a

BASIC
ELECTRIC
MACHINES
a
rs©>=

oe
S3
;
i

paige.
_-
oe ae

é :
eee
 Basic
Electric
Machines
 —
<1

18 ;
. ’ sro ge, ie.
Hea .
1 rad
ges
ay '
Wiss = i

a ation
 Basic
Flectric
Machines

Vincent Del Toro

Professor Emeritus
Department of Electrical Engineering
City University of New York

HK
07632
Prentice Hall, Englewood Cliffs, New Jersey

ATHENS. REGIONAL LIBRAReY, rA9»,err9a

ATHENS, GEORGIA cast

42ab000
Del Toro, Vincent. (date)
Basic electric machines / Vincent Del Toro.
p. cm.
Bibliography: p.
Includes index.
ISBN 0-13-060146-2
1. Electric machinery. 2. Electromechanical devices. I. Title.
TK2000.D44 1990
89-8714
621.31'042—dc20 CIP

Editorial/production supervision
and interior design: Sophie Papanikolaou
Cover design: Ben Santora
Manufacturing buyer: Bob Anderson

© 1990 by Prentice-Hall, Inc.

A Division of Simon & Schuster
Englewood Cliffs, New Jersey 07632

All rights reserved. No part of this book may be

reproduced, in any form or by any means,
without permission in writing from the publisher.

Printed in the United States of America

LOO S Bel yOu Sede2aL

ESBN O-13-O601L4G-e

Prentice-Hall International (UK) Limited, London

Prentice-Hall of Australia Pty. Limited, Sydney
Prentice-Hall Canada Inc., Toronto
Prentice-Hall Hispanoamericana, $.A., Mexico
Prentice-Hall of India Private Limited, New Delhi
Prentice-Hall of Japan, Inc., Tokyo
Simon & Schuster Asia Ptc. Ltd., Singapore °
Editora Prentice-Hall do Brasil, Ltda., Rio de Janeiro
 To
Franca
 , -

F; r hae
a
"
\ vetines bigaeni, rt

1, on
<7 + 7 [
4

’ : aa
oa
+e z

a. reste fetta os yi o ee - eee

a <6 ‘iia
ee |
& oa nisruehanie ee mer sok
» a ae

ae
es

Se” eon
f

SE
at,

r ae ae ea
2 - - fy

” ah a

: re a
¢s
od oe
7 é.
: ‘its

a i tas
Ne, vee Ltr} Lees - ; - ~

row UE Az deaks Ney Cane ’ i

ae
(ena Chaat: )2. Fagus
} reortl o>WRATH : 8, Hh
y br ll aetin Iria taut ae :
Bagi } Foun a
=e
a % ‘ p _ + ure
26 ¢

‘ jared 2o Acodh tie A

poe
 Contents

Preface xiii

PART I: BASIC TOPICS

1 Magnetic Theory and Circuits

ile Ampere’s Law—Definition of Magnetic Quantities 2
|- Theory of Magnetism 11
i Magnetization Curves of Ferromagnetic Materials I3
1- The Magnetic Circuit: Concept and Analogies 16
ile Units for Magnetic Circuit Calculations 21
1- Magnetic Circuit Computations 21
[= The Inductance Parameter 29
Hysteresis and Eddy Current Losses
ie OAAINDUNHRWN
in Ferromagnetic Materials 34
1-9 Relays—An Applicatio n of Magnetic Force 38
Problems 44

vii
 Contents
viii
1
2 Transformers elopment
2-1 Theory of Operation and Dev
of Phasor Diagra ms 5
The Equivalent Circuit 65
2-2
No- Loa d Test s Te
2-3 Parameters from
on 80
2-4 Efficiency and Voltage Regulati
Mutual Inductance 83
2-5 86
The Equivale nt Circ uit: Coupled Circuit Viewpoint
2-6
The Autotransformer 89
2-7
Per-Unit Calc ulat ions 98
2-8 104
Transformers for Thr ee- Pha se Circuits
2-9
Problems 114

3 Fundamentals of Electromechanical
122
Energy Conversion
Electromagnetic Torque [23
3-1
3-2 Induced Voltages 128
Machines 132
3-3 Construction Features of Electric 140
ue and Volt age Form ulas
3-4 Practical Forms of Torq
3-5 Final Remarks 148
Problems 149

153
4 Three-Phase Induction Motors
154
4-1 The Revolving Magnetic Field
4-2 Induction Motor Slip 161
164
4-3 The Equivalent Circuit
168
4-4 Computation of Perf orma nce
Moto r Oper ation
4-5 Correlation of Inductio n
173
with the Basic Torq ue Equa tion s
4-6 Torque-Speed Characte rist ic:
77
Starting and Maximum Torques 182
mete rs from No-Load Tests
4-7 Equivalent Circuit Para
4-8 Speed Control 188
196
4-9 Electronic Method of Speed Control
ions of Thre e-Ph ase Induction
4-10 Ratings and Applicat
Motors 204
Motors on |
4-11 Controllers for Three-Phase Induction
Problems PIS)

rs 224
5 Three-Phase Synchronous Generato 225
5-1 Generation of a Three-Phase Voltage
226
5-2 Linear Analysis by the General Method
Contents

Linear Analysis by the Synchronous

Reactance Method: The Equivalent Circuit oad
Nonlinear Machine Analysis UA
Experimental Determination of Leakage Reactance x;
and Armature Reaction MMF A 242
Theory for Salient-Pole Machines 245
Power 255
Parallel Operation of Synchronous Generators 260
Problems 265

6 Three-Phase Synchronous Motors 272

6-1 Phasor Diagram and Equivalent Circuit pa:
6-2 Performance Computation 2/5
6-3 Power-Factor Control 283
6-4 Synchronous Motor Applications 285
6-5 Field Excitation by Rectified Sources 286
6-6 Correlation of Synchronous Motor Operation
with the Basic Torque Equation 295
Problems 299

7 Direct-Current Generators 303

74 Generation of Unidirectional Voltages 304
72 Direct-Current Generator Types 308
fBe Demagnetizing Effect of the Armature Winding MMF oil
7-4 External Generator Characteristics 313
7-5 Computation of Generator Performance 320
7-6 Commutation 325 .
Problems 33]

8 Direct-Current Motors 335

8-1 Motor Speed-Torque Characteristics 33a
8-2 Equivalent Circuits and the Power Flow Diagram 345
8-3 Speed Control 348
8-4 Speed Reversal 363
8-5 Permanent Magnet (PM) DC Motors 365
8-6 Operating Features and Applicat ions 369
8-7 Starters and Controllers for DC Motors 370
8-8 Speed Control by Electronic Means 373
Problems 384
 Contents

394
9 Single-Phase Induction Motors 394
9-1 How the Rotating Field Is Obtained
ving
9-2 Equivalent Circuit by the Double-Revol
Field Theory 397
Via Symmetrical Components
401
9-3 Equivalent Circu it
406
9-4 Torque-Speed Characteristic
9-5 Performance Analy sis 408
alent
9-6 Approximate Determination of the Equiv
Circuit Parameters 412
9-7 Single-Pha se Moto r Type s 417
res and Typic al Applications 426
9-8 Operating Featu
Problems 427

PART Il: TOPICS FOR FURTHER STUDY

10 Unbalanced Two-Phase Motors:

431
Servomotors
10-1 Symmetrical Components 432
10-2 Symmetri cal Comp onen ts Appli ed
to Unbalanced Two-Phase Motors 438
10-3 Equivalent Circuit and Performance
Based on Symmetrical Components 44]
Servomotor Torque-Speed Curves 449
10-4
Problems 453

45¢€
1G Stepper Motors
11-1 Construction Features 459
11-2 Method of Operation 459
Drive Amplifiers and Translator Logic 465
11-3
hing Sequence 469
11-4 Half-Stepping and the Required Switc
11-5 Reluctance-Type Stepper Motor 47]
Ratings and Other Characteristics 473
11-6
Problems 474

47
12 Synchros
12-1 Construction Features 476
12-2 Voltage Relations 479
12-3 Applications 482
487
12-4 The Advantage of Sinusoidally Distributed Windings
12-5 Stator MME of the Control Transformer
in the CX-CT Mode 489
Contents xi

12-6 Displacement Errors Caused by CT Space Harmonics 493

12-7 Static Errors and Residual Voltages 499
Problems S01

13 Dynamics of Electric Machines 505

13-1 The Transfer Function 506
13-2 Dynamic Behavior of the DC Motor
with Connected Mechanical Load 510
13-3 DC Generator Dynamics 518
13-4 Dynamics of a Generator-Motor System 524
13-5 DC Motor Transfer Function 529
13-6 Two-Phase Servomotor Transfer Function 535
13-7 Amplidyne Generator 540
13-8 A Feedback Voltage Control System 546
Problems bay

Appendix to Chapter 3
3A Developed Torque for Sinusoidal Flux Density
and Ampere-Conductor Distribution:
The Case of AC Machines 561
3B Developed Torque for Nonsinusoidal Flux Density
and Uniform Ampere-Conductor Distribution:
The Case of DC Machines 569
a¢ The General Torque Equation 373

Appendices
Units and Conversion Factors 586
Pitch and Distribution Factors for Machine Windings 588
Derivation of the Peak MMF per Pole
of a Polyphase Winding 593
Reduction Factors for Polyphase Windings 597
ODD
mo Saturated Synchronou s Reactance Method Based
on Resultant Flux 600
Slip Test for Salient-Pol e Synchronous Machines 602
Om A Short Bibliography 604

Answers to Selected Problems 606

Index 613
ro

on
ea He
mies! i ok

mee
ao | ode ements apne

. Senet
ae Nad) "alison
ww. am Pesetur hae pig:
ae nie «ia ;
ase ALP idea's amen T wie
ahuonT soronip ese vent wT i
*e. padkee :
sur toyhteae) are
Rit ils 2
wer MRE ide : at. ¢
<- argv

ww niitoncsd Tare-Piate Mater

Ereteaet oFsone eT
“gpatrotvre at a
tae a: asoontadi weap at anidos
(th? sqodideretirailsotma pantie Ai
a eam ar ~
QIg | ahead BL of
giant A babiduanmcneh® tel Spe” yonteegnce _
3 sre A ae se “ar
cope ee teil poeta
Yehr ome ater cone ~~
4 .
Las sbbiop slaes? lavond

1f apie Mate oe
epee eel won ears
. rei) Lee tae
bog apeinnth’ nue
‘ \ ghore

Ser rahi
is naval iailiniali
ALS © sei
to) onich at ates - Sd
2 27 fw) _ gly:
r cabal ae .
a
“] iy ‘ w% ‘

Pascoe ~eset
=5¢ A:

Ry be wae
ts Sone fe Satie } i '; L}
ae :
 Preface

ion of topics
This book is divided into two parts. Part I is devoted to an exposit
te the basic subject matter of a core course in
that can be considered to constitu
ed with a series of eclectic topics that easily
electric machinery. Part II is concern
hed by the basic topics of the first part. In most
builds on the foundation establis
can be fashion ed by a syllabus that includes
instances a suitable semester course
necessarily be part
Chapters | to 9. Of course not all topics in each chapter need
nt often lends itself to selectiv e exclusion of
of such a syllabus. The treatme
should time be a pressing factor. Moreover, the
specialty topics in each chapter
an eye on flexibili ty in order to allow instructors,
Table of Contents was cast with
in the syllabus such focused topics as stepper
who are so inclined, to include
be achieved by avoiding
motors, synchros, and machine dynamics. This can
chapters.
some of the more in-depth subject matter of the early
of Part I serve to establi sh the fundamental back-
The first three chapters
tional electric machines—
ground which makes it possible to analyze the conven
phase induct ion motors , de generators and mo-
namely, single-phase and three-
d to simplest terms, one
tors, and synchronous generators and motors. Reduce
machin ery involve s two basic laws: Faraday’s
can say that the study of electric
which descri bes the genera tion of voltage in a
law of electromagnetic induction,
and Ampere’s law, which
winding by a moving or time-varying magnetic field;

xiii
 Preface
xiv

or ina magnetic
describes the production of torque on a current-carrying conduct
in these laws, the book
field. Because of the preeminence of the magnetic field
and their circuits. A
begins in Chapter 1 with a description of magnetic theory
and to delinea te their
primary goal here is to define the various magnetic quantities
are constru cted to an
roles in magnetic circuit theory. After all, electric machines
c field. The focus in
important degree to provide suitable paths for the magneti
for the constru ction
Chapter 2 is on the application of Faraday’s law as the basis
many useful
of a very useful device—the transformer. The transformer fulfills
compre hensio n of its
purposes in many fields of electrical engineering, and a
operati on of all electri-
principle of operation permits an easy understanding of the
a physica l under-
cal machines. Here the treatment is heavily biased towards
d even in the derivat ion
standing of transformer operation. This attitude is reflecte
who prefer more
of the equivalent circuit. However, to satisfy those individuals
3 effort
mathematically oriented treatments, such are included as well. In Chapter
is directed to an analysis of torque production in an element ary machin e by the
how the ensuing electric al
use of the two fundamental laws and to a description of
cal quantiti es, T and a,
quantities, e and i, are related to the associated mechani
through the Law of Conservation of Energy. More detailed analyse s of torque
production and generation of voltage in situations associated with practical ma-
r
chine configurations appear, for the sake of simplicity, in the Appendix to Chapte
in this appendi x are used in
3 at the end of the book. However, the results found
the last section of Chapter 3 to derive the practical forms of the torque and voltage
equations as they apply to ac and de machines. These latter forms are then often
invoked throughout the remaining chapters of the book.
By drawing on the general background established from a study of Chapters
1 through 3, the study of the various types of electric machines becomes a routine
matter. Attention is first directed in Chapter 4 to the three-phase induction mo-
tor. This choice is made not only because this motor is the most popular found in
industry, but also because the analysis is very closely related to the static trans-
former despite the fact that a rotating element is involved. The equivalent circuit
is exactly the same in form as that of the transformer. The significance of some of
the circuit parameters, however, is different, and this distinction is drawn quite
effectively by continuing with an emphasis on physical interpretation while simul-
taneously pursuing rigorous mathematical formulations. The power flow diagram
is introduced in this chapter for the first time and its usefulness in helping to
evaluate motor performance is amply illustrated. Speed control of these motors is
an important application feature and, accordingly, receives a good deal of atten-
tion with special emphasis on electronic means of control. The chapter closes
with a listing of operating features and areas of applications as well as with a
discussion of controllers.
Chapters 5 and 6 deal with the subject matter of three-phase synchronous
generators and motors, respectively. Both linear and nonlinear analysis by the
general method and the synchronous reactance method are considered and these
procedures are applied to cylindrical-rotor as well as salient-pole machines. The
Preface XV

static stability of the synchronous motor under load, use of the synchronous
motor for power factor control, and excitation of its field winding by rectified
sources are topics that receive deserved attention. Again in these chapters the
theme of stressing the physical interpretations of the mathematical results is con-
tinued.
The operation and performance of dc generators and motors are treated in
Chapters 7 and 8. Because the dc machine is equipped with a mechanical rectifier
(the commutator) to convert the induced ac coil voltages to unidirectional forms,
the need for good commutation is imperative, and, for consistency, this topic is
allotted corresponding attention. Also important to a more complete understand-
ing of dc machine performance is an appreciation of the demagnetizing effect
caused by the armature winding magnetomotive force. Hence, this topic too is
singled out for special treatment. Moreover, because of the importance that the
slope and shape of the speed-torque characteristics of the various types of dc
motors have on their application suitability, an extensive analysis is given of these
characteristics for the shunt, compound, and series motors. This study is also
helpful in understanding the various methods of speed control of these motors. In
addition to a treatment of speed control by the conventional schemes of field
control, armature circuit resistance adjustment, and armature voltage adjustment,
the use of electronic procedures based on thyristor and chopper drives is also
discussed at length.
Chapter 9 deals with the single-phase induction motor, which is the most
complex machine treated in Part I. Two procedures are described to analyze this
motor. One is the double-revolving field theory, which requires no new theory
beyond that appearing in preceding chapters. The other is the use of symmetrical
components, which does call for new theory. However, in the interest of simplic-
ity and continuity, the derivation of this new theory is placed in Chapter 10 but the
results are used in Chapter 9 to derive an equivalent circuit which is shown to be
identical to that obtained by the double-revolving field theory. The remainder of
Chapter 9 is devoted to a description of the construction and operating features of
the various types of single-phase motors that are frequently encountered in com-
mercial and home environments.
The unbalanced two-phase motor is treated in Chapter 10, which is the first
chapter of Part II. It contains a detailed analysis of four-phase symmetrical com-
ponents which can be modified to furnish the results that are applicable to single-
The
phase motors or to unbalanced two-phase motors such as the servomotor.
with the servomoto r, which is
second half of this chapter is concerned exclusively
found in many servomechanisms.
on
Stepper motors are introduced in Chapter 11 through a detailed descripti
with their drive amplifier s
of the permanent magnet version of these motors along
motor is also discusse d
and translator logic devices. The reluctance-type stepper
The con-
briefly. Chapter 12 deals with another specialty topic—the synchro.
along with their voltage
struction features of the various types are amply detailed
relationships, applications, errors, and residual voltages.
 Preface
xvi

behavior.
All analysis of the machines of Part I is restricted to steady-state
is placed on formulat ions that allow
By contrast, in Chapter 13 the emphasis
be describe d. These results are
dynamic as well as steady-state behavior to
or by a state-va riable formula-
achieved either through a transfer-function analysis
tion. Both procedures are treated and applied to dc and ac machines.
 PART 1
Basic Topics

Magnetic Theory
and Circuits

An understanding of electromagnetism is essential to the study of electrical engi-

neering because it is the key to the operation of a great part of the electrical
apparatus found in industry as well as the home. All electric motors and genera-
tors, ranging in size from the fractional horsepower units found in home appli-
ances to the 25,000-hp giants used in some industries, depend upon the electro-
magnetic field as the coupling device permitting interchange of energy between an
electrical system and a mechanical system and vice versa. Similarly, static trans-
formers provide the means for converting energy from one electrical system to
another through the medium of a magnetic field. Transformers are to be found in
such varied applications as radio and television receivers and electrical power
distribution circuits. Other important devices—for example, circuit breakers, au-
tomatic switches, relays, and magnetic amplifiers—require the presence of a con-
fined magnetic field for their proper operation. This chapter provides the reader
with background so that he or she can identify a magnetic field and its salient
characteristics and more readily understand the function of the magnetic field in
electrical equipment.
The science of electrical engineering is founded on a few fundamental laws
derived from basic experiments. In the area of electromagnetism it is Ampere’s
 Magnetic Theory and Circuits | Chap. 1
2
starting point of our treatment.t
law that concerns us and, in fact, serves as the
in 1820, in his experiments on the
On the basis of the results obtained by Ampere
tors, such quantities as mag-
forces existing between two current-carrying conduc
bility, and magnetic flux are
netic flux density, magnetic field intensity, permea
attention is then directed to a
readily defined. Once this base is established,
n useful engineering materials as
discussion of the magnetic properties of certai
help simplify the computations
well as to the idea of a ‘“‘magnetic circuit” to
involved in analyzing magnetic devices.

QUANTITIES
1-1 AMPERE’S LAW—DEFINITION OF MAGNETIC
’s experiment. The
Appearing in Fig. 1-1 is a simplified modification of Ampere
g a constant-magnitude
configuration consists of a very long conductor 1 carryin
carryin g a constant-magnitude
current J, and an elemental conductor of length /
togeth er the elemental con-
current J, in a direction opposite to /;. When taken
t Ihl. The elemental conduc-
ductor and the current J) constitute a current elemen
but for simplicity and
tor 2 is actually part of a closed circuit in which I flows,
except for the length /. More-
convenience the details of the circuit are omitted
horizontal plane and are
over, it is assumed that conductors 1 and 2 lie in the same
found that with this
parallel to each other. In accordance with Ampere’s law it is
conduc tor directed to the
configuration there exists a force on the elemental
to be directl y proportional
right. Furthermore, the magnitude of the force is found
tors as well as inversely
to I,, I, 1, and the medium surrounding the conduc
units? the magnit ude of this
proportional to the distance between them. In mks
force can be shown to be given by

newtons (N) (1-1)

a property of
where J, and J, are expressed in amperes, / and r in meters, and yp is
is that, if the
the medium. A further interesting revelation about this experiment
points in space
elemental conductor 2 is used as an exploring device to find those
, the locus is
where the force is of constant magnitude and outwardly directed
or 1. In other
found to be acircle of radius r and centered along the axis of conduct
In this
words it is possible to identify a field having constant lines of force.
connection it is useful at this point to rewrite Eq. (1-1) as follows:
F=TIB (1-2)
a quantitative
+It should not be here inferred that this is the only starting point in developing
circuit. Faraday’s law of induction is equally valid as a starting point and is
theory of the magnetic
rather than a theory
preferred when the goal is the development of an electromagnetic wave theory
leading to the treatment of electromechanical energy conversion.
tMeter-kilogram-second system of units.
Sec. 1-1 Ampere’s Law—Definition of Magnetic Quantities 3

Current
element 2 rat
((f FL
Sin ach
Conductor
1 Figure 1-1 [Illustrating the force exist-
ing between a current element J,/ and a
very long conductor carrying current J,
as described by Ampere’s law.

where

_ bh
2ar i
Here for reasons that become clearer presently, B is definedt as the magnetic field
or, still better, the magnetic flux density existing in the region where the elemental
conductor 2 lies and has the units of webers/meter? or teslas (T). As Eq. (1-2)
reveals, the magnetic flux density is defined as a measured force F per known
current element J>/. Equation (1-3) is merely a means of identifying the manner in
which the current J, influences the force field about the current element. It is
significant to note that, as long as J, is not zero, the force field and the magnetic
field have the same characteristics—both have a circular locus and both are
vector quantities possessing magnitude and direction. However, because of the
way B is defined, the magnetic field exists as long as J; is not zero irrespective of
the value of J,. Magnetic flux density thus represents the effects caused by the
moving charges that make up J.
In our study of electrical machinery, which comes later, conductor 1 will be
referred to as the field winding because it sets up the working magnetic field,
whereas the total circuit of which the elemental conductor is a part is called the
armature winding.
The direction of the magnetic field is readily determined by the right-hand
rule which states that, if the field winding conductor (1 in this case) is grasped in
the right hand with the thumb pointing in the direction of current flow, the lines of
flux (or flux density) will be in the direction in which the fingers wrap around the
conductor.

+This definition of B is correct for the configuration of Fig. 1-1. For other configurations the
forms will be similar, but the constants will differ generally.
 Magnetic Theory and Circuits Chap. 1
4

Permeability

are known in this equation with the

A glance at Eq. (1-1) shows that all the factors
which is a characteristic of the sur-
exception of the proportionality factor pw,
iment of Fig. 1-1 in iron rather than
rounding medium. Upon repeating the exper
r for the same values of 11, hb, 1,
air, it is found that the force is many times greate
d from Eq. (1-1) for various media
andr. Therefore, it follows that 1 may be define
ver, becau se of the way magnetic
because it is the only unknown quantity. Moreo
that the effect of the surrounding
flux density was defined, Eq. (1-3) indicates
to which it increa ses or decreases
medium may be described in terms of the degree
t J;. Thus when iron rather than
the magnetic flux density for a specified curren
iron provid es a greate r penetration
free space is the medium it can be said that the
there is a greate r flux density. This
of the magnetic field in a given region, i.e.,
the condu ctors are embedded is
property of the surrounding medium in which
called permeability.
in a vacuum (free
When the conductors of Fig. 1-1 are assumed placed
of 1), 2, /, andr, the solution
space) and the force is measured for specified values
Eq. (1-1) and expressed in mks
for the permeability of free space obtained from
units comes out to be

po = 4 X 1077 (1-4)

The unit of permeability also follows from Eq. (1-1). Thus

newtons
a
Mo “ ampere?

where « denotes ‘‘is proportional to.’” But

newton-meter = joule = volt x ampere x second

or
. volt X ampere X second volt x second
te ee ee
,0 ampere? X meter ampere X meter
ed in henrys.
However, volt X seconds/ampere is the unit of inductance express
Accordingly, permeability is expressed in units of henrys/meter (H/m).
the
In those cases where the surrounding medium is other than free space,
from Eq. (1-1). A compari son with
absolute permeability is again readily found
permea-
the result obtained for free space then leads to a quantity called relative
bility, u,. Expressed mathematically we have

p= _ba (1-5)
Sec. 1-1 Ampere’s Law—Definition of Magnetic Quantities 5

Equation (1-5) clearly indicates that relative permeability is simply a numeric

which expresses the degree to which the magnetic flux density is increased or
decreased over that of free space. For some materials, such as Deltamax, the
value of , can exceed one hundred thousand. Most ferromagnetic materials,
however, have values of yu, in the hundreds or thousands.

Magnetic Flux ®

Before considering the manner in which this quantity is defined, let us first investi-
gate more closely the dimensions of the magnetic flux density B as described by its
defining expression Eq. (1-2). Thus
a newtons
amperes X meter

. volts X seconds X amperes _ volts x seconds

4 meter (1-6)
amperes X meter?
A study of this expression reveals that B is in fact a density because it involves
meters squared in the denominator. Accordingly, it is reasonable to expect that,
since B is defined as a magnetic flux density, then multiplication by the effective
area that B penetrates should yield the total magnetic flux. To illustrate this point
refer to Fig. 1-2, which shows a coil of area ab lying in the same horizontal plane
containing conductor 1. We already know that when a current J; flows through
this conductor a magnetic field is created in space and specifically is described by
Eq. (1-3). To find the total flux penetrating the coil it is necessary merely to
perform an integration of B over the surface area involved. Of course if B were a
constant over the area of concern, the flux would be simply the product of B and

ere Conductor 1

(a) Top view (b) Front view

flux.
Figure 1-2 Associating an area with a magnetic field B to identify a magnetic
 Magnetic Theory and Circuits Chap. 1
6

f 60°

(+) B, Figure 1-3 Same as Fig. 1-2(b) except

n
Conductor 1 60° that the coil is tilted 60° relative to the
B horizontal plane.

the area ab. Next consider that the plane of the coil is tilted with respect to the
plane of conductor 1 by 60°, as depicted in Fig. 1-3. Clearly now the total flux
penetrating the coil is reduced by a factor of 3. If the coil is oriented to a position
of 90° with respect to the horizontal plane, no flux threads the coil.
On the basis of these observations, then, the magnetic flux through any
surface is more rigorously defined as the surface integral of the normal component
of the vector magnetic field B. Expressing this mathematically we have

where s stands for surface integral, A represents the area of the coil, and B,, is the
normal component of B to the coil area. From expression (1-6) we know that
magnetic flux must have the dimensions of volt-seconds. However, this is more
commonly called webers and denoted by Wb. The volt-seconds unit of flux is
better understood in terms of Faraday’s law of induction (see Eq. (2-8)).

Magnetic Field Intensity H

Often in magnetic circuit computations it is helpful to work with a quantity repre-

senting the magnetic field which is independent of the medium in which the
magnetic flux exists. This is especially true in situations such as are found in
electrical machinery where a common flux penetrates several different materials,
including air. A glance at Eq. (1-3) reveals that division of B by w identifies such a
quantity. Accordingly, magnetic field intensity is defined as

B
H=z—Fi (1-8)
1-8

and has the units of

newtons
amperes X meter amperes A
newtons meter. m
ampere?
Sec. 1-1 Ampere’s Law—Definition of Magnetic Quantities 7

Thus H is dependent upon the current that produces it and also on the geometry of
the configuration but not the medium. For the system of Fig. 1-1 the value of the
magnetic field intensity immediately follows from Eq. (1-3) and is given by

sefe Ii.
207
Because H is independent of the medium it is frequently looked upon as the
intensity that is responsible for driving the flux density through the medium.
Actually, though, H is a derived result.
More generally the units for H are ampere-turns/meter rather than amperes/
meter. A little thought should make this apparent whenever the field winding is
made up of more than just a single conductor.

Ampere’s Circuital Law

Now that the magnetic field intensity has been defined and shown to have dimen-
sions of ampere-turns per meter, we can develop a very useful relationship.
Recall that H is a vector having the same direction as the magnetic field B. For the
configuration of Fig. 1-1 H has the same circular locus as B. A line integration of
H along any given closed circular path proves interesting. Of course the line
integral is considered because H involves a per unit length dimension. Thus

¢Hdl = ihezo d=; amperes (1-9)

(Again keep in mind that the units here would be ampere-turns if more than one
conductor were involved in Fig. 1-1.) Equation (1-9) states that the closed line
integral of the magnetic field intensity is equal to the enclosed current (or ampere-
turns) that produces the magnetic field lines. This relationship is called Ampere’s
circuital law and is more generally written as

where ¥ denotes the ampere-turns enclosed by the assumed closed flux line path.
The quantity ¥ is also known as the magnetomotive force and frequently abbrevi-
ated mmf. This relationship is useful in the study of electromagnetic devices and
is referred to in subsequent chapters.

Derived Relationships

In the preceding pages the fundamental magnetic quantities—flux density, flux,

field intensity, and permeability—are defined starting with Ampere’s basic experi-
ment involving two current-carrying conductors. By the appropriate manipula-
tion of these quantities additional useful results can be obtained.
 Magnetic Theory and Circuits Chap. 1
8

for
Equation (1-8) is a vector equation describing the magnetic field intensity
total path length of a flux line is assumed to
a given geometry and current. If the
(mmf) associat ed with the specified flux
be J, then the total magnetomotive force
line is
B
F = Hl = = (1-11)
bh
known area
Now in those situations where B is a constant and penetrates a fixed,
flux may be written from Eq. (1-7) as
A, the corresponding magnetic

Inserting Eq. (1-12) into Eq. (1-11) yields

F = Hl=@ Ge (1-13)
pA
a
The quantity in parentheses in this last expression is interesting because it bears
very strong resemblance to the definition of resistance? in an electric circuit.
Recall that the resistance in an electric circuit represents an impediment to the
flow of current under the influence of a driving voltage. An examination of Eq.
(1-13) provides a similar interpretation for the magnetic circuit. We are already
aware that & is the driving mmf which creates the flux ® penetrating the specified
cross-sectional area A. However, this flux is limited in value by what is called the
reluctance of the magnetic circuit, which is defined as

(1-14)

No specific name is given to the dimension of reluctance except to refer to it as so

many units of reluctance.
Equation (1-14) reveals that the impediment to the flow of flux which a
magnetic circuit presents is directly proportional to the length and inversely pro-
portional to the permeability and the cross-sectional area—results which are en-
tirely consistent with physical reasoning.
Inserting Eq. (1-14) into Eq. (1-13) yields

F = OR (1-15)

which is often referred to as Ohm’s law of the magnetic circuit. It is important to

keep in mind, however, that these manipulations are permissible as long as B and
A are fixed quantities.

+The resistance of a coper wire of length | and cross-sectional area A is given by R = pl/A,
where p is the resistivity of the copper.
 Sec. 1-1 Ampere’s Law—Definition of Magnetic Quantities 9

e e e + + +
B-field directed out B-field directed into paper
of paper here
Io F
I, i Ky

Current Figure 1-4 Showing the direction of

Conductor 1 element the force when the current element is
e fe) e + + no longer located parallel to conductor
1 but remains in the same plane.

In most practical ferromagnetic materials the functional relationship be-

tween ¥ and © (or the corresponding quantities H = ¥/] and B = ®/A) is nonlin-
ear, which indicates that the reluctance & varies with the applied mmf. A glance
at Fig. 1-10 makes this apparent. In some applications it is the change in reluc-
tance about an operating point that is important rather than the total value of the
reluctance. This leads naturally to the notion of differential reluctance, which is
defined as
foe sani) od, l
Ra = Gq ~ dBA) AGB wsA
(1-16)

Here jz is the differential permeability and is represented graphically as the slope

of a tangent drawn to the B versus H curve.

Ampere’s Law for Various Orientations of the

Current Element

parallel
In Fig. 1-1 the assumption was made that the current element was located
this orientati on was sufficient
to conductor 1 and lying in the same plane. Because
es—it was
for the purpose at hand—to define the fundamental magnetic quantiti
interest of furnishin g a more
pursued as a matter of convenience. However, in the
the effect on the force
complete picture of the experiment we shall now consider
different orientati ons. Con-
of placing the current element J,/ in two additional
parallel to conducto r 1 but
sider first that the current element is no longer placed
Refer to Fig. 1-4. The dots
continues to be located in the same horizontal plane.
outward on the left side of
in this figure indicate that the magnetic field is directed
paper) on the right side of
conductor 1 and inward (with respect to the plane of the
results of this experime nt
the conductor as revealed by the right-hand rule. The
that found by using the
show that the magnitude of the force is the same as
g because the value of
configuration of Fig. 1-1.+ This conclusion is not surprisin
ed so that Eq. (1-2) is
the magnetic flux density as well as I, and / remain unchang

element, the value of B is assumed to be

+Because of the infinitesimal nature of the current
constant about J,/ irrespective of its orientation.
 Magnetic Theory and Circuits Chap. 1
10

force.
still valid in describing the force. The only change is the direction of the
current
However, as Fig. 1-4 indicates, the force continues to be normal to the
for
element. It is worthwhile to keep this point in mind. Presently a general rule
establishing the force direction for all configurations is described .
Next let us consider that orientation of the current element which places it
parallel to conductor 1 but inclined at an angle @ = 30° with respect to the verti-
cal. A side-view projection of the configuration is depicted in Fig. 1-5. Note that
the magnetic field is directed downward along the vertical for this view. Actually,
of course, the locus of B is circular, but in Fig. 1-5 we are looking at just that small
portion of the B field about the plane containing conductor 1. With this configura-
tion the force on the current element is found to have the same direction but one-
half the magnitude of that obtained with the orientation of Fig. 1-1. It follows,
then, that the angle between the current vector 1,1 and the flux density B affects
the magnitude of the force. As a matter of fact, further experimentation reveals
that the general expression for the force is

ene sind ula mun (1-17)

Equation (1-17) conveys information solely about the magnitude of the force
and not its direction. It is possible, however, by employing the notation of vector
analysis, to rewrite Eq. (1-17) so that information about magnitude as well as
direction is present. This result is readily accomplished by the use of the cross-
product notation between two vectors, yielding a third vector having magnitude
and direction. Thus Eq. (1-17) is more completely expressed as

F=1L1xB N (1-18)

The cross notation must always be understood to involve the sine of the angle
between the two vectors B and / (or the direction of J), which is determined by the
orientation of /). Moreover, wherever the cross product is involved, the direction
of the resultant vector is always normal to the plane containing the vectors B and /
in the sense determined by the direction of advance of a right-hand screw as lis °

Ip
Conductor 1

Current ; E 5 ae
Figure 1-5 Side-view projection of
Fig. 1-1 but with the current element
tilted relative to the horizontal plane.
The force is directed out of the paper.
 Sec. 1-2 Theory of Magnetism 11

turned into B through the smaller of the two angles made by the vectors. Accord-
ingly, in the configuration of Fig. 1-5 the direction of the force is found by turning
I,l into B and then noting that this would cause a right-hand screw to advance out
of the plane of the paper. Hence the force is directed outward, and this corre-
sponds with the experimentally established result.
It is also possible to determine the direction of the force by means of another
right-hand rule, which requires that the forefinger be put in the direction of the
current and the middle finger in the direction of B, with the two assumed to be
lying in the same plane. The thumb of the right hand then points in the direction of
the force when placed perpendicular to the other two fingers.

1-2 THEORY OF MAGNETISM

To understand the magnetic behavior of materials, it is necessary to take a micro-

scopic view of matter. A suitable starting point is the composition of the atom,
which Bohr described as consisting of a heavy nucleus and a number of electrons
moving around the nucleus in specific orbits. Closer investigation reveals that the
atom of any substance experiences a torque when placed in a magnetic field; this
is called a magnetic moment. The resultant magnetic moment of an atom depends
upon three factors—the positive charge of the nucleus spinning on its axis, the
negative charge of the electron spinning on its axis, and the effect of the electrons
moving in their orbits. The magnetic moment of the spin and orbital motions of
the electron far exceeds that of the spinning proton. However, this magnetic
moment can be affected by the presence of an adjacent atom. Accordingly, if two
hydrogen atoms are combined to form a hydrogen molecule, it is found that the
electron spins, the proton spins, and the orbital motions of the electrons of each
atom oppose each other so that a resultant magnetic moment of zero should be
expected. Although this is almost the case, experiment reveals that the relative
permeability of hydrogen is not equal to 1 but rather is very slightly less than
unity. In other words, the molecular reaction is such that when hydrogen is the
medium there is a slight decrease in the magnetic field compared with free space.
This behavior occurs because there is a precessional motion of all rotating charges
field
about the field direction, and the effect of this precession is to set up a
opposed to the applied field regardless of the direction of spin or orbital motion.
for obvi-
Materials in which this behavior manifests itself are called diamagnetic
are
ous reasons. Besides hydrogen, other materials possessing this characteristic
silver and copper.
is
Continuing further with the hydrogen molecule, let us assume next that it
complete neu-
made to lose an electron, thus yielding the hydrogen ion. Clearly,
the spin and orbital electron motions no longer takes place. In fact,
tralization of
magnetic
when a magnetic field is applied, the ion is so oriented that its net
in flux den-
moment aligns itself with the field, thereby causing a slight increase
istic of such
sity. This behavior is described as paramagnetism and is character
 Magnetic Theory and Circuits Chap. 1
12

crystal: (a) unmagnetized and

Figure 1-6 Representation of a ferromagnetic
(b) fully magnetized by the field H.

materials as aluminum and platinum. Paramagnetic materials have a relative per-

meability slightly in excess of unity.
So far we have considered those elements whose magnetic properties differ
only very slightly from those of free space. As a matter of fact the vast majority of
materials fall within this category. However, there is one class of materials—
principally iron and its alloys with nickel, cobalt, and aluminum—for which the
relative permeability is very many times greater than that of free space. These
materials are called ferromagnetic and are of great importance in electrical engi-
neering. We may ask at this point why iron (and its alloys) is so very much more
magnetic than other elements. Essentially, the answer is provided by the domain
theory of magnetism. Like all metals, iron is crystalline in structure with the
atoms arranged in a space lattice. However, domains are subcrystalline particles
of varying sizes and shapes containing about 10'° atoms in a volume of approxi-
mately 10~° cubic centimeters. The distinguishing feature of the domain is that
the magnetic moments of its constituent atoms are all aligned in the same direc-
tion. Thus in a ferromagnetic material, not only must there exist a magnetic
moment due to a nonneutralized spin of an electron in an inner orbit, but also the
resultant spin of all neighboring atoms in the domain must be parallel.
It would seem by the explanation so far that, if iron is composed of com-
pletely magnetized domains, then the iron should be in a state of complete magne-
tization throughout the body of material even without the application of a magne-
tizing force. Actually, this is not the case, because the domains act independently
of each other, and for a specimen of unmagnetized iron these domains are aligned
haphazardly in all directions so that the net magnetic moment is zero over the
specimen. Figure 1-6 illustrates the situation diagrammatically in a simplified

+Derived from the Latin word for iron—ferrum.

 Sec. 1-3 Magnetization Curves of Ferromagnetic Materials 13

fashion. Becausé of the crystal lattice structure of iron the ‘‘easy’’ direction of
domain alignment can take place in any one of six directions—left, right, up,
down, out, or in—depending upon the direction of the applied magnetizing force.
Figure 1-6(a) shows the unmagnetized configuration. Figure 1-6(b) depicts the
result of applying a force from left to right of such magnitude as to effect alignment
of all the domains. When this state is reached the iron is said to be saturated—
there is no further increase in flux density over that of free space for further
increases in magnetizing force.
Large increases in the temperature of a magnetized piece of iron bring about
a decrease in its magnetizing capability. The temperature increase enforces the
agitation existing between atoms until at a temperature of 750°C the agitation is so
severe that it destroys the parallelism existing between the magnetic moments of
the neighboring atoms of the domain and thereby causes it to lose its magnetic
property. The temperature at which this occurs is called the curie point.

1-3 MAGNETIZATION CURVES OF FERROMAGNETIC MATERIALS

for
If the experiment of Fig. 1-1 is repeated with iron or steel as the medium
current J; and the correspo nding values of wu
increasing values of the field winding
permeabil ity varies considera bly with the
computed, it is found that the relative
variation
magnetizing force that establishes the operating flux density. A typical
Here p, is plotted versus flux density rather
of 1, for cast steel appears in Fig. 1-7.
is the onset of the realignme nt of more and
than the magnetizing force because it
change in permeabil ity. Unfortuna tely, the
more domains that brings about the
magnetis m is not so far advanced that it
state of development of the theory of
propertie s of a material on a purely theoreti-
allows the prediction of the magnetic
For exam-
cal basis even though the exact composition of the material is known.
to say exactly what the flux density
ple, with the present theory it is not possible
specified value of the magnetizi ng force.
will be in a given specimen of iron for a
informati on by consultin g technical and
Rather, it is customary to obtain this

Hr 1400
1200
1000
800
600

0.8 4 1.6 ;
O 0.4
Figure 1-7 Graph of relative permea-
B(teslas) bility versus flux density.
14 Magnetic Theory and Circuits Chap. 1

descriptive bulletins where the measured magnetic properties of a representative

sample of the specimen are published. These bulletins are made available to users
by the manufacturers of magnetic steels, and they include information on such
varied shapes and forms as sheets, wires, bars, and even castings weighing up to
hundreds of tons.
Usually the published magnetic characteristics of the various iron and steel
samples are presented as plots of flux density B as a function of the magnetic field
intensity H. In the interest of presenting a complete graphical picture of the
functional relationship existing between these two quantities as well as to define
additional terms used in this connection, refer to Fig. 1-8. Assume that the steel
specimen is initially unmagnetized and is in the form of a toroidal ring with a coil
of N turns wrapped around it. Assume too that the coil can be energized from a
variable voltage source capable of furnishing current flow in either direction in the
coil. As the current J is increased from zero in the positive direction (current
flowing into the top terminal of the coil), an increasing magnetic flux ® can be
measured taking place within the body of the toroid in the clockwise direction.
For any fixed value of J there is a specific value of flux. Then by Eq. (1-12) the
corresponding flux density is determined since the toroidal cross-sectional area is
known. Moreover, the magnetomotive force NIJ can be replaced by HI,, in accor-
dance with Ampere’s circuital law [Eq. (1-10)], where /,, is the mean length of path
of the toroid (as shown by the broken-line circle in Fig. 1-8). The two fundamental
quantities involved in this arrangement then are B in webers/meter? and H ex-
pressed in ampere-turns per meter. H is the quantity we want to deal with rather
than magnetomotive force because, for the same flux density, doubling the mean
magnetic length will not change H but will require doubling the magnetomotive
force. The conclusion to be drawn is that a plot of B versus H is a universal plot
for the given material because it can be extended to any geometry of cross-
sectional area and length. In contrast, a plot of ® versus NIJ is limited to a single
geometrical configuration. Therefore, a plot of the magnetic characteristics of a
material always involves plotting B versus H. For the virgin sample of Fig. 1-8 the
graph of B versus H follows the curve Oa of Fig. 1-9 for field intensities up to H,.
Take note of the nonlinear relationship existing between these two quantities.
Another interesting characteristic of ferromagnetic materials is revealed
when the field intensity, having been increased to some value, say H,, is subse-
quently decreased. It is found that the material opposes demagnetization and,
accordingly, does not retrace along the magnetizing curve Oa but rather along a
curve located above Oa. See curve ab in Fig. 1-9. Furthermore, it is seen that,
when the field intensity is returned to zero, the flux density is no longer zero as
was the case with the virgin sample. This happens because some of the domains
remain oriented in the direction of the originally applied field. The value of B that
remains after the field intensity H is removed is called residual flux density.
Moreover, its value varies with the extent to which the material is magnetized.
The maximum possible value of the residual flux density is called retentivity and
results whenever values of H are used that cause complete saturation.
Sec. 1-3 Magnetization Curves of Ferromagnetic Materials 15

D
Cross _section

Sample steel specimen

Figure 1-8 Obtaining the magnetization curve of a sample steel specimen.

Frequently, in engineering applications of ferromagnetic materials, the steel

is subjected to cyclically varying values of H having the same positive and nega-
tive limits. As H varies through many identical cycles, the graph of B versus H
gradually approaches a fixed closed curve as depicted in Fig. 1-9. The loop is
always traversed in the direction indicated by the arrows. Since time is the im-
plicit variable for these loops, note that B is always lagging behind H. Thus, when
is
H is zero, B is finite and positive, as at point b, and when B is zero, as at c, H
finite and negative, and so on. This tendency of the flux density to lag behind the
field intensity when the ferromagnetic material is in a symmetrically cyclically

B (webers )

Normal magnetization curve

Retentivity
Residual flux
density

Hysteresis loop corresponding to

value of H large enough to saturate
the steel specimen fully

curve.
Figure 1-9 Typical hysteresis loops and normal magnetization
 16 Magnetic Theory and Circuits Chap. 1

magnetized condition is called hysteresis +and the closed curve abcdea is called a
hysteresis loop. Moreover, when the material is in this cyclic condition the
amount of magnetic field intensity required to reduce the residual flux density to
zero is called the coercive force. Usually, the larger the residual flux density, the
larger must be the coercive force. The maximum value of the coercive force is
called the coercivity.
A glance at the hysteresis loops of Fig. 1-9 makes it quite evident that the
flux density corresponding to a particular field intensity is not single-valued. Its
value lies between certain limits depending upon the previous history of the ferro-
magnetic material. However, since in many situations involving magnetic devices
this previous history is unknown, a compromise procedure is used in making
magnetic calculations by working with a single-valued curve called the normal
magnetization curve. This curve is found by drawing a curve through the tips of a
group of hysteresis loops generated while in a cyclic condition. Such a curve is
Oafg in Fig. 1-9. Typical normal magnetization curves of commonly used ferro-
magnetic materials appear in Fig. 1-10.
A final observation is in order at this point. By Eq. (1-8) the permeability of
a material may be expressed as a ratio of B to H. Coupling this with the nonlinear
variation existing between B and H (see Fig. 1-9) bears out the variation of perme-
ability with flux density as already cited in connection with Fig. 1-7. As a matter
of fact, for a material in a cyclic condition the permeability is nothing more than
the ratio of B to H for the various points along the hysteresis loop.

1-4 THE MAGNETIC CIRCUIT: CONCEPT AND ANALOGIES

In general, problems involving magnetic devices are basically field problems be-
cause they are concerned with quantities such as ® and B which occupy three-
dimensional space. Fortunately, however, in most instances the bulk of the space
of interest to the engineer is occupied by ferromagnetic materials except for small
air gaps which are present either by intention or by necessity. For example, in
electromechanical energy-conversion devices the magnetic flux must penetrate a
stationary as well as a rotating mass of ferromagnetic material, thus making an air
gap indispensable. On the other hand, in other devices an air gap may be inten-
tionally inserted to mask the nonlinear relationship existing between B and H.
But in spite of the presence of air gaps it happens that the space occupied by the
magnetic field and the space occupied by the ferromagnetic material are practi-
cally the same. Usually this is because air gaps are made as small as mechanical
clearance between rotating and stationary members will allow and also because
the iron by virtue of its high permeability confines the flux to itself as copper wire
confines electric current or a pipe restricts water. On this basis the three-dimen-
sional field problem becomes a one-dimensional circuit problem and in accordance

}Derives from the Greek hysterein meaning to be behind or to lag.

 A-t/m
intensity,
field
Magnetic

I00
80
60
materials.
ferromagnetic
typical
of
Magnetizatio
1-10
Figure
curves
(S)
|

es |.
t+

oe ie)

Oe ©
1 ‘Ajisuap xni} dijauboy
 Magnetic Theory and Circuits Chap. 1
18

Iron

Air gap
Figure 1-11 Typical magnetic circuit
involving iron and air.

Figure 1-12 Single-line equivalent

circuit of Fig. 1-11.

with Eq. (1-15) leads to the idea of a magnetic circuit. Thus we can look upon the
magnetic circuit as consisting predominantly of iron paths of specified geometry
which serves to confine the flux; air gaps may be included. Figure 1-11 shows a
typical magnetic circuit consisting chiefly of iron. Note that the mmf of the coil
produces a flux which is confined to the iron and to that part of the air having
effectively the same cross-sectional area as the iron. Furthermore, a little thought
reveals that this magnetic circuit may be replaced by a single-line equivalent
circuit as depicted in Fig. 1-12. As suggested by Eqs. (1-14) and (1-15) the equiva-
lent circuit consists of the magnetomotive force driving flux through two series-
connected reluctances—;, the reluctance of the iron, and R,, the reluctance of
the air.
This analogy of the magnetic circuit with the electric circuit carries through
in many other respects. For the sake of completeness these details are presented
next for the case of a2 toroidal copper ring and a toroidal iron ring having the same
mean radius r and cross-sectional area A.
Electric Case Magnetic Case

The toroidal copper ring is as- The toroidal iron ring is assumed
sumed open by an infinitesimal to be wound with N turns of wire
amount with the ends connected to so that with a current 7 flowing
a battery; a current of J amperes through it the mmf creates the
flows through the ring. flux ®.
Sec. 1-4 The Magnetic Circuit: Concept and Analogies 19

Driving Force

applied battery voltage = E applied ampere-turns = #

Response
-
sight: driving for 2
re Meath
magnetic reluctance
electric resistance

E F (1-19)
Beers
soak
Impedance

Impedance is a general term used to indicate the impediment to a driving

force in establishing a response.
; l l
resistance = R = p mi eluctance aa)= R
reluct wi
OF a) ae (1-20)

where | = 27r, the mean length of where / = 27rr, the mean length of
turn of the toroid, and A is the to- turn of the toroid, and A is the to-
roidal cross-sectional area. roidal cross-sectional area.

Equivalent Circuit

@ (4-21)

E=IR F= OR

Electric Field Intensity Magnetic Field Intensity

With the application of the When a magnetomotive force

voltage E to the homogeneous cop- is applied to the homogeneous iron
per toroid, there is produced within toroid, there is produced within the
the material an electric potential material a magnetic potential gradi-
gradient given by ent given by

(1-22)
Ofe OE

=—— V/m H= a i
27r
A-t/m

This electric field must occur As already pointed out in con-

in a closed path if it is to be main- nection with Ampere’s circuital
tained. It then follows that the law, the closed line integral of H
closed line integral of ¢ is equal to equals the enclosed mmf. Thus
the battery voltage E. Thus
$edl=E $ Hdl =F (1-23)
20 Magnetic Theory and Circuits Chap. 1

Electric Potential Difference Magnetic Potential Difference

If it is desired to find the volt- This is the mmf drop appear-

age drop occurring between two ing between two points when flux
points a and b of the copper toroid, flows. Thus the portion of the total
we may write applied mmf appearing between
points a and b is found similarly:

Fe OR
:
Vr = fed =" f'dt= ly Use = |” Hdl = = las = las

soles: lab =_ ht
iA
ee ee eee
= 7 P| |ab = Ip A TRap

Le.
ar = TRap Ua a DRap (1-24)

where R,, is the resistance of the where &,, is the reluctance of the
copper toroid between points a iron toroid between points a and b.
and b.

Current Density Flux Density

By definition, current density Flux density is expressed as

is the number of amperes per unit webers per unit area. Thus
area. Thus

feed elingore FS ae ee
nm Ait AR © ASA »p =A AR AdiaAy ©
or or

B
e=pJ H=— fe (1-25)

This last expression is often re-

ferred to as the microscopic form
of Ohm’s law.

It should not be inferred from the foregoing that electric and magnetic cir-
cuits are analogous in all respects. For example, there are no magnetic insulators
analogous to those known to exist for electric circuits. Also, when a direct cur-
rent is established and maintained in an electric circuit, energy must be continu-
ously supplied. An analogous situation does not prevail in the magnetic case,
where a flux is established and maintained constant.
 Sec. 1-6 Magnetic Circuit Computations 21

1-5 UNITS FOR MAGNETIC CIRCUIT CALCULATIONS

Magnetic circuit calculations can be carried out by use of any one of several
different systems of units. These various systems arose initially because it was
thought that the phenomena of electricity and magnetism were unrelated—which
lead to the development of a separate system of units for each—and, second,
because there was a need to deal with practical values of the units once the
relationship was discovered. Up to now attention has been given exclusively to
the mks (meter-kilogram-second) system of units as developed by Giorgi about the
turn of the twentieth century. This policy is prompted by the acceptance in 1960
of the mks system of units as the standard for scientific work and now referred to
as SI units (Systéme International Unités). However, a good part of the past
literature is written in the units of the cgs (centimeter-gram-second) system.
Furthermore, many of the present-day computations are carried on in the mixed
system employing such units as ampere-turns/inch, maxwells/inch*, and ampere-
turns because of the convenience they offer in dealing with dimensions that are
expressed in inches. For these reasons the units of all three systems are shown in
Table 1-1.
The weber, which is the unit of flux in the mks system, is equal to 108
maxwells (or lines), where the maxwell is the unit of flux in the cgs system. The
gilbert is the cgs unit for mmf and is equal to 0.47 times the number of ampere-
turns. The cgs unit for magnetic field intensity H is the oersted (or gilbert/cm) and
the cgs unit for flux density B is the gauss (or lines/cm?). The relationships
existing for the same quantity among the various systems of units are given in the
last column of the table.

1-6 MAGNETIC CIRCUIT COMPUTATIONS

s fall into
Basically magnetic circuit calculations involving ferromagnetic material
value of the flux is known and it is required to find the
two classes. In the first the
the situation typical of the design of ac and de electro-
mmf to produce it. This is
On the basis of the desired voltage rating of an
mechanical energy converters.
information about the
electric generator or the torque rating of an electric motor
obtained . Then with this knowledge and the
required magnetic flux is readily
the total mmf needed to establish the flux is
configuration of the magnetic circuit
In the second case it is the flux for which we must
determined straightforwardly.
the magneti c circuit and the applied mmf. An
solve, knowing the geometry of
this situation prevails is the magneti c amplifier,
engineering application in which
the resultant magneti c flux caused by one or
where it is often necessary to find
lity) of the ferromag-
more control windings. Because the reluctance (or permeabi
of this problem is considerably more
netic material is not constant, the solution
illustrat ed by the example s that follow.
involved than that of the first type as
 pas,fagomeet pei
RIso} |= z1a}our
sassnes (QO'O] 12qaM |
RISo} | = _ _lojou sossne3
19}0UI
cyouysoul] QOS'r9 — 19qam | HdW=¢q = ,W9/soul] Ayisuop xnj4
H'‘W6re=4 youl /soul] Hod =g L S19qoM
2youl/soul] ¢p’9 = ssnes |
pajsis0Rid oor _ YON I
COG? aN:
: (AS(Gm lB Ayisuoyul
pa}s190 C6h'0 = 5 = =
270 = TV Spojs190 PIPY
1 bi,
$pd}S120 QO] = Poysssovld | IN~+'0 = wo/qis douse]
19}0W/1-Y 9°6L = P23S190 |

S[[aMxeul

soul]
V,-01Ly'77
soul]

=
ee l =© = sjjemxeul =O an Jaqom
SURO

=
ee
‘nl
[sma
ealy
7Sayoul OSS] = -1o}ou
Soyoul p'6¢ = Jo}oW
19]9U [Q'°0 = Jo}OWNUIO | Ty
“wed
“WIOU “SqY
aoeds 901
Ayriqeoulled

-ouadwie
ulin}
spoqyts IN=8
LST = SHOQIIs 4p" = 1-V IN“¥'0 = £ spaqy!3

=£

ry
rv

IN
swags QO] = Moqyseid Pea.
-
$10}DBJ UOISISAUO,)
SLINN DILANDVWsL-L ATEaVL

22
Sec. 1-6 Magnetic Circuit Computations 23

Figure 1-13 Toroid composed of three

different materials.

Example 1-1
A toroid is composed of three ferromagnetic materials and is equipped with a coil
having 100 turns as depicted in Fig. 1-13. Material a is a nickel-iron alloy having a
mean arc length J, of 0.3 m. Material b is medium silicon steel and has a mean arc
length J, of 0.2 m. Material c is of cast steel having a mean arc length equal to 0.1 m.
Each material has a cross-sectional area of 0.001 m’.
(a) Find the magnetomotive force needed to establish a magnetic flux of ® =
6 X 10-4 Wb = 60,000 lines.
(b) What current must be made to flow through the coil?
(c) Compute the relative permeability and reluctance of each ferromagnetic
material.

Solution (a) To obtain the total mmf of the coil all we need to do is to apply
Ampere’s circuital law. Thus
F = OF ar U, + U. = Agly + Ayl, + Aol.

from a
The unknown quantities here are H,, H,, and H,. These can readily be found
because the
knowledge of the flux density, which here is the same for each section
flux is common and the cross-sectional areas are the same. Hence

® 0.0006
= 0.6T
By = By= Be =A = 0.001
alloy of Fig. 1-10
Now H, is found by entering the B-H curve of the nickel-iron
corresponding to B, = 0.6. This yields
H, = 10 A-t/m

Similarly
H, = 77 A-t/m
H, = 320 A-t/m

Accordingly, the total required mmf is

§ = Hilo Aglyt Hele
Il 10(0.3) + 77(0.2) + 320(0.1)
3+ 15.4 + 32.0 = 50.4 A-t
24 Magnetic Theory and Circuits Chap. 1

Note that, although the path length of cast steel is the smallest, it nonetheless
requires the greatest portion of the mmf to force the specified flux through. This
happens because of its much lower permeability as shown in part (c).
(b) In the SI system the mmf is equal to the number of ampere-turns. Hence

Fe 50.4
T= 3 = 709 = 05044

(c) From Eq. (1-8)

Be 0.6" 2
[i a = i024 0.06 H/m

Also

Mrabo — Ma

y ie e Bee oe OW Oy
pon 45 10-*
Wanyeae
Furthermore, from Eq. (1-13) the reluctance is found to be

OFF 3
Bee x10 = 5000 rationalized mks units of reluctance

Proceeding in a similar fashion for materials b and c leads to the following

results:
rp = 6207 Ry, = 25,667

bre 1492 R, = 53,333

Next we consider the more difficult problem: that of finding the flux in a
given magnetic circuit corresponding to a specified mmf. The solution cannot be
arrived at directly because, as a result of the nonlinear relationship between B and
H, there are too many unknowns. The easiest way of finding the solution is to
employ a cut-and-try procedure guided by a knowledge of the permeability char-
acteristics of the materials such as appears in their magnetization curves. The
following example illustrates the technique involved.
Example 1-2
For the toroid of Example 1-1, shown in Fig. 1-13, find the magnetic flux produced by
an applied magnetomotive force of ¥ = 35 A-t.
The solution cannot be determined directly because to do so we must know the
reluctance of each part of the magnetic circuit, which can be known only if the flux
density is known—which means that ® must be known right at the start. This is
clearly impossible.
To obtain the solution by the cut-and-try procedure, we begin by first assuming
that all of the applied mmf appears across the material having the highest reluc-
tance. This yields an approximate value of ® which can subsequently be refined. A
glance at Fig. 1-10 shows that the poorest magnetic ‘‘conductor’’ is cast steel.
Sec. 1-6 Magnetic Circuit Computations 25

Hence by assuming the entire mmf to appear across material c we can find H, from
which B follows, which in turn yields ®. Thus

Un? 2 4.3)
HA. = in Sh 700 7 350 A-t/m

From Fig. 1-10

B. = 0.65 T
®, = B.A, = 0.65(0.001) = 0.00065 Wb
This value represents the first approximation for the flux as indicated by the sub-
script.
Also, since the cross-sectional area is the same for each material, it follows
that
B, = By = B. = 0.65 T
Reference to the nickel-iron magnetization curve reveals that the value of H, corre-
sponding to B, is negligibly small compared with H,. Hence for all practical pur-
poses its effect can be neglected. However, note that for medium silicon steel the
value of H;, is almost 81 A-t/m. This, coupled with the fact that /, = 2l., indicates
that material b takes about half as much mmf as material c in maintaining the flow of
flux. In other words, at this point in our analysis we can make a refinement on our
original assumption of assigning the entire mmf to material c. Now we see that about
50% of that assigned to c should be assigned to b. Thus
U.+ U, = & (assumed)

but
U, = 0.5U, (assumed)

Hence
1.5U, = # = 35
U, = 23.3 A-t

Accordingly a second approximation for the solution can be obtained. There-

fore

H, = oe = 233 A-t/m

which in turn yields

B, = 0.4T

so that the value of the flux now becomes

®, = B.A, = 0.0004 Wb
at this point
To determine whether or not this is the correct answer we must
whether they yield a value
compute the mmf drops for each material and add to see
 Magnetic Theory and Circuits Chap. 1
26

until
equal to the applied mmf. If not, the foregoing procedure must be repeated
circuital law is satisfied. Making this check for the second approxima tion
Ampere’s
we have
H, = 62 A-t/m corresponding to B, = 0.4 T

and
H, = 5.7 A-t/m for B, = 0.4 T

Accordingly |
mmf = Ayla F Ayly re H, |,

Il 5.70.3) + 62(0.2) + 233(0.1)

1:7 3 12,4. 423,32 = 37.4 At

Obviously this is too high by about 7%. Hence, as a third try, reduce the
biggest contributor to the mmf by a factor of 5%. That is, assume that now
U. = 22 At
Then
H, = 220 A-t/m and B, = 0.375 T
®, = 0.000375 Wb

Corresponding to this flux we find

H,=59 and H,=5

Hence
mmf = 5(0.3) + 59(0.2) + 22 = 35.3 A-t
Since this summation of mmf’s agrees with the applied mmf of 35 A-t, the
correct solution for the flux is

® = 0.000375 Wb = 37,500 lines

When making magnetic circuit computations of the kind just illustrated, it is

common practice to accept as valid any solution that comes within +5% of the
exact solution. The reason is that we are dealing with normal magnetization
curves which neglect hysteresis and which are after all only typical of the material
actually being used in the specified circuit. Deviations can and often do exist.
As a final example to illustrate magnetic circuit computations, we shall solve
a problem involving parallel magnetic paths as well as the presence of an air gap.
Moreover, we shall consider only the first type where the mmf needed to establish
a specified flux is to be found. This is justified not only because the solution is
straightforward but also because it is far more representative of the kind of mag-
netic circuit problem the engineer is likely to be concerned with.
Sec. 1-6 Magnetic Circuit Computations 27

Bos ys
mo Tog
Figure 1-14 Magnetic circuit for
eRe 1-3. All dimensions are in
=. 0075 Lo. Bi Ace [ie

Example 1-3
A magnetic circuit having the configuration and dimensions shown in Fig. 1-14 is
made of cast steel having a thickness of 0.05 m and an air gap of 0.002 m length
appearing between points g and h. The problem is to find the mmf to be produced by
the coil in order to establish an air-gap flux of 4 x 10-4 Wb (or 40,000 lines).
The method of solution can be readily ascertained by referring to the equiva-
lent circuit of this magnetic circuit as shown in Fig. 1-15. Knowledge of ®, enables
us to find the mmf drop appearing across b and c. From this information the flux in
leg bc can be determined and, upon adding it to ®,, we find the flux in leg cab. In
turn, the mmf needed to maintain the total flux in leg cab can be computed, and when
we add it to the mmf drop across bc we obtain the resultant mmf.
The computations involved for the various parts of the magnetic circuit are as
follows:
Part gh. This is the air gap for which the flux is specified as 4 x 10-4 Wb. The
cross-sectional area of the gap is 0.05(0.05) = 0.0025 m2. Normally, however, this
the
area is slightly higher because of the tendency of the flux to bulge outward along
edges of the air gap—which is often referred to as fringing. For convenience this
effect is neglected. Thus, the air-gap flux density is found to be
®, 4x 104
Be= 4 = “pos ~ 167
have
Since the permeability of air is practically the same as that of free space, we
B, 0.16
(Qe ts = an x 107 = 127,300 A-t/m

“
Reod b dy

(F) Rotke
Figure 1-15 Equivalent circuit of
Cc Fig. 1-14.
 Magnetic Theory and Circuits Chap. 1

and
Ue Hele = 127,300(0.002) = 255 A-t

Part bg and hc. The length of ferromagnetic material involved here is

l
big ar bie = 2 (isaae (he Ge lat 2) = i

= 2(0.025 + 0.1 + 0.025 + 0.1) — 0.002

= 0.498 m
the field
Moreover, corresponding to a flux density in the cast steel of 0.16 T,
intensity H is found to be 125 A-t/m. Hence
Ube+he = 125(0.498) = 62.2 A-t

Part bc. Because path bfkc is in parallel with path bc, the total mmf across
path bfkc also appears across path be. Hence
peo 205 022 = 3112 At

Also
Ine = 0.1 + 0.075 = 0.175 m

wg ie
= 1810 A-t/m
iol a)Mi)
And from Fig. 1-10 for cast steel the corresponding flux density is found to be
By. = 1.38 T

Hence
®,. = 1.38(0.0025) = 0.00345 Wb

Part cab. Accordingly, the total flux existing in leg cab is

Dab = Doc + Py = 0.00345 + 0.0004 = 0.00385 Wh

Knowledge of this flux then leads to determination of the mmf needed in leg cab to
sustain it. Thus

Bag = 2385 =. on6 7

From which
Heap = 690 A-t/m

Hence
U-ab = Aviles = 690(0.5) = 345 A-t

Therefore the total mmf required, to produce the desired air-gap flux is
F = Ucgn + Use = 345 + 317.2 = 662.2 A-t
 Sec. 1-7 The Inductance Parameter 29

1-7 THE INDUCTANCE PARAMETER

Inductance is a characteristic of magnetic fields, and it was first discovered by

Faraday in his renowned experiments of 1831. Ina general way inductance can be
characterized as that property of a circuit element by which energy is capable of
being stored in a magnetic flux field. A significant and distinguishing feature of
inductance, however, is that it makes itself felt in a circuit only when there is a
changing current or flux. Thus, although a circuit element may have inductance
by virtue of its geometrical and magnetic properties, its presence in the circuit is
not exhibited unless there is a time rate of change of current. This aspect of
inductance is particularly stressed when we consider it from the circuit view-
point. However, for the sake of completeness, inductance is also treated from an
energy and a physical view.

Circuit Viewpoint

The current-voltage relationship involving the inductance parameter is expressed

by
di
uv, = L (1-26)
dt
In general both v, and i are functions of time. Figure 1-16 depicts the potential
difference v; appearing across the terminals of the inductance parameter when a
changing current flows into terminal c. Note that the arrowhead on the v,; quan-
tity is shown at terminal c, indicating that this terminal is instantaneously positive
with respect to terminal d. In turn, this means that current is increasing in the
positive sense, i.e., the slope di/dt in Eq. (1-26) is positive. Any circuit element
that exhibits the property of inductance is called an inductor and is denoted by the
symbolism shown in Fig. 1-16. In the ideal sense the inductor is considered to be
resistanceless, although practically it must contain the wire resistance out of
which the inductor coil is formed.
It follows from Eg. (1-26) that an appropriate defining equation for induc-
tance is

(1-27)

Figure 1-16 Circuit with pure induc-

tance parameter.
 Magnetic Theory and Circuits Chap. 1
30

given time instant across the

Thus by recording the potential difference at a
ponding derivative of the cur-
terminals of an inductor and dividing by the corres
ter. Note that the units of
rent time function we determine the inductance parame
city this is commonly called
inductance are volt-seconds per ampere. For simpli
the henry.
ter is independent
A linear inductor is one for which the inductance parame
a space flux. When
of current. As current flows through an inductor it creates
and flux prevails so
this flux permeates air, strict proportionality between current
of current . A plot of
that the inductance parameter stays constant for all values
derivat ive of the current
the potential difference across the coil as a function of the
(1-27). When the flux is
then appears as shown in Fig. 1-17, which is a plot of Eq.
current s to upset the
made to penetrate iron, however, it is possible for large
it produc es. In sucha
proportional relationship between the current and the flux
will then no longer be
case the inductor is called nonlinear and a plot of Eq. (1-27)
a straight line.
express either
For the resistance parameter Ohm’s law can be written to
The same procedure
voltage in terms of current or current in terms of voltage.
already expresses
may be followed for the inductance parameter. Equation (1-26)
the current in terms
the voltage as a function of the current. However, to express
be transpo sed to
of the potential difference across the inductor, Eq. (1-26) must
read as follows:
1 (1-28)
di = aL dt

In integral form this becomes

[e ai => [iv at
HO As 1 fe
(1-29)
or

he ah wee eG) (1-30)

di
=) amp/sec :
at) Figure 1-17 Graphical representation
of the inductance parameter L, from
the circuit viewpoint.
 Sec. 1-7 The Inductance Parameter 31

Equation (1-30) thus reveals that the current in an inductor is dependent upon the
integral of the voltage across its terminals as well as the initial current in the coil at
the start of integration.
An examination of Eqs. (1-26) and (1-30) reveals an important property of
inductance: The current in an inductor cannot change abruptly in zero time. This
is made apparent from Eq. (1-26) by noting that a finite change in current in zero
time calls for an infinite voltage to appear across the inductor, which is physically
impossible. On the other hand, Eq. (1-30) shows that in zero time the contribution
to the inductor current from the integral term is zero so that the current immedi-
ately before and after application of voltage to the inductor is the same. In this
sense, then, we may look upon inductance as exhibiting the property of inertia.
Example 1-4
An inductor has a current passing through it which varies in time in the manner
depicted in Fig. 1-18(a). Find the corresponding time variation of the voltage drop
appearing across the inductor terminals if it is assumed that the inductance of the coil
is 0.1 H.
The solution appears in Fig. 1-18(b). Note that, in the interval from 0 to 0.1 s,
di/dt = 100 A/s. Hence the voltage across the coil is then a constant given by
d
wy=La = = 0.11100) = 10V for0<1<0.1

In the time range from 0.1 to 0.3 s the slope of the current curve is —50. Hence the
voltage across the coil is —-5 V. Finally, in the interval from 0.3 to 0.6 s the current
wave shape is sinusoidal so that the corresponding voltage wave is a cosine. In
drawing the cosine wave it is assumed that the maximum slope of the sine wave
exceeds 100 A/s.

i,amp

Time, sec

Figure 1-18 (a) Input current

waveshape to an inductor; (b) cor-
Time, sec
responding voltage variation across the
(b) inductor terminal.
 Magnetic Theory and Circuits Chap. 1
32

in an inductor, the
It is interesting to note in Fig. 1-18 that, unlike the current
r is allowed to change discont inuousl y.
voltage across the inducto

Energy Viewpoint

a current i is made to
Assume that an inductor has zero initial current. Then, if
nce vz, the total
flow through the coil across which appears the potential differe
energy received in the time interval from 0 to tis

w= vidt joules (J) (1-31)

Inserting Eq. (1-26) leads to

w= | (z4 jdt = |Lidi (1-32)

or

W = 3Li? J (1-33)
Continuing with the assumption that the inductor has no winding resistance, Eq.
(1-33) states that the inductor absorbs an amount of energy which is proportional
to the inductance parameter L as well as the square of the instantaneous value of
the current. Thus energy is stored by the inductor in a magnetic field. It is of
finite value and retrievable. As the current is increased, so too is the stored
is
magnetic energy. Note, however, that the energy is zero whenever the current
zero. Because the energy associated with the inductanc e parameter increases and
decreases with the current, we can properly conclude that the inductor has the
property of being capable of returning energy to the source from which it re-
ceives it.
A glance at Eq. (1-33) indicates that an alternative way of identifying the
inductance parameter is in terms of the amount of energy stored in its magnetic
field corresponding to its instantaneous current. Thus, in mathematical form we
can write

H (1-34)

This is an energy description of the inductance parameter.

There is one final point worthy of note. It has already been demonstrated
that for a potential difference to exist across the inductor terminals the current
must be changing. A constant current results in a zero voltage drop across the
ideal inductor. This is not true, however, about the energy absorbed and stored in
the magnetic field of the inductor. Equation (1-33) readily verifies this fact. A
constant current results in a fixed energy storage. Any attempt to alter this energy
Sec. 1-7 The Inductance Parameter 33

state is firmly resisted by the effects of the initial energy storage. This again
reflects the inertial aspect of inductance.

Physical Viewpoint

The voltage across the terminals of an inductor may be expressed from a circuit
viewpoint by Eq. (1-26). However, this same voltage may be described by Fara-
day’s law in terms of the flux produced by the current and the number of turns N
of the inductor coil. Accordingly, we may write
ioe nll races AD
of = L yer Taye (1-35)

It then follows that

_
LN do
i (1-36)

In those cases, where the flux ¢ is directly proportional to current i for all values
(i.e., for linear inductors), the last expression becomes
_N® Wb-t
ie or H (1-37)
i A
Here the inductance parameter has a hybrid representation because it is in part
expressed in terms of the circuit variable i and in part in terms of the field variable
®. To avoid this we replace flux by its equivalent, i.e.,
mmf moi
(1-38)
~ magnetic reluctance
where mmf is the magnetomotive force which produces the flux @ in the magnetic
circuit having a reluctance R. Appearing in Fig. 1-19 is an inductor consisting of
N turns wound about a circular iron core. If the core is assumed to have a mean
length of / meters and a cross-sectional area of A» meters’, then the magnetic
reluctance can be written as

Di l (1-39)

where y is the permeability, a physical property of the magnetic material.

Upon substituting Eqs. (1-38) and (1-39) in (1-37) there results the expression
for the inductance parameter of the circuit of Fig. 1-19. Thus

_ N?*Am _ N?* (1-40)

2 i R
A study of Eq. (1-40) reveals some interesting and useful facts about the
inductance parameter which are not readily available when this quantity is defined
 Magnetic Theory and Circuits Chap. 1
34

Figure 1-19 Linear inductor with iron

core.

is the fact that

either from the circuit or the energy viewpoint. Most impressive
the geometr y of physical dimen-
inductance, like resistance, is dependent upon
This is significa nt because it tells
sions and the magnetic property of the medium.
Thus, for the inductor illustrat ed in
us what can be done to change the value of L.
d in any one of four ways:
Fig. 1-19, the inductance parameter may be increase
reducing
increasing the number of turns, using an iron core of higher permeability,
ctional area of the iron core.
the length of the iron core, and increasing the cross-se
It is interesting to note that neither the circuit viewpoint nor the energy
viewpoint could tell us these things, because essentially they deal with the effects
associated with a given inductor geometry. It is emphasized that all three view-
points are needed to complete the picture of this circuit element.

1-8 HYSTERESIS AND EDDY-CURRENT LOSSES IN

FERROMAGNETIC MATERIALS

The process of magnetization and demagnetization of a ferromagnetic material in

a symmetrically cyclic condition involves a storage and release of energy which is
not completely reversible. As the material is magnetized during each half-cycle, it
is found that the amount of energy stored in the magnetic field exceeds that which
is released upon demagnetization. The background for understanding this behav-
ior was provided in Sec. 1-3. There the hysteresis loop was identified as the

Area abda represents

energy stored in magnetic b
field during positive = PeReVees>
half cycle of H

Area bde represents energy

released by magnetic field
during positive half cycle

Area abca represents

hysteresis loss per half
cycle

Figure 1-20 Hysteresis loop and en-

ergy relationships per half-cycle.
Sec. 1-8 Hysteresis and Eddy-Current Losses in Ferromagnetic Materials 35

variation of flux density as a function of the magnetic field intensity for a ferro-
magnetic material in a cyclic condition. The salient feature of the hysteresis loop
is the delayed reorientation of the domains in response to a cyclically varying
magnetizing force. A single hysteresis loop is depicted in Fig. 1-20. The direction
of the arrows on this curve indicates the manner in which B changes as H varies
from zero to a positive maximum through zero to a negative maximum and back to
zero again, thus completing the loop. To appreciate the meaning of the various
shaded areas shown in Fig. 1-20, let us look at the units associated with the
product of B and H. Thus
Bet ey Te amperes newtons __ newtons _ N
meter ampere-meter meter? m2

But
newton-meter = joule

Hence
joules J"
units of HBo= os ap

which is clearly recognized as an energy density. Therefore, in dealing with areas

involving B and H in connection with a hysteresis loop we are really dealing with
energy densities expressed on a per-cycle basis because the hysteresis loop is
repeatable for each cyclic variation of H.
The foregoing is more precisely described by dealing with the energy as the
integral of power with respect to time. Thus,

Wenergy in J) = i p at

But,

tt
p=el pag n ea(2)
Wiki

so that
~ bb > dd rs dp Fi
w= [ni (2) ar = : Ni db (1-41)

Also,

aeae
ape ae
ba a nd @=BA
= for Afixed f

Inserting these last expressions into Eq. (1-41) yields

We in N (=) A dB = Al {, H dB
By Hl Bp
(1-42)
 Magnetic Theory and Circuits Chap. 1
36

Hence the energy density is

WwW Bp
ye = —_>- =
es H dB (1-43)
-4

The energy stored in the magnetic field during that portion of the cyclic
variation of H when it increases from zero to its positive maximum value (assum-
ing the material is already in a cyclic state) is given specifically by

w=] HaB Sim (1-44)

when mks units are used; i.e., H must be expressed in A-t/m and B in webers per
square meter. Note that the axes of Fig. 1-20 are so labeled. Moreover, during
that portion of its cyclic variation when H decreases from its positive maximum
value to zero (as it follows along curve bc of the hysteresis loop), energy is being
released by the magnetic field and returned to the source, and this quantity can be
represented as

he irHdB Mm (1-45)
In this equation, since B, > B,, the quantity w2 will be negative, indicating that the
energy is being released rather than stored by the magnetic field.
A graphical interpretation of Eq. (1-44) leads to the result that the energy
absorbed by the field when A is increasing in the positive direction can be repre-
sented by the area abdca. Similarly the energy released by the field as H varies
from Hynax to zero can be represented by area bdcb. The difference between these
two energy densities represents the amount of energy which is not returned to the
source but, rather, is dissipated as heat as the domains are realigned in response to
the changing magnetic field intensity. This dissipation of energy is called hystere-
sis loss. Keep in mind that Fig. 1-20 depicts this energy density loss for a half-
cycle variation of H. Hence area abca represents the hysteresis loss per half-
cycle. It certainly follows from symmetry that upon completion of the negative
half-cycle variation of H an equal energy loss occurs. Therefore, as H varies over
one complete cycle, the total energy loss per cubic meter is represented by the
area of the hysteresis loop. More specifically this energy loss per cycle can be
expressed mathematically as

Wh = (area of hysteresis loop) J/m?/cycle (1-46)

where mks units are used for H and B.

It is frequently desirable to express the hysteresis loss of ferromagnetic
materials in watts—the unit of power. A little thought about the units of w, in Eq.
(1-46) shows how this can be directly accomplished. Thus
Sec. 1-8 Hysteresis and Eddy-Current Losses in Ferromagnetic Materials 37

2 energy power X seconds power

Ww, = FF ”>”TCOCECL'-ESX)
vol x cycles vol X cycles vol. X cycles/second
Now let P, = power loss, watts (W)
v = volume of ferromagnetic material
f = hertz (Hz) = frequency of variation of 1

Then Eq. (1-47) becomes

Wp, h = Pr
uf 1-48
( “3 )

or

P;, = wpvf (1-49)

where w,—the energy density loss—is determined from Eq. (1-46).

To obviate the need of finding the area of the hysteresis loop in order to
compute the hysteresis loss in watts from Eq. (1-49), Steinmetz obtained an em-
pirical formula for w, based on a large number of measurements for various
ferromagnetic materials. He expressed the hysteresis power loss as

where B,,, is the maximum value of the flux density and n lies in the range 1.5 =
n < 2.5 depending upon the material used. The parameter K;, also depends upon
the material. Some typical values are: cast steel 0.025, silicon sheet steel 0.001,
and permalloy 0.0001.
In addition to the hysteresis power loss, another important loss occurs in
ferromagnetic materials that are subjected to time-varying magnetic fluxes—the
eddy-current loss. This term is used to describe the power loss associated with
the circulating currents that are found to exist in closed paths within the body of a
ferromagnetic material and cause an undesirable heat loss. These circulating cur-
rents are created by the differences in magnetic potential existing throughout the
body of the material owing to the action of the changing flux. If the magnetic
circuit is composed of solid iron, the ensuing power loss is appreciable because
the circulating currents encounter relatively little resistance. To increase signifi-
is
cantly the resistance encountered by these eddy currents the magnetic circuit
invariably composed of very thin laminations (usually 14 to 25 mils thick) when-
in normal
ever the electromagnetic device is such that a varying flux permeates it
all ac electric motors and gener-
operation. This is the case with transformers and
ators. An empirical equation for the eddy-current loss is

Pr=Kef Bary W (1-51)

 Magnetic Theory and Circuits Chap. 1 |
38

where K, = a constant dependent upon the material

f ll= frequency of variation of flux, Hz
B,, = maximum flux density
7 = lamination thickness
vy = total volume of the material
rrent losses
A comparison of this equation with Eq. (1-50) reveals that eddy-cu
the hysteres is loss varies directly
vary as the square of the frequency, whereas
with the frequency.
te what is
Taken together the hysteresis and eddy-current losses constitu
magneti c devices that involve time-
frequently called the core losses of electro
just passing attentio n is devoted to
varying fluxes for their operation. More than
losses have an importa nt bearing
these losses here because, as will be seen, core
on temperature rise, efficiency, and rating of electromagnetic devices.

1-9 RELAYS—AN APPLICATION OF MAGNETIC FORCE

A relay is an electromagnetic device which can often be activated by relatively
little energy, causing a movable ferromagnetic armature to open or close one or
several pairs of electrical contact points located in another control circuit or a
main circuit handling large amounts of energy. Alternating- and direct-current
motor starters are equipped with relays designed to ensure proper operation of
motors during starting and running conditions. These devices are found in many
applications in all fields of engineering, especially in situations where control of a
process or machine is involved. Our objective in this section is to describe the
principles that underlie the operation of these electromagnetic devices. Besides
the knowledge it offers, this treatment gives the motivation for studying the theory
of magnetic fields and circuits.
The derivation of a magnetic force equation is our primary interest because it
shows how it is possible to do mechanical work—moving a relay armature—by
abstracting energy from that stored in the magnetic field. Moreover, since the
emphasis is on the principles involved, the simplifying assumptions of no satura-
tion and no losses are imposed; i.e., linear analysis is used throughout. Accord-
ingly, the magnetization curve of the ferromagnetic material is assumed to be a
straight line as depicted in Fig. 1-21(a). (Only the curve corresponding to positive
values of H is shown.) Now, by appropriately modifying the axes of the curve
plotted in Fig. 1-21(a), a significant and useful thing happens. First recall that the
area between the B axis and the magnetization curve represents the energy ab-
sorbed from the source and stored in the magnetic field on a per-unit-volume
basis. Equation (1-44) states this result mathematically.

w= | Hd Mm (1-52)
In the simplified situation of Fig. 1-21(a), Eq. (1-52) becomes merely the area of
Sec. 1-9 Relays—An Application of Magnetic Force 39

Flux, Wb
Closed

HH (AT/m)
(a) (b)
gives the energy
Figure 1-21 Linear magnetization curve of a relay. (a) Area OaB
stored magnetic
density corresponding to fixed H. (b) Area OaB gives the total
field energy correspon ding to fixed #.

triangle OaB. Thus

wr = 2BH J/m? (1-53)

where H is assumed fixed at the value shown.

energy stored in the
Moreover, since wy is an energy density, the total
(1-44) by the volume. Thus
magnetic field is found by multiplying Eq.
Wr=wev=wAl J (1-54)
of the magnetic circuit.
where / is the length and A the cross-sectional area
Inserting Eq. (1-53) into Eq. (1-54) yields

W, = 3(BA)HI = 20% J (1-55)

to identify the flux ® and H is

Note that in this equation A is combined with B
cal representation of Eq. (1-55)
combined with / to identify the mmf *. A graphi
Fig. 1-21(b) derives from Fig.
appears in Fig. 1-21(b). It should be apparent that
and the abscissa axis by /, thus
1-21(a) by multiplying the ordinate axis by A
area OaB of Fig. 1-21(a) gives the
plotting ® versus #. Then for a fixed H (or #)
Oad of Fig. 1-21(b) gives the total
energy density, whereas the corresponding area
energy stored in the magnetic field.
done by the abstraction of
To understand how mechanical work can be
the circuitry appearing in Fig. 1-22,
energy stored in the magnetic field, consider
electromagnetic relay. It consists of an
which depicts the basic composition of an ele-
ferromagnetic core equipped with a movable
exciting coil placed on a fixed
 Magnetic Theory and Circuits Chap. 1
40

Relay
armature

Figure 1-22 Basic

;
composition of an
Confinin
q electromagnetic relay.
surface

ment called the relay armature. The relay is energized from a constant voltage
source through an adjustable resistor R. To begin with, consider that R is fixed at
that value which makes the coil mmf equal to ¥ and produces the flux ® as shown
in Fig. 1-21(b) and that initially the relay armature is prevented from movement by
being held fast. If the relay armature is now allowed to move to a smaller air gap
position, the magnetization curve, Oa, moves to a steeper position, Ob, consistent
with the lower reluctance of the magnetic circuit. It is important to understand
that in the arrangement of Fig. 1-22, operation of the relay from the open to the
closed position occurs under conditions of constant mmf because for a constant-
source voltage the coil current is determined by the circuit resistance and not by
the reluctance of the magnetic circuit. The lower reluctance of the magnetic
circuit in the closed’armature position simply means that the given mmf produces
more flux. The increase in flux is indicated in Fig. 1-21(b) by the quantity d¢.
The area of rectangle abcd in Fig. 1-21(b) represents an increase of energy
(dW;) to the relay as it goes from the open to the closed state. Quantitatively, it
can be described by
dW; = & do (1-56)

Energy that is applied to the relay through the input circuit can be distributed in
two forms: as stored magnetic field energy and as work done by the relay in
moving from an open to a closed state. Of the total change in relay input energy,
how much is expended to do the mechanical work in moving from the open to the
closed position and how much is used to increase the initial field energy to a new
total value? The latter quantity is readily determined by the area of triangle Obc
in Fig. 1-21(b). Thus

(Wy )retay closed a 2F(p uF do = 3Fh cs 3F do (1-57)

Since the first term on the right side is the field energy corresponding to the open
position, it follows that half of the energy represented by Eq. (1-56) goes to
augment the stored field energy. Of course, the remaining half serves to do me-
chanical work which can be represented in differential form by (F dx) where x
denotes distance. It is instructive to note that this mechanical work is represented
in Fig. 1-21(b) by triangle Oab. An expression for the relay force can be readily
derived by equating these two terms. Thus,
Sec. 1-9 Relays—An Application of Magnetic Force 41

Fe 1
F dx =43# do =3#d (=) = 4% d (=) (1-58)

For convenience introduce

1
= HR ~ Permeance of the magnetic circuit (1-59)

Then,
F-dx = 3F* dP?
Or,

N (1-60)

Equation (1-60) states that the force on the relay armature always acts to increase
the permeance of the magnetic circuit, which is equivalent to saying that F acts to
reduce the reluctance.
When the relay is energized from an ac source, the flux must remain essen-
tially constant irrespective of the reluctance of the magnetic circuit (see Chap. 3).
In such a situation the current automatically adjusts to meet the magnetomotive
needs of the magnetic circuit. As the relay armature moves in circumstances that
require operation at constant flux, the B-H curve for the magnetic circuit at the
open and closed relay positions can now be depicted by Fig. 1-23. Here too the
energy to move the relay armature from the open to the closed position is indi-
cated by the area of triangle Oab in Fig. 1-23. Expressed mathematically we have
mechanical work on relay armature = reduced magnetic field energy
= area of triangle Oab

Flux, Wb Closea

Figure 1-23 Flux and mmf conditions

at open and closed positions of the
relay armature when the relay is ex-
cited from an ac source that demands
operation at constant flux.
 Magnetic Theory and Circuits Chap. 1
42

Or
F dg = 36(F, — Fa) (1-61)
where
= Ra and F, = oR, (1-62)

and where 9%, is the reluctance of the magnetic circuit with the relay in the open
position and %, denotes the reluctance in the closed position. Moreover, Ra >
R,. Inserting Eq. (1-62) into Eq. (1-61) yields

F dg = 36(bRy — bRa) = 2b6°(Rs — Ra) (1-63)

We can now introduce
Aa Rf he) (1-64)

as the change in reluctance from the open to the closed position. The negative
sign is in recognition of the fact that a decrease in reluctance occurs. Substituting
Eq. (1-64) into Eq. (1-63) yields finally

fees yom N (1-65)

Equation (1-65) states that the action of the stored magnetic field energy is to
reduce the reluctance of the magnetic circuit, a result which is consistent with the
conclusion in Eq. (1-60).

Example 1-5
In the relay circuit of Fig. 1-22 assume the cross-sectional area of the fixed core and
the relay armature to be A and the circuit mmf to be #. Neglecting the reluctance of
the iron, find the expression for the magnetic force existing on the relay armature.

Solution The force is easily found by invoking Eq. (1-60). Thus,

where

Sak gape
Ri 8
Hence,

Next introduce
Sec. 1-9 Relays—An Application of Magnetic Force 43

since movement of the relay armature reduces the gap length. Accordingly, the
force on the relay becomes

An alternative expression for this force can be found by rewriting the last
equation as

F = 397 (#2A\ er
1 aritu
eee R21
PARSEEE Ge 1
ji Ce ee ee io, Gage |Pg
rte eee o? N
< moAg — MoA
Example 1-6
The cross-sectional view of a cylindrical plunger magnet appears in Fig. 1-24. The
plunger (or armature) is free to move inside a nonferromagnetic guide around which
the coil is surrounded by a cylindrically shaped steel shell. The plunger is separated
from the shell by an air gap of length g.
(a) Derive the expression for the magnetic force exerted on the plunger when it
is in the position shown in Fig. 1-24. The length of the plunger is at least equal to that
of the shell. Also it has a radius of a meters. Neglect the reluctance of the steel.
(b) Find the magnitude of the force when the mmf is 1414 A-t and the plunger
magnet dimensions are
x = 0.025 m = 0.05 m

a = 0.025 m g = 0.00125 m
that
Solution Before proceeding with the calculations let us review the principle
assumed, a typical flux path
gives rise to the force. For the direction of coil current
It is
is that shown in Fig. 1-24. Note that the flux path must cross the air gap twice.
to note too that the reluctance ‘‘seen’’ by the flux as it crosses
particularly important
at the upper
the bottom gap is less than that seen by the same flux crossing the gap
the magnetic circuit is smaller
part of the plunger because the cross-sectional area of
part of the plunger. As a matter of fact, at the
at the upper part than it is at the lower
the cross-sect ional area seen by the flux is fixed for all
lower part of the plunger

Cylindrical Figure 1-24 Plunger magnet for Exam-

plunger ple 1-6.
 Magnetic Theory and Circuits Chap. 1

positions of the plunger lying in the range 0 = x = hand specifically is equal to 277ah
(assuming g is small compared to a). Accordingly, a force is created on the plunger
so directed as to decrease the reluctance. This means that the magnetic force acts to
move the plunger upward.
(a) The solution of this part is obtained from Eq. (1-65), but first we need to
identify the correct expression for the reluctance as seen by the flux. The reluctance
of the steel is negligible. Hence the reluctance associated with the typical flux shown
in Fig. 1-24 is merely the sum of the reluctances associated with each air gap. Thus

cee 2 g (= + *) (1-66)
fo27rah x bo27ax ‘s 27apo \h x

Then

ak __ 8 (- +)
ax 2qapo\ x? G62)
Also
F F xh
= = Se (1-68)
RR glugdmah+ x
Inserting Eqs. (1-67) and (1-68) into Eq. (1-65) yields

ra ra 2 (L Fh? 1 2
) = 3.94 x te
2
Fe? (.: 7
2
N_

(b) Inserting the specified dimensions into Eq. (1-69) gives the magnetic force
(1-69)

as

F = 70N = 15.74 lb

PROBLEMS

1-1. A long straight wire located in air carries a current of 4 A. Assume the relative
permeability of air is unity.
(a) Find the value of the magnetic field intensity at a distance of 0.5 m from the
center of the wire.
(b) A second long straight wire carrying a current of 2 A is placed parallel to the first
one at a distance of 0.5 m with the current flowing in the same direction. Find
the direction and magnitude of the force per meter existing between the wires.
(c) Repeat part (b) for wires that are embedded in iron having a relative permeability
of 10,000 and a spacing of 0.05 m.
1-2. The wires shown in Fig. P1-2 are long, straight, and parallel and are completely
embedded in iron having a relative permeability of 1000. Each wire carries a current
of 10 A.
(a) Compute the magnitude and direction of the resultant force per meter on the wire
in which J, flows.
Chap. 1 Problems 45

(b) Compute the magnitude and direction of the force per meter on the wire in which
I, flows.
(c) Repeat part (a) for the third wire.

I, Ip I;

Wires imbedded
in iron having
pH = 1000
O.Im O.lm
Figure P1-2

1-3. Repeat Prob. 1-2 for current J, flowing opposite to J; and J;.
1-4. A uniform magnetic field of 0.7 Wb/m* in the iron is applied to the configuration of
Fig. P1-2.
(a) Compute the direction and magnitude of the resultant force per meter when the
magnetic field is directed perpendicularly into the plane of the paper.
(b) What is the value of this resultant force per meter when the magnetic field is
applied in the plane of the wires and directed from right to left?
(c) Compute the resultant force per meter when the magnetic field is applied at an
angle of 45° to the plane of the paper and directed into it from right to left.
(d) What is the resultant force per meter when the magnetic field is applied at an
angle of 60° to the plane of the paper and directed into it from top to bottom?
1-5. In the configuration of Fig. P1-5, the current J, has a value of 40 A. Find the value of
I, that causes the magnetic field intensity at point P to disappear.

iy Ig

10cm
20cm
Figure P1-5

1-6. A circular loop of wire of radius r meters and consisting of a single turn carries the
current J as shown in Fig. P1-6. Derive the expression for the magnetic field inten-
sity at the center.

Figure P1-6

1-7. A magnetic circuit composed of silicon sheet steel has the square construction shown
in Fig. P1-7.
 Magnetic Theory and Circuits Chap. 1
46

Wb.
(a) Find the mmf required to produce a core flux of 25 x 10-4
how much current must be made to flow through the coil?
(b) If the coil has 80 turns,

10 cm

l<——40 cm—>|
Figure P1-7

1-8. The magnetic circuit of Prob. 1-7 has an air gap of 0.1 cm cut in the right leg. Fora
coil having 100 turns find the current that must be allowed to flow in order for the
core flux to be 0.0025 Wb.
1-9. In the magnetic circuit of Fig. P1-9 determine the coil mmf needed to produce a flux
of 0.0014 Wb in the right leg. The thickness of the magnetic circuit is 0.04 m and is
uniform throughout. Medium silicon steel is used.

}<———0,2 m 0.2m——> Figure P1-9

1-10. Repeat Prob. 1-9 for a coil placed on the center leg.
1-11. In the magnetic circuit of Prob. 1-7 find the core flux produced by a coil current of
200 A-t.
1-12. In the magnetic circuit of Prob. 1-8 find the core flux produced by a coil current of
600 A-t.
1-13. The core shown in Fig. P1-13 has a uniform cross-sectional area of 2 in.? and a mean
length of 12 in. Also, coil A has 200 turns and carries 0.5 A, coil B has 400 turns and
carries 0.75 A, and coil C carries 1.00 A. How many turns must coil C have for the
core flux to be 120,000 lines? The coil currents have the directions indicated in the
figure. The core is made of silicon sheet steel.

Figure P1-13
Chap. 1 Problems 47

1-14. In the magnetic circuit shown in Fig. P1-14 the coil F; is supplied with 350 A-t in the
direction indicated. Find the direction and magnitude of the mmf required in coil F,
for the air-gap flux to be 180,000 lines. The core has an effective cross-sectional area
of 9 in.? and is made of silicon sheet steel. The length of the air gap is 0.05 in.
9 Fp

ks in: Tee: aed nele—75 peaks: Be Figure P1-14

1-15. In the magnetic circuit of Fig. P1-14 the coil F; is supplied with 200 A-t in the
direction shown. Find the direction and magnitude of the mmf required of coil F> for
the air-gap flux to be 90,000 lines.
1-16. Find the inductance of a coil in which
(a) A current of 0.1 A yields an energy storage of 0.05 J.
(b) A current increases linearly from zero to 0.1 A in 0.2 s producing a voltage
of 5 V.
(c) A current of 0.1 A increasing at the rate of 0.5 A/s represents a power flow
of 3 W.
1-17. The coil in the configuration of Fig. 1-19 is equipped with 100 turns. Moreover, the
mean length of turn of the magnetic core is known to be 0.2 m and the cross-sectional
area is 0.01 m2. The value of the permeability of the iron is 107°.
(a) Find the inductance of the coil.
(b) When a dc voltage is applied to the inductor it is found to draw a current of
0.1 A. How much energy is stored in the magnetic field?
1-18. When a dc voltage is applied to the coil of Prob. 1-17, the current is found to vary in
accordance with i = *(1 — e~>‘) expressed in amperes. Moreover, it is found that,
after the elapse of 0.2 s, the voltage induced in the coilis 0.16 V. Find the inductance
of the coil.
1-19. A sample of iron having a volume of 16.4 cm} is subjected to a magnetizing force
is
sinusoidally varying at a frequency of 400 Hz. The area of the hysteresis loop
found to be 64.5 cm? with flux density plotted in kilolines per square inch and
1 in. =
magnetizing force in ampere-turns per inch. The scale factors used are
5 kilolines/in.2 and 1 in. = 12 A-t/in. Find the hysteresis loss in watts.

1-20. A ring of ferromagnetic material has a rectangular cross section. The inner diameter
of 600
is 7.4 in., the outer diameter is 9 in., and the thickness is 0.8 in. There is a coil
turns wound on the ring. When the coil carries a current of 2.5 A, the flux produced
units:
in the ring is 1.2 x 10°? Wb. Find the following quantities expressed in mks
 Magnetic Theory and Circuits Chap. 1
48

(a) magnetomotive force;

(b) magnetic field intensity;
(c) flux density;
(d) reluctance;
(e) permeability;
(f)relative permeability.
10 A-t/in., and
1-21. In plotting a hysteresis loop the following scales are used: 1 cm =
loop for a certain material is found to be
1 cm = 20 kilolines/in.? The area of the
is loss in joules per cycle for the specime n tested if
6.2 cm.? Calculate the hysteres
the volume is 400 cm’.
cy of 500 Hz. The
1-22. The flux in a magnetic core is alternating sinusoidally at a frequen
loss then amounts to
maximum flux density is 50 kilolines/in.’. The eddy-current
cy is 750 Hz and the
14 W. Find the eddy-current loss in this core when the frequen
flux density is 40 kilolines/in’.
of magnetic sheet
1-23. The total core loss (hysteresis plus eddy current) for a specimen
constant and the
steel is found to be 1800 W at 60 Hz. If the flux density is kept
d 50%, the total core loss is found to be 3000 W.
frequency of the supply increase
hysteresi s and eddy-cur rent losses at both frequenc ies.
Compute the separate
in Fig. P1-24.
1-24. A flux ¢ penetrates the complete volume of the iron bars as shown
assumed separated by g meters, find the expressi on for the force
When the bars are
between the two parallel plane faces. Neglect the reluctanc e of the iron.
existing

ei
Figure P1-24

1-25. A magnetic flux ¢ penetrates a movable magnetic core which is vertically misaligned
relative to the north and south poles of an electromagnet as depicted in Fig. P1-25.
The depth dimension for the core and the electromagnet is b. Determine the expres-
sion for the force that acts to bring the core into vertical alignment. Express the
result in terms of the air-gap flux density and the physical dimensions. Neglect the
reluctance of the iron.

elle Figure P1-25

Chap. 1 Problems 49

1-26. The magnetization curve of a relay is shown in Fig. P1-26.

(a) Find the energy stored in the field with the relay in the open position.
(b) Assume that the relay armature moves rapidly under conditions of constant
flux. Compute the work done in going from the open to the closed position.
(c) Calculate the force in newtons exerted on the armature in part (b).
(d) Assume the relay armature moves slowly at constant mmf. Compute the work
done in going from the open to the closed position.
(e) Does the energy of part (d) come from the original stored field energy? Explain.

D-C
voltage Relay

Figure P1-26

1-27. In the magnetic circuit of Fig. P1-27 the cross-sectional area of the air gap is 0.0025
m? and its length is 0.05 cm. Moreover, the cast steel core has a mean length of 0.2 m
and it is equipped with 1000 turns. It is desirable to operate the circuit with an air-
gap density of 1 T.
(a) Determine the coil current needed to establish this gap density.
(b) Compute the energy stored in the air gap.
(c) Find the energy stored in the cast steel section of the circuit.
(d) What is the inductance in henrys of this magnetic circuit?

¢-
Figure P1-27
 Z

Transformers

echanical
An understanding of the transformer is essential to the study of electrom
echanica l energy conversi on involves in-
energy conversion. Although electrom
system and a mechanic al system,
terchange of energy between an electrical
or more
whereas the transformer involves the interchange of energy between two
electrical systems, nonetheless the coupling device in both cases is the magnetic
we find
field, and its behavior in each case is fundamentally the same. As a result
that many of the pertinent equations and conclusi ons of transfor mer theory have
equal applicability in the analysis of ac machiner y and some aspects of dc machin-
ery. For example, the equivalent circuit of single-phase and three-phase induc-
tion motors is found to be identical in form to that of the transformer. Other
examples can be cited, as a glance through the following chapters readily reveals.
In addition to serving as a worthwhile prelude to the study of electromechan-
ical energy conversion, an understanding of transformer theory is important in its
own right because of the many useful functions the transformer performs in prom-
inent areas of electrical engineering. In communication systems ranging in fre-
quency from audio to radio to video, transformers are found fulfilling widely
varied purposes. Input transformers (to connect a microphone output to the first
stage of an electronic amplifier), interstage transformers, and output transformers
are to be found in radio and television circuits. Transformers are also used in
communication circuits as impedance transformation devices which allow maxi-

50
 Sec. 2-1 Theory of Operation and Development of Phasor Diagrams 51

mum transfer of power from the input circuit to the coupled circuit. Telephone
lines and control circuits are two more areas in which the transformer is used
extensively. In electric power distribution systems the transformer makes it pos-
sible to convert electric power from a generated voltage of about 15 to 20 kV (as
determined by generator design limitations) to values of 380 to 750 kV, thus
permitting transmission over long distances to appropriate distribution points
(e.g., urban areas) at tremendous savings in the cost of copper as well as in power
losses in the transmission lines. Then at the distribution points the transformer is
the means by which these dangerously high voltages are reduced to a safe level
(208 to 120 V) for use in homes, offices, shops, and so on. In short, the trans-
former is a useful device to be found in many phases of electrical engineering and
therefore merits the attention given to it in this chapter.

2-1 THEORY OF OPERATION AND DEVELOPMENT

OF PHASOR DIAGRAMS

In the interest of clarity and motivation the theory of operation of the transformer
is developed in five steps, starting with the ideal iron-core reactor and ending with
the transformer phasor diagram under load. In the process the important trans-
former equations are developed along with a physical explanation of how the
transformer operates. Emphasis on complete understanding of the physical be-
havior of a device is stressed, because, once this is grasped, the theory as well as
the associated mathematical descriptions readily follow.
In its simplest form a transformer consists of two coils which are mutually
coupled. When the coupling is provided through a ferromagnetic ring (circular or
otherwise), the transformer is called an iron-core transformer. When there is no
ferromagnetic material but only air, the device is described as an air-core trans-
former. In this chapter attention is confined exclusively to the iron-core type. Of
the two coils the one that receives electric power is called the primary winding,
whereas the other, which can deliver electric power to a suitable load circuit, is
called the secondary winding.

Ideal Iron-Core Reactor

As the first step in the development of the theory of transformers refer to Fig. 2-1,
which depicts an ideal iron-core reactor energized from a sinusoidal voltage
source v,. The reactor is ideal because the coil resistance and the iron losses are
assumed to be zero and the magnetization curve is assumed to be linear.
Our interest at this point is to develop the phasor diagram that applies to the
circuitry of Fig. 2-1, because thereby much is revealed about the theory. As v,
increases in its sinusoidal variation, it causes a sinusoidal magnetizing current igt

+Lowercase letters are used to represent instantaneous values.

 Transformers Chap. 2
52

Figure 2-1 Ideal iron-core reactor:

zero winding resistance and no core
losses. Polarities are instantaneous.

s a sinusoidally
to flow through the N, turns of the coil, which in turn produce
the current is
varying flux in time phase with the current. In other words, when
maximum value, so
zero, the flux is zero and, when the current is at its positive
we have
too is the flux. Applying Kirchhoff’s voltage law to the circuit
vy, —e, = 9 (2-1)
the coil
where e, is the voltage drop associated with the flow of current through
and can be written in terms of ig as

Cab = Ef — eer (2-2)

Eq. (2-1)
where L is the inductance of the coil and i, = i,». Inserting Eq. (2-2) into
and rearranging yields
di (2-3)
eee L a er,

We also know that the magnetizing current is sinusoidal because the applied
voltage is sinusoidal and the magnetization curve is assumed linear. Conse-
quently we can write for the magnetizing current the expression
igo V2 gsiniot (2-4)
where J, is the rms value of the current.
Introducing Eq. (2-4) into Eq. (2-3) and performing the differentiation called
for yields

v, = e) = V2 (@L)Iy cos wt = V2 (@L)Ig sin (1 + z) (2-5)

The maximum value of this quantity occurs when the cos wt has the value of
unity. Then
Bie DIge a 2 (2-6)
where
E, = wLI, = rms value of the inductive reactance drop (2-7)
Sec. 2-1 Theory of Operation and Development of Phasor Diagrams 53

E,o=-E, 90° 60 Y=E,= Tgwl

= voltage rise _ =E, = voltage
Ig drop

(b) (a)

Figure 2-2 Phasor diagrams of the ideal iron-core reactor: (a) circuit viewpoint;
(b) field viewpoint; (c) combined diagram.

A comparison of Eqs. (2-4) and (2-5) shows that the magnetizing current lags
the voltage appearing across the coil by 90 electrical degrees. The phasor diagram
of Fig. 2-2(a) depicts this condition. Since all the quantities involved here are
sinusoidal, the phasors shown represent the rms value of the quantities.
Another approach that can be used in arriving at the phasor diagram for the
ideal iron-core reactor is worth considering because of its use later. The approach
just described can be referred to as a circuit viewpoint because it deals solely with
the electric circuit in which J, flows as well as the circuit parameter L. No direct
use is made of the magnetic flux @, which is a space (or field) quantity. In contrast
the second approach starts with the magnetic field as it appears in Faraday’s law
of induction. This is one of the fundamental laws upon which the science of
electrical engineering is based. Faraday’s law states that the emf induced in a coil
is proportional to the number of turns linking the flux as well as the time rate of
change of the linking flux. In applying Faraday’s law to the circuit of Fig. 2-1 the
expression may be written either as a voltage drop (a to b) or as a voltage rise (b to
a): By the former description we can write

Cap = C1 = + Aan
dd (2-8)

Expressed as a voltage rise, the induced emf equation becomes

em = —e) = N= (2-9)
In either case, the sign is attributable to Lenz’s law. The minus sign states that
the emf induced by a changing flux is always in the direction in which current
would have to flow to oppose the changing flux. On the other hand, the plus sign
in Eq. (2-8) denotes that the polarity of a is positive whenever the time rate of
change of flux is positive as determined by the direction of current flow.
Equation (2-9) then represents the starting point for the second approach in
analyzing the circuit of Fig. 2-1. Corresponding to the sinusoidal variation of the
magnetizing current as expressed by Eq. (2-4), there is also a sinusoidal variation
of flux which may be expressed as
bd = @,, sin wt (2-10)
 Transformers Chap. 2
54

Putting this expres-

where ©®,, denotes the maximum value of the magnetic flux.
leads to
sion into Eq. (2-9) and performing the required differentiation

Cbg = —C1 = —N, “ = —N,®,,@ COS wi = Ey max sin (ws=a z) (2-11)

where
Et max = Ni®no = V2 E (2-12)
and
E, = rms value of the induced emf
emf —
A comparison of Eqs. (2-10) and (2-11) makes it clear that the reaction
it by 90°. This situation is depicted in Fig.
e, lags the changing flux that produces
a voltage rise (or generate d emf) which has such
2-2(b). Basically, the emf —e, is
the
a direction that, if it were free to act, would cause a current to flow opposing
in terms of the action-r eaction law consider any
action of J,. To verify this solely
two adjacent points such as a’ and b’ on coil Nj in Fig. 2-1. If the flux is assumed
increasing in the direction shown, then the reaction in the coil must be such that
current would have to flow from point b’ to a’ to oppose the increasing flux.
Recall that by the right-hand rule any current assumed to be flowing from b' to a’
produces flux lines opposing the increasing flux. For this condition to prevail, it is
therefore necessary for point a’ to be positive with respect to b’ for the instant
being considered. This line of reasoning when extended to the full coil, leads to
the polarity markings shown in Fig. 2-1. The interesting thing to note here is that
on the basis of the flux viewpoint the emf — e, in Fig. 2-1 is treated as a reaction (or
generated) voltage rise in progressing from b’ to a '. On the other hand, from the
circuit viewpoint the voltage e,, and its polarity when considered in the direction
of flow of the current J,, appears as a voltage drop. The voltage e; remains the
same; it is merely the point of view that changes.
Because J, and the flux ¢ are in time phase, it is customary to combine the
two viewpoints into a single phasor diagram as depicted in Fig. 2-2(c). Since
E,» = E; represents the voltage drop, it follows that the same induced emf viewed
on the basis of a reaction to the changing flux, being a voltage rise, is equal and
opposite to E,. That is, Ey. = —E,. Another way of describing this is to say that
the terminal voltage must always contain a component equal and opposite to the
voltage rise.
It should be apparent up to this point that the reaction voltage viewed from
b' to a’ in Fig. 2-1 does not succeed in actually establishing a reverse current to
oppose the increasing flux. This never happens in the primary winding of a two-
winding transformer, but it does happen in the secondary winding as is presently
described.
In the work that follows in this chapter preference is given to treating the
induced emf as a voltage drop. That is, in applying Faraday’s law, the version
described by Eq. (2-8) will be used. Due regard is given to Lenz’s law in this
Sec. 2-1 Theory of Operation and Development of Phasor Diagrams 55

equation by noting that, when the time rate of change of flux in the configuration
of Fig. 2-1 is positive, the polarity of point a is also positive.

Practical lron-Core Reactor

As the second step in the development of transformer theory, let us consider how
the phasor diagram of Fig. 2-2(a) must be modified to account for the fact that a
practical reactor has a winding-resistance loss as well as hysteresis and eddy-
current losses. A linear magnetization curve will continue to be assumed because
the effect of the nonlinearity is to cause higher harmonics of fundamental fre-
quency to exist in the magnetizing current—which, of course, cannot be included
in the phasor diagram. Only quantities of the same frequency can be shown.
To develop the phasor diagram in this case we start with the flux phasor ¢
and then draw the induced emf E, 90° leading. This much is shown in Fig. 2-3(a).
Next we must consider the location of the current. Keep in mind that the iron is in
a cyclic condition, i.e., the flux is sinusoidally varying with time. Moreover, this
cyclic variation takes place along a hysteresis loop of finite area because the core
losses are no longer assumed zero. Now, as pointed out in Chapter 1, it is charac-
teristic of ferromagnetic materials in a cyclic condition to have the flux lag behind
the magnetizing current by an amount called the hysteretic angle y. See Fig.
2-3(b). Since the positive direction of rotation is counterclockwise in a phasor
diagram, J, is shown leading ¢ to bear out the physical fact. Note, too, that the
magnetizing current I, is now considered made up of two fictitious components J,
and /.. The quantity J, is the purely reactive component and is the current that

Ey

(a)

Figure 2-3 Development of the phasor

diagram of a practical iron-core reactor:
(a) building on the flux phasor; (b)
accounting for core losses; (c) account-
ing for core losses and winding resis-
tance.
 Transformers Chap. 2
56

would flow if the core losses were zero. The quantity J, is that fictitious quantity
which, when multiplied by the voltage E, produced by the changing flux, repre-
sents exactly the power needed to supply the core losses caused by the same
changing flux. Thus we can write
LE ey. (2-13)
Or

eee:
h=z A (2-14)

where P, denotes the sum of the hysteresis and eddy-current losses. This tech-
nique for handling the core loss of an electromagnetic device is a useful one and is
frequently applied in the study of ac machinery.
To complete the phasor diagram for the practical iron-core reactor, it is only
necessary to account for the winding-resistance drop. This is readily accom-
plished by applying Kirchhoff’s voltage law to the circuit. Hence
V; — Del + E, V (2-15)

where r; is the winding resistance in ohms. Since r; is a constant, it follows that

the I,,r, drop must be in phase with I, or located parallel to [,, as depicted in Fig.
2-3(c). Note that the component of J, that is in phase with the applied voltage Vi
exceeds J, as well it should because this component must supply not only the
core losses but the winding copper losses too. Thus, the input power may be
expressed as
Pi, = Vile COS 0) = [27 Pe W (2-16)

where 6) denotes the power-factor angle.

The Two-Winding Transformer

By placing a second winding on the core of the reactor of Fig. 2-1, we obtain the
simplest form of a transformer. This is depicted in Fig. 2-4. The rms value of
the induced emf E, appearing in the primary winding ab readily follows from
Eq. (2-12) as

Figure 2-4 Two-winding transformer.

Sec. 2-1 Theory of Operation and Development of Phasor Diagrams 57

ee N,®,,@ auf
E1 05) ~ WE P,,N; == 4.44f®@,,Ni ‘
(2-17)

Keep in mind that this result derives directly from Faraday’s law and is very
important in machinery analysis. One significant result deducible from this equa-
tion is that for transformers having relatively small winding resistance (i.e., FE, ~
V,), the value of the maximum flux is determined by the applied voltage.
The induced emf appearing across the secondary winding terminals is pro-
duced by the same flux that causes E,. Hence the only difference in the rms
values is brought about by the difference in the number of turns. Thus we can
write

E, = 4.44f®,,N> (2.18)

Dividing Eq. (2-17) by (2-18) leads to

= a = ratio of transformation (2-19)

Another fact worth noting is that E, and E, are in time phase because they are
induced by the same changing flux.

Transformer Phasor Diagram at No-Load

When a transformer is at no-load, no secondary current flows. Figure 2-5 depicts

the situation. A study of Fig. 2-5 should make it apparent that the development of
the phasor diagram follows from Fig. 2-3(c) by making two modifications. One is
to identify the secondary induced emf in the diagram. But, as already observed in
the preceding step, the emf induced in the secondary winding must be in phase
with the corresponding voltage E, of the primary winding. For convenience call
this secondary induced emf E). Then £ and E, must both lead the flux phasor by
90° as shown in Fig. 2-6. For convenience the turns ratio a is assumed to be
unity. The second modification is concerned with accounting for the fact that not
all of the flux produced by the primary winding links the secondary winding. The

Figure 2-5 Two-winding transformer

at no-load; S open.
 Transformers Chap. 2
58

Figure 2-6 Phasor diagram of a transformer at no-load.

difference between the total flux linking the primary winding and the mutual flux
linking both windings is called the primary leakage flux and is denoted by gj.
Note that some of the paths drawn to represent ;; encircle only several turns and
not all N; turns. As long as such closed paths encircle an mmf different from zero,
such a flux path does in fact exist. This is borne out by Ampere’s circuital law.
An important distinction exists between the mutual flux ¢ and the primary
leakage flux @;,. It is this. The mutual flux exists wholly in iron and so involves a
hysteresis loop of finite area. The leakage flux, on the other hand, always in-
volves appreciable air paths, and, although some iron is included in the closed
path, the reluctance experienced by the leakage flux is practically that of the air.
Consequently the cyclic variation of the leakage flux involves no hysteresis (or
lagging) effect. Hence in the phasor diagram the primary leakage flux must be
placed in phase with the primary winding current as shown in Fig. 2-6. Now this
leakage flux induces a voltage E;), which must lead ¢;; by 90°. This emf due to
leakage flux may be replaced by an equivalent primary leakage reactance drop.
Thus

This quantity must lead /,, by 90°. The phasor addition of this drop and the
primary winding resistance drop as well as the voltage drop E, associated with the
mutual flux yields the primary applied voltage V,; as depicted in Fig. 2-6.
The quantity x; of Eq. (2-20) is called the primary leakage reactance; itis a
fictitious quantity introduced as a convenience in representing the effects of the
primary leakage flux.

Transformer under Load

In this last step in the development of the theory of operation of the transformer,
we direct attention first to the phasor diagram representation of Kirchhoff’s volt-
age law as it applies to the secondary circuit. For simplicity consider that the load
appearing in the secondary circuit of Fig. 2-7 is purely resistive and further that
the flux has the direction indicated with the flux increasing. By Lenz’s law there
is emf induced in the secondary winding which instantaneously makes terminal c
positive with respect to terminal d. When switch S is closed, a current J; flows
Sec. 2-1 Theory of Operation and Development of Phasor Diagrams 59

Figure 2-7 Two-winding transformer

under load; S closed.

instantaneously from c through the load to d. For convenience the load resistor iS
assumed adjusted to cause rated secondary current to flow. E> establishes the
secondary load voltage besides accounting for other voltage drops existing in the
secondary circuit. When equilibrium is established after the load switch is closed,
the appropriate phasor diagram must show V, and J in phase as depicted in Fig.
2-8. Since it is physically impossible for the secondary winding to occupy the
same space as the primary winding, a flux is produced by the secondary current
which does not link with the primary winding. This is called the secondary leak-
age flux. The voltage E,. induced by this secondary leakage flux leads ¢) by 90°.
Again as a matter of convenience this secondary leakage flux voltage is replaced
by a secondary leakage reactance drop. Thus

where x) is the fictitious secondary leakage reactance which represents the effects
pre-
of the secondary leakage flux. Also, because leakage flux has its reluctance
Thus, doubling the current
dominantly in air, Eq. (2-21) is a linear equation.
ed
doubles the leakage flux, which doubles En, and by Eq. (2-21) this is represent
of the terminal voltage, the second-
by a doubled reactance drop. The phasor sum
adds up to
ary winding resistance drop, and the secondary leakage reactance drop
the source voltage E, as demonstrated in Fig. 2-8.
that
When a phasor diagram shows V, and J, in phase, the implication is
V7J,. Clearly this power must come
power is being supplied to the load equal to

load for the case

Figure 2-8 Complete phasor diagram of a transformer under
where a = N,/N2 > 1.
 Transformers Chap. 2
60

about? To
from the primary circuit. But by what mechanism does this come
to this question, consider the situation depicted in Fig.
understand the answer
g in the direction shown, the secondar y emf has the
2-7. With the flux increasin
when the switch is closed the action of the secondar y
polarity indicated, so that
in the mutual flux. This tendency for the flux to
current is to cause a decrease
of NJ, causes the voltage drop FE; [or e; = N i(dd/
decrease owing to the action
slight
dt)] to decrease. However, since the applied voltage V, is constant, the
apprecia ble increase in primary current. In fact the
decrease in E; creates an
primary current is allowed to increase to that value which, when flowing through
the primary turns N,, represents sufficient mmf to neutralize the demagnetizing
action of the secondary mmf. Expressed mathematically we have
I,N; — Nol, = 0 (2-22)

where /, denotes the increase in primary current needed to cancel out the effect of
the secondary mmf to reduce the flux. I, is often called the balance current. This
is above and beyond the value /,, needed to establish the operating flux. Equation
(2-22) reveals that N; I, must be oppositely directed to N>I,. In short, then, it is
through the medium of the flux and its small attendant changes that power is
transferred from the primary to the secondary circuit of a transformer. Of course,
the total primary current is the phasor sum of J, and J,,. In this case the primary
leakage flux is in phase with /; rather than J,,, which is the case at no-load.
For transformers in which the winding resistance drop and the leakage reac-
tance drop are negligibly small, the magnetizing mmf is the same at no-load as
under load. In other words, the phasor sum of the total primary mmf N; I, and the
secondary mmf N2I> must yield NiJ,. Thus

Nt oe Nol = Nin (2-23)

Dividing through by N, leads to

N
Ke De Lin (2-24)

Inserting Eq. (2-22) then yields

imo dy = In = 1, — 1, (2-25)

where J} is equivalent to J,. These equations are readily verified in the phasor
diagram of Fig. 2-8.

Transformer Construction Features

Contrary to the schematic representation of Fig. 2-7 it should not be implied that
the total number of primary winding turns are placed on one leg of the ferromag-
netic core and all the turns of the secondary winding are placed on the other leg.
 Sec. 2-1 Theory of Operation and Development of Phasor Diagrams 61

Such an arrangement leads to an excessive amount of leakage flux, so that for a

given source voltage the mutual flux is correspondingly smaller. The usual proce-
dure in transformer construction is to split the primary and secondary turns. This
keeps the leakage flux within a few percent of the mutual flux. Of course still
greater reduction in leakage flux can be achieved by further subdividing and
sandwiching the primary and secondary turns, but this is clearly obtained at an
appreciable increase in the cost of assembly.
The two principal types of construction used for single-phase transformers
are the core type and the shell type. In the core type the windings are made to
surround the iron in the fashion illustrated in Fig. 2-9. In the typical configuration
one-half the turns of the low-voltage coil are placed on each leg nearest the iron
separated by an appropriate layer of insulation. Then following another layer of
suitable insulation half the turns of the high-voltage winding are placed over the
low-voltage coil. The process is repeated for the second leg as well. The core-
type construction is favored for high-voltage transformers because the problem of
insulating is easier. In the shell-type transformer it is the iron that appears to
surround the coils as depicted in Fig. 2-10. The coils in this configuration are
generally of a pancake form as distinguished from the cylindrical forms used in the
core-type transformer. Often the low-voltage winding is divided into three sec-
tions with the two outer sections each having a quarter of the turns and the

Insulation

Pancake-type wire
coils

Laminated
shell iron

Figure 2-9 Construction features of the core- Figure 2-10 Construction features of the
type transformer. H denotes the number of shell-type transformer. H denotes the num-
turns of the high voltage coil; L denotes the ber of turns of the high voltage coil; L denotes
number of turns of the low voltage coil. the number of turns of the low voltage coil.
62 Transformers Chap. 2

remaining half is sandwiched in between the two halves of the high-voltage wind-
ing as indicated in the diagram. The shell-type construction is favored for power
transformers where large currents are made to flow. The iron shell provides
better mechanical protection to the windings in such circumstances.

Effect of Saturation on Magnetizing Current

All the quantities that appear in a phasor diagram such as the one depicted in Fig.
2-8 must necessarily be at the same frequency. Consequently, the magnetizing
component, J,,; of the current I, has to be treated as a quantity with a sinusoidal
shape when so used. However, the true shape of the actual magnetizing current is
other than sinusoidal chiefly because of the influence of the magnetic core. Two
factors are important in this consideration. One is the state of saturation of the
magnetic iron core and the other is the combined effect of hysteresis and core
losses.
Frequently, transformers function with time flux variations that place opera-
tion just above the knee of the saturation curve. It is instructive to examine the
consequences of such a mode of operation. This is best accomplished by referring
to Fig. 2-11. At the outset it is important to understand that the situation de-
scribed in this diagram is a transitory one. We consider it in order to put into
perspective the forces that create the final shape of the magnetizing current.
The voltage waveshape that is applied’to the primary winding of the trans-
former is assumed to be a pure sinusoid. Accordingly, this voltage source is
capable solely of supplying a sinusoidal magnetizing current to the primary coil of
the transformer. Figure 2-11 shows this magnetizing current drawn below the
magnetization curve, which in turn is depicted as possessing a fair degree of
saturation but no hysteresis. The effect of hysteresis in the magnetic core mate-
rial is being neglected for the moment in the interest of simplicity. The projection
of the sinusoidally varying magnetizing current onto the magnetization curve
produces a flat-topped flux wave as shown to the right of the magnetization
curve. Moreover, this flat-topped flux wave contains primarily two harmonic
components: a large fundamental component, ¢,, and a substantial negative third
harmonic component, @;. The fundamental flux component induces a voltage, e,,
in the primary coil which is equal and opposite to the applied primary voltage, v;,
on the assumption of negligible primary winding resistance. However, the nega-
tive third harmonic flux component also induces a voltage in the primary coil, e3.
But this voltage does not encounter an equal and opposite component in the
primary source voltage because it is a purely sinusoidal quantity. As a result, the
coil-generated third harmonic voltage is free to act in the primary circuit to pro-
duce a negative third harmonic current which then combines with the positive
fundamental component supplied by the source voltage. Thus, the resultant mag-
netizing current is the superposition of these two components. The result is a
waveshape for the magnetizing current that is characteristically peaked as de-
picted in Fig. 2-12. The peakirig effect is associated with the fact that the third
Sec. 2-1 Theory of Operation and Development of Phasor Diagrams 63

Figure 2-11 Illustrating the effect of a sinusoidal magnetizing current: (a) magne-
tization curve showing saturation at high magnetizing forces and no hysteresis; (b)
sinusoidal magnetizing current assumed to flow from source voltage to primary
coil; (c) resultant flat-topped flux wave produced by the current of (b).

harmonic component is made to flow in a direction opposite to the fundamental

component because it originates from the primary coil.
It is interesting to observe that the action of the negative third harmonic
current can also be interpreted as producing a primary-coil mmf that in turn acts to
third
produce a negative third harmonic flux which serves to cancel the original
harmonic flux component of the flat-toppe d flux wave. Ultimately , of course,
Kirchhoff’s voltage law demands that the resultant core flux must be sinusoidal
the
because only such a flux produces an induced coil voltage that balances
applied sinusoidal source voltage.
The peaked nature of the resultant magnetizing current can also be demon-
can be
strated by starting with the condition that the sinusoidal source voltage
64 Transformers Chap. 2

Magnetizing
current

'¢

Figure 2-12 Transformer magnetizing current waveshape neglecting hysteresis

and core losses. Here is; denotes the component supplied by the source voltage
and i; is the component supplied by the transformer coil arising from saturation
of the magnetic core.

balanced in the primary circuit only by the presence of a sinusoidal core flux. The
subsequent projection of such a sinusoidal flux variation onto the saturated mag-
netization curve then leads directly to the peaked waveshape of the magnetization
current. However, such an approach does not identify the source of the third
harmonic current component, which we now know is not directly the sinusoidal
source voltage. j
The final waveshape of the magnetizing current is obtained by adding to the
peaked curve of Fig. 2-12 a fundamental sinusoidal component (in phase with v,; or

Figure 2-13. Final waveshape of mag-

netizing current i,, that results upon
adding a sinusoidal fundamental com-
ponent (in phase with the source volt-
age) to iy of Fig. 2-12. The in-phase
component takes care of the hysteresis
and eddy current losses.
 Sec. 2-2. The Equivalent Circuit 65

e;) to account for the hysteresis and eddy-current losses. Figure 2-13 shows the
result.

2-2 THE EQUIVALENT CIRCUIT

Derivation

It is customary in analyzing devices in electrical engineering to represent the

device by means of an appropriate equivalent circuit. In this way further analysis
and design as well as computational accuracy is facilitated by the direct applica-
tion of the techniques of electric circuit theory. This procedure is followed when-
ever new devices are investigated. Accordingly, equivalent circuits are derived
not only for the transformer but for all ac and de motors as well as important
electronic devices such as the transistor. Generally, the equivalent circuit is
merely a circuit interpretation of the equation(s) that describe the behavior of the
device. For the case at hand there are two equations—Kirchhoff’s voltage equa-
tions for the primary winding and for the secondary winding. Thus, for the pri-
mary winding we have
Vi = hn + jx; == E, (2-26)

and for the secondary winding

E= V> of Dr, af jbx (2-27)

Both equations are represented in the phasor diagram of Fig. 2-8, and the corre-
sponding circuit interpretation appears in Fig. 2-14(a). The dot notation is used to
indicate which sides of the coils instantaneously carry the same polarity. The
schematic diagram of Fig. 2-14(a) can be replaced by the equivalent circuit of Fig.
2-14(b) by recognizing that J; is composed of I, and I,,. Moreover, since by
definition J,, consists of two components /, and I,, it can bé considered splitting
into two parallel branches. One branch is purely resistive and carries the current
I., which was defined as in-phase with E,. The other branch must be purely
reactive because it carries J,, which was defined as 90° lagging E;, as depicted in
Fig. 2-3(b). The ideal transformer is included in Fig. 2-14(b) to account for the
transformation of voltage and current that occurs between primary and secondary
windings. Thus, with the ratio of transformation a = 2, the ideal transformer
conveys the information that E, = 4E, and that , = 2/,. The fact that the actual
transformer has primary winding resistance and leakage flux and must carry a
magnetizing current is taken care of by the circuitry preceding the symbol for the
ideal transformer in Fig. 2-14(b).
The resistive element r, must have such a value that, when the voltage EF,
appears across it, it permits the current J. to flow. Thus

we (2-28)
 Transformers Chap. 2
66

(a)

Ideal transformer
a+!

(b)
at Meee
a*r, —jx2=j0°X2

Ideal transformer
a=1
(c)

Figure 2-14 Development of the exact

equivalent circuit: (a) schematic dia-
gram; (b) two-winding equivalent cir-
cuit; (c) replacing the actual secondary
winding with an equivalent winding
having N, turns; (d) final complete
(d) form.
Sec. 2-2 The Equivalent Circuit 67

The second form of this equation is obtained by replacing E; with the expression
of Eq. (2-13). Hence, r, may also be looked upon as that resistance which ac-
counts for the transformer core loss. Similarly, the magnetizing reactance of the
transformer may be defined as that reactance which when E, appears across it
causes the current J, to flow. Thus

(2-29)

A single-line equivalent circuit of the transformer is a desirable goal. How-

ever, as long as E> # E, this is impossible to accomplish because the potential
distribution along the primary and secondary windings can never be identical.
Only when a = 1 is this condition possible. Therefore, in the interest of achieving
the desired simplicity in the equivalent circuit, let us investigate replacing the
actual secondary winding with an equivalent winding having the same number of
turns as the primary winding. A little thought should make it reasonable to expect
that, if an equivalent secondary winding having N, turns is made to handle the
same kilovolt-amperes, the same copper loss and leakage flux, and the same
output power as the actual secondary winding, then insofar as the applied voltage
source is concerned it can detect no difference between the actual secondary
winding of N> turns and the equivalent secondary winding of N, turns. That is,
one winding can replace the other without causing any changes in the primary
quantities.
Because the equivalent secondary winding is assumed to have N, turns, the
corresponding induced emf must be different by the factor a. We have

where the prime notation is used to refer to the equivalent winding having N,
turns. For the kilovolt-amperes of the equivalent secondary winding to be the
same as that of the actual secondary winding the following equation must be
satisfied:

EjIy = Eyl, (2-31)

Inserting Eq. (2-30) into the last equation shows that the current rating of the
equivalent secondary winding must be

(2-32)

Note that this current is the balance current J,.

 Transformers Chap. 2
68

nce of the equiva-

Since in general a # 1, it follows that the winding resista
loss is the same in both.
lent secondary must be adjusted so that the copper
Accordingly
(hry = Lr (2-33)
of the winding
This equation, together with Eq. (2-32), establishes the value
resistance of the equivalent secondary. Thus

n=(2) n= en
! lh ¢ 9
(2-34)
I,

referred to the
Often the quantity r; is described as the secondary resistance
are dealing
primary. The phrase ‘‘referred to the primary”’ is used because we
seconda ry winding having the same number of turns as the
with an equivalent
primary winding.
By proceeding in a similar fashion the secondary leakage reactance referred
to the primary can be expressed as

XPS Xs (2-35)

Moreover, since the kilovolt-ampere load must be the same for the two
windings we have .
Vily = Volo (2-36)

from which it follows that the load terminal voltage in the equivalent secondary
winding is

V2 =—7 V2 = aV2 (2-37)

Appearing in Fig. 2-14(c) are all of the foregoing results in addition to an

ideal transformer having a unity turns ratio. Keep in mind that Fig. 2-14(c) is
drawn with the equivalent secondary winding having N; turns, whereas Fig.
2-14(b) is drawn with the actual secondary winding. As far as the primary is
concerned it cannot distinguish between the two. For the ideal transformer of
Fig. 2-14(c) note that joining point a to a’, b to b’, and c to c’ causes no distur-
bance because instantaneously these points are at the same potential. Accord-
ingly, this ideal transformer may be entirely dispensed with, in which case Fig.
2-14(c) becomes Fig. 2-14(d), which is called the complete equivalent circuit of the
transformer. This is the single-line diagram we set out to derive.

Approximate Equivalent Circuit

In most constant-voltage, fixed-frequency transformers such as are used in power

and distribution applications, the magnitude of the magnetizing current I, is only a
few percent (2 to 5) of the rated winding current. This happens because high-
Sec. 2-2 The Equivalent Circuit 69

permeability steel is used for the magnetic core and so keeps Jy very small.
Furthermore, steel with a few percent of silicon added is often used and this helps
to reduce the core losses, thereby keeping [, small. Consequently, little error is
introduced and considerable simplification is achieved by assuming J/,, so small as
to be negligible. This makes the equivalent circuit of Fig. 2-14(d) take on the
configuration shown in Fig. 2-15. Note that the primary current /, is now equal to

Re =ntre X= jx1+jxo

Figure 2-15 Approximate equivalent

circuit referred to the primary.

the referred secondary current /;. Also we can now speak of an equivalent resis-
tance referred to the primary

Re =f; +rn=n + ar, (2-38)

and an equivalent leakage reactance referred to the primary

Xo =X, +42 = x1 + BX (2-39)

The subscript | denotes the primary.

The corresponding phasor diagram drawn for the approximate equivalent
circuit appears in Fig. 2-16. Note the considerable simplification over the phasor
diagram of the exact equivalent circuit which appears in Fig. 2-8.

Figure 2-16 Phasor diagram of the

approximate equivalent circuit for a
lagging power factor angle 6).

The analysis of the preceding pages concerns itself with replacing the actual
secondary winding by an equivalent winding having the same number of turns as
the primary winding. A little thought should make it apparent that a similar proce-
dure may be employed to draw a single-line equivalent circuit with all quantities
referred to the secondary winding. In other words, the actual primary winding
can be replaced by an equivalent primary winding having N> turns. The resulting
equivalent circuit then assumes the form depicted in Fig. 2-17. It is interesting to
note that, to refer the primary winding resistance r; to the secondary, it is neces-
sary to divide by a*. Thus
r= (2-40)
 Transformers Chap. 2
70

= ah : ae
T, Reo ge *'2 iXen=iga + ike
a

Re

a: ideal transformer

Figure 2-17 Approximate equivalent cir- Figure 2-18 Approximate equivalent

cuit referred to the secondary. circuit drawn with the ideal transformer.

Furthermore, the equivalent winding resistance referred to the secondary is

(2-41)

and the equivalent leakage reactance referred to the secondary is

(2-42)

Note too that the ideal transformer is included in Fig. 2-17 merely to emphasize
the transformation that takes place from the primary side to the secondary side.
Normally, the primary circuit is omitted entirely leaving only the single-line sec-
ondary circuit.

Meaning of Transformer Nameplate Rating

A typical transformer carries a nameplate bearing the following information:

10 kVA 2200/110 V. What are the meanings of these numbers in the terminology
and analysis developed so far? To understand the answer refer to Fig. 2-18, which
is really the approximate equivalent circuit drawn to include the ideal transformer
to bring into evidence the transformation action existing between the primary and
secondary circuits. The number 110 refers to the rated secondary voltage. It is
the voltage appearing across the load when rated current flows at rated kVA. We
call this V>. The 2200 number refers to the primary winding voltage rating, which
is obtained by taking the rated secondary voltage and multiplying by the turns
ratio a between primary and secondary. In our terminology this is V;. By speci-
fying the ratio V/V. = 2200/110 on the nameplate, information about the turns
ratio is given. Finally, the kilovolt-ampere figure always refers to the output
kilovolt-amperes appearing at the secondary load terminals. Of course the input
kilovolt-amperes are slightly different because of the effects of R.; and X.;.
Sec. 2-2. The Equivalent Circuit 71

Transfer Function

The transfer function of the transformer may be specified with respect either to
voltage or to current. The voltage transfer function is defined as the ratio of the
output voltage V> to the input voltage V; with the load connected. Thus
a ogVs V
Ty = 2 = — ~- 1 (2-43)
Vi @Vo PDR tjbXn, 4
Equation (2-43) indicates that when the winding resistances and leakage fluxes are
negligibly small the voltage transfer function Ty is approximately the reciprocal of
the ratio of transformation. When these quantities cannot be neglected, Ty will be
a complex number which varies with the load.
The current transfer function 7; is defined as the output current divided by
the input current. For the approximate equivalent circuit of Fig. 2-18 we have
Se awl:
T; = = =a (2-44)
I,
Here the current transfer function is exactly equal to a because the magnetizing
current is assumed negligible. When /,, is not negligible, 7; will be somewhat
smaller than a and may be complex.

Input Impedance

The input impedance of a transformer is the impedance measured at the input

terminals of the primary winding. In the case of the approximate equivalent cir-
cuit shown in Fig. 2-18 we have

7 = la = R,, + aRy t+ jXe1 (2-45)

rf
A glance at this result indicates that the input impedance is dependent upon the
load impedance. In the case of power applications the input impedance is of no
particular concern because the internal impedance of the source is negligibly
from a
small. In communication applications, however, where V, is obtained
is a useful quantity because it describes the effect that
transistor, input impedance
the transformer has on the voltage source.

Frequency Response

y response of
One of the things we are interested in, when referring to the frequenc
in which the voltage transfer function varies fora
a transformer, is the manner
the source voltage V, is varied. The amplitud e of V,
fixed load as the frequency of
mers frequenc y response is of no serious
is assumed fixed. For power transfor
operated at a single fixed frequen cy—usua lly 60
interest because these units are
 Transformers Chap. 2
72

Hz and sometimes 50 Hz. However, in communication circuits the frequency of

the source voltage is likely to vary over a wide range. For example, in the case of
an output transformer which couples the power stage of an audio amplifier to a
loudspeaker the frequency may vary over the entire audio range. What, then, is
the effect of this varying frequency on the voltage transfer function? The answer
follows from an inspection of Fig. 2-14(d). Keep in mind that inductive reactance
is directly proportional to the frequency, i.e.,
x= ob = 2nfh (2-46)

Hence at low frequencies the reactance x4, which normally is so high that it may
be removed from the circuit, can now no longer be neglected. As a matter of fact
this low magnetizing reactance can effectively shunt the fixed load impedance
thereby causing a severe drop-off of output voltage as depicted in Fig. 2-20.
Figure 2-19(a) shows the configuration of the equivalent circuit as it applies at very
low frequencies. Note that x; and x: are so small at these frequencies that they
may be omitted entirely from the diagram.
In the intermediate frequency range the design of these transformers for
communication systems is such that x; and x} are still quite small while xg is
sufficiently large that it may be omitted. The appropriate circuitry appears in Fig.
2-19(b). Note that for a resistive load the equivalent circuit consists solely of
resistive elements so that the transfer function remains constant over the band of
intermediate frequencies. Refer to Fig. 2-20.
At very high frequencies x, and x} are no longer negligible and therefore must
be included in the circuit. However, xs may continue to be omitted. The equiva-
lent circuit now takes the form depicted in Fig. 2-19(c). A little thought reveals
that, as the frequency is allowed to get higher and higher, more and more of the

z
"4 lo

Ri =07R, Ry =a*R,

(b)

" IX4 lo jx

(M1) Ri =02R,

(c)
Figure 2-19 Equivalent circuit for various portions of the frequency spectrum:
(a) low frequencies; (b) intermediate frequencies; (c) high frequencies.
Sec. 2-2. The Equivalent Circuit 73

Intermediate
frequency

function
transfer
voltage
Ty= 10 100 1000 107 10° Figure 2-20 Voltage transfer function
Frequency, Hz V,/V, versus frequency.

source voltage V; appears across x; and x} and less and less appears across the
fixed load resistance. Therefore again a drop-off occurs in the value of the voltage
transfer function as depicted in Fig. 2-20. In communication transformers a suit-
able frequency response such as that shown in Fig. 2-20 is very often the most
important characteristic of the device.
Example 2-1
Hz,
A distribution transformer with the nameplate rating of 100 kVA, 1 100/220 V, 60
(Q) and a leakage reactance of
has a high-voltage winding resistance of 0.1 ohm
resistance is 0.004 © and the leakage reactance is
0.3 . The low-voltage winding
0.012 ©. The source is applied to the high-voltag e side.
(a) Find the equivalent winding resistance and reactance referred to the high-
voltage side and the low-voltage side.
in volts
(b) Compute the equivalent resistance and equivalent reactance drops
of the rated winding voltages expressed in terms of the primary
and in percent
quantities.
(c) Repeat (b) for quantities referred to the low-voltage side.
referred to
(d) Calculate the equivalent leakage impedances of the transformer
the primary and secondar y sides.

Solution From the nameplate data the turns ratio is a = 1100/220 = 5.

primary by Eq.
(a) Hence the equivalent winding resistance referred to the
(2-38) is
Re = + @ry = 0.1 + (5)?(0.004) = 0.2 0

Also
Xo = x1 + ax, = 0.3 + 25(0.012) = 0.69
y are
The corresponding quantities referred to the secondar

Re = "3 + 0.004 = 0.008 2

Xo = ‘2 + 0.012 = 0.024 0

x3. This is a very common

These computations reveal that r; = r5 and x; =
ormers . A well-d esigne d transformer is one that
occurrence in well-designed transf
74 Transformers Chap. 2

uses a minimum amount of iron and copper for a specified output. Note, too, that the
resistance of the secondary winding is considerably smaller than that of the primary
winding. This is consistent with the larger current rating of the former.
(b) The primary winding current rating in this case is

, 100kVA _
ted kV eitdus
Hence
I5Re; = 9100.2) = 18.2 V

Expressed as a percentage of the rated voltage of the high-tension winding this is

Re 18.2 Ak
Vv; 100 = T100 100 = 1.65%

Also
I5X,., = 9110.6) = 54.6 V

and

1X1 SOPH
aye 100= T100 = 4.96%

(c) The secondary winding current rating is

I, = aly = 591) = 455 A

Hence

LR.2 = 455(0.008) = 3.64 V

and

LR.» 3.64
V; 100 = 790 100= 1.65%

Note that the percentage value of the winding resistance drop on the secondary side
is identical with the value on the primary side. Similarly, for the leakage reactance,

IX. = 455(0.024) = 10.9 V

Pi
7." 100 10g
= 55 100=4.96%
(d) The equivalent leakage impedance referred to the primary is

Ze = Rey + jXe1 = 0.2 + j0.6 = 0.634/71.6°

Referred to the secondary this is

sae IZ = 0.0253 /71.6°

Ze = i
 Sec. 2-3 Parameters from No-Load Tests 75

2-3 PARAMETERS FROM NO-LOAD TESTS

The exact equivalent circuit of the transformer has a total of six parameters as Fig.
2-14(d) shows. A knowledge of these parameters allows us to compute the perfor-
mance of the transformer under all operating conditions. When the complete
design data of a transformer are available, the parameters may be calculated from
the dimensions and properties of the materials involved. Thus, the primary wind-
ing resistance r; may be found from the resistivity of copper, the total winding
length, and the cross-sectional area. In a similar fashion a parameter such as the
magnetizing reactance x, may be determined from the number of primary turns,
the reluctance of the magnetic path, and the frequency of operation. The calcula-
tion of the leakage reactance is a bit more complicated because it involves ac-
counting for partial flux linkages. However, formulas are available for making a
reliable determination of these quantities.
A more direct and far easier way of determining the transformer parameters
is from tests that involve very little power consumption, called no-load tests. The
power consumption is merely that which is required to supply the appropriate
losses involved. The primary winding (or secondary winding) resistance is readily
determined by applying a small dc voltage to the winding in the manner shown in
Fig. 2-21. The voltage appearing across the winding must be large enough to
cause approximately rated current to flow. Then the ratio of the voltage drop
across the winding as recorded by the voltmeter, VD, divided by the current
flowing through it as recorded by the ammeter, AD, yields the value of the winding
resistance.

Open-Circuit Test

Information about the core-loss resistor r, and the magnetizing reactance x, is

obtained from an open-circuit test, the circuitry of which appears in Fig. 2-22.
Note that the secondary winding is open. Therefore, as Fig. 2-14(d) indicates, the
input impedance consists of the primary leakage impedance and the magnetizing
the
impedance. In this case the ammeter and voltmeter are ac instruments and
ac ammeter. The
second letter A is used to emphasize this. Thus AA denotes

Current-limiting
resistor

source F |
|

‘Low High
side _ side

Figure 2-22 Wiring diagram for the trans-

Figure 2-21 Circuitry for finding wind-
ing resistance. former open-circuit test.
76 Transformers Chap. 2

wattmeter shown in Fig. 2-22 is used to measure the power drawn by the trans-
former. The open-circuit test is performed by applying voltage either to the high-
voltage side or the low-voltage side, depending upon which is more convenient.
Thus, if the transformer of Example 2-1 is to be tested, the voltage would be
applied to the low-tension side because a source of 220 V is more readily available
than one of 1100 V. The core loss is the same whether 220 V is applied to the
winding having the smaller number of turns or 1100 V is applied to the winding
having the larger number of turns. The maximum value of the flux, upon which
the core loss depends, is the same in either case as indicated by Eq. (2-17). Now
for convenience assume that the instruments used in the open-circuit test yield the
following readings:
wattmeter reading = P
ammeter reading = I),

voltmeter reading = V;
The voltmeter reading carries the subscript L to emphasize that the test is per-
formed with the source and instruments on the low-voltage side.
Neglecting instrument losses, the wattmeter reading may be taken entirely
equal to the core loss, i.e., P. = P. This is because the attendant copper loss is
negligibly small. Moreover, since /,, is very small, the primary leakage impedance
drop may be neglected, so that for all practical purposes the induced emf is equal
to the applied voltage, i.e., E; = V_. In accordance with these simplifications the
open-circuit phasor diagram is represented by Fig. 2-3(b). The no-load power-
factor angle 4) is computed from

6) = cos"! VI, (2-47)

It then follows that

12 = Ip COs 06 (2-48)
and
Ig = Im Sin (2-49)
It is important to keep in mind that these currents are referred to the low-tension
side. The corresponding values on the high-voltage side would be less by the
factor 1/a. The low-side core-loss resistor is
P P
oT Gea (2-50)
The corresponding high-side core-loss resistor is

We nes (2-51)
The magnetizing reactance referred to the low side follows from Eq. (2-29). Thus
Sec. 2-3 Parameters from No-Load Tests 77

is VE
Eee Tey 2 Leasimn OG eres
On the high side this becomes
X¢H = aX or (2-53)

Although the manner of finding the magnetizing impedance is treated in the

foregoing, the motivation is not so much necessity as the desire for complete-
ness. Invariably, the use of the approximate equivalent circuit is sufficient for all
but a few special applications of transformers. Accordingly, the useful informa-
tion obtained from the open-circuit test is the value of the core loss, which helps to
determine the efficiency of the transformer.

Short-Circuit Test

Of the six parameters of the exact equivalent circuit two remain to be determined:
the primary and secondary leakage reactances x, and x2. This information is
obtained from a short-circuit test which involves placing a smail ac voltage on one
winding and a short-circuit on the other as depicted in Fig. 2-23(a). Appearing in

WM Xen Rey

Variable
voltage
A-C supply

Figure 2-23 Transformer short-circuit test: (a) wiring diagram; (b) equivalent
circuit.

secondary winding
Fig. 2-23(b) is the approximate equivalent circuit with the
ce is merely the equivalent imped-
short-circuited. Note that the input impedan
ance.
is applied to the
When performing the short-circuit test, a reduced ac voltage
to do so. From the computations
high side usually because it is more convenient
nce drop is only about 5% of the
of Example 2-1 recall that the leakage impeda
is 55 V, is easier and more
winding voltage rating. Hence 5% of 1100 V, which
Also, the wattmeter
accurate to deal with than 5% of 220 V, which is 11 V.
entirely equal to the winding
reading in the circuitry of Fig. 2-23(a) can be taken
greatly reduced voltage used in
copper losses. This follows from the fact that the
small. Calling the wattmeter
the short-circuit test makes the core loss negligibly
ne the equivalent winding
reading P,. and the ammeter reading /;., we can determi
resistance as
78 Transformers Chap. 2

ee
Ruz = (2-54)
— en
8

The subscript H is used in place of L since it is known from the wiring diagram
that the primary side is the high-voltage side. The equivalent resistance referred
to the low side is then

Re = -G (2-55)

where

_ Nu _ high-side turns
(2-56)
“"N, low-side turns

Before the equivalent reactance can be determined, it is first necessary to

find the equivalent impedance, which a glance at Fig. 2-23(b) reveals to be

(2-57)

During the test the applied voltage V,. is adjusted so that J,, is at least equal to the
rated winding current. The equivalent reactance follows from Eqs. (2-54) and
(2-57). Thus

Se Ae (2-58)

It is important to note that this computation provides the sum of the primary
and secondary leakage reactances. It gives no information about the individual
breakdown between x; and x.. Whenever the approximate equivalent circuit is
used in analysis, such a breakdown is unnecessary. On the few occasions when
the individual breakdown is needed, it is customary to assume that

xX) = x3 (2-59)

This statement is based on the assumption that the transformer is well-designed.

One final note. It should be apparent from the theory of transformer action
that the establishment of rated current in the high-side winding requires a corre-
sponding flow of rated current in the secondary circuit. It is for this reason that
the total winding copper loss is measured in the short-circuit test.
Sec. 2-3 Parameters from No-Load Tests 79

Example 2-2
The following data were obtained on a 50-kVA, 2400/120-V transformer:
Open-circuit test, instruments on low side:
wattmeter reading = 396 W

ammeter reading = 9.65 A

voltmeter reading = 120 V

Short-circuit test, instruments on high side:

wattmeter reading = 810 W

ammeter reading = 20.8 A

voltmeter reading= 92 V

Compute the six parameters of the equivalent circuit referred to the high and low
sides.
Solution From the open-circuit test
iE 396
8) == cos =) Vil,
ee cos a
7200. 65) = 70 te)
' ee
ee

Ter, = Im COS 09 = 9.65(0.342)

= 3.3 A
= Im Sin 0) = 9.65 sin 70°= 9.07A
Tgp

Hence
ler = = 36.40

This can also be found from

ee
E,
as
_ Vi
Tea_ 120 = Be

On the high side this quantity becomes

ren = Grex = 400(36.4) = 14,560 0

The magnetizing reactance referred to the low side is

PEN eg, 120.

+i aw Rees Sor
The corresponding high-side value is
X¢H — axe. — 400(13.2) = 5280 0

y
From the data of the short-circuit test we get directl

+ Ze =a400 = 0.0110
4.42
Z eH ~
2 440,
20.8
 80 Transformers Chap. 2

Also

and

Xen = VZ2, — Roy = V442 — 18P = 40

é.

Xet = qn = 9.01 1

2-4 EFFICIENCY AND VOLTAGE REGULATION

The manner of describing the performance of a transformer depends upon the

application for which it is designed. It has already been pointed out that in com-
munication circuits the frequency response of the transformer is very often of
prime importance. Another important characteristic of such transformers is to
provide matching of a source to a load for maximum transfer of power. The
efficiency and regulation is usually of secondary significance. This is not so,
however, with power and distribution transformers, which are designed to operate
under conditions of constant rms voltage and frequency. Accordingly, the mate-
rial that follows has relevance chiefly to these transformers.
As is the case with other devices, the efficiency of a transformer is defined as
the ratio of the useful output power to the input power. Thus
_ output watts
input watts co)
Usually the efficiency is found through measurement of the losses as obtained
from the no-load tests. Therefore by recognizing that
output watts = input watts — > losses (2-61)
and inserting into Eq. (2-60), a more useful form of the expression for efficiency
results. Thus

> losses
ae input watts (2-62)

where

> losses = core loss + copper loss = P, + 13R,2 (2-63)

Equation (2-62) also leads to a more accurate value of the efficiency.

Power and distribution transformers very often supply electrical power to
loads that are designed to operate at essentially constant voltage regardless of
whether little or full-load current is being drawn. For example, as more and more
Sec. 2-4 Efficiency and Voltage Regulation 81

light bulbs are placed across a supply line, which very likely originates from a
distribution transformer, it is important that the increased current being drawn
from the supply transformer does not cause a significant drop (more than 10%) in
the load voltage. If this happens the illumination output from the bulbs is sharply
reduced. Similar undesirable effects occur in television sets that are connected
across the same supply lines. The reduced voltage means reduced picture size.
An appreciable drop in line voltage with increasing load demands can also cause
harmful effects in connected electrical motors such as those found in refrigerators,
washing machines, and so on. Continued operation at low voltage can cause these
units to overheat and eventually burn out. The way to prevent the drop in supply
voltage with increasing load in distribution circuits is to use a distribution trans-
former designed to have small leakage impedance. The figure of merit used to
identify this characteristic is the voltage regulation.
The voltage regulation of a transformer is defined as the change in magni-
tude of the secondary voltage as the current changes from full-load to no-load with
the primary voltage held fixed. Hence in equation form,

Vi =
voltage
eee |V;|
regulation =
[V3
|
) pa. i fal?! (2-64)
[V2 [V2
where the symbols denote the quantities previously defined, absolute signs are
used in order to emphasize that it is the change in magnitude that is important.
Equation (2-64) expresses the voltage regulation on a per unit basis. To convert to
percentage, it is necessary to multiply by 100. The smaller the value of the voltage
regulation, the better suited is the transformer for supplying power to constant-
voltage loads (constant voltage is characteristic of most commercial and industrial
loads). ,
In determining the voltage regulation it is customary to assume that V; is
adjusted to that value which allows rated voltage to appear across the load when
rated current flows through it. Under these conditions the necessary magnitude of
V, is found from Kirchhoff’s voltage law as it applies to the approximate equiva-
lent circuit. Refer to Figs. 2-15 and 2-16. For the assumed lagging power-factor
angle 02 of the secondary circuit the expression for V, is

Vi = V4/0° + 15 /-02Zc1 [8 (2-65)

where

0, == tan“! tan! B 3 2-66

Xo
(2-66)
el

secondary terminal
Equation (2-65) is written using the referred value of rated
voltage as the reference.
 Transformers Chap. 2
82

Example 2-3
m-
For the transformer of Example 2-2(a) find the efficiency when rated kilovolt-a
(b) compute the
peres are delivered to a load having a power factor of 0.8 lagging;
voltage regulation.

Solution (a) The losses at rated load current and rated voltage are

> losses = 396 + 810 = 1206 W = 1.2 kW

Also
output kW = 50(0.8) = 40 kW

input kW = 40 + 1.2 = 41.2 kW

Hence

y= 1 ~ OSE = 1 — gy = 1 — 0.029 = 0.971

or

n = 97.1%
(b) From Eq. (2-65) we have
Vi = Vn = 2400 + 20.8 /—37° 4.42 /65°
where

X oH by =| 4 s °
6, = tan“!
Rect aides ca
Therefore

V, = Vy = 2400 + 92/28° = 2400 + 81.1 + 43.1

Vi ~ 2481.1
Note that the j-component is negligible in comparison to the in-phase component.
This calculation indicates that, in order for rated voltage to appear across the load
when rated current is drawn, the primary voltage V; must be slightly different from
the primary winding voltage rating of 2400 V. The percentage of voltage regulation is
accordingly found to be
2481.1 — 2400
percent of f voltage regulation
ion = 7400 100 = 3.38%

Example 2-4
In this example our aim is to find the voltage regulation by assuming that the trans-
former of Example 2-3 is operated at the input terminals with the rated voltage of
2400 V. To facilitate the comparison of the ensuing result with that obtained in
Example 2-3, use is made of the same impedance which yields the rated output
current when rated voltage appears at the secondary terminals; in other words, the
load impedance has the value Zy = 115.4 /30° = 100 + 757.8 ohms referred to the
high side.
 Sec. 2-5 Mutual Inductance 83

Solution By Kirchhoff’s voltage law for the primary circuit we have

V, = 2400/0° = TLR + Ry + (Xe + XDI

= 1,[1.87 + 100 + j(4 + 57.8)]
= 2400 2400
eG cueraoke 11915/-31.28 0 ee
V5 = TZ), = 20.14 /-31.24° (115.4 /30°) = 2324.465 /—1.24°
2400 — 2324.465
percent voltage regulation =
7324.465 100 = 3.25%

2-5 MUTUAL INDUCTANCE

Analysis of the two-winding transformer by the methods of electric circuit theory

involves the use of self- and mutual-inductance. So far, Kirchhoff’s voltage equa-
tion applied to the primary winding has led to Eq. (2-26). In the material that
follows this result will be shown to be identical with that obtained by the classical
approach.
For simplicity assume the ferromagnetic material has a linear magnetization
curve with no core losses. Refer to Fig. 2-24 where for the indicated current
directions Kirchhoff’s equation becomes

ny =int LG -Me (2-67)

where lowercase letters for current and voltage denote instantaneous values.
Here the quantity L, di,/dt denotes the voltage drop associated with the total flux
produced by i; and linking the primary turns N,. Similarly, the quantity —M di,/
dt refers to the voltage rise associated with the flux produced by i, and linking the
primary turns. The existence of a voltage rise in the primary circuit for the as-
sumed direction of i> is easily verified. Note that the secondary mmf N2i2, if
allowed to act freely, produces a flux which threads counterclockwise in the
configuration of Fig. 2-24. Then the reaction in the primary coil between two
points such as a and b by Lenz’s law requires that point b should be at a higher
re-
potential than a. Accordingly, this induced emf becomes a voltage rise with
spect to the specified direction of i.

Figure 2-24 Elementary transformer

for analysis by the classical approach.
 Transformers Chap. 2
84

Assuming that the applied primary voltage is a sinusoid, Eq. (2-67) may be
written in rms quantities as
V, =fin + joLyl, — joMh (2-68)
Inserting the relationship
ip="als (2-69)
wherein a is the ratio of transformation, we get
V, =fin + joLi, — joaMls (2-70)
But by Eq. (2-24)
Dati In = =a

so that
V, = (yr, + jo(L, — aM)l, + joaMI, (2-71)
Equation (2-71) is now in a form where correspondence with Eq. (2-26) is possible
after some algebraic manipulation.
The expression for the self-inductance of the primary winding may be writ-
ten as
[= do,
l,=M Tee N, ®,
z (2-72)

The equivalence of the differential form with the total quantities in this last ex-
pression is allowed by the linearity condition. Recalling that the total flux linking
the primary coil consists of the mutual flux ®,, plus the primary leakage flux Dj,
Eq. (2-72) can be rewritten as

L, = N, ———- (2-73)

Also,

®,, =—— and ®, = — (2-74)

where &,, is the reluctance experienced by the mutual flux and consists exclu-
sively of iron and &,, is the net reluctance experienced by the leakage flux of the
primary in air. Accordingly, Eq. (2-73) becomes

i=
ONG
R,. a Ri Ly ey (2-75)

where L,, denotes an inductance associated with the mutual flux and L,); is the
inductance associated with the primary leakage flux. It is important to note here
that L,, is not the mutual inductance.t

+L, is in fact the mutual inductance referred to the primary winding.

Sec. 2-5 Mutual Inductance 85

The classical definition of mutual inductance is given in this case by

My, = Np “Sn!
1
= Ny, 2m!yy
d m ®P,
(2-76)
Thus the mutual inductance is related to the number of turns of winding 2 and the
change of flux with respect to the current in winding 1 which produces the mutual
flux ®,,;. Upon inserting N,i; = ®n;R, into Eq. (2-76), there results

N2N,
My = (2-77)
Rin
The expression for mutual inductance may also be found by considering a change
in mutual flux corresponding to a change in secondary current i, linking N, turns.
Thus

dbm2 = O12
My = N diy Ni pe (2-78)

But
O2Rm = Noir (2-79)

Inserting this into Eq. (2-78) yields

N,N.
Me a 2=M (2-80)

A comparison of Eq. (2-80) with Eq. (2-77) shows them to be the same, and from
here on either expression will be denoted by M.
Returning to Eq. (2-71) we are now in a position to evaluate the significance
of the term in parentheses. Thus, by Eqs. (2-75) and (2-77)
TINTON ONT NON Ni _
Li Mats Ge i Ri N> “abe - Ri me bi cor

Clearly this quantity is the primary leakage inductance, and it represents the effect
of the primary leakage flux. Of course the total quantity
w(L; on aM)I, = oLyl = xt; (2-82)

is nothing more than the primary leakage reactance drop and is the counterpart of
the second term of Eq. (2-26).
Let us now examine the last term of Eq. (2-71). Thus

waMI, a) re Rp, Ig = oN, R, (2-83)

The fraction quantity on the right side of the last equation is the expression for
mutual flux. Specifically, however, since it involves the reluctance and the rms
of
value I, of the magnetizing current, this expression represents the magnitude
the maximum value of the mutual flux divided by V2. That is,
 Transformers Chap. 2
86

Mg _ im
® (2-84)
Te Ne
(2-83)
where ®,, denotes the maximum value of the mutual flux. Accordingly, Eq.
becomes
® 27
aMI, = wN, — = —fNiPn = E (2-85)
eae iD) van
parameter so
It therefore follows from Eq. (2-85) that the mutual inductance
for
commonly encountered in linear circuit analysis is implied in the expression
both the induced emf and the mutual inductanc e are
the induced emf. Clearly
related by the same mutual flux.

2-6 THE EQUIVALENT CIRCUIT: COUPLED CIRCUIT VIEWPOINT

In Sec. 2-2, the equivalent circuit of the transformer is derived from the field
viewpoint based on interacting mmf’s and fluxes. This approach is often favored
because it cultivates a better physical understanding of the mechanism that under-
lies the operation and behavior of this device. However, it is entirely feasible to
arrive at the same equivalent circuit by applying the techniques of circuit analysis
to the coupled circuit, which is what the transformer is, after all. The usual
starting point in such an analysis is Eq. (2-67), which is repeated here for conven-
ience. Thus

nant Lome (2-86)

Keep in mind that L; denotes the self-inductance of the primary winding that is
associated with Nj; on the other hand, M represents the mutual inductance be-
tween the primary and secondary windings and involves the product of N; and N>
as indicated by Eq. (2-80). Figure 2-25 illustrates the schematic arrangement from
which it is obvious that Eq. (2-86) describes the situation that prevails in the
primary circuit in a manner consistent with Kirchhoff’s voltage law. For ease of

"4
é M
'4 , ea ts

V4 Ly
Figure 2-25 Schematic diagram of the
two-winding transformer expressed in
terms of the mutual and self-induc-
tances.
Sec. 2-6 The Equivalent Circuit: Coupled Circuit Viewpoint 87

handling, let us rewrite the last equation by replacing differentiation with the
derivative operator s. Accordingly,
OS ir} ae Ly si = sMiy (2-87)

Moreover, because the secondary emf is related to the primary emf by the turns
ratio a and because power is preserved assuming negligible core losses, it follows
that
1 = Alp (2-88)

where i, is the required balance current needed in the primary winding to ensure
power balance. In other words,

errsek = —
iCina:
(aip) (2-89)
where e, and e» are, respectively, the instantaneous voltages induced in the pri-
mary and secondary windings. Inserting Eq. (2-88) into Eq. (2-87) yields
Up = ir ts Lisi + aMsi, (2-90)

Next, we add zero to the right side of this equation in the form: aMsi, — aMsi,.
Hence we get
UN= ir ar (L; a aM)si, a aMs(i, = ip) (2-91a)

For sinusoidal sources the differential operator s can be replaced by jw so that the
last equation can be rewritten as
US ir] + jo(L; aa aM)i, + joaM(i, ay ip) (2-91b)

Before proceeding with a circuit interpretation of this equation, let us repeat the
foregoing analysis for the secondary circuit in Fig. 2-25.
In accordance with the dot notation, when the current i, takes the direction
shown in Fig. 2-25, the transformer action is such that the dotted side of the
secondary coil is positive and the current i, assumes the clockwise direction
indicated. Kirchhoff’s voltage law then leads to
di diy 2-92
=pnt+u+
M —Apa? 2 l=
BP (2-92)

the
In this expression the voltage drops appear on the right side. The use of
derivative operator then allows us to rewrite this equation as

= inl + Uy ae Ly Sir (2-93)

Msi,

Now multiply both sides of Eq. (2-93) by the turns ratio a to yield
aMsi, = av2 + aigr, + aLzSi (2-94)

Inserting Eq. (2-88) then leads to

aMsi; = av2 + i,(a?rr) + (a7L2) sip (2-95)
 Transformers Chap. 2
88

It is now necessary to manipulate this expression in such a manner that the

ensuing interpretation will be entirely consistent with Eq. (2-91), which after all is
part of the same equipment with the difference that Eq. (2-91) provides a descrip-
tion in the primary circuit while the latter does so in the secondary circuit.
Because the mutual inductance is the circuit parameter that links the primary and
secondary circuits, it should play the key role in providing a linkage between Eqs.
(2-91) and (2-95). A study of Eq. (2-91) suggests that the last term on the right,
namely, aMs(i, — i,), is the term that represents this linkage. This conclusion
becomes even more apparent when (i; — i,) is replaced by i,, as called for by Eq.
(2-25) because the magnetizing current and the mutual inductance go hand in
hand. Therefore, to establish this term in Eq. (2-95), it is necessary merely to add
zero to the left side in the form: aMsi, — aMsi,. Upon so doing and then rearrang-
ing, there results
aMs(i; — iy) = av2 + i,(a?r2) + (a7L2 — aM)Sip (2-96a)

Then by replacing s by jw we get

awM(i, — i,) = avr + a?rrip + jo(a*L2 — aM)ip (2-96b)

The circuit interpretation of Eqs. (2-91b) and (2-96b) taken together leads to
the T configuration depicted in Fig. 2-26. Several important points are worth
noting here. First, the quantity aM is really the mutual inductance referred to an
equivalent winding having N, turns. This is apparent from

aM = —~ (
1 Nj (2-97)
he, oR,

Moreover, the quantity (L; — aM) must necessarily represent the primary leakage
flux. Recall that L, represents the total flux that links the primary coil which
includes the leakage flux as well as the mutual flux whereas aM represents the
mutual flux referred to primary turns. It then follows that w(L; — aM) denotes the
primary leakage reactance. A similar situation prevails in the secondary part of
this circuit except that all quantities are expressed in terms of an equivalence
based on replacing the actual secondary turns, N2, by the primary turns Nj.

|
jo(L] -aM) jw(a2Lg —aM) aero

Hef

Ideal transformer

Figure 2-26 Equivalent circuit of the two-winding transformer derived from the
coupled-circuit viewpoint.
 Sec. 2-7 The Autotransformer 89

Thus, although the load voltage v2 is derived from a secondary winding having N>
turns, multiplication by a in Eq. (2-96) means that the level is raised to that of a
winding having N, turns. Similarly, the self-inductance of the secondary winding,
which is generally expressed as L2 = N 3/Rm, iS Converted to an equivalent induc-
tance of N; turns when it is multiplied by a? as shown in Eq. (2-96). It follows
then that the quantity (a2L, — aM) can be considered to be an inductance that
represents the secondary leakage flux referred to the primary. The resistance
parameter too is modified in a like manner. The secondary winding resistance
appears in Eq. (2-96) as well as in the circuit of Fig. 2-26 raised to a level of ar.
This is consistent with the fact that the current is reduced by 1/a, i.e., i, = ip/a, as
required by the changed voltage level, av,. The result is to keep the copper loss
the same for power cannot be changed by the transformation of a winding of N>
turns to one of N; turns. An ideal transformer is included in the equivalent circuit
of Fig. 2-26 to provide the proper transition to the load where the voltage v2
appears. Once the configuration of Fig. 2-26 is obtained, the matter of accounting
for the core losses can readily be resolved by inserting a resistor in series with the
mutual inductance parameter, aM, of such a value that it equals the core losses
when multiplied by the square of the rms value of magnetizing current for sinusoi-
dal sources.

2-7 THE AUTOTRANSFORMER

One of the principal features of the ordinary configurations of the transformer (see
Fig. 2-24) is the electrical isolation that exists between the primary and secondary
windings. Thus in the case of a distribution transformer that transforms power
at
from 4400 V_to220-V, there does not exist the danger of the 4400 V appearing
are involved,
the secondary terminals. Whenever large transformation ratios
where
safety demands that ordinary transformers be used. However, in instances
hazard), a trans-
the transformation ratio is close to unity (which minimizes the
notable advantage s.
former employing a single tapped winding may be used with
arrange ment of the
To understand some of these advantages, consider the
transformer is
conventional transformer appearing in Fig. 2-27(a). Assume the
at a terminal voltage of
ideal for simplicity. A resistive load is shown to draw 12 A
nt observa tion to
100 V. The corresponding primary current is 10 A. An importa
between points
make in this arrangement is that, all along the secondary winding
b and c, it is
b' and c’ and all along that section of the primary winding between
potentia l. If b and b’ are
possible to identify points having the same instantaneous
Extendi ng this proce-
two such points, they may be joined without adverse effect.
the entire length of winding s be and b'c' leads to the simplified
dure throughout
is called the autotrans-
configuration appearing in Fig. 2-27(b). This arrangement
former.
ary winding b'c’ of
Several things are worth noting here. First, the second
and in its place an appro-
the ordinary transformer can be completely eliminated,
 Transformers | Chap. 2
90

20 v

(a) (b)

Figure 2-27 (a) Ordinary transformer arrangement; (b) derived autotransformer

arrangement. Comparison is on the basis of same output kVA.

priate tap point, i.e., b, can be used to provide the required secondary voltage.
Since the output volt-amperes remain intact in the arrangement of Fig. 2-27(b), it
follows that the use of the autotransformer to provide the needed kVA at the load
leads to a saving in cost. Less copper is required for the same output. As a matter
of fact the saving in copper is even more pronounced than already indicated. The
current that flows in the common winding of the autotransformer is always the
difference between the primary and secondary currents of the ordinary trans-
former. In the case at hand this current is 2 A. Accordingly, the cross-sectional
area of the copper of winding bc of the autotransformer can be much smaller than
that of the primary of the ordinary transformer for the same design value of
current density. Second, the efficiency of the autotransformer for the same out-
put is greater. This result immediately follows from the smaller copper losses.
Third, the voltage regulation of the autotransformer is superior because of the
reduced resistance drop and lower leakage reactance drop. Keep in mind that one
winding has been completely eliminated. Hence the resistance and leakage flux of
this winding are zero. Fourth, the autotransformer is smaller in size for the same
output. This too follows from the reduced copper need.
There is another way in which the autotransformer of Fig. 2-27(b) may be
derived from an ordinary transformer. The second approach is predicated on the
fact that it uses, completely, the primary and secondary windings of an appropri-
ate ordinary transformer. The first approach is based on the same output for the
ordinary transformer and the autotransformer. In contrast the second approach is
based on the use of the same copper and iron; we should expect therefore that the
autotransformer will have a greater output rating than the corresponding ordinary
transformer. (The latter in this case is not the same as the ordinary transformer
used in the first approach.)
A close examination of Fig. 2-27(b) reveals that the autotransformer can be
derived by placing the winding voltages of two separate coils ab and bc in series
aiding. The coils ab and bc can then be treated as the primary and secondary
windings, respectively, of an equivalent ordinary transformer such as that illus-
trated in Fig. 2-28(a). Note that the volt-ampere rating of the equivalent ordinary
Sec. 2-7 The Autotransformer 91

transformer is 200, compared with 1200 for the derived autotransformer. In a

practical situation coils ab and b’c appear on the same magnetic core. Refer to
Fig. 2-28(b). The series-aid connection of the coils can be achieved by first ener-
gizing one of the coils, say ab, with rated voltage, then joining one terminal of the
first coil, say b, to one terminal of the second coil, say b’, and finally checking that
a voltmeter placed between the remaining two terminals a and c yields a reading
that is the sum of the voltage ratings of the individual coils. If the voltmeter reads
the difference, the connection to the second coil must be reversed. For the config-
uration of Fig. 2-28(b) the sense of the winding b’c has been so taken that the
series-aid connection results when b is joined to Db’.

Current and Voltage Relations for Step-Down

Autotransformer

the
A step-down autotransformer is one in which the primary voltage is larger than
voltage to the full winding abc in Fig.
secondary voltage. By applying a source
2-28(b) and placing the load across bc, operation as a step-down autotransformer
ing
results. In this configuration the applied primary voltage causes a magnetiz
In view of its small magnitud e com-
current to flow through the full winding ac.
analysis,
pared with the rated input current, for the sake of ease and simplicity of
this current will be considered equal to zero from here on.
trans-
The same basic principle that underlies the operation of the ordinary
the autotran sformer, i.e., the need for
former is found to govern the operation of
point in calculat ing the current and
balancing mmf’s. We use this as the starting
notation, use is
voltage relationships for the autotransformer. To simplify the
When the load switch S is closed in the
made of the symbols depicted in Fig. 2-29.

10 amp,

20v
b b

2 apt \00v

(a) (b)
rmer; (b) its derived autotrans-
Figure 2-28 (a) Equivalent ordinary transfo
copper and iron.
former. Comparison is on the basis of the same
92 Transformers Chap. 2

Figure 2-29 Symbolism for analysis of

the step-down autotransformer.

secondary of the autotransformer of Fig. 2-28(b), a current J begins to flow in the

common coil bc. It attempts to reduce the operating flux, but as soon as this
begins to happen there occurs an increase in primary current J; to such a value
that the mmf in winding ab neutralizes the mmf in winding bc. That is,
ny = NiI (2-98)
or

(Nu — Ni)In = Ni (2-99)

Hence,

(2-100)

where

a=
NPN
ane. Na anal (2-101)

Since the right side of Eq. (2-100) is a pure number, it follows that the currents in
coils ab and cb are in-phase. The opposing action of the mmf’s is derived rather
from the sense of the windings.
Applying Kirchhoff’s current law at point b yields
i, = Tr “Tr I (2-102)

But since J, and J are both in time phase, it follows that Eq. (2-102) is valid also
when only magnitudes are used. Thus,

Inserting J = I, — Iy into Eq. (2-100) yields the relationship between the low- and
the high-side currents. Hence

I - In _ ay = Il
inne
Sec. 2-7 The Autotransformer 93

or

I,
Te = a4 (2-104)

Finally, inserting Jy = I,/a, into Eq. (2-100) gives

alba a ie
(2-105)
te aa

The induced voltages in windings ab and bc are also in time phase because they
are caused by the same flux. Letting Ey = Eac = Ey, Ex = Enc = Ex, and E,, = Eq,
we then have

fe = SHE = ay— (2-106)

An inspection of Eqs. (2-100) and (2-106) shows that they bear out mathe-
matically the equivalence described in Fig. 2-28. These equations indicate that
the ratio of the voltages and currents in the coils ab and bc are the same as if the
turns n = N,» formed the primary and the turns Nz = Noe formed the secondary of
of
an ordinary transformer having a ratio of transformation of a, — 1. The value
a, for the autotransformer of Fig. 2-28(b) is 1.2. Hence the equivalent ordinary
Fig. 2-28(a)
transformer should have a turns ratio of a, — 1, or 1/5. A glance at
corroborates the result.
One advantage of treating the autotransformer as an equivalent ordinary
established
transformer is that use may be made immediately of previously
autotransformer
results. For example, to get the equivalent resistance of the
nt resistance of
referred to the high side all that is necessary is to write the equivale
primary and wind-
the equivalent ordinary transformer, treating winding ab as the
ordinary trans-
ing bc as the secondary. From our previous knowledge of the
former we can then write

lie a (a4 a 1)r2 (2-107)

Rex = lab Tr (aa at 1)°Pbe SS

= rp-, the resistance of winding

where r, = rap, the resistance of winding ab, and rz
high side becomes
bc. Similarly the equivalent reactance referred to the

Xn ar (a4 as 1)*x2 (2-108)

Xen = Xab ax (a, = Lene =

the expression
The result expressed in Eq. (2-107) can be verified by writing
load termina ls are short-c ircuited. If
for the total copper loss when the secondary
throug h bc is J, the copper loss at
the current flowing through ab is Jy and that
 Transformers Chap. 2
94

short circuit is then given by

Pe = ioe = Pry. (2-109)

An equivalent resistance as seen from the high side may be derived from Eq.
(2-109) provided J can be expressed in terms of [7. This is readily accomplished
by Eq. (2-100). Thus,

Pou = Vira + (aa — Thr oc

or
Parlors + (aa Tere) (2-110)

The quantity in brackets can be interpreted as that equivalent high-side resistance

which when multiplied by the high-side current squared yields the copper loss.

Power and Volt-Ampere Relations

Unlike the conventional transformer, the autotransformer has two kinds of volt-
ampere power associated with it. It is the result of the direct copper connection
existing between the source and the load. Refer to Fig. 2-29 to distinguish the
two. Since the transformer action present in the autotransformer takes place in
windings ab and bc, it follows that the number of volt-amperes in winding ab is
equal to the number of volt-amperes in winding bc and specifically represents the
transformed apparent power. Expressing this mathematically we have
transformed VA = Egplan = Enclye (2-111)

where VA denotes volt-amperes. If for simplicity we assume negligible winding

resistance and leakage reactance, Eq. (2-111) can be written in terms of the termi-
nal voltages. Thus
transformed VA = Vaplan = Voclyc (2-112)

For the values specified in Fig. 2-28 note that the transformed volt-amperes are
20(10) = 2(100) = 200.
However, the input corresponding to the specified transformer current is
V.clap = 120(10) = 1200 VA. This means that the number of volt-amperes that is
not transformed is

Vael ap i, Vanday TT (Vee i Vip) Lov Fi Viclas (2-113)

The specific value for this example is then

Vila, = 100110) = 1000 VA
The quantity represented by Eq. (2-113) is called the conducted volt-amperes.
The reason for the description follows from the fact that the input current J,, flows
conductively to the load at the potential V,.. Although J,, starts at the potential
Vac, it experiences a drop V,, through transformation and then proceeds to the
Sec. 2-7 The Autotransformer 95

load at the reduced potential. However, it is important to keep in mind that the
reduced potential at which J,» flows to the load is compensated for exactly by an
increase in load current originating from the transformer action.
In the notation of Fig. 2-29 we can write for the step-down autotransformer

transformed VA = V,Jy = VI (2-114)

conducted VA = V_Iy (2-115)

An interesting comparison to make at this point is the ratio of the output volt-
amperes of the autotransformer to the transformed volt-amperes. The latter quan-
tity in essence is the output of an ordinary transformer having the same current
density and magnetic density ratings. Accordingly,
output VA of autotransformer _ output VA of autotransformer
transformed VA output VA of equiv. ord. transformer

Lebar V thin ogyvhal Cdl

GA (2-116)

last term in
where ay is the ratio of transformation of the autotransformer. The
Eq. (2-105). Applying Eq. (2-116) to the
the preceding equation follows from
of 6, which emphasiz es that for many
configuration of Fig. 2-28 yields a factor
unity) most of the power delivered to the
autotransformer applications (a, close to
and just a small portion is derived through
load occurs through direct conduction
transformation.
Phasor diagram. The phasor diagram of the step-down autotransformer
to the high-
can be conveniently developed by applying Kirchhoff’s voltage law
nt ordinary transfor mer in establis hing
side circuit and making use of the equivale
Fig. 2-27 Kirchhof f’s voltage law yields
the quantity V., = V,. Accordingly, in
Vi = Var + Vico = Vn + VE (2-117)
transformer, it
By treating winding ab as the primary of an equivalent ordinary
follows that
— 1)?Xvc]
Vip = Vn = (aa — DV + Tala + (aa — 1)’ ve] + jlylxan + (aa
(2-118)

or
Va I(Ren + jXen) = (aa — 1)Vi + InZen (2-119)
oe Ve = aa
Inserting Eq. (2-119) into Eq. (2-117) yields
Vu = (aa — Vt + TnZen + Vi (2-120)

or more simply
Vi = aaVz + TnZen (2-121)
 Transformers Chap. 2
96

jl4%eH Figure 2-30 Phasor diagram of the

step-down autotransformer: (a) inter-
= pretation in terms of Eq. (2-120);
(b) interpretation in terms of Eq.
(b) (2-121).

Appearing in Fig. 2-30(a) is the phasor diagram representation of Eq. (2-120). The
simplified version is depicted in Fig. 2-30(b).

Step-Up Autotransformer

The step-up autotransformer may be analyzed in a manner similar to that used for
the step-down configuration. In the step-up case the source is applied to winding
bc and the load is placed across the full winding ac. Refer to Fig. 2-31. The
functional relationships between the input, output, and common winding currents
can be derived by starting with the balancing mmf condition that must prevail in
coils ab and bc. If this is done and if the ratio of transformation of the step-up
case is defined as greater than unity, i.e., a4 = Nac/Nbe = Vy/V_, the following
results are obtained:

var Gyaeel (2-122)

H

t, ay
t= (2-123)
H

Pau
2 Saul 2-124
(2-124)

TL,=Int+] (2-125)

and
E.b = En = Encla, — 1) = Ex(aa = 1) (2-126)

A comparison of these equations ‘with Eq. (2-100) and Eqs. (2-103) to (2-104)
reveals that the results for the step-up case can be obtained from those of the step-
Sec. 2-7 The Autotransformer 97

Figure 2-31 Configuration and symbol-

ism for the step-up autotransformer.

down case by merely interchanging the roles of the currents J; and I provided
that a, is defined the same way in both cases.
Example 2-5
motor in order to
A step-down autotransformer is to be used to start an induction
to an acceptabl e level which is known to be
limit the magnitude of the starting current
of 230 V. At starting, the autotran sformer is to be set at
34 A for a line voltage rating
an 80% tap point.
arrangement
(a) Show the schematic diagram of the autotransformer starter
with voltages clearly indicated.
(b) Find the autotransformer input current.
(c) What is the value of the current in the common coil?
(d) Determine the transformed VA.
(e) What is the value of the conducted VA?
ns specified.
(f) Find the VA rating of the autotransformer for the conditio
Identify the rating of an equivale nt ordinary transfor mer that can be used as
(g)
the autotransformer.
A (See Fig. 2-32.)
Solution (a) Vy = 230, Vz = 0.8(230) = 184 V, I, = 34
(b) First find a4 = Nac/Noc = 1/0.8 = 1.25. Then by Eq. (2-104)

I 34
‘ig Gapiae »gee Sade Sa

(c)
1=1, — In = 34 -— 27.22 =68A

(d)
Transformed VA = V,IJ = 184(6.8) = Daite2

= (Vy — Vi)Iy = 46(27.2) = 1251-2

(e)
Conducted VA = IV, = 27.2(184) = 5004.8
= 230(27.2) = 6256
(f) VA rating of autotransformer = Vyly
ormer: 46/184 V, 1251.2 VA
(g) VA rating of equivalent ordinary transf
 98 Transformers Chap. 2

Figure 2-32 Solution to part (a) of

Example 2-5.

2-8 PER-UNIT CALCULATIONS

Computations in per-unit values first found wide use in transmission-line calcula-

tions often involving several generators and many transformers. One feature of
the per-unit notation is that it makes it unnecessary to refer quantities on the high
or low side of transformers. The per-unit value of a circuit element such as
primary winding resistance is the same on either side. The bookkeeping proce-
dure is therefore appreciably simplified by such a notation. However, the per-unit
notation bears another noteworthy advantage. It reveals important information
about the design and performance characteristics of the equipment to which it is
applied. It is this feature which is emphasized in this section.
Application of the per-unit system of notation depends upon the establish-
ment of a set of specific base quantities by which other similar quantities can be
expressed. For example, all voltage drops in the low-voltage circuit of a trans-
former can be expressed as a ratio of the rated low-side voltage where the latter
quantity is identified as the base, or reference, voltage. In fact it is customary to
use the rated quantities of a device as the base values whenever per-unit calcula-
tions are to be performed. There is nothing that restricts this selection. Rather,
the choice of using the rated quantities as the base values is made solely because
of the significance that can be attached to the resulting numbers. Directing atten-
tion now specifically to the two-winding transformer we identify the following
base quantities:
base voltage on high side = rated voltage on high side = Vr
base current on high side = rated current on high side = Ip
., ., _ rated high-side voltage — Vp
base impedance on high side = ratedhightaideaurreni ia = I, = Zen

rated high-side current Ir

base admittance on high side = fated highaden wleawe annie Yor
Sec. 2-8 Per-Unit Calculations 99

base output VA = rated output VA =Vrlr = Valr

base voltage on low side = rated voltage on low side = Vr
base current on low side = rated current on low side = Ip

’ : rated low-side voltage V wh

base impedance on low = =—* — aco
P : oie rated low-side current ie 23
: rated low-side current I
base admittance on low side = ——-——.._,,___ = k= ="
ce on low side ~~eed low-side voltage Vie Va Ce
It is worthwhile to note that base quantities have units. This is not true of
the per-unit values of these quantities because by definition the per-unit value of a
quantity is a ratio, or the quotient, of the actual value divided by the correspond-
ing base value. Thus to obtain the per-unit value of the rated low-side voltage of a
transformer we write
Vr (volts) Apes
pu Vr iiaee voltage on low side (volts) Vr ~ : ete’

Clearly, if the rated quantities of a device are chosen to be the base values, then
the per-unit values of all base quantities are equal to unity. A similar procedure is
followed to find the per-unit value of other quantities. To illustrate further, let us
find the per-unit value of the equivalent resistance referred to the high side of a
two-winding transformer, R.7. Here we write :

R.y (ohms)
p-u Rez = (2-128)
ZBH (ohms)

The result is a numeric value—a quantity without units. However, considerable

in
significance can be attached to this number by multiplying Eq. (2-128) by unity
the form of Ip/Ip. Equation (2-128) then becomes

p-u Rey =
Reulr ay Rent ES TrRen
(2-129)
ZealeomValizix Ve
that J rp and Vp denote the rated values of the current and voltage,
Recalling
value of
respectively, it follows from the right side of Eq. (2-129) that the per-unit
voltage that appears across
R.n gives information about the portion of the applied
when rated current flows.
the equivalent winding resistance as a voltage drop

Example 2-6
e resistance of 0.1 Q and
A 10-kVA, 200/100-V transformer has an equivalent high-sid
©. Determi ne (a) the per-unit value
an equivalent high-side leakage reactance of 0.16
voltage absorbe d by the equivalent winding
of R. and (b) the percent of rated
resistance.

Solution (a) From the transformer rating

100 Transformers Chap. 2

Tk = 75 = SOA

is se
Zon = 7 50 7 42

Hence by Eq. (2-128)

pu R= ee oor eho

(b) By Eq. (2-129) it follows that

0.025Vr = 0.025(200) = 5 V
or 2.5% of the rated voltage appears across the equivalent winding resistance.

Still more significance can be attached to the per-unit value of R.y by multi-
plying Eq. (2-129) once again by unity in the form of [p/Ip. Thus

p-u Ren =
TrRen Ip _ [RRen
Vrodr, ~ Vale
__ total copper loss at rated value
volt-amperes (2-130)
"rated
This last expression reveals that the per-unit value of R.q also conveys informa-
tion about the portion of the rated volt-amperes that is consumed as a copper loss
at rated current. Thus for the transformer of Example 2-4 it can be said that, when
rated current is delivered to the load, 2.5% of the transformer rating is dissipated
as copper loss. In watts, this represents a loss of 0.025(10,000) = 250 W. Note
how much more meaningful it is to deal in the per-unit value of R,, than merely to
cite the absolute value.
If the equivalent resistance of a transformer is specified with reference to the
low side, the per unit value is found by using the low-side base impedance. Thus

Ret Ret = TrRet TRRet

Dalal 57. oth Valdion 6VAY oe mabe
A glance at this expression reveals that the per-unit value of R,,; carries the same
interpretation as the per-unit value of R.,. Moreover, the numerical values are
identical for percent of voltage drop and percent of copper loss.
Computing the performance of a transformer—regulation and efficiency—
can be accomplished quite conveniently in per-unit values. For example, in find-
ing the voltage regulation for a lagging power factor, the expression for the re-
quired primary voltage we know to be

Vi II Vr + TpZe1 a Vr os Tr(cos 0 eh sin O)(Rei + JXe1)

Vp/0° + IrRei cos 0 + [rXe sin 6 + fURXe1 cos 6 — IRR.y sin 6) (2-132)
Sec. 2-8 — Per-Unit Calculations 101

Dividing through both sides of this equation by Vaz yields

Vi TrRei ja eipal q+ J (4
TeX IR
trie |.
_ = sin 6
:
(2-133)
a ~=1+ Vi“cos 6+ Vi<s in Vi cos 0

Keeping in mind that Iz and Vp denote rated quantities, we recognize the coeffi-
cients of the trigonometric terms as the per-unit values of the equivalent resis-
tance and equivalent reactance. For ease of handling let
ee (2-134)
Rl = pu Re = &
Vr
love. (2-135)
Roel = pu Xe = &
Vr
and
V :
b= pu V; (2-136)
Vr
Equation (2-133) can then be expressed in a shorthand per-unit notation as
p-u V; = 1 + «, cos 0 + e, sin 6 + j(e, cos 6 — «, sin 6) Cris)

The corresponding phasor diagram for this equation appears in Fig. 2-33. Since
for practical transformers the per-unit values of ¢, and e, rarely exceed 0.1, the
contribution of the j part of Eq. (2-137) in determining the per-unit value of V; is
negligibly small. Hence for all practical purposes Eq. (2-137) may be more simply
expressed as
p-u V; = 1+ &,cos @ + &, sin 0 (2-138)

The equation for the voltage regulation in per-unit notation then becomes
_ |p-u Vi| — |p-u Val _ 1 + €, cos 6 + & sin 6— 1
4 |p-u Va :
or

s an accurate means of
Equation (2-139) is significant not only because it provide
e it emphasizes the factors
computing voltage regulation but even more becaus

p-u Vy

Figure 2-33 Phasor diagram of a

transformer for lagging power factor,
with all quantities expressed in per-unit
p-ulp=140@ notation.
 Transformers Chap. 2
102

that are responsible for its existence, i.e., the winding resistances, the leakage
reactances, and the influence of the power-factor angle.
When the transformer delivers power to a leading power-factor load, 6 is
introduced as a negative quantity in Eq. (2-139). The expression for the per-unit
voltage regulation is then found to be

A comparison with Eq. (2-139) makes it clear that a transformer is more capable of
maintaining constant load voltage in the presence of a leading power-factor than a
lagging one.
The efficiency of a transformer can also be expressed in the per-unit nota-
tion. However, it is helpful first to identify the core loss as a per-unit quantity.
Equation (2-28) shows the core loss to be given by P. = [2r.. If the core-loss
current J. is replaced by E;/r, ~ Vr/r-, the core loss may be rewritten as

P, ~ = (Vr)8c (2-141)

where
1
8: = — = core-loss conductance (2-142)

The core-loss conductance can then readily be expressed as a per-unit quantity by

Be _ 8c __ Vr&e _ (Vr)°8e _ core loss (2-143)

p-U & =
Yeu IplVr Ip Varlr rated VA

Equation (2-143) indicates that the per-unit conductance representing core loss
may be found either from a knowledge of the core loss itself expressed in watts or
from a knowledge of the circuit parameter (g, or r.) associated with the core loss.
The efficiency of a transformer is affected by the presence of core losses and
copper losses. Since we now know how to express both losses in the per-unit
notation, let us return to the general expression for the efficiency, Eq. (2-62), and
appropriately modify it so that the per-unit notation is brought into evidence.
Thus

core loss + copper loss ‘

= So oo
(2-144)
UN =

rated output + core loss + copper loss

Assuming operation at rated voltage, rated current, and a power-factor of cos 6,

we have

(VrV8e + Ur) Ret (2-145)

=1- SP
nS ae Ge A On
iH Vrlp cos 6 + (Vrge + Ur)? Ret
Sec. 2-8 Per-Unit Calculations 103

Rearranging yields
(Vr)’ge , Ur) Ret
1 has Vedaet.
ons Vader
79 VE
.. (2-146)
2
ees
te ga (Va) 8c 4 Ue) Re
Vrlr Vrlr

and by Eqs. (2-130) and (2-143) this simplifies to

n=l
-ug. t+
peewee
p-u R.
(2-147)
~ cos 6+ p-ug. + puR,
was
Note that the second subscript for equivalent resistance in the last equation
value is the same irrespect ive of high or low side. It
dropped because the per-unit
that Eq. (2-147) is valid for finding the efficiency at
is important to understand
R, must be
rated load only. For any other load condition the value of per-unit
of the per-unit value of the current. Thus to find the
modified by the square
efficiency at load current [5 Eq. (2-147) becomes

p-u g. + (p-u R.)US /TRy (2-148)

ap elie a 5 f} ey
I;
— cos 0 + p-U g, +
TR

d that operation at
The per-unit value of g, remains intact because it is assume
rated voltage continues.

Example 2-7
efficiency when rated current is
For the transformer of Example 2-3 compute the
delivered to a 0.8 power-factor load.

Solution
core loss 396 W
P-U 8c = tated volt-amperes 50,000 VA Seige
copper loss at rated current _ 81055 = 0.0162
ae ae rated volt-amperes 50,000
the desired result:
Inserting these values into Eq. (2-147) yields
0.0162 _tS
0.00792 +pat z 0.9707
fe
Hence
n in percent = 97.07 ~ 97.1
computed in Example 2-3.
which compares favorably with the result
 104 Transformers Chap. 2

2-9 TRANSFORMERS FOR THREE-PHASE CIRCUITS

For reasons of efficiency and considerable economy in the use of copper, the
generation, transmission, and distribution of electric power takes place on a three-
phase basis rather than single-phase. The process begins with the electric genera-
tor, which is designed to develop a three-phase voltage. Characteristically, the
three-phase voltage system is distinguished by the fact that each phase voltage has
the same maximum value but is displaced in time from the other phase quantities
by 120 electrical degrees. The original energy source may be steam originating
from a boiler, water falling through an appropriate head, or the combustion of
gasoline or diesel fuel. Frequently, the output voltage of the electric generator is
limited by physical considerations to approximately 25 kV. The transmission of
large amounts of electrical energy over considerable distances at such relatively
low voltages would entail huge losses, thereby making it impractical to do so. A
common solution is to introduce transformers between the generator and the
transmission lines to boost the transmitting voltage to levels where the associated
line losses are manageable. Often the transmitting line voltage is raised to a level
of 230 kV and higher by means of step-up transformers. At the receiving end of
the transmission line transformers are again needed to reduce the voltages at the
substations to more reasonable levels. This is often followed by a further reduc-
tion at many places along a distribution network where the electrical energy is
finally consumed. Throughout the process, transformers are used in a three-phase
mode to provide the required transformations.
The three-phase mode of operation can be achieved either through use of a
single three-phase transformer or by use of three single-phase transformers, which
are in turn appropriately connected for three-phase operation. The single-unit
three-phase transformer consists of a three-legged construction with each leg
carrying the primary and secondary coils of one phase. When compared to three
single-phase transformers equipped to deal with the same volt-amperes, the single
three-phase unit is lighter, cheaper, requires less space, and is slightly more
efficient. Some cost-conscious installations prefer them because of the smaller
initial investment. However, a selection in favor of three single transformers has
some notable advantages. The larger total surface area, for example, makes for
better cooling and greater ease of repair. Moreover, in the initial development of
a power system, it may be entirely possible to operate with just two of the trans-
formers, albeit at reduced capacity. The third unit can be added at some future
date when the full capacity of the system is needed. Also, if a reserve unit is
needed to meet emergency conditions, only a single smaller unit is often sufficient
to have on hand as compared to the need to have a complete single-unit three-
phase transformer. In addition, the latter will require more space for storage and
is more difficult to move around. In view of these pros and cons, it is obvious that
the choice in a specific situation will be strongly influenced by the special circum-
stances surrounding that particular case. Because the operation of the transform-
ers in either mode is virtually identical, the development of the subject matter of
Sec. 2-9 Transformers for Three-Phase Circuits 105

The
this section is continued by dealing with the use of three single transformers.
choice is made without prejudice.

The A-A Connection

n, care must
When connecting three single transformers for three-phase operatio
system of
be taken to ensure that the secondary terminals provide a balanced
e the point,
voltages. The primary coils may be arbitrarily connected. To illustrat
depicted in Fig. 2-34. The objectiv e is to arrange the
refer to the arrangement
so that the line voltages will be balanced and no
secondary coil connections
distinguishing the
circulating current takes place in the secondary coils. To aid in
are used for the
primary windings from the secondary windings, uppercase letters
of common letters
former and lowercase letters for the latter. Moreover, the use
the coils are associated
(whether upper- or lowercase) is to be taken to mean that
primary winding and
with the same core. Thus, in Fig. 2-34(a) coil AB denotes the
transformer. The
coil ab denotes the secondary winding of the same single-phase
the three transformers
generator that supplies the voltages to the primary coils of
is shown in Fig. 2-34(b).
is assumed to furnish the balanced three-phase set that
in the closed form. Note,
Observe that this set is represented by a phasor diagram
cted. These connec-
too, that in Fig. 2-34(a), the secondary coils are left unconne
tions must now be made in a prescribed manner.
coil bc. The magni-
Begin by joining terminal b of coil ab to terminal b of
to one another, i.e., Vag =
tudes of the primary phase voltages are each equal
phase voltage. Similarly, the
Vac = Vea = Voi, where Vp1 denotes rms primary
to each other based on the
values of the secondary phase voltages are also equal
Thus Va, = Voc = Ve, = Vo2.
assumption that the transformers are identical.
It is helpful to be aware that
where V,2 denotes rms secondary phase voltage.
same, they do differ in phase.
although the magnitudes of these voltages are the
the open terminals ac in the
Consequently, when a voltmeter is placed across
be one of two choices. Either it
manner illustrated in Fig. 2-34(c), the reading will
to V,,. Keep in mind that the
will be V3 V,, or it will be equal in magnitude
at an angle of 60° (as
phasor voltage V;, can be added to the phasor voltage V.»
m of Fig. 2-34(c) or at an angle of
shown by the dashed line in the phasor diagra
values are dictated by the relation-
—120° (as indicated by the solid line). These
higher of the two values is found upon
ship of Vac to Vag on the primaries. If the
connection reversed. It is important to
testing, it should be rejected and the coil
phase voltages are equal to the line
keep in mind that in a A arrangement, the
voltages.
terminal c of coil ac is joined to c
In the last step of this phasing procedure,
opened except for the insertion of a
of coil bc, with the gap in Fig. 2-34(d) left
of two possibilities: either the reading
voltmeter. This voltage, too, presents one
result depends upon whether the voltage
will be twice V,, or it will be zero. The
tion or a direction of 120°. Clearly, that
of coil ac is directed along a —60° direc ge. Surely, we
that yields a zero gap volta
connection of coil ac must be chosen
 106 Transformers Chap. 2

Three- phase
generator

bs Vas
Vea

Vec
(b)
4
if
/
= a=,
a b Vab /—Vp¢

¢ |
Voc

(c)

Qa b % Vab

Vea
a &
Vbe

C
\
—Vea
(d)
& Figure 2-34 Transformer phasing
Vab procedure: (a) excitation of the pri-
maries; secondaries unconnected;
-, is (b) phasor diagram of primary voltages;
Vea Vbc (c) phasing two secondary coils; (d)
phasing the third coil for balanced
voltages; (e) phasor diagram of the
(e) balanced secondary voltages.
would not want to connect the two a terminals except when the potentials of these
terminals are the same. This avoids fireworks. Before leaving this matter, how-
ever, it is worthwhile to note that transformer manufacturers do use markings that
lead to the proper phasing of three-phase transformers. But obviously the phasing
procedure must be done by someone. Figure 2-34(e) shows the final phasor dia-
gram as it pertains to the balanced secondary voltages.
Sec. 2-9 Transformers for Three-Phase Circuits 107

pri-
The salient characteristics between the line and phase quantities from
d next.
mary to secondary for the A-A connection may be summarized as describe
Introduce

secondary phase
where the quantities on the right side represent the primary and
to the secondary
voltages. Assume that a balanced three-phase load is applied
windings. Then

Ratio of phase voltages: = =a

ab

Ratio of line voltages: yay VARs a

Viz “Vab
ary line voltages. The pri-
where V;; and V;» represent the primary and second
voltag es are also respectively in phase with
mary and secondary phase and line is in
with V;., and Vea
one another: i.e., Vaz is in phase with Vay, Vac is in phase
comparison of Fig. 2-34(b) with Fig. 2-34(e) makes this state-
phase with Vig. A
ment self-evident. Also,

: I 1
Ratio of phase currents: AB —
thas a

epelo/3- 1), Te
Ratio of line currents: 7— =
Tr2 3 Ip. Lab
t in a balanced A-connection is V3
Here use is made of the fact that the line curren
times the phase current.
are used in these arrangements
It is important that the transformers that
minimize the existence of circulating
have equal transformation ratios in order to
ge impedances of the transformers
currents. Furthermore, the equivalent leaka
to ensure proper load sharing during
should be as nearly identical as possible
makes only one voltage available with
operation. The use of a A secondary also
no possibility of including a neutral wire.
the magnetizing current in these
Finally, a word is in order concerning med
Fig. 2-34(a) a sinusoidal voltage is assu
arrangements. In the configuration of Beca use of its sinu-
three transforme rs.
to be applied to the primary coils of the
can deliv er to the transformer a magnetizing current
soidal nature, the generator flat-
oidal current, however, produces a
which is itself sinusoidal. Such a sinus of the trans forme r.
ation in the iron core
topped flux wave as a result of the satur toppe d flux wave con-
component, a flat-
In addition to the principal fundamental s
part, which , in turn, serves to induce, by Faraday’
tains a large third harmonic each of the prima ry
action takes place in
law, a third harmonic voltage. This s
third harm onic voltage induced in coil BC occur
transformers. Furthermore, the
108 Transformers Chap. 2

at angle of 3(—120°) = —360° or in-phase with the component appearing in phase

AB. Similarly, in coil CA, the phase of the third harmonic voltage is 3(120) = 360°,
so that it too is in-phase with the third harmonic voltage in the other two coils.
Since the sinusoidal generator contains no such component in its voltage cycle,
neutralizations of the third harmonic voltages cannot take place. Accordingly, the
additive effect of these third harmonic voltages is to cause a third harmonic
circulating current to exist in the three transformer primaries. This current, then,
serves the useful purpose of changing the magnetizing current from a sinusoidal
variation to one which is peaked to the extent that is necessary to ensure a
sinusoidal flux variation. The sinusoidal flux variation in turn provides a sinusoi-
dal voltage variation at the primary terminals of the transformers, thereby satisfy-
ing Kirchhoff’s voltage law in the generator-transformer primary circuit.

A-Y Connection

The delta-wye arrangement of transformers is commonly found at the transmitting

end of transmission lines as well as at the distribution end. At the transmitting
end, the A-Y connection offers the advantage of an increased step-up ratio beyond
that associated with the individual transformers. At the distribution end the inclu-
sion of a neutral wire allows a choice of two voltages for the customer’s use. A
popular combination that is available from the electric power companies to homes
and commercial establishments is the 208/120-V system. The larger figure is
available between any two line terminals of the Y-connection, whereas the lower
figure represents the voltage from line to neutral.
The phasing procedure for a A-Y connection proceeds differently from that
of the A-A connection insofar as the secondary connections are concerned. The
primary coils in each case can be wired arbitrarily in a closed delta arrangement.
Because the wye secondaries involve a neutral terminal, a modified terminal
marking is used for the individual transformers, as illustrated in Fig. 2-35(a). Of
course, when the primary windings are joined to form a closed circuit, the n
markings may be dropped as shown in Fig. 2-35(b). However, in the phasor
representation of the primary voltages that is depicted in Fig. 2-35(c) the n nota-
tion is carried along because it makes it easier to understand what is happening at
the secondaries.
We begin the phasing process at the secondaries by joining n of coil na to n
of coil nb. A voltmeter placed across the free terminals will indicate one of two
voltages, as was the case with the A-A connection. The specific reading depends
upon the relative sense of the winding. The connection causes the voltage in coil
an to be added to the voltage in coil nb along a —60° direction in the phasor
diagram or a 120° direction. These are the only two choices and the one that
prevails depends upon the relative sense of the windings when the two terminals
are connected. If the voltmeter reads a value for the line voltage V,,, that is equal
to V,», it means the addition is occurring along the 120° direction [see the dashed-
line component in Fig. 2-35(d)] and must be rejected. To correct this situation, it
Sec. 2-9 Transformers for Three-Phase Circuits 109

Figure 2-35 Phasing procedure for the

A-Y connection: (a) terminal markings
for the primary and secondary coils of
transformers A, B, and C; (b) primaries
in A mode with generator excitation;
(c) phasor diagram of primary line and
phase voltages; (d) phasing secondary
coils a and b; (e) phasing coil c for
balanced line-to-line and line-to-neutral
secondary voltages.
110 Transformers Chap. 2

is merely necessary to reverse one of the coil connections. Now the voltmeter
will yield for V,, a value of V3 V,». This is the proper value for the line-to-line
voltage of the wye connection. The line voltage must always be greater than the
phase voltage by a factor of 73. When the third coil c is joined in the manner of
Fig. 2-35(e), the correct phasing is determined when the voltages between each
pair of terminals is equal in magnitude to \/3 times the phase voltage. (What are
the voltage readings when coil c is incorrectly phased?)
A comparison of the phasor diagram of a properly phased A-Y connection
such as appears in Fig. 2-35(e) reveals that the line-to-line voltages at the second-
aries are displaced by 30 electrical degrees from the line-to-line voltages at the
primaries. Thus, for the situation depicted in Fig. 2-35, the line voltage between
terminals a and b at the secondaries is V,, = Va» /—30°, whereas the line voltage
between terminals A and B at the primaries is Vag = V4g/0°. Accordingly, there is
a lag of 30 electrical degrees in this arrangement. It will be recalled that this
condition did not occur in the A-A connection where it was found that the line-to-
line voltages between primaries and secondaries were all in phase. The impor-
tance of this distinction lies in the conclusion that a bank of A-A connected
transformers cannot be placed in parallel with a bank of A-Y transformers even if
the voltage magnitudes correspond. The phase angle discrepancy which is inher-
ent with these connections will result in wasteful circulating currents.
At the generator end of a transmission line, the A-Y connection is used to
step up the generator voltage to a suitable transmission line voltage. Let the step-
up voltage ratio be represented by

VA Voi
— ve — Vo (2-149)

where a is less than unity for a step-up arrangement and V,, denotes the rms
primary value of the phase voltage and V,. is the rms secondary value of the phase
voltage. The ratio of the line-to-line voltages at the secondaries to the line-to-line
voltages at the primaries can then be expressed as

(Vity _ Var _ V3 Vp
comin )
Upon replacing V,, by its equivalence from Eq. (2-149), there results

(VeEira Vigae TN3 Vga

gr nes > (2-151)
(Vira Vas aVn2 a
Thus, if a is assigned a value of one-tenth, the line-to-line voltage at the second-
aries will be 17.32 times greater than the line-to-line voltages at the primary. The
attractive feature about this arrangement, of course, is that insulation for the
secondary windings need be provided only to the extent of a factor of 10 over that
of the primary voltage rather than 17.32. Furthermore, in view of the fact that
output volt-amperes at the secondaries of the Y must equal the input volt-amperes
Sec. 2-9 Transformers for Three-Phase Circuits 111

at the primaries of the A under the assumption of negligible losses, it follows that
the ratio of the line current at the secondaries (/;)y to the line current at the
is,
primaries (J,), for a balanced load is the reciprocal of Eq. (2-151). That

U1)y [2
= —— = Iy2
= —a 2-152
Gr oy Vii V3 pie
a

ely.
where J,; and I, denote the primary and secondary phase currents, respectiv
lly, the correspo nding transmis-
Therefore, by raising the line voltage substantia
sion line currents can be markedly reduced, which in turn can serve to improve
the efficiency of transmission appreciably.
allows a
The presence of a delta arrangement on the primary side, which
nt to flow, ensures that the line-to-
third harmonic magnetizing current compone
sinusoida l. Of course, the foregoin g
neutral voltages on the wye side will be
the source voltage that energize s the A-
statement is based on the premise that
connected primaries is itself purely sinusoidal.

Y-Y Connection

, the wye-wye connection

When equipped with a primary and secondary neutral
of two voltages at the secondary
offers two advantages: (1) it provides a choice
for an insulat ion that guards against
side; and (2) transformers need to be specified
than line-to -line voltage. Also, if
the magnitude of the line-to-phase voltage rather
second ary phase voltage is a, this
the ratio of the primary phase voltage to the
On the other hand, the ratio of the
same ratio obtains for the line-to-line voltages.
phase or line curren ts is I/a.
primary phase or line currents to the secondary nearest
phase displa ced from the
Moreover, on either side, the line voltage is
degree s.
phase voltage by a magnitude of 30 electrical
Any attemp t to operat e a Y-Y arran gemen t of transformers without the
will lead to difficulty and potential
presence of a primary neutral connection
no-load conditions. To illustrate the
failure. In fact, trouble occurs even under
is energized from a sinusoidal
point, assume that the primary wye connection
tor delive rs a sinusoidal magnetizing
three-phase generator. Since such a genera
ion curve of the transformer core Causes
current, the typical saturating magnetizat
this flat flux wave contains a large
the flux variation to be flat-topped. In turn,
an apprec iable third harmonic voltage in
third harmonic component which induces
this voltag e depen ds upon the degree of
the transformer coils. The magnitude of nic voltage is not
ces, this third harmo
saturation of the core and in most instan the A-connected
was noted to exist in
negligible. Recall that this same condition ant difference.
there is an import
primaries. But with the Y-Y connection, e was presented
the third harmo nic voltag
Whereas in the A-connected primaries not
current to flow, this situation does
with a closed path that allowed a circulating
neutral. As a matter of fact, because the
happen with the Y connection without a
112 Transformers Chap. 2

= /
ee ae Vno3 a Vnaz [0°

Vix Say 0° Figure 2-36 The line-to-neutral volt-

ae i Lo° ages caused by the third harmonic flat-
ay topped flux wave present in a Y-Y
Vae3nc3 = Vna3na3 LO” ion without
connection wi a primary neutral.

third harmonic phase voltages are each in phase with one another, they fail to
make their presence felt at the line-to-line terminals. Figure 2-36 shows the phase
and magnitudes of the third harmonic rms voltages from line-to-neutral in each
phase. Observe that their magnitudes and phase angles are identical. To obtain
the effect of these voltages between line terminals, it is necessary to perform the
phasor addition

(Vab)3 ma Van3 ate Vinb3 = —Vna3 “i Vnb3 = 0

which clearly adds to zero. The significance of this result in the generator-primary
transformer circuit is that Kirchhoff’s voltage law is satisfied despite the presence
of a flat-topped (nonsinusoidal) flux wave in the transformer core. One conse-
quence of this condition is that the resultant line-to-neutral voltage can now
readily exceed the rated transformer voltage, so that harmful insulation stresses
can take place. Of course, when a primary neutral connection is established, the
third harmonic voltage that originates in the transformer from line-to-neutral is
now provided with a path that allows the magnetizing current once again to
become peak-shaped, thereby restoring a sinusoidal flux variation. In turn, this
virtually eliminates the third harmonic voltages in the transformer coils. In those
instances where the ensuing third harmonic neutral ground currents cause tele-
phone or other related interference, it may be necessary to resort to an alternate
transformer connection involving the A arrangement on the primary or secondary
side or else to purchase transformers equipped with tertiary windings which can
then be connected in the A mode.
Another serious shortcoming of a Y-Y connection without a primary neutral
is that it makes it practically impossible to supply power to a load that is placed
from line-to-neutral on the secondary side. For energy to be delivered to the load
in such an arrangement, the generator source, which supplies the primary wind-
ings, must do so through the two unloaded primaries. However, as we learned
earlier in this chapter, an unloaded transformer presents an impedance which is
essentially the magnetizing reactance. This high impedance mitigates against the
flow of significant currents. In this regard it is interesting to note that even if a
short-circuit were placed from one line-to-neutral on the secondary, very little
current would flow. Because such a short-circuit has the effect of placing the
neutral at the potential of one line terminal, it follows that the line-to-neutral
voltages appearing across each of the unloaded transformers is increased almost
by a factor of V3. Such a high overvoltage can readily cause damage to the
Sec. 2-9 Transformers for Three-Phase Circuits 113

insulation. In conclusion, it ought to be obvious that a Y-Y connection should not

be used without the primary neutral connected to the source.

Y-A Connection

The wye-delta arrangement of single-phase transformers is often found at the

receiving end of a transmission line. The reason is that it provides a large step-
down voltage ratio from primary line voltage to secondary line voltage. If we let a
denote the ratio of the primary phase voltage (Y side) to the secondary phase
voltage (A), it follows that the ratio of the line voltage from primary to secondary
is V3 a. Correspondingly, the ratio of the primary line current to the secondary
line current is 1/V3 a.
Third harmonic voltage stresses do not exist in the wye-delta arrangement
because the closed path provided by the secondary A connection permits the third
harmonic magnetizing current to exist. In turn, this current acts to virtually elimi-
nate the third harmonic component in the flat-topped flux wave, thus ensuring a
sinusoidal flux variation. Unbalanced secondary loads, however, can present
some difficulties, as explained for the Y-Y connection, without a primary neu-
tral. To avoid problems, it is best to establish the primary neutral to the source.
Example 2-8
a loaded
A A-Y connection is used to connect a large three-phase generator to
line. The turns ratio of each single-pha se transforme r from the primary
transmission
trans-
side to the secondary side is 0.1 (see Fig. 2-37). The three-phase load on the

(Wy

(Ws i Pa
(/,) pak 240 kV |3-phase
3-phase is bs 139 kV aoe
generator

Figure 2-37 Schematic diagram for Example 2-8.

The load power factor is

mission line is 10 MVA at a line-to-line voltage of 240 kV.
0.8 lagging.
s on both the high
(a) Find the magnitude of the transformer winding current
and low sides. Assume all losses are negligib le.
d by the generator.
(b) Determine the magnitude of the line currents supplie
describ ed by
(c) If the line-to-line voltages on the high side are
kV
Vio = 240 /0° kV, Vic = 240/—120° kV, Vea = 240 /—240°
s on the A side. Assume
find the corresponding expressions of the line-to-line voltage
all leakage impedances are negligible.
114 Transformers Chap. 2

Solution (a) Working with the expression for the volt-amperes on the high side of
the transmission line we can write

(ieee ee
“V3 (Vi), V3 (240)103 “
(Ip), = Gs = * = 240 4

(b) Here,

(Iz)a = V3 Up)a = V3 (240) = 415.7 A

Or, by Eq. (2-152),
V3), V3 ¢ Hepes
Ui) = ras 0.1
(c) As per the discussion relating to Fig. 2-35(e), the corresponding line volt-
ages of the A side lead the respective line voltages on the Y side by 30°. Hence,

Vaz = 24/30° KV, Vac = 24/-90° kV, V = 24/-210° kV

PROBLEMS

2-1. A reactor coil with an iron core has 400 turns. It is connected across the 115-V, 60-
Hz power line.
(a) Neglecting the resistance voltage drop, calculate the maximum value of the
operating flux.
(b) If the flux density is not to exceed 75 kilolines/in.?, what must be the cross-
sectional area of the core?
2-2. A fixed sinusoidal voltage is applied to the circuit shown in Fig. P2-2. If the voltage
remains connected and the shaded portion of the iron removed, state what happens
to the maximum value of the flux and the maximum value of the magnetizing cur-
rent. Justify your answer. Assume negligible leakage impedance and winding resis-
tance.

Figure P2-2

2-3. Repeat Prob. 2-2 assuming the voltage applied to the core is a fixed dc quantity.
Assume dc current is limited by an external resistor.
2-4. A transformer coil rated at 200 V and having 100 turns is equipped with a 0.5 tap. If
50 V is applied to half the number of turns, find the value of the maximum flux. The
frequency is 60 Hz.
Chap. 2 Problems 115

2-5. Refer to the transformer coil of Prob. 2-4.

(a) Determine the maximum value of the operating flux when 150 V is applied across
half the number of turns.
(b) What is the required change in magnetizing current in part (a)? Neglect satura-
tion.
is
(c) Briefly describe the effect on the magnetizing current in part (b) if saturation
not neglected.
(d) What must be the frequency of the 150-V source in part (a) if the flux is to remain
unchanged?
2-6. In an open-circuit test of a 25-kVA, 2400/240-V transformer made on the low side the
corrected readings of amperes, volts, and watts are, respectively, 1.6, 240, and 114.
and
In the short-circuit test the low side is short-circuited and the current, voltage,
y, 10.4 A, 55 V, and 360 W.
power to the high side are measured to be, respectivel
(a) Find the core loss.
(b) What is the full-load copper loss?
(c) Find the efficiency for a full-load of 0.8 power factor leading.
(d) Compute the percent voltage regulation for part (c).
:
2-7. A transformer with a rating of 25 kVA, 2400/240 V has the following parameters
at rated voltage is 114 W.
Rex = 3.33 O and Z.4 = 5.28 0. The core loss
pf lagging.
(a) Find the efficiency when the transformer delivers rated kVA at 0.8
(b) What is the value of the high-side voltage in part (a)?
(c) Determine the percent voltage regulation in part (a).
resistance of
. A 50-kVA, 2300/230-V, 60-Hz transformer has a high-voltage winding
ry tests showed
0.65 Q and a low-voltage winding resistance of 0.0065 (2. Laborato
the following results:

Ve—soO0NVE P= 5.1 JA; P = 190 W

Open-circuit test:
Ve= 45 V; T= 21ST 8AS P = none used
Short-circuit test:

rated secondary voltage

(a) Compute the value of primary voltage needed to give
and is deliveri ng 50 kVA at a power
when the transformer is connected step-up
factor of 0.8 lagging.
part (a).
(b) Compute the efficiency under the condition of
ed to have a core loss of 1250 W at rated volt-
. A 100-kVA transformer was measur
was perfor med at 125% of rated current and the input power
age. A short-circuit test
ncy of this transformer when it
was measured to be 2875 W. Calculate the efficie
delivers rated kVA at a pf of 0.9 lagging.
2200/440-V, 60-Hz transformer:
2-10. The following test data were taken ona 15 kVA,

P = 620 W, I= 404, Via 25,N,

Short-circuit test:
P = 320 W, =a) AS V = 440 V
Open-circuit test:

ormer when it supplies full load at

(a) Calculate the voltage regulation of this transf
0.8 pf lagging. Neglect the magnetizing current.
(a).
(b) Find the efficiency at the load condition of part
116 Transformers Chap. 2

2-11. The following test data were taken on a 110-kVA, 4400/440-V, 60-Hz transformer:

Short-circuit test: P = 2000 W, I = 200 A, V=18V

Open-circuit test: P = 1200 W, (ees (UPd jr V = 4400 V

Calculate the voltage regulation of this transformer when it supplies rated current at
0.8 pf lagging. Neglect the magnetizing current.
2-12. A 50-kVA, 2300/230-V, 60-Hz distribution transformer takes 360 W at a power factor
of 0.4 with 2300 V applied to the high-voltage winding and the low-voltage winding
open-circuited. If this transformer has 230 V impressed on the low side with no load
on the high side, what will be the current in the low-voltage winding? Neglect
saturation.
2-13. A 25-kVA, 2400/240-V transformer draws 254 W at a pf of 0.15 when 240 V is
impressed on the low-voltage side with the high-voltage side open-circuited. Find
the current that is drawn from the line when 2400 V is applied to the high side with
the low side open-circuited.
2-14. A 10-kVA, 2400/240-V transformer draws 165 W at a pf of 0.2 when 220 V is applied
to the low side with the high side open. Find the current drawn by the high side when
2400 V is applied to the high side with the low side open. Neglect saturation.
2-15. A 200/100-V, 60-Hz transformer has an impedance of 0.3 + j0.8 0 in the 200-V
winding and an impedance of 0.1 + j0.25 in the 100-V winding. What are the
currents on the high and low side if a short-circuit occurs on the 100-V side with
200 V applied to the high side?
2-16. A 60-Hz, three-winding transformer is rated at 2300 V on the high side with a total of
300 turns. Of the two secondary windings, each designed to handle 200 kVA, one is
rated at 575 V and the other at 230 V. determine the primary current when rated
current in the 230-V winding is at unity power factor and the rated current in the
575-V winding is at 0.5 pf lagging. Neglect all leakage impedance drops and magne-
tizing current.
2-17. A transformer is equipped with three windings. The high side (primary) is rated at
2300 V and has a total of 3000 turns. Each of the low-side secondary windings has a
kVA rating of 200 and voltage ratings of 460 V and 230 V. When rated current at
unity pf flows through the 460-V winding and rated current simultaneously is made to
flow through the 230-V winding, the primary current is found to be 122.98 [45° A.
Describe the type of load that appears across the 230-V winding. Neglect all leakage
impedance and magnetizing current.
2-18. An “‘ideal’”’ transformer has a secondary winding tapped at point b. The number of
primary turns is 100 and the number of turns between a and 5 is 300 and between
b and c is 200. The transformer supplies a resistive load connected between a and
c and drawing 7.5 kW. Moreover, a load impedance of 10 [45° Q) is connected be-
tween a and b. The primary voltage is 1000 V. Find the primary current.
2-19. Draw a neat phasor diagram of a transformer operating at rated conditions. Assume
that N,/N, = 2 and

Tyr, = 0.1E,, Kee Ihr, = 0.1 Vp, I, = 0.31,

Ix, = 0.2E,, Ix, — 0.2 Vp, De = 0.17,

Chap. 2 Problems 117

Consider the load power factor to be 0.6 leading. Use Vz as the reference phasor and
show all currents and voltages drawn to scale.
2-20. A 10-kVA, 500/100-V transformer has R, = 0.3 2 and X,y = 5.2, and it is used to
supply power to a load having a lagging power factor. When supplying power to this
load an ammeter, wattmeter, and voltmeter placed in the high-side circuit read as
follows:

I, = 20 A, V, = 500 V, WM
= 8 kW

For this condition calculate what a voltmeter would read if placed across the second-
ary load terminals. Assume the magnetizing current to be negligibly small.
2-21. A 10-kVA, 460/115-V transformer has a high-side winding resistance of 0.4 9 and a
low-side winding resistance of 0.02 . The equivalent leakage reactance on the high
side is 3.2 Q. When this transformer is used to supply power to a passive load, it
draws a lagging current of 21.7 A at 460 V and 8 kW. Determine the resistive and
reactive components of the load impedance. Neglect the magnetizing current.
2-22. The following data are taken on a 30-kVA, 2400/240-V, 60-Hz transformer:

Short-circuit test: V=70V, T= 18.8 A, P = 1050 W

Open-circuit test: V = 240 V, I=3.0A, P = 230 W

the low-
(a) Determine the primary voltage when 12.5 A at 240 V is taken from
voltage side supplying a load of 0.8 pf lagging.
(b) Compute the efficiency in part (a).
2-23. A 30-kVA, 240/120-V, 60-Hz transformer has the following data:
r, = 0.140, r, = 0.035 0
x, = 0.22 0, x, = 0.055 0

e to the primary
It is desired to have the primary induced emf equal in magnitud
the transfor mer carries the full-load current. Neglect the
terminal voltage when
ing current. How must the transfor mer be loaded to achieve this result?
magnetiz
have resistances of
. The two windings of a 48-kVA, 2400/240-V, 60-Hz transformer
0.025 © for the high- and low-volt age windings , respectiv ely. This trans-
0.6 Q and
tage coil in order that rated
former requires that 238 V be impressed on the high-vol
low-vol tage winding .
current be circulated in the short-circuited
the high side.
(a) Calculate the equivalent leakage reactance referred to
power is needed to circulat e rated current on short-circuit?
(b) How much
factor is 0.8 lagging. Assume
(c) Compute the efficiency at full load when the power
that the core loss equals the copper loss.
turns, and is rated for 200 V.
2-25. A transformer coil is equipped with a 0.5 tap, has 100
, the magneti zing current is 1 A. If
When this voltage is applied to the full winding
the number of turns, what will be the magnetizing
now 50 V is applied to half
current? Assume no satura tion.
rmer with a rating of 50 kVA,
2-26. A short-circuit test is to be performed on a transfo
de and low-sid e resista nces are known to be 0.06 © and
460/115 V. The high-si
short-circuit conditions is 0.3.
0.004 Q, respectively. Moreover, the power factor at
118 Transformers Chap. 2

Compute the input voltage that is required to force rated current through the low-side
coil at short-circuit.
2-27. A coil located on an iron core has a constant 60-Hz voltage applied to it. The total
inductance is found to be 10 H. A second identical core is placed in parallel with the
first one and is so wound that it produces aiding flux. What is the total inductance of
the parallel combination? Neglect the winding resistance and leakage flux of each
winding.
2-28. Repeat Prob. 2-27 for the case where the second identical coil is placed in series with
the first coil. Assume that aiding flux is produced.
2-29. Show the derivation that allows Eq. (2-68) to be written from Eq. (2-67) for a sinu-
soidal forcing function and no saturation.
2-30. A two-winding transformer has a high-voltage coil equipped with 1000 turns and a
low-voltage coil equipped with 100 turns. The self-inductance of the high voltage
primary is known to be 20 H. Moreover, the primary winding has a leakage induc-
tance of 1 H. The secondary leakage inductance is 0.01 H.
(a) Find the mutual inductance in henrys.
(b) Determine the value in henrys of the self-inductance of the secondary winding.
2-31. A conventional transformer with a rating of 10 kVA, 2000/1000 V is available.
(a) Show how to arrange this transformer for use as a step-down autotransformer
with a secondary voltage of 2000 V.
(b) For a 0.8 pf lagging load, what is the maximum current that can be delivered
without overloading the transformer coils?
(c) At the load conditions of part (b), express the conducted kVA of the autotrans-
former as a percentage of the rating of the conventional transformer.
(d) List the advantages which the autotransformer has over an ordinary transformer
of equal kVA and voltage rating for doing the same job.
2-32. A 10-kVA, 2300/230-V transformer has an iron loss of 190 W and a copper loss of 196
W at rated voltage and rated current. The secondary winding is made up of two
equal sections, each section series-connected for 230-V operation. It is desired to
reconnect this conventional transformer as an autotransformer in order to raise the
2300-V supply line to 2415 V and to deliver maximum power without overheating the
coils. Using all windings of the transformer, determine for rated current:
(a) kVA delivered to the load.
(b) kVA conducted.
(c) Efficiency for a load pf of 0.8 leading.
(d) The rating (voltage and kVA) of an ordinary transformer to do the same job.
What are the advantages of using an autotransformer?
2-33. The following data were found during a short-circuit test on an autotransformer rated
at 22 kVA, 440/330 V. With the low side short-circuited and voltage applied to the
high side, the readings were:

P = 320 W, V=20V, I=50A

(a) Find the voltage regulation of this autotransformer at full-load output and unity
power factor.
Chap. 2 Problems 119

(b) If the noncommon portion of the winding has a resistance of 0.064 ©, determine
the resistance in ohms of the common portion.
(c) Find the transformed and conducted power for full-load output at unity power
factor.
(d) What must be the rating (kVA and voltages) of an ordinary transformer that can
be used to do the same job?
2-34. The expression for the equivalent resistance of an autotransformer looking into the
high-side terminals is given by Eq. (2-107). Derive the expression for the equivalent
resistance of the autotransformer looking into the low-side terminals, with the high
side short-circuited.
2-35. When the transformer of Prob. 2-15 is used as a step-up autotransformer (200/300),
what are the line currents on the high and low sides if a short circuit occurs on the
300-V side with 200 V applied to the low side?
2-36. An autotransformer has N,- = 100 turns and N,, = 60 turns, where N, represents
the full winding. A short-circuit test is performed by short-circuiting winding ab and
applying a reduced voltage to winding bc. The equivalent impedance looking into
winding bc is found to be: 1.5 + j4.5.
(a) Determine the equivalent impedance looking into winding ac for the condition
where winding bc is short-circuited.
(b) Find the equivalent resistance looking into bc when ac is short-circuited and
voltage is applied to bc.
2-37. A single-phase autotransformer rated at 40 kVA supplies a load impedance of 4 [36.9°
Q at a terminal voltage of 200 V from a 125-V supply. All power losses and leakage
common
reactances are negligible. Calculate the magnitudes of the currents in the
taking into account the magnetiz-
and noncommon parts of the transformer winding,
ing current of 0.075 per unit.
booster
2-38. A 25-kVA, 2400/240-V single-phase transformer is connected to act as a
the maximum kVA load that it is possible to supply
from 2400 to 2640 V. With
without overloading the transformer coils, determine:
(a) kVA received and delivered.
(b) kVA transformed.
(c) kVA conducted.
e.
Neglect all losses, magnetizing current, and leakage impedanc
where the secondar y coil is reversed so that the output
2-39. Repeat Prob. 2-38 for the case
(load) voltage is 2160 V.
W when operating at
2-40. A transformer rated at 40 kVA has a total copper loss of 250
Determi ne the per-unit value of the equivale nt resistance.
50% of rated current.
2-41. Refer to the configuration of Fig. P2-41.
ng in the positive
(a) Assume that the current that enters terminal a is increasi
n. Identify the instanta neous polarity marking s of each terminal.
directio
e load that draws 20 A is
(b) The coils have the number of turns indicated. A resistiv
placed across terminals c and d. A purely capaciti ve load drawing 40 A is placed
current. Neglect all leakage impedan ces and magne-
across ef. Find the source
tizing current.
 120 Transformers Chap. 2

(c) Disregard winding ef. Assume winding ab has a voltage rating of 400 V and a
current rating of 50 A. A short-circuit test on this transformer yields the follow-
ing data:

WM = 600 W, VM = SOV, AM
= 20A
Find the voltage regulation at unity pf.

Figure P2-41

2-42. The high-side and low-side winding resistances of a 100-kKVA, 440/110-

V single-
phase transformer are 0.04 2 and 0.0025 0, respectively.
(a) Find the per-unit values of the high-voltage and low-voltage winding
resistances.
(b) If the core loss at rated voltage is 1200 W, find the transformer
efficiency when it
delivers half-rated kVA at unity power factor.
2-43. A transformer rated at 2500 kVA, 10,000/2000 V, 60 Hz is designed
for maximum
efficiency at 80% of rated load. The per-unit equivalent impedan
ce of this trans-
former is 0.02 + j0.06. For a resistive load and operation at
maximum efficiency,
find the losses and the change in voltage from 80% load (at
rated voltage) to no load.
2-44, A transformer rated at 1100 kVA has a short-circuit test conduct
ed at a current equal
to 75% of rated value, and takes an amount of power equal
to 7030 W. An open-
circuit test is also conducted and at rated voltage the input power
is measured to be
9850 W.
(a) Find the per-unit value of the equivalent resistance.
(b) Find the per-unit value of the conductance.
2-45. If the per-unit value of the equivalent resistance of the
conventional transformer of
Prob. 2-31 is 0.02, find the value of the equivalent
resistance of the autotransformer
in ohms referred to the high side.
2-46. Describe the predominate waveshape of the primar
y magnetizing current of a Y-A
three-phase transformer connection employing single-
phase transformers.
A three-phase transformer bank is connected with
the primaries in delta and the
2-47.

secondaries in wye. The line voltage on the primary

side is 120 V and on the second-
ary side 240 V. Find the ratio of primary to secondary
turns for each of the single-
phase transformers.
Chap. 2 Problems 121

2-48. Three single-phase transformers with ratings of 2400/120 V are used to step down a
distribution line voltage from 4160 V to a lower value.
(a) Assuming the use of a Y-A arrangement, find the value of the secondary A-side
voltage.
(b) If the maximum value of the magnetic flux puts operation beyond the knee of the
magnetization curve, does a third harmonic voltage exist between the line and
neutral terminals on the Y-side? Explain.
2-49, Refer to the arrangement shown in Fig. P2-49. The single-phase transformers are
rated for 2400/120 V. Consider that the three coils for the secondary delta connec-
tion are properly joined with regard to the fundamental induced voltage component.
Assume further that the magnitude of the third harmonic induced voltage is one-third
that of the fundamental.
(a) Find the gap voltage when a sinusoidal voltage of 4160 V rms is applied to the
line terminals of the wye.
(b) Calculate the line-to-neutral voltage at the Y connection.
(c) When the gap is closed, describe the role played by the voltage found in part (a)
in the operation of these transformers. Be specific.

Figure P2-49

2-50. A Y-Y connection of transformers consists of single-phase transformers which carry

a rating of 69/6.9 kV. A sinusoidal transmission line voltage of 120 kV is applied to
the high side. The third harmonic voltage, when it exists, is known to be one-third
the value of the fundamental voltage. Neither the high nor the low side is provided
with a ground neutral.
(a) Determine the rms value of the line-to-neutral voltage on each side.
voltage
(b) What recommendation can you make to eliminate the third harmonic
from the line-to- neutral voltage?
 3
Fundamentals of
Electromechanical
Energy Conversion

Electromechanical energy conversion involves the interchange of energy between

an electrical system and a mechanical system through the medium of a coupling
magnetic field. The process is essentially reversible except for a small amount
which is lost as heat energy. When the conversion takes place from electrical to
mechanical form the device is called a motor. When mechanical energy is con-
verted foeléctrieal energy the device is called a cenehator. MGieoveT-Wherthe_
electrical system is characterized by alternating current the devices are referred to
as ac motors and ac generators, respectively. Similarly, when the electrical sys-
tem is characterized by direct current the electromechanical conversion devices
are called dc motors and dc generators.
The same fundamental principles underlie the operation of both ac and dc
machines, and they are governed by the same basic laws. Thus, in the computa-
tion of the developed torque of an electromechanical energy-conversion device,
one basic torque formula (which derives directly from Ampere’s law) applies
whether the machine is ac or dc. The ultimate forms of the torque equations
appear different for the two types of machines only because the details of mechan-
ical construction differ. In other words, starting with the same basic principles for
the production of electromagnetic torque, the final forms of the torque equations
differ to the extent that the mechanical details differ. These comments apply
equally in generating emf in the armature winding of a machine whether it is ac or

122
 Sec. 3-1 Electromagnetic Torque 123

dc. Once again a simple basic relationship (Faraday’s law) governs the voltage
induced. The final forms of the voltage equations differ merely as a reflection of
the differences in the construction of the machines. It is important that this is
understood at the outset, because a prime objective of the treatment of the subject
matter as it unfolds in this chapter is to impress upon the reader the fact that ac
machines are not fundamentally different from dc machines. They differ merely
in construction details; the underlying principles are the same.
This chapter is arranged to give still greater emphasis to this theme. Atten-
tion first is given to a basic analysis of the production of electromagnetic torque as
well as the generation of voltages starting with the classical fundamental laws.
Then the construction features of the various types of electric machines are exam-
ined with the purpose of indicating how the conditions for the production of
torque and voltage are fulfilled and also to point out why differences are neces-
sary. Once this background is established, the fundamental equations for devel-
oped torque and induced voltage are modified to make these expressions consis-
tent with the particular construction features of the machines involved. In this
way the equations are put in a more useful form for purposes of analysis and
design. Attention is next directed in the following chapters to an analysis of each
found
of the major types of electromechanical energy-conversion devices to be
industry and in homes, shops, and of-
performing innumerable tasks throughout
fices everywhere.

3-1 ELECTROMAGNETIC TORQUE

of energy
Since electromechanical energy conversion involves the interchange
primary quantiti es involved in
between an electrical and a mechanical system, the
the analogou s quantiti es in the
the mechanical system are torque and speed, while
3-1 represents the
electrical system are current and voltage respectively. Figure
when the electrical system
situation diagrammatically. Motor action results
are placed in a magnetic field.
causes a current i to flow through conductors that
so that, if the conductors
Then by Eq. (1-2) a force is produced on each conductor
an electrom agnetic torque T
are located on a structure that is free to rotate,
velocity w,,. Moreover, with
results, which in turn manifests itself as an angular
conducto rs cut through the
the development of this motor action the revolving
is really a reaction voltage
magnetic field and thereby undergo an emf e, which

Electrical Coupling Mechanical

system field system

Pe e Figure 3-1 Block representation of

<< electromechanical energy conversion.
124 Fundamentals of Electromechanical Energy Conversion Chap. 3

analogous to the induced emf in the primary winding of a transformer. Note that
the coupling field is involved in establishing the electromagnetic torque T as well
as the reaction induced emf e. In the case of generator action the reverse process
takes place. Here the rotating member—the rotor—is driven by a prime mover
(steam turbine, gasoline engine, and so on), causing an induced voltage e to
appear across the armature winding terminals. Upon the application of an electri-
cal load to these terminals, a current i is made to flow, delivering electrical power
to the load. Of course the flow of this current through the armature conductors
interacts with the magnetic field to produce a reaction torque opposing the applied
torque originating from the prime mover.
In the interest of emphasizing further the character of this motor and genera-
tor action and its relationship to the coupling magnetic field, a mathematical
description is now undertaken. Consider that we have an electromechanical en-
ergy-conversion device which consists of a stationary member—the stator—and
the rotating member, the rotor. Refer to Fig. 3-2(a).
Assume that the stator is equipped with a coil 1-2, which is energized to
cause a constant current to flow in the direction shown by the dot-cross notation.
By Flemmings’ right-hand rule, the flux lines that are produced by the coil am-
pere-turns leave the stator from the upper half, cross their air gap, penetrate the
rotor iron, and then once again cross the air gap to form the closed path illustrated
in Fig. 3-2(a). If the rotor and stator surfaces were to be cut and laid out in a flat
manner, the resulting configuration would appear as shown in Fig. 3-2(b). If the
reluctance of the iron is neglected, half the coil mmf is used for each crossing of
the air gap. Accordingly, the value of the flux density that appears between coil
sides 1-2 in Fig. 3-2(b) is one-half the coil mmf divided by the air-gap reluctance
and the cross-sectional area per pole. Since these quantities are fixed, it follows
that the flux density, too, is constant from coil side 1 to 2 and 2 to 1. This situation
is depicted in Fig. 3-2(c) in a manner that emphasizes that the flux that leaves the
stator (in the upper section between coil sides 1-2) behaves like a north-pole flux,
while the flux that enters the stator (in the lower section between coil sides 2-1)
behaves like a south-pole flux.
With the uniformly distributed field now established, the focus of attention is
shifted to the rotor, which is assumed to be equipped with a total of Z conductors,
only two of which (a and b) are shown in Fig. 3-2(a). It is further assumed that
conductor a is joined to conductor b to form a coil in the fashion illustrated in Fig.
3-2(d). Because the principle of electromechanical energy conversion is con-
cerned with the movement of a coil relative to a magnetic field, the analysis now
proceeds from the arrangement shown in Fig. 3-2(e), where only the presence of
the field density B is given importance and not its source. It is helpful to observe
that initially the flux that links rotor coil a-b is zero. In Fig. 3-2(e), this situation is
represented by noting that coil a-b links one-half of the positive (north-pole) air-
gap flux in addition to one-half of the negative (south-pole) air-gap flux for a
resultant of zero. By Faraday’s law, displacement of the coil by a differential
Sec. 3-1 Electromagnetic Torