fig. 8.1. Physical and schematic arrangements:

(a) Physical. (b) Schematic 1. (c) Schematic.2.

armature windings have been discussed in chapter 2 where it has been noted
that the commutator brush is normally situated on the centre line of a main
pole, although it is connected to a coil in the interpolar gap. It is usual to use
a schematic representation of a d.c. machine in which the brushes are shown
in the position of the coil to which they are connected, and the physical
arrangement together with the two schematic diagrams to be used in this
text are shown in Fig. 8.1 for the particular case of a generator. The torque
138 THE D.C. MACHINE

angle (} defined in chapter 1 will normally be 90° so that, under steady state
conditions, any difference between generated e.m.f. and terminal voltage is
governed by armature resistance voltage drop. The generated e.m.f. E is
given from (3.35) as
pz
E = - Cl>w = KCI>w (8.1)
2na

and the electromagnetic torque, T., from (3.51) as

pz
T. = -2 Cl>/a = KCI>Ia. (8.2)
na

In the case of a motor, the terminal voltage will always be numerically

greater than the generated e.m.f. and the machine torque will produce
rotation against a load. For a generator, the terminal voltage will be less than
the generated e.m.f. and the machine torque will oppose that applied to the
shaft by the prime mover.

8.1. Armature reaction

Any current in the armature conductors must produce an m.m.f. with an

associated flux, and this flux, the so-called flux of armature reaction will
combine with the field flux to produce the resultant flux of the machine. A
developed diagram showing the component and total m.m.f.s and fluxes in a
2-pole machine is given in Fig. 8.2. The m.m.f. of the armature current will be
of a stepped form shown in Fig. 8.2(b) which can be approximated to the
triangular form shown. The resulting armature reaction flux is shown in
Fig. 8.2(b) and is low in the interpolar gaps because of the high reluctance of
this area. The form of the resultant flux is shown in Fig. 8.2(c) and its
magnitude will be less than that produced by the field alone. It can be seen
from Fig. 8.2(c) that the effect of armature reaction is to distort the flux
wave and to shift the position of the magnetic neutral axis (m.n.a.) in the
direction of rotation for the generating case and against the direction of
rotation for the motoring case.
The effect of total flux reduction by armature reaction is known as the
demagnetizing effect and this effect can be completely removed by the use of
a distributed compensating winding arranged in slots in the main poles so that
the effects of the m.m.f. set up by currents in armature conductors situated
under the main poles is cancelled. Thus the compensating winding is con-
nected in series with the armature winding and the total m.m.f. per pole of this
winding must equal the m.m.f. of that portion of the armature under any one
pole, i.e., be equal to the number of armature conductors per pole per
parallel path times the ratio pole arc-pole pitch.
 THE COMMUTATION PROCESS 139

Pole piece Pole piece

-Rotation---
Motor Generator
(b)

(c)

Fig. 8.2. Component and total m.m.f. and flux:

(a) Field alone. (b) Armature alone. (c) Resultant.

8.2. The commutation process

When the commutator bar which is connected to a particular armature
coil rotates past a fixed commutator brush it follows that the coil current
must reverse. In many cases, a brush will span more than one commutator
segment, so that, during commutation, several coils in series will be short-
circuited. The reversal of the coil current induces a voltage of self-inductance,
the so-called reactance voltage, which opposes the change of current so that
a spark could appear at the trailing edge of the brush. The ideal process of
commutation is illustrated in Fig. 8.3 and the reactance voltage will produce
the condition known as 'under commutation' shown in Fig. 8.4.
It can be seen from Fig. 8.2 that one of the reasons for poor commutation
is the effect of armature reaction flux in shifting the position of the magnetic
neutral axis. This flux distortion and the reactance voltage will always exist,
140 THE D.C. MACHINE

but their effects on commutation can be countered by the introduction of a

so-called 'commutation voltage'. In the case of small machines, this voltage
can be introduced by shifting the brush position but the most general method
is to insert small auxiliary poles centred on the interpolar gap, known as
compoles, connected in series with the armature. Since the compole must

I
I
I
j-/c
Timet
I
I
Timet

I 1I \\ ,...._... Sparking
1
I
I I
I
1---T---l
I I l
1--T---l
Fig. 8.3. Ideal commutation. Fig. 8.4. Under commutation.

produce flux proportional to armature current, it must operate with an

unsaturated magnetic circuit, and the air-gap between the armature and the
compole will normally be greater than that between the armature and the
main-pole. The arrangement of a 2-pole generator with compoles is shown in
Fig. 8.5 and the compote has the same polarity as the main pole ahead in the
direction of rotation. In the case of a motor, the arrangement is shown in
Fig. 8.6 and the oompole has the same polarity as the main pole behind in
the direction of rotation. In these circumstances, with the correct design of

Fig. 8.5. Generator with compotes. Fig. 8.6. Motor with compotes.
 GENERATOR PERFORMANCE CHARACTERISTICS 141

compole, a d.c. machine can be operated as either a motor or a generator

with either direction of rotation.
A developed diagram showing the component m.m.f. and flux patterns
for a generator fitted with a compensating winding and compoles is shown in
Fig. 8. 7, and it can be seen that the total flux is almost the same as that of the
field winding acting alone.
Com pole
I 0 0 0
Main pole
0 0 0 0 ol U lo Main pole
?OOOOO~
I

Field alone

~ Armature

=:__// ~e
Compole
v alone

~ ~ting
~ winding alone

Fig. 8.7. Component and total m.m.f. and flux for machine with compotes and compensating
winding.

8.3. Generator performance characteristics

The terminal voltage V of a d. c. generator is related to the armature current
Ia and the generated e.m.f. E, by
(8.3)
142 THE D.C. MACHINE

where R. is the total internal armature resistance. The value of the e.m.f. E, is
given by (8.1) and is governed by the field flux <I> and the angular velocity w
of the rotor. A curve relating the e.m.f. to the field current at constant speed
is known as the magnetization curve or open-circuit characteristic. In the case
of a separately excited machine, the field voltage is independent of the
armature voltage; for a shunt connected machine the field is in parallel with
the armature and for a series connected machine, the field is in series with the
armature. It follows immediately that, for a shunt or series connected
generator, the armature voltage will only build up if residual flux is present.
In general, the three characteristics which specify the performance of a
d.c. generator are:
(i) the open-circuit characteristic,
(ii) the external characteristic which gives the relationship between the
terminal voltage and load current at constant speed,
(iii) the load characteristic which gives the relationship between the terminal
voltage and the field current, with constant armature current and speed.
All other characteristics depend on the form of the open-circuit characteris-
tic, the load and the method of connection.

8.4. The separately excited generator

The schematic connection diagram for a separately excited generator is
shown in Fig. 8.8 and the form of the open-circuit and load characteristics
are shown in Fig. 8.9. When the voltage drop in the armature circuit, for a
I.
+

v.

Fig. 8.8. Schematic diagram for separately excited generator.

given armature current, is added to the load characteristic the curve C shown
in Fig. 8.9 is obtained. Then, at a given field current OA, the distance CD
represents the voltage drop arising from armature reaction. The form of the
external characteristic is shown also in Fig. 8.9, and the terminal voltage
falls slightly as the load current increases.
Voltage regulation is defined as the percentage change in terminal voltage
when full load is removed, so that, from Fig. 8.9
E-V
Voltage regulation= - V - x 100%. (8.4)
 THE SHUNT GENERATOR 143

Since the separately excited generator requires a separate d.c. field supply,
its use is limited to applications where a wide range of controlled voltage is
essential.
E

0 Load current rated

Field load
Fig. 8.9. Open-circuit and load characteristics.

8.5. The shunt generator

This machine operates with the field connected in parallel with the
armature in the manner shown in Fig. 8.10. The open-circuit characteristic
has the same form as that of the separately excited machine and, since this

v.

Fig. 8.10. Schematic diagram for shunt generator.

144 THE D.C. MACHINE

curve shows the relationship between the armature voltage V, and field
current Ir, a field resistance line slope Rr = V/ Ir can be incorporated on the
characteristic. Figure 8.11 shows the form of the characteristic with a field
resistance Rr = OA/OB so that the open-circuit voltage will build up to a
value OA. It follows that there must be a critical value of field resistance, such
that the field resistance line is coincident with the linear part of the character-
istic. This value of resistance, known as the critical field resistance, is shown

Rc R

Field current
Fig. 8.11. Open-circuit characteristic.

as the line 0 Rc in Fig. 8.11 and is the maximum permissible value of field
resistance if the armature voltage is to build up. For a value of field current
given by OC in Fig. 8.11, the field resistance line OR can be drawn and the
field voltage and, hence the terminal voltage, are given by CD. The generated
e.m.f. is given by CF and hence the armature resistance drop is given by DF.
Various graphical techniques are available to determine the variation of
terminal voltage with armature current when the form of the open-circuit
characteristic, the field resistance and turns, the demagnetizing ampere turns
and the armature resistance are known. One such technique is illustrated in
Fig. 8.12. For an assumed value of armature current O'L, the effective
reduction in total ampere turns produced by demagnetization is known.
Then the equivalent reduction in field current produced by demagnetization
can be set up as the length OA in Fig. 8.12, such that the length OA is the
demagnetization ampere turns divided by the field turns. The armature
resistance drop for the assumed value of armature current is given by the
length AB in Fig. 8.12. A line through B parallel to the field resistance line
OR is now drawn to meet the O.C.C. at the points C and D. Horizontal lines
CM and DN are drawn to meet the vertical from L at points M and N. Then
M and N are points on the terminal voltage-armature current characteristic
 PERFORMANCE CHARACTERISTICS OF MOTORS 145

and, if this process is repeated for several different values of armature current,
the complete characteristic in the form shown in Fig. 8.12 can be obtained.
The external characteristic can then be obtained by subtracting the field
current Ir from the armature current for each point on the voltage-armature
current characteristic. It can be seen from Fig. 8.12 that there is a maximum
value of armature current so that, beyond this point, a further decrease in
load resistance will produce a decrease in armature current. Under short-
circuit conditions, the armature current will be the ratio of the residual voltage
to the internal armature resistance. When the demagnetizing effect is neg-
lected the distance OA in Fig. 8.12 becomes zero and the construction for the
external characteristic is simplified.
R

Field current L Armature current

Fig. 8.12. Graphical technique to find V.-I. curve.

A shunt generator maintains approximately constant voltage on load and

finds wide application as an exciter for the field circuit of a large generator
and as a tachogenerator when a signal proportional to motor speed is
required for control or display purposes.

8.6. Performance characteristics of motors

An electric motor must produce rotation of a shaft against a load torque
TL and the motor must, therefore, produce torque. The electromagnetic
torque T. of a d.c. motor is given by (8.2) as
pz
T. = -2 <1>/a = K<I>Ia (8.5)
na
and it follows immediately from (8.5) that the torque of a d.c. motor can be
controlled by variations in the armature current. In the case of a d.c. shunt
motor, the field current can simply be controlled by the use of a variable
resistance in series with the field winding. For a series motor, the armature
and field circuits are connected in series and it is somewhat difficult to control
the armature and field currents independently.
146 THE D.C. MACHINE

The terminal voltage Vis given by

V = E +laRa (8.6)

where, by (8.1 ), E = pz <Dw.

2na
It immediately follows that
(8.7)
where Ela is the gross output power.
The forms of the torque-armature current, speed-torque, and speed-
power characteristics for a shunt connected, series connected, and cumu-
latively compounded d.c. motor are illustrated in Fig. 8.13. Most motors

Armature current Torque Power

Fig. 8.1 3. Characteristic curves for motors.

will be designed to develop a given horse-power at a specified speed and it

follows from (8.1) and (8.6) that the angular velocity w can be written

W= (8.8)

Thus the speed of a d.c. motor depends on the values of the applied voltage
V, the armature current Ia and resistance Ra, and the field flux per pole <D.

8.7. The shunt motor

The arrangement of a shunt motor is given in schematic form in Fig. 8.14
and, if the applied voltage is constant, the field current and hence the flux <D
will be constant. Then, from (8.2) the torque Te is given by
(8.9)
In a similar manner, the generated e.m.f. is given by
E = (K<D)w = Kmw (8.10)
 THE SERIES MOTOR 147

so that, from (8.6), (8.9), and (8.1 0)

VKm Km 2
T ------w (8.11)
e- Ra Ra

and, from (8. 7) and (8.11)

VK K 2 2
El =wT =~-~- (8.12)
a e Ra Ra

The form of (8.11) and (8.12) confirm the form of the characteristic curves
for the shunt motor shown in Fig. 8.13.

va

Fig. 8.14. Schematic diagram for shunt motor.

The shunt motor is essentially a constant speed machine with a low speed
regulation. It can be seen from (8.8) that the speed is inversely proportional
to the field flux, so that its speed can be varied by control of field flux. When,
however, the motor operates at very low values of field flux, the speed will be
high and, if the field becomes open-circuited the speed will rise rapidly beyond
the permissible limit governed by the mechanical structure. If a shunt motor
is designed to operate with a low value of shunt field flux, it will usually be
fitted with a small cumulative series winding, known as a stabilizing winding,
to limit the speed to a safe value.

8.8. The series motor

A schematic representation of the series motor is given in Fig. 8.15. The
field flux is determined directly by the armature current and, if there is no
saturation, flux is proportional to armature current so that, from (8.2)
T. = KtJ>Ia = KI/. (8.13)
From (8.6), with negligible armature resistance,
V = E = KtJ>w = Klaw (8.14)
148 THE D.C. MACHINE

and it follows directly from (8.13) and (8.14) that

Te = KI. 2 = -vz-2 • (8.15)

Kw

Thus Pg = wTe
vz
=- (8.16)
Kw
and the speed-power curve is a rectangular hyperbola. The forms of (8.13),
(8.15), and (8.16) confirm the shape of the characteristic curves of the series
motor shown in Fig. 8.13.

v.

Fig. 8.15. Schematic diagram for series motor.

Figure 8.13 shows that the no-load speed is very high and care must be
taken to ensure that the machine always operates on load. In practice, the
series machine will normally have a small shunt field winding to limit the no-
load speed. The assumption that flux is proportional to armature current is
only valid on light load and, in general, performance characteristics of the
series motor must be obtained using the magnetization curve.
This machine is ideally suited to traction where large torques are required
at low speeds and relatively low torques are required at high speeds.
EXAMPLE 8.1. A 200 V d.c. shunt motor has a field resistance of 160 nand an
armature resistance of 0·5 n. The motor operates on no-load with full field
flux at its base speed of 800 r.p.m. with an armature current of 4 A. If the
machine drives a load requiring a torque of 98 N.m, calculate the armature
current and speed of the machine.
The machine is required to develop 7·5 h.p. at 1000 r.p.m. What is the
required value of external series resistance in the field circuit?
Magnetic saturation and armature reaction can be neglected.

Solution: Full field current Ir = 200/160 = 1·25 A.

On no-load, E = v. - I.R. = 200 - 4 x 0·5 = 198 V. Also E = Klrw.
198
K = = 1·895.
1·25(~~ 800)
 THE SERIES MOTOR 149

On load, T. = Klcla, that is, 98 = 1·895 x 1·25 x /a.

Armature current /a = 41·1 A.
Now V = E +laRa, i.e., E = 200- 41·4 x 0·5 = 179·3 V. ButE = Klcm,
179·3
:. w = 1·895 x 1·25 = 75·7 rad/s.

Load speed = 724 r.p.m.

For 7·5 h.p. at 1000 r.p.m.,
7·5 X 746
T. = 2n = 53·2 N.m.
60 1000

Then 53·2 = 1·895/rla, or lela = 28·1.

28·1
= 198·5/r + 0·5/a = 198·5/r + 0·5~·
f

Therefore, Ic = 0·935 A or 0·075 A. Thus /a= 28·1/Ir = 30 A or 375 A.

The value of Ir = 0·075 A will produce very high armature currents and
will not be considered.
200
When Ic = 0·935, Rc = 0.935 = 214 Q.

External resistance required = 214 - 160 = 54 n.

EXAMPLE 8.2. A 200 V d.c. series motor has the following O.C.C. measured
at 1000 r.p.m.
Field current (A) 5 10 15 20 25 30
Open-circuit voltage (V) 80 160 202 222 236 244
The armature resistance is 0·25 Q and the series field resistance is 0·25 Q.
Find the speed of the machine when (a) the armature current is 22·5 A,
(b) the gross torque is 36 N.m.

Solution: The O.C.C. is plotted in Fig. 8.16.

In general, E = Kfl)w, so that
E1 f1)1w1 f1)1N1
E2 = f1)2W2 = f1)2N2.
!50 THE D.C. MACHINE

(a) When Ia = 22·5 A, E 1 = 230 Vat N1 = 1000 r.p.m. from Fig. 8.16.
Now E = V - JaR = 200 - 22·5 x 0·5 = 188·75 V. Then
230 1000 188·75
N and Speed N = 23Q x 1000 = 819 r.p.m.
188·75

10
Armature current (A)
Fig. 8.16. Open-circuit characteristic for Example 8.2.

(b) When T. = 36 N.m, T. = Ela/w. Then, at 1000 r.p.m.,

E/a = 36G~ X 1000) = 3765.

The rectangular hyperbola E/a = 3765 is plotted on the O.C.C. in the

manner shown in Fig. 8.16. At the point of intersection,

/a = 17·4, £ 1 = 212·5 Vat 1000 r.p.m., E = 200 - 17·4 x 0·5 = 191·3 V.

191·3
Then Speed N = 212 _5 x 1000 = 902 r.p.m.
 SPEED CONTROL OF MOTORS 151

8.9. The compound motor

A schematic diagram for a cumulatively compounded d.c. motor is shown
in Fig. 8.17. The operating characteristics of this machine lie between those
of the shunt and series motor. A series motor with a shunt limiting winding
is, of course, a compound motor with essentially series characteristics, while
a shunt motor with a series stabilizing winding is a compound motor with
shunt characteristics.

• +

v.

Fig. 8.17. Schematic diagram for compound motor.

The differentially compounded motor operates with the flux set up by

current in the series winding opposing that set up by current in the shunt
winding. On light load, the effect of the series winding will be small. On high
loads, however, its effect will be to continuously reduce the total flux as the
load increases. This has the effect of increasing the speed with a consequent
increase in armature current, decrease in flux, and further increase in speed.
Such an arrangement will therefore be unstable and, in these circumstances,
the differentially compounded motor has little application.

8.10. Speed control of motors

It has been shown in section 8.6 that the speed of a motor can be varied by
control of the armature applied voltage, the armature resistance, or the field
flux. The base speed of the machine is defined as the speed with rated armature
voltage and normal armature resistance and field flux.
Speed control above the base value can be obtained by variation of field
flux. In the case of a machine with a high resistance shunt field winding, speed
control over a wide range above the base speed can be obtained by inserting a
series resistance in the field circuit. It is important to note that a reduction in
field flux produces a corresponding increase in speed, so that the generated
e.m.f. does not change appreciably as the speed is increased. However, the
machine torque is reduced as the field flux is reduced, so that this method of
speed control is suited to applications where the load torque falls as the speed
increases. In the case of a machine with a series field, speed control above the
152 THE D.C. MACHINE

base value can be obtained by placing a diverting resistance in parallel with

the series winding so that the field current is less than the armature current.
When speed control below the base value of speed is required, the effective
armature resistance can be increased by inserting external resistance in series
with the armature, and this method can be applied to shunt series or com-
pound motors. It has the disadvantage however, that the resistances required
must take full armature current and there will, therefore, be a large external
power loss with an associated reduction in overall efficiency. The speed ofthe
machine will be governed by the value of the voltage drop in the series resistor
and will therefore be a function of the load on the machine. In these cir-
cumstances, the application of this method of control is limited although it
has some use where reduced speed is required intermittently.

8.11. The Ward-Leonard method of speed control

This system is one of the most versatile methods of speed control available
and involves the use of a separate motor-generator set to supply variable
voltage to the armature of the machine under control. The arrangement is
illustrated in Fig. 8.18 in which machine A is usually a 3-phase, squirrel-cage

Fig. 8.18. Ward-LeoiUlrd system.

induction motor driving the separately excited d.c. generator B whose arma-
ture is electrically connected to that of the separately excited d.c. motor M
under control. The field of machine B is supplied with a variable voltage and
that of the main motor M with a fixed voltage. These field voltages can be
obtained from a small excitor in the form of a shunt generator driven by
machine A. As the excitation of machine B is varied, so its generated e.m.f.
and, hence, the speed of the main motor varies. Thus, continuous speed
control is available over very wide ranges with the major advantages that large
torques are available at low speeds. This arrangement also has the advantage
that, when the armature voltage of the main motor is reduced to produce a
decrease in speed, the motor M will momentarily operate as a generator and
 MOTOR STARTERS 153

machine B will operate as a motor driving machine A as a generator. In this

manner, the change in kinetic energy of the main motor during retardation
can be converted to electrical energy and returned to the power system. Such
a process is known as regeneration.

8.12. Motor starters

When voltage is applied to the armature of a d.c. motor with the machine
stationary, there will be no generated e.m.f. and the armature current will
only be limited by the internal armature resistance of the machine. It is,
therefore, usual to incorporate a variable resistance in series with the arma-
ture which is gradually decreased to zero as the motor accelerates. Typical
forms of manual starters for d.c. motors are shown in Fig. 8.19. The three-

Three-point Four-point
(a) (b)
Fig. 8.19. Typical forms of manual starter:
(a) Threepoint. (b) Fourpoint.

point starter shown in Fig. 8.19(a) has the hold-coil connected in series with
the field circuit so that the starter arm, which is spring loaded, drops back to
the off position if the field circuit is open-circuited. In the case of the four-
point starter shown in Fig. 8.19(b), the hold-coil is connected directly across
the supply so that the starter arm drops back to the off position if the voltage
supply to the machine is removed. In both cases, the field circuit is completed
when the moving arm is connected to the first stud of the variable resistor and
it is usual to include a relay in the armature circuit to give overload protection.
If facilities for changing the field current are incorporated within the starter,
it is known as a controller. It is usual to incorporate mechanical interlocks in
the controller to ensure that the machine can only be started with full field
154 THE D.C. MACHINE

flux. The drum controller for a series motor will provide facilities for starting,
speed control, and reversal of the direction of rotation and it follows that the
external series resistor must be rated to carry full-load current continuously.

8.13. Losses and efficiency

The efficiency of a d.c. machine can be written as
.
Effic1ency Output Input - Losses Losses
= = = 1 ---· (8.17)
Input Input Input
The losses are made up of copper loss in the armature and field circuits,
brush contact resistance, iron losses, friction and windage losses, and stray
load losses. It is usual to determine the efficiency by some method based on
the measurement of losses.
One such method is known as Swinburne's method in which the machine
is run on no-load at rated speed with rated applied voltage. The armature
resistance Ra is measured and the armature copper loss is calculated for an
assumed armature current /a. The remainder of the losses are assumed to be
constant and independent of the armature current. For a no-load applied
voltage V, armature current / 0 , and field current Ir, the efficiency as a motor
is given by
Effi . - 1 V(/r + /o) + 1/Ra
c1ency - - V(/c + /a) · (8.18)

The corresponding efficiency as a generator is given by

.
Effic1ency = 1 - V(/r + /o) + /a 2 Ra · (8.19)
V(/a + /o) + /a 2 Ra
The value of efficiency obtained by this method is usually greater than the
actual efficiency.
A more accurate method of predicting the efficiency is the Kapp-Hopkin-
son Test for which two similar machines mechanically coupled and electrically
connected back-to-back are required. The method of connection is illustrated
in Fig. 8.20. Machine A is run up to normal speed as a motor with machine
B unexcited. Excitation is then applied to machine B and increased until the
open-circuit voltage of this machine is equal to the system voltage. Machine
B is then connected to the system and, if the excitation of machine B is
increased when that of machine A is decreased so that the speed is unchanged,
machine A will motor while machine B generates. In these circumstances the
supply system will provide the total losses.
Then, from Fig. 8.20,
Generator input = Motor output = x x Motor input = xVIA
where x is the efficiency of each machine.
 LOSSES AND EFFICIENCY !55

Generator output = V/8 = x x Generator input, so that x 2 VIA = V/8 .

Jlu
Then,
.
Effictency x = -· (8.20)
/A
It can be seen from Fig. 8.20 that the quantity P = Vh - (/A 2RA +
/ 82 R 8 ) represents the total fixed loss of the system so that the fixed loss of

v.

Fig. 8.20. Kapp-Hopkinson test.

each machine can be assumed to be tP. The input power to the motor is
VIA + VI. and the output power from the generator is V/8 . Then

Motor efficiency = 1
+ IA 2 RA + VI.
(P/2)
(8.21)
V(/A + I.)
(P/2) + Il R 8 + Vlb
and Generator efficiency = 1 - _ __:_-'--'----=-----=----=-=-- (8.22)
2
V(I8 +/b) + (P/2) + / 8 R 8
The values of efficiency found from (8.21) and (8.22) will be close to the
actual efficiency.

Fig. 8.21. Separation offixed losses.

When it is necessary to obtain measured values for the separate iron losses
and friction and windage losses for a machine, the arrangement shown in
Fig. 8.21 can be used. The machine under test, G, is driven by an auxiliary
d.c. motor M, the excitation of M is set to produce the rated speed for G, and
the excitation of G is set to produce rated voltage on open-circuit. The speed
!56 THE D.C. MACHINE

of the set is then increased to approximately 125% of rated speed by increasing

the field resistance of the motor and the two excitations are then maintained
constant. Increasing armature resistance is used to decrease the speed of the
set in steps and the quantity W1 = VIM - /M 2 RM (where RM is the armature
resistance of the motor) is measured at each step. This process is repeated with
the generator unexcited to obtain W 2 , and with the generator uncoupled to
obtain W3 • The iron losses ofthe generator are then equal to W1 - W2 and
the friction and windage losses of the generator are w2 - w3.

8.14. Dynamics of separately excited motors

The transient performance of d.c. generators and motors is of some
importance when considering the operation of closed loop control systems,
and many position control servomechanisms utilize a separately excited d.c.
motor as the main control element.
A schematic diagram representing this machine is shown in Fig. 8.22.
The machine will normally operate with constant field current Ir so that the
generated e.m.f. e can be written
(8.23)
The torque t. can be written
(8.24)
where Km is known as the motor torque constant and has the dimensions of

Fig. 8.22. The separately excited d.c. motor.

N.m per ampere. The voltage--<:urrent relationship for the armature can be
written
Va = e + (ra + L.P)ia. (8.25)
In many cases, the armature inductance can be neglected.
 DYNAMICS OF SEPARATELY EXCITED MOTORS 157

Consider the operation of the machine when a step voltage V H(t) is

applied to the armature with constant field current. Then, from (8.23) and
(8.25)
(8.26)
Now from (8.24)
te = Kmia = Jpw + TL (8.27)
so that, from (8.26) and (8.27) with the motor on no-load,

K m [ VH(t)- Kmw] =
ra
T
.ipW, Or Km[~
ra s
_ -] =
KmW T-
JSW.

Then

On solution, (8.28)

where r m = Jr j Km 2 is the motor time constant. It follows from (8.28) that

the speed rises exponentially when a step voltage is applied to the armature
with the motor on no load.
When the load torque TL can be written in the form TL = a + bw, where
a and b are constants, the above method of solution can be directly applied
to obtain the speed as a function of time. When, however, the load torque TL
is given by TL = a + bw + cw 2 + · · · + nw 2 , where a, b, c, ... are con-
stants and n is an integer, the resulting equations are non-linear differential
equations. The solution of such equations is beyond the scope of this text,
although graphical techniques leading to the solution of this type of problem
are available.
When the separately excited d.c. motor is rheostatically braked, the
arrangement can be represented by the schematic diagram shown in Fig.
8.23. With the motor initially on no-load, the operating equations can be
written
(8.29)
where rL is the external braking resistance.
The machine torque te is given by (8.27) as te = Kmia = Jpw. Then, from
(8.27) and (8.29)

(8.30)

If the machine is initially operating at a speed w 0 , it follows from (8.30) that

- K 2w
m = J(sw - Wo), where r = ra + rL.
r
158 THE D.C. MACHINE

Then w(Js + K; 2 ) = Jwo, or ()) = Wo e-tK,2JJr. (8.31)

Thus, from (8.31), the speed of the machine falls exponentially.

When the load is a friction load in the form TL = bw, (8.27) can be written
(8.32)
Then from (8.29) and (8.32), -Km 2 w/r = Jpw + bw so that -Km2 wjr =
J(sw - w 0 ) + bw.
On solution, w = w0 exp - ( +b K; 2
) }· (8.33)

Comparison of (8.31) and (8.33) shows that the effect of the friction load
is to decrease the time taken for the motor to stop.

Fig. 8.23. Rheostatic braking.

The process of plugging a motor involves reversing the polarity of the

supply to the armature of the machine. With the machine on no-load, the
operating equations can be written
Jra
Va = - pw + Kmw. (8.34)
Km
The initial speed w 0 is given by w 0 = V/ Km, where Vis the armature voltage.
When the machine is plugged, the armature voltage is reduced to zero and
then reversed. This corresponds to a step voltage of- V H(t) with an initial
speed of w0 = V/Km. Then, from (8.34),
-V Jr
- - = _a (sw - w0 ) + Kmm
Km
-(slra + Km ).
S

so that Jra Wo
- -
V=
- OJ -
Km S Km
 DYNAMICS OF SEPARATELY EXCITED MOTORS !59

On solution, with w 0 = V/ Km

w
v
= K (2 e-tt•m - 1). (8.35)
m

When t = 0, w = V/Km = w 0 ; when t = oo, w = - V/Km = -w 0 •

An alternative method of analysis under plugging conditions is to use the
method of superimposed change. Initially, with the machine on no-load
t'8 = V, i 8 = 0, W = Wo = V/Km.
Finally,va = - V,i 8 = O,w = wF,sothat,forthechange,v3 = -2VH(t),
w = wl" Then

-2V
On solution, ro>
"'-'1=~
(1 - e-t/<m).
m

When t = oo, w1 = 2V/Km = -2w 0 .

Final velocity wF = w 0 + w1 = - V/ Km = - w 0 •

EXAMPLE 8.3. A separately excited 200 V d.c. motor operates at rated voltage
with constant excitation on no-load. The torque constant of the motor is
2 N.m/A, its armature resistance is 0·5 Q and the total moment of inertia of
the rotating parts is 5 kg.m 2 •
Derive an expression for the speed of the machine after plugging and find
the time taken for the machine to stop. Fixed losses can be neglected.

Solution: On no-load, E = V 8 = 200 = Kmw 0 • No-load speed w 0 =

100 radjs.
Plugging corresponds to applying a step voltage -200 H(t) to the armature
of the machine with, as initial speed, w 0 = 100. Then

raJ
so that V3 = Kmw + Kpw
m

= (2 + 0·5 2x 5 s)w- 0·5 2x 5 x 100

- 100 100 100

or W= --+ .
s + 1·6 s s + 1·6
Motor speed w = 200 e- 1 . 61 - 100.
160 THE D.C. MACHINE

Check: t = 0, OJ = OJ 0 = 100; t = oo, OJ = -OJ 0 = -100.

When OJ = 0; 0 = 200 e- 1 "6 ' - 100, so that 1·6t = log. 2.
Time for motor to stop = 0· 347 sec.

Tutorial Problems
I. The following data were obtained for the magnetization curve of a 4-pole, shunt
generator with compoles when run at 1500 r.p.m.
Field
current (A) 0 0·1 0·4 0·6 0·8 1·0 1-14 1·32 1·56 1·92 2-4 3·04
Open-circuit
voltage (V) 6 20 80 120 160 200 220 240 260 280 300 320
Draw the magnetization curve and hence obtain (a) the voltage to which the
machine will build up as a self-excited generator, if the total shunt-field resistance
is 125 ohms; (b) the critical value of shunt field resistance; (c) the magnetization
curve for a speed of 1000 r.p.m.; (d) the open-circuit voltage of the generator
when run at 1000 r.p.m. with a shunt field resistance of 125 ohms; (e) the speed
for which a field resistance of 125 ohms is critical.
(Answer: 298 V; 200 ohms; 155 V; 938 r.p.m.)

2. If the armature resistance of the machine, whose magnetization curve at 1500

r.p.m. as given in Problem 1, is 0·4 ohm, calculate the terminal voltage of the
machine when the armature current is 40 A and the field resistance is 125 ohms.
Neglect armature reaction.
If armature reaction is equivalent to a reduction in field current of 0·0025/8
where I. is the armature current, plot the terminal voltage-armature current
relationship.

3. A long shunt cumulative compound d.c. generator has a shunt field of resistance
75 ohms with 700 turns per pole. The series winding of resistance 0·1 ohm has 6
turns per pole and the armature resistance is 0·25 ohm. Armature reaction is
equivalent to a reduction of 0·0021•. The O.C.C. is given by
v.c = 190 + 32(ADc
where v.c is the generated e.m.f. and (A T>r is the field ampere turns for 4000 <
(AT)r < 7000.
Plot the terminal voltage-armature current relationship.

4. A d.c. generator has an open-circuit characteristic given by v.c = 200/rf(K + Ir)

such that the open-circuit voltage v.c is 150 V when the field current Iris 1·5 A.
Find the critical field resistance and the open-circuit voltage when the field
resistance is 200 ohms.
(Answer: 400 ohms; 100 V)

5. If the armature resistance of the machine described in Problem 4 is 0·5 ohm,

calculate the terminal voltage and load current when the armature is connected
to a load of resistance 10 ohms with a field resistance of 100 ohms.
(Answer: 138 V; 13-8 A)
 TUTORIAL PROBLEVlS 161

6. A 200 V, 10 kW, 4-pole, long-shunt compound d.c. generator has armature,

com pole, series field, and shunt field resistances of 0·1, 0·03, 0·07, and I 00 ohms
respectively. Calculate the value of the generated e.m.f. when the machine is
operating at full load. If the full-load speed is 1500 r.p.m. and the armature is
lap-wound with 720 conductors, determine the flux per pole under the above
conditions.
(Answer: 210·4 V; ll·67 mWb)

7. Two shunt generators operate in parallel to supply a total load current of 3000 A.
Each machine has an armature resistance of 0·05 ohm and a field resistance of
100 ohms. If the generated e.m.f.s are 200 V and 210 V respectively, find the load
voltage and the armature current of each machine.
(Answer: 1401 A; 1601 A; 130 V)

8. The external characteristics of two shunt generators in parallel are as follows:

Load
current (A) 0 5 10 15 20 25 30
Terminal I 270 263 254 240 222
voltage (V) II 280 277 270 263 253 243 228

Find the load current and terminal voltage of each machine when
(i) the generators supply a load of resistance 6 ohms
(ii) the generators supply a battery of e.m.f. 200 V and resistance 1·5 ohms.
(Answer: 240 V, 15 A, 26 A; 248 V, 12·2 A, 22 A)

9. A d.c. shunt machine with an armature resistance ofO·l ohm and a field resistance
of I 00 ohms operates from a 200 V supply. If the machine operates with a line
current of 60 A, find the ratio of the speed as a generator to that as a motor.
Saturation can be neglected.
(Answer: I : 1·06)

10. Two similar shunt machines are tested by the Hopkinson method. When con-
nected to a 240 V d.c. supply, and the generator is delivering an output of 42 kW,
the current taken from the supply is 40 A. The shunt-field circuit resistance values
of the motor and generator respectively are 300 and 240 ohms; the armature
circuit resistance of each machine is 0·03 ohm. Calculate the efficiency of the
generator, stating any assumption made.
(Answer: 0·916 approximate; 0·894 exact)

II. A 250 V shunt motor with an armature resistance of 0·25 ohm and a shunt field
resistance of 250 ohms takes 4 A when running light. Calculate the efficiency
when the armature current is 60 A.
(Answer: 0·873)

12. A 200 V, 10 h.p., d.c. shunt motor has an armature resistance of0·35 ohms and a
shunt field resistance of 200 ohms. The machine runs at 1750 r.p.m. with an
efficiency of 87% on full-load. Calculate the no-load armature current and speed.
If a 2 ohm resistor is placed in the armature circuit to lower the speed, calculate
the speed when the motor is developing full-load torque.
(Answer: 1885 r.p.m.; 964 r.p.m.)
162 THE D.C. MACHINE

13. A d.c. machine is separately excited from a variable voltage source and its
armature is supplied through a resistor R ohms from a separate constant voltage
source E 0 volts. The load on the machine has a torque requirement T N.m pro-
portional to the speed N r.p.m. raised to the power p. Derive an expression relat-
ing the speed of the machine to the voltage across its field, and show that the speed
will be a maximum when the armature current I. is equal to E 0 /2R. Saturation can
be ne)!;lected.

14. A series motor has a total resistance of0·25 ohm. When the motor is connected
across a 230 V, d.c. supply it runs at 700 r.p.m. and develops a gross torque ofl50
N.m.
Calculate the speed at which it will run and the gross torque it will develop when
it is connected to a 240 V d.c. supply and the armature current is 35 A.
(Answer: 1045 r.p.m.; 76·2 N.m)

15. A 200 V series motor has an O.C.C. given by Voc = 1·2N[I/I + 20)] where N is
the speed in r.p.m. and /is the armature current. The internal resistance is 0·5 ohm
and the armature current on full load is 30 A. Calculate the torque and speed at
full load and the speed and armature current at half full load.
(Answer: 255 r.p.m.; 208 N.m.; 243 r.p.m.; ll·5 A)

16. A d.c. series motor has a rated output of 500 b.h.p. at 750 r.p.m. At full load the
input current is 660 A at 600 V. Calculate the torque developed by the motor
when the speed increases to cause the input current to rise to 725 A assuming
that the resultant flux rises 40% as much as the current. If the resistance of the
armature and compoles of this machine is 0·5 ohm, calculate the value of the
external resistance to be added to the armature circuit if full-load torque is to be
developed at standstill. Assume that flux is proportional to current.
(Answer: 2290 N.m; 0·409 ohm)

17. A 220 V, d.c. series motor takes 60 A when developing its rated output at 1000
r.p.m. The armature resistance is 0·12 ohm and the field resistance is 0·08 ohm.
(a) Calculate the torque developed at rated output.
(b) If the maximum current on starting is not to exceed 100 A find the resistance
to be connected in series at starting and the starting torque.
(c) If a diverter of 0·16 ohm is connected in parallel across the field when the
motor is running at its rated output power find the new speed. (Assume flux
is proportional to field current.)
(Answer: ll9 N.m; 2 ohms; 331 N.m; 1525 r.p.m.)

18. A 200 V, d.c. shunt motor has an armature resistance of 0·5 ohm and a field
resistance of 180 ohms. The machine operates on full load at 950 r.p.m. with an
armature current of 40 A. If the speed is to be decreased to 700 r.p.m., calculate
the required value of series armature resistance when the load torque is propor-
tional to speed.
(Answer: 1·77 ohms)

19. A 220 V, 50 h.p., d.c. shunt motor has the following open-circuit characteristics at
a speed of 650 r.p.m.:
Field current (A) 0·5 1·0 1·5 2·0 2·5 3·0 3·5
Generated e.m.f. (V) 54 108 152 185 207 223 234
 TUTORIAL PROBLEMS 163

The armature resistance is 0·1 ohm and the shunt field resistance is 100 ohms.
Calculate:
(a) The no-load speed with full field flux.
(b) The additional field resistance required for a no-load speed of 850 r.p.m.
(c) The full-load speed when the total field resistance is 110 ohms and the
corresponding armature current. The fixed losses amount to 270 W.
(Answer: 732 r.p.m.; 29 ohms; 713 r.p.m.; 200 A)
20. A 200 V, compound d.c. motor has a total armature circuit resistance of 1·8 ohms,
excluding the resistance of the series winding, the resistance of the series winding
being 0·3 ohm; when the full-load armature current of 15 A flows in the series
winding the proportion of series ampere-turns to shunt ampere-turns is 0·25: 1.
When run as a shunt machine the speed on no-load is 1600 r.p.m. Calculate the
full-load speed
(i) as a shunt motor,
(ii) as a long shunt cumulative compound motor.
It may be assumed that, on no-load, the armature circuit volt-drop and the
series ampere-turns are both negligible. It may also be assumed that the effective
flux is proportional to the field m.m.f.
(Answer: 1384 r.p.m.; 1080 r.p.m.)
21. Show that, when a step voltage V H(t) is applied to the armature of a separately
excited d.c. motor with fixed excitation supplying a constant torque load, the
energy dissipated in the armature in bringing the motor up to its steady-state
speed is given by
(
Energy = V + - 1-
KmTL)
2 J
2Km 2 ·

22. A separately excited d.c. motor with fixed excitation drives a separately excited
d.c. generator with fixed excitation which is connected to a load of resistance
3.75 Q. The armature resistance of the generator is 0·25 Q, that of the motor is
0·5 Q and both armature inductances can be neglected. The generated voltage
constants for the motor and generator are both equal to 1 volt.secjrad and the
total moment of inertia of the rotating parts is 2 kg.m 2 • If, with the system initially
at rest, a step of 50 V is applied to the motor armature, find an expression for the
resulting motor armature current as a function of time. Fixed losses can be
neglected.
(Answer: 11-1(1 + 8 e- 1·13'))
23. A 400 V d.c. shunt motor with an armature resistance of0·037 Q and inductance
0·0078 H drives a constant torque load of inertia 40 kg.m 2 at a speed of9·8 r.p.m.
The torque constant is 4 N.m per ampere.
Find the time taken for the motor to stop when it is dynamically braked with a
total braking resistance of 0·5 Q. What value of external braking resistance is
required to make the motion during braking critically damped?
(Answer: 1·003 sec; 0·075 Q)
24. A separately excited 200 V, d.c. motor operates at rated voltage with constant
excitation on no-load. The torque constant of the motor is 4 N.m per ampere of
armature current, the armature resistance is 0·5 Q and the total moment of inertia
of the rotating parts is 5 kg.m 2 .
Derive an expression, as a function of time, for the speed of the machine after
plugging. Hence, or otherwise, find the time taken for the motor to stop. Fixed
losses and armature inductance can be neglected.
(Answer: w = -50 + 100 e- 6 -4 '; 0·108 sec)
9. Semiconductor control of motors

The thyristor is now well established as the power control element in many
forms of static electrical energy conversion system and, in particular, is
widely used for the control of speed of both d. c. and a.c. motors. In the par-
ticular case of the d.c. machine, variation of the magnitude of the voltage
applied to the armature and, in some cases, the field of the motor is required.
When considering the control of the speed of an a.c. motor, variation in both
the magnitude and frequency of the a.c. voltage applied to the motor will
generally be required. These two different requirements must always be
borne in mind when considering general thyristor circuits and the particular
cases of a.c./d.c. conversion (controlled rectification) and d.c./a.c. conver-
sion (controlled inversion) are the most widely used modes of operation for
the control of motor speed.

9.1. The thyristor as a power device

The thyristor is a silicon semiconductor device made in a 4-layer sandwich
of pnpn construction such that its three terminals known as the anode,
cathode and gate are normally biased in the manner shown in Fig. 9.1. Since
the thyristor operates in a switching mode, the power loss is inherently low.

Gate
Ia

Cathode

Fig. 9.1. Norma/junction bias.

 THYRISTOR FIRING MECHANISMS 165

In practice, the losses represent only a few per cent of the maximum load
power under control and, in these circumstances, the value of the load
which can be controlled by a thyristor is generally limited by its voltage and
current ratings, rather than the internal power loss of the device.
The thermal capacity of the semiconductor element forming a thyristor is
minute and it must therefore be effectively coupled to a heat sink with large
thermal capacity in order to keep the mean junction temperature to a safe
value. Junction temperature affects the value of breakover voltage and cur-
rent capability of the device. Excessive values of junction temperature can
cause device failure and it is therefore the most important single factor con-
trolling thyristor performance. The form of the static anode characteristic
of a typical thyristor is given in Fig. 9.2.

Conduction

Forward

IG Increasing

Reverse

Fig. 9.2. Thyristor static anode characteristic.

9.2. Thyristor firing mechanisms

A thyristor can be turned on by an increase in temperature, by the applica-
tion of gate current or by rapid change in anode voltage. Gate firing is the
most useful since it utilizes the high power gain of the device and readily
results in repeatable modes of operation.
When the gate current is increased, transistor action results in an increase
in cathode current and firing occurs at the point of instability. The general
form of the gate input characteristic is given in Fig. 9.3 and is very similar
in form to the base-emitter characteristic of a transistor. Once the thyristor
has fired, the gate input appears as a voltage source of the order of0·6 V with
a non-linear output resistance. This simplified treatment assumes that the
thyristor is a one-dimensional device with the whole width turned on
166 SEMICONDUCTOR CONTROL OF MOTORS

instantaneously. In practice, only a portion of the device near the gate con-
nection is turned on instantaneously and the conducting area spreads
laterally at high speed. The action is regenerative and will therefore propa-
gate under its own influence. Initially, however, since current flow is only
through a small section of the device, there will be a large local power dissi-
pation. This effect, which is limited by the load to some extent, can be con-
siderably reduced by using a value of gate current well above the threshold
value. Normally a current pulse or pulse chain is applied to the gate and the
rise-time of the pulse must be considerably less than the turn-on time of the
thyristor if satisfactory performance is to be achieved.
IG (rnA)

Forward

50

-10 -8 -6 -4 -2 4

Increasing

Fig. 9.3. Gate input characteristic.

A thyristor can also be fired by a rapid change in anode voltage since

such a positive change results in a transient current flowing towards the gate
junction. This effect momentarily increases the cathode current so that,
under certain conditions, the thyristor will fire. Such an effect can be pro-
duced by supply switching surges and could result in refiring immediately
after turn-off. However, susceptibility to this form of spurious firing can be
reduced by including a small inductor in series with the anode and a capacitor
across the thyristor. Alternatively a low resistor can be positioned between
the gate and the cathode or the gate circuit can be reverse-biased immediately
after gate firing has taken place.
In order to turn off a thyristor, the anode current must be reduced to a
value less than the holding value and turn-off is inherently a slow process.
In the conducting states, all junctions are injecting current and all the cor-
responding charge carriers must be removed.
The turn-off time is clearly a function of anode current and can be con-
siderably reduced by the application of reverse anode voltage. The process of
turn-off is generally known as commutation and when reverse anode voltage
 GATE FIRING 167

occurs naturally as in a.c. circuits, the process of turn-off is known as natural

commutation. When, however, the supply system is d.c., no such natural
reversal of anode voltage takes place and the voltage must be forced to
reverse by connecting a pre-charged capacitor to the anode. Such a process
is known as forced commutation and requires the use of an auxiliary circuit
known as a commutation circuit. In many applications the efficiency of
power conversion will be relatively low if the commutation energy in a
forced commutation system is dissipated, and in these circumstances special
energy recovery circuits are needed.

9.3. Gate firing

The gate characteristic of a thyristor before firing can lie between very
wide limits set by temperature variations and the form of individual devices
and typical extremes are given in Fig. 9.4 with limits set by maximum voltage,
current and power ratings and maximum and minimum gate impedance
lines. The voltage limit is virtually temperature independent but the current
limit does not change with temperature. It is therefore usual to only consider
the current limit for minimum temperature and also to include a lower limit
in voltage below which no device will fire.

Max. Ia
Area of Uncertain
no firing firing

Fig. 9.4. Limits on gate characteristic for certain firing.

168 SEMICONDUCfOR CONTROL OF MOTORS

Any firing source can be represented by an e.m.f. E with a source resistance

R5 so that:
E- V
Ia = G (9.1)
Rs
The graphical representation of this law is known as a load-line and the
design of the firing circuit should position the load-line so that it exceeds
the minimum conditions by a large margin but does not exceed any of the
maximum ratings. The applied gate drive should always exceed the minimum
conditions if certain and continuous firing is to be achieved. The rise-time
must be shorter than the anode turn-on time and, in general, pulses shorter
than 10 J1S are not acceptable. In some applications, single-pulse firing of
relatively short duration is possible although, in general, maintained trigger-
ing in the form of a continuous pulse or, more usually, as a pulse train con-
sisting of short pulses (30 J1S) at a high repetition rate is used.
It has previously been noted that a negative bias makes the thyristor less
sensitive to unwanted turn-on and also improves the turn-off time. Small
thyristors can be stabilized using an external gate-<:athode resistance, but
this method is not particularly useful for large devices since the gate only

JL

-111-----t ~--1 1 - - - + - - - - - - - '

vb
Fig. 9.5. Gate firing circuit.
 SERIES AND PARALLEL OPERATION 169

directly controls that area of the junction immediately adjacent to it. A more
practical method is illustrated by the circuit of Fig. 9.5. The condition
required for the thyristor leakage current / 0 to pass into the gate, thereby
bypassing the cathode, is that the diode Dl in Fig. 9.5 is always forward
biased, i.e. /b > / 0 . Then approximately:
vb- vo >
_____:::_ _=.. Io. (9.2)
Rb
The battery Vb can be replaced by a suitable Zener diode or capacitor. It is
generally considered unwise to allow the gate voltage to go positive while
the anode voltage is negative and this condition can be readily met by the
addition of a second diode D2 and a resistor Ra to the circuit of Fig. 9.5.
Operation of the firing source tends to reverse-bias diode D2 to allow the
gate to go positive but the current in Ra keeps Dl conducting while the anode
voltage is negative.

9.4. Series and parallel operation

In many practical circuits, it becomes necessary to operate several devices
in series in order to obtain a sufficiently high safety factor for voltage tran-
sients. If a series chain of thyristors contains, say, two fast and one slow
device, the full system voltage could appear across the slow thyristor for
several microseconds producing very high local switching losses and this
condition must be avoided. If thyristors are randomly connected in a series
chain, the system voltage will divide unevenly between them according to
their individual leakage characteristics. It is therefore normal practice to
shunt each thyristor in a series chain with a resistor which passes a current
much larger than the highest leakage current. However, such a chain of
resistors is not capable of accurate voltage division at the high frequencies
met in transient operation and it is usual to shunt each resistor with the
capacitor to obtain proper sharing of transient over-voltage. A small resistor
will normally be connected in series with each capacitor to limit the dis-
charge current when the thyristor is fired. It is apparent that, under these
conditions, the gates of the individual thyristors are at different potentials and
special techniques for gate connection must be used. The most commonly
used method is known as slave firing and the basis of this method is illustrated
in Fig. 9.6. Thyristor THl is fired in the usual manner and the resulting fall in
its anode voltage results in an increase in the anode-cathode voltage across
thyristor TH2. A charging current then passes through the capacitor C2
and, since this current must pass through the gate circuit, thyristor TH2
will fire. This process is then repeated for all other thyristors in the series
chain. The capacitor Cl in Fig. 9.6 is present to prevent spurious firingofeach
thyristor which may be produced by transient overvoltages. More elegant
forms of this simple circuit have been in established use for many years.
170 SEMICONDUCfOR CONTROL OF MOTORS

Cl

Fig. 9.6. Series operation with sla11e firing.

In practice, the power rating of a piece of thyristor equipment may exceed

that of the biggest available device and it becomes necessary to operate
thyristors is parallel. The voltage-current characteristic of a thyristor when
conducting is non-linear and in these circumstances a random selection of
devices operating in parallel will not share the total load in proportion to
their individual ratings. If these unbalanced operating conditions are allowed
to exist it is probable that either one thyristor would overheat and fail or
the complete stack would be derated and some means of ensuring even
current sharing must be found.
Perhaps the simplest method is to use matched devices in parallel which
have been selected by the device manufacturer as having equal voltage drops
on full load, but such devices are obviously expensive and may be difficult
to replace in the event of a device failure. As a simple economical alternative
a resistor can be inserted in series with each device in order to make the
voltage drop more uniform but the extra heat to be dissipated may be a
problem.
A more reliable method is to use current sharing reactors in which the
out-of-balance ampere-turns induce a voltage between the anodes of the
thyristors connected in parallel, thereby tending to reduce the out-of-
balance currents.
There can be problems associated with the reliable firing of thyristors
operating in parallel, in that the first thyristor to fire will deprive the remain-
ing thyristors of their anode voltage. Thus some devices may not even fire
and will not then take their share of the load current. The first two methods
 OVERCURRENT PROTECTION 171

of load sharing previously mentioned, those of selected devices and series

resistors, are particularly prone to this effect. The third method, that of
sharing reactors, overcomes this problem since the device that fires first
tends to increase the anode voltage of the other devices, thereby helping with
their firing.

9.5. Overcurrent protection

Thyristors, like all semiconductor devices, have very short thermal time
constants and will react to overloads through the resulting increase in junc-
tion temperature. A small increase in the temperature of the junction will
produce loss of the forward voltage blocking characteristic of the device
which, while not destructive in itself, could result in circuit malfunction.
A large increase in junction temperature will result in loss of the reverse
blocking capability and the device will be damaged, usually destroyed, by
reapplication of reverse voltage. This form of failure is the most common
form of failure in practice and forms the basis of overload ratings of thyris-
tors.
Since semiconductor devices change from zero-impedance to infinite-
impedance devices in a very short period of time, they have non-linear
characteristics and it follows that they must be protected against high cur-
rents under short-circuit conditions and high voltage under open-circuit
conditions. Any fuse used to protect a thyristor must be capable of quickly
interrupting the fault current in the faulty device without producing voltages
beyond the capability of other thyristors in the circuit which become effec-
tively open-circuited when the fuse blows. For moderate overloads in a
circuit operating from an a.c. supply, the fusing arc is extinguished at the
first current zero after its formation and it is then possible to compare the
prospective current for a given fuse pre-arcing time with the permitted device
overload current for the same time period when producing a fuse rating. It
is important to note that, over long time periods, the fuse characteristic is
flatter than the device rating curve. In these circumstances, the fuse will not
protect against long-term overloads unless the device is well-rated. In the
case of a heavy overload, when the current is likely to exceed the one-cycle
rating of the thyristor, the fuse must also perform a current-limiting function
and manufacturers' data are generally available for the calculation of various
combinations of fuse ratings and current overloads. The maximum instan-
taneous current that the fuse will carry must not exceed the fault current
rating of the thyristor it is designed to protect, and it is usual to design the
supply impedance to control the potential fault current. In many cases the
arc voltage developed across a fuse will also appear across an operating
thyristor and, in these circumstances, the voltage rating of the thyristor
must at least equal the arc voltage. In general, fuses can be used to either
protect against external faults, or perhaps of more importance, to disconnect
172 SEMICONDUCTOR CONTROL OF MOTORS

an individual thyristor that has failed in order to maintain overall system

operation.
Many practical circuits will be fitted with automatic current limiting used
to satisfy two different requirements. Firstly, the load itself must be protected
against high currents (for example, limiting the charge current of a battery).
Secondly, the thyristors and associated semiconductor devices must be
protected against sustained overloads (for example, protecting a battery
charge under short-circuit conditions). Any current limiting circuit needs a
current-dependent control signal which can be obtained in several different
manners. The first and most simple method is to use a shunt such that the
voltage developed across the shunt is amplified and then compared with a
reference voltage. In d.c. circuits where electrical isolation is necessary, a d.c.
current transformer can be used to replace the shunt. An a.c. current trans-
former can be used to provide a feedback signal from the a.c. lines to a con-
trolled rectifier or in the output of an a.c. circuit.

9.6. Natural commutation

In any thyristor circuit operating from an a.c. supply commutation occurs
naturally at voltage reversal for a resistive load and, under these circumstan-
ces, it is possible to use a thyristor as a controlled rectifier producing a d.c.
voltage from an a.c. source. This configuration is illustrated in Fig. 9.7 for a
resistive load. If a gate signal is applied at 150° after the start of the positive
half-cycle, the thyristor will turn on and the resistor will be connected to the
supply for the remaining 30° of the half-cycle. At the end of the positive
half-cycle, the voltage and current will fall to zero together and the thyristor
will cease to conduct. As the voltage builds up during the negative half-cycle
the thyristor operates in a blocking condition and the load voltage and cur-
rent will be zero during this period. Thus, as the instant in time during the
positive half-cycle at which a signal is applied to the gate is varied, so the
period of time for which the resistive load is connected to the supply, and
hence the value of load voltage current, is varied. This process is illustrated
in Fig. 9.7.
When, however, the load is a combination of inductance and resistance,

Fig. 9.7. Controlled rectification with resistirJe load.

 NATURAL COMMUTATION 173

the load current cannot change instantaneously. There will be an exponential

change in current at the instant of firing and the load current will not be
zero at the instant when the supply voltage is zero. This process is illustrated
by Fig. 9.8 and it can be seen that the average value ofload current is reduced
by the extension of conduction into the negative half-cycle. In the extreme
theoretical case of a pure inductive load, the volt-second areas above and

Fig. 9.8. Controlled rectification with inductive load.

below the zero voltage axis will be equal. The reduction in load voltage
introduced by an inductive load can be overcome by placing a rectifier,
known as a 'free-wheeling diode', across the load in a manner shown in
Fig. 9.9. At the end of a positive half-cycle, the current in the load cannot be
coerced to zero by the negative-going supply voltage and will continue to
flow through the now low forward impedance of the free-wheeling diode.
h

Fig. 9.9. Action offree-wheeling diode.

The current then decays at a rate governed by the time constant of the load
and the forward voltage drop of the diode. In many practical cases, it will
not have reached zero by the time the thyristor is refined in the next half-
cycle. In these circumstances, the current is always finite. This mode of
operation is known as continuous conduction and is illustrated in Fig. 9.10.
In the practical case of a thyristor circuit designed to control the armature
voltage of a d.c. motor, the presence of the generated e.m.f. of the motor
174 SEMICONDUCTOR CONTROL OF MOTORS

Fig. 9.10. JloltiiiJe tuUl current wt111ejorms with free-wheeling diode.

Fig. 9.11. ControUed rectification with actirJe lotul.

further complicates matters. The armature circuit presents an RL load with

a superimposed e.m.f. and the general arrangement is illustrated by Fig. 9.11
where the e.m.f. is assumed to be equal to half the peak value of the system
voltage and the action of the free-wheeling diode has not been included. The
supply voltage is positive with respect to the load for a range of angles be-
tween 30° and 150° after zero in the positive half-cycle and this range of
angles gives the limits of control. If, for example, the thyristor is fired at
120° after zero, conduction is possible for a maximum of 60° and the current
will be relatively low. As the firing angle is brought closer to zero, the con-
duction angle is extended and the current increases until ultimately, when
the thyristor is fired at 30° after zero, the maximum conduction angle with
the largest current is produced. It is apparent that the range of possible firing
points for the thyristor decreases as the generated e.m.f. of the motor in-
creases. In these circumstances, the continuous current mode of operation
may not be possible for certain loading conditions even in the presence of a
free-wheeling diode.
The process of natural commutation is not easy to obtain in a circuit
operating from a d.c. supply, but can, however, be induced by resonant
 NATURAL COMMUTATION 175

action. This method is illustrated in Fig. 9.12 which is drawn for a simple
series converter. Corresponding waveforms of anode current /A, thyristor
voltage VT and capacitor voltage Vc are given in Fig. 9.13. The thyristor is
fired at t = 0 with zero current and, if resistance is neglected, the anode

....
.J!s

Fig. 9.12. Natural commutation with d.c. supply.

Time
tVT
v.
Time

Time

Fig. 9.13. Walleformsfor natural commutation with d.c. supply.

176 SEMICONDUCTOR CONTROL OF MOTORS

current varies on a sinewave of half-period nJLc

seconds as shown. The
capacitor then charges up to a voltage of +2V. (approx) and the thyristor
turns off when the current through it falls below the holding value. When
the thyristor has turned off, the capacitor discharges exponentially on the
time constant CR to zero voltage and the cycle can be restarted. A fundamen-
tal disadvantage of this type of circuit is that the capacitor must recover to its
initial state before the thyristor can be fired again.

9.7. Forced commutation

When negative anode-cathode voltage does not occur naturally in a
thyristor circuit, forced commutation must be introduced. This is usually

Fig. 9.14. Forced commutation with d.c. supply.

Time

Time

~------.---------------
( rr t fr

Fig. 9.15. War~eformsfor forced commutation circuit.

 SPEED CONTROL OF D.C. MOTORS 177

effected by switching a charged capacitor across the conducting thyristor in

order to divert current from it and, preferably, to apply reverse voltage to it
for a period of time greater than its turn-off time. A simple circuit operating
from a d.c. source and supplying an inductive load which utilizes forced
commutation is given in Fig. 9.14 in which the capacitor Cis precharged to
a voltage V1 in the sense shown. Approximate waveforms of load current h
and thyristor voltage Vr are given in Fig. 9.15. At timet 1 the switch s (which
may be an auxiliary thyristor) is closed. During the reverse recovery period
trr = t 2 - t 1 , excess charge is being flushed from the thyristor which is
still presenting a low anode-cathode impedance and the load current is
assumed to be constant during this period. When the thyristor begins to
recover to the blocking state, the load current passes through the capacitor
thereby resulting in the capacitor being recharged to the given potential of
the opposite polarity. The thyristor passes reverse current during much
of this time and this reverse current is limited by the presence of the induct-
ance L 1 • Positive anode voltage must not be reapplied until the thyristor has
reached the forward blocking state. The total turn-off time increases slightly
as the time rate of change of reapplied voltage increases and considerably
with increase in junction temperature. In order to keep the total turn-off
time to a reasonable value, an appreciable reverse voltage must be applied
during commutation.
In practice, the main thyristor will often be shunted by a diode, such that
at time t 1 , with the thyristor conducting, when the switch s is closed, the
diode now clamps the thyristor voltage to nearly zero as the thyristor begins
to recover blocking capability. The capacitor current rises as the capacitor
recharges and, ultimately, the diode becomes open-circuited. The use of
this diode can shorten the effective turn-off time.

9.8. Speed control of d.c. motors

The most common requirement for a drive is that giving speed control
from zero to full-load speed with a load torque which is approximately
constant or increasing with speed. In such an application it is necessary to
supply the armature with variable voltage, and a controlled rectification
configuration operating from an a.c. source is admirably suited to this
application.
The simplest and often the cheapest configuration is that of the half-wave
rectifier applied to either a shunt or series d.c. motor in the manner shown
in Fig. 9.16. The number of components required is minimal and, in its
simplest form, it is possible to dispense with current limiting circuits. The
current build-up in the positive half-cycle is then limited by the armature
inductance. This system suffers from the major disadvantage that the form
factor is bad (at least 2: 1), leading to a significant de-rating of the motor.
The torque is pulsating and this may produce objectionable results,
178 SEMICONDUCTOR CONTROL OF MOTORS

particularly at low speeds. There will also be a large a.c. component in the
armature current which can produce poor commutation. Such a system
injects a d.c. component into the supply system and there will, in general,
be a practical limit to the rating of such a system. In practice, this arrange-
ment will be restricted to ratings below 1 h.p.

Shunt motor Series motor

Fig. 9.16. Half-waJJe controller for d.c. motor.

Some improvement in form factor can be made by adding a flywheel diode

across the armature. If an inductance is based in series with the armature, its
effect will be to smooth out the armature current. In practice, however, a
large value of inductance is necessary making such a component large and
expensive. It if often cheaper to use a full-wave arrangement.
When regeneration is not needed, it is possible to use the half-controlled
bridge shown in Fig. 9.17. This circuit has thyristors on two arms with
cathodes common to the positive terminal and rectifiers in the other two
arms. It produces full-wave rectification but has one major disadvantage
in that it is essential to ensure that each thyristor is extinguished before the
start of its positive half-cycle of the system voltage. This condition is im-
portant because it is possible for the armature current to flywheel through
the conducting thyristor and its adjoining rectifier during the negative half-
cycle. If this current is present when the system voltage becomes positive

Fig. 9.17. FuU-waJJe controller for d.c. motor.

 SPEED CONTROL OF D.C. MOTORS 179

again, a full half-cycle of the supply voltage will appear across the motor
armature, resulting in a large surge of current. This effect can be prevented
by either firing the second thyristor towards the end of the negative half-
cycle or by using a flywheel diode across the armature so that an alternative
low impedance path is provided for the current.

Fig. 9.18. AlterMtive full-wtnJe controller.

It is, however, possible to rearrange the components in the bridge in the

manner shown in Fig. 9.18, to remove the problem of failure to commutate
the armature current from the conducting thyristor. In the arrangement
shown in Fig. 9.18, the thyristors are commoned to a supply terminal so
that the armature current circulates around the rectifiers. In these circum-
stances the thyristors will be extinguished by reversal of the system voltage.
Thus it is possible that the current in the rectifiers will exceed the current in
the thyristors and that the armature current will exceed the mains current.
It then becomes necessary to use larger rectifiers or to use protective current
limiting operated from the armature circuit. The circuit shown in Fig. 9.17
has the advantage that, if the rectifier bridge is completed, a supply is avail-
able for the motor field.
An alternative full-wave configuration with some useful features is shown
in Fig. 9.19. This arrangement consists of a full-wave uncontrolled rectifier

Fig. 9.19. Simple fuU-wave thyristor controller.

180 SEMICONDUCTOR CONTROL OF MOTORS

supplying a single thyristor which can now conduct on both half-cycles of

the system voltage. Its original merit in the days when thyristors were very
expensive was that it made full use of a single thyristor. It also has the ad-
vantage that a supply for the motor field is readily available and that, for a
multi-machine system, only one main uncontrolled rectifier is necessary.
Only one thyristor firing circuit is necessary and this ensures an even balance
of current pulses between alternate half-cycles of the supply. Its major dis-
advantage is that some reliable method of extinguishing the thyristor at
each zero of the mains voltage must be found so that the thyristor can enter
a blocking mode to regain control during the next half-cycle.
All the circuits so far described have operated from a single-phase supply
with either line to neutral or line to line voltage. The normal acceptable
loading limit for a single-phase system is about 7 kW and for ratings above
this, it is necessary to use a 3-phase supply with the corresponding thyristor
configuration.
The simplest 3-phase arrangement is the half-controlled bridge with
thyristors in one half and rectifiers in the other half of the bridge, connected
in the manner shown in Fig. 9.20. Such an arrangement suffers from the
same commutation trouble as the single-phase bridge. In these circumstances
either a minimum firing angle unit must be used or a flywheel diode must be
fitted across the armature.

Fig. 9.20. Half-wave three-phase controller.

In some applications, a d.c. motor will be used to speed control an over-

hauling load such that, under certain circumstances, power transfer from the
load to the motor is possible. Under these conditions the motor will act as a
generator and a thyristor power converter capable of transferring power
into the a.c. mains must be used. The basic circuit of such a regenerative
system is shown in Fig. 9.21. Under regenerative conditions current must
flow out of the positive motor terminal into the supply system and the thyris-
tor THl in Fig. 9.21 controls the amount of regenerative power returned
to the mains while TH2 controls the system under motoring conditions. Care
must obviously be taken to ensure that both thyristors are not conducting
at the same time and, as previously noted, a thyristor can only be turned off
by having its current reduced to below the holding value. In a regenerative
 SPEED CONTROL OF A.C. MOTORS 181

Fig. 9.21. RegeneratirJe controller for d.c. motor.

II
v

Fig. 9.22. Bi-phase regeneratirJe controUer.

circuit of this type it is usual to ensure that the net circuit e.m.f. always acts
in a direction such that the rate of change of current is negative and to ensure
that the e.m.f. is present until zero current is reached.
This simple circuit suffers from several major disadvantages and, in par-
ticular, a severe restriction must be placed on the value of the firing angle
during regeneration if uncontrolled conduction is to be avoided. This type of
restriction can be overcome by the use of the hi-phase, half-wave system
illustrated in Fig. 9.22, for which continuous conduction is possible with the
current being forcibly transferred from one thyristor to the other because
the cathode voltage of the incoming device is lower than that of the outgoing
device at the instant of firing.
A fuller treatment of more complex multi-phase thyristor circuits, includ-
ing regenerative controllers can be found in ref. 1.

9.9. Speed control of a.c. motors

Most a.c. motors operate at constant speed and speed control can be ob-
tained by varying the frequency of the applied voltage. In many cases the
182 SEMICONDUCfOR CONTROL OF MOTORS

magnitude of the applied voltage will also be varied in direct proportion to

the frequency in order to maintain the flux in the machine at a constant
value. In general, a static power converter producing a variable frequency,
variable magnitude polyphase output voltage from fixed polyphase a.c.
mains is required and this can be achieved in one of two ways. Firstly, a
direct conversion (a.c. to a.c.) can be made using the so-called cycloconverter
principle and a full discussion of such circuits can be found in ref. l. The
second and more common way is to convert the fixed a.c. to variable d.c.
and to then reconvert this d.c. voltage to the required variable a.c. system;
such a method uses a d.c. link. In this case negative anode-cathode voltage
does not occur naturally in the d.c. to a.c. thyristor converter and the pro-
cess of forced commutation must be used. Two basic methods of forced
commutation inverters, the so-called parallel inverter and the pulse-width
modulated inverter are in common use.
A simplified diagram representing a single-phase parallel inverter is given
in Fig. 9.23 in which the inductance L between the source and the thyristors
THl

Fig. 9.23. Single-pluzse parallel inf!erter.

acts as a current limiter. With thyristor THl in Fig. 9.23 conducting, the
supply voltage V appears across one half of the primary of the output trans-
former T and load current flows. The voltage across the whole of the transfor-
mer primary winding is then 2 V and the capacitor Cis charged to the voltage
2 V. When thyristor TH2 is fired, the capacitor discharges through the two
thyristors and TH 1 is reverse biased until it turns off. Thyristor TH2 is
then in a conducting state and the supply voltage V appears across the other
half of the transformer primary in an opposite sense. The output voltage
across the secondary then reverses and is therefore an approximate square
wave whose frequency is controlled by the firing pulses applied to the two
thyristors. Filtering can be introduced at the output if a sinusoidal outr ut
voltage is required.
The general principles of pulse-width modulation are illustrated by the
circuit of Fig. 9.24. Thyristor THl is the main circuit device and thyristor
TH2, capacitor C and resistor R form the commutating circuit for the main
thyristor. The presence of the diode D is now essential in that it provides a
 SPEED CONTROL OF A.C. MOTORS 183

path for load current and allows the capacitor to discharge. Initially neither
thyristor is conducting so that both points A and B in Fig. 9.24 are at the
potential of the negative rail. IfTHI is fired the point Bin Fig. 9.24 will now
be at the potential of the positive rail. Current will build up in the load on an
exponential of time constant L/r and, at the same time, the capacitor C will
be charged to a voltage + V, through the resistor R, with the potential at
point B positive with respect to that at point A. If after some convenient
time, thyristor TH2 is fired, the potential at point A will rise to + V volts,
the presence of the charged capacitor will result in the potential at point B,
in Fig. 9.24, rising to + 2 V volts. The main thyristor THl will then be reverse

Time
v

Time

Fig. 9.24. Pulse-width modulation controller.

biased and can turn off. Load current must still be flowing, because of the
presence of load inductance and this current is supplied by the charge stored
in the capacitor. The capacitor will thus be discharged resonantly until
point B goes to a negative potential, at which point the diode D will conduct,
thereby clamping the load voltage at approximately zero and also providing a
flywheel path for the load current. Capacitor C will now have a voltage of
+ V across it with point A positive and the thyristor TH I will be conducting
with a small current passing through the resistor R. The load current will fall
exponentially until thyristor TH2 is fired again. This will cause the potential
at point B to rise to + V so that the potential at point A rises to + 2 V. The
thyristor THI is thus extinguished and the capacitor C recharges through R
with the point B again at the positive potential, ready for the next cycle of
operation.
More elegant forms of the basic circuit of Fig. 9.24 can be used to generate
a sinusoid of output voltage with a superimposed high frequency ripple. The
ripple frequency is controlled by the maximum permissible switching rate for
the thyristors and the allowable commutation loss which takes place as
184 SEMICONDUCTOR CONTROL OF MOTORS

switching occurs. The frequency of the output voltage is readily controlled

by control of the pulse repetition rate of the gate signal.
A full treatment of semiconductor control of a.c. machines is given in
ref. 2.

References
1. Pelly, B. R., Thyristor Phase-controlled Convertors and Cyc/oconverters, Wiley,
1971.
2. Murphy, J. M. D., Thyristor Control of Alternating Current Motors, Pergamon,
1973.
Index

a.c. commutator machines 94-7 damper winding 134

basis of operation 13 d.c. machines 137-63
doubly fed motor 15 armature reaction 138
frequency changer 13, 97 basis of operation 10
Leblanc system 14 commutation 138
repulsion motor II 0 compensating winding 138
Scherbius machine 96 compoles 140
Schrage motor 15 efficiency 154
series motor I 09 equalizers 27
air-gap 2 excitation, compound !51
air-gap line 118 separate 142, 156
armature reaction 114, 138 series 147
armature winding 20-9 shunt 143, 146
chording 22, 39 flux distribution 139, 141
clock diagram 23, 24 generator, characteristics 141-5
distribution factor 39 voltage regulation 142
double-layer 20, 25 Kapp-Hopkinson test !55
dummy coil 26 losses 154
equalizer 27 motor, characteristics 145-51
fractional pitch 40 dynamics 156--60
integral slot 23 speed control !51, 152, 177-81
lap connection 2, 20, 25 starting 153
multiplex 25 reactance voltage 139
phase belts 21 Swinburne test !54
pitch factor 39 Ward-Leonard 152
simplex 25 dot convention 17
wave connection 20, 28 doubly fed motor 15
winding factor 39
eddy-current loss 50, 61
back-to-back test 57, 154 energy conversion 30-7
energy sink 15
energy source 16
capacitor motor 105, 107
circuit conventions 15
field construction
co-energy 32 field winding I
commutation 138, 172-6
frequency changer 13, 97
commutator 2
compensating winding 138
compoles 140 general machine 6-15
coupling field I, 30 generator of e.m.f., a. c. machine 38
cylindrical rotor 2, 113, 137 d.c. machine 38
186 INDEX

harmonics, current 61 repulsion motor 110, 129

e.m.f. 66 rheostatic braking !57
hysteresis 50, 64 rotating m.m.f. 4

induction motor 74--112 salient poles I, 113, 137

dynamics 97 Scherbius machine 96
polyphase 74--100 Schrage motor 15
basis of operation 11 self-impedance voltage 34
equivalent circuits 74 shaded-pole motor 108
locked-rotor test 85 slip 11, 74
pole-amplitude modulation 90 slip frequency 74
power balance 77 slip test 122
power relations 81 slot angle 39
rotor resistance variation 80, 94 speed voltage 34
running-light test 84 symmetrical components 101, 102
slip 11, 74 synchronous machine 113-36
speed control 181-4 air-gap line 118
starting methods 87-90 armature reaction 114, 116, 120
torque relations 78 basis of operation 8
single-phase 101-12 construction, cylindrical rotor 113
analysis 101 salient-pole 113
basis of operation 12 damper circuits 134
capacitor motors 105, 107 direct axis 120
locus diagrams I 06 reactance 121
phase converter 101 equivalent circuits 116
shaded pole 108 generator, operating chart 131
split phase I 05, I 06 operation on busbars 125, 129
parallel operation 130
lap connection 2, 20, 25 paralleling 124
Leblanc system 14 leakage reactance 115
load angle 9, 36, 113 load angle 9, 36, 114
loss separation 50, !55 maximum lagging current 123
motor, operation on busbars 126, 129
magnetization curve 31, 62, 142 starting 134
magnetizing current 45, 61 open-circuit characteristic 118
maximum power transfer 78 quadrature axis 120
m.m.f., distribution 5 reactance 121
pulsating 12 short-circuit characteristic 119
rotating 4 slip test 122
synchronous impedance 118
negative sequence 13 synchronous reactance 118
normalization 18 two-reaction theory 120
vee-curves 128
parallel operation, generators 130 voltage regulation, e.m.f. method 123
thyristors 170 m.m.f. method 123
transformers 58, 70 synchronous condenser 116
percentage values 19 synchronous induction motor 134
per-unit values 19 synchronous speed 4,6,8, II
phase-sequence 124 synchroscope 124
plugging 158
pole-amplitude modulation 90 Thevenin's theorem 78
positive sequence 12 thyristors, anode characteristic 165
power-factor control 116 construction 164
production of torque, a.c. machine 40 control ofa.c. motors 181-4
d.c. machine 41 control of d.c. motors 177-81
current limiting 172
reluctance motor 35 firing mechanisms 165-7
reluctance torque 36 forced commutation 176
 INDEX 187

free-wheel diode 173 ideal transformer 43

fuses 171 magnetization curve 62
gate firing 167-9 main groups 66
inverters 182 no-load waveform 61
natural commutation 172-5 open-circuit test 50
overcurrent protection 171 parallel operation 58, 70
parallel operation 170 practical transformer 45
regeneration 180 referred values 48
series operation 169 short-circuit test 51
slave firing 169 terminal markings 66
tertiary winding 69
transformer voltage 34 vee connection 67
transformers 43-73 voltage regulation 55
approximate voltage drop 53
auto-connection 70 universal motor 110
back-to-back test 57
connections 66 voltage regulation, d.c. generator 142
construction, core-type 43 synchronous machine 123
shell-type 43 transformer 55
efficiency 56
e.m.f. 44 Ward-Leonard system 152
equivalent circuits, approximate 49 wave connection 20, 28
exact 49 winding factor 39
hysteresis loop 64 windings see armature and field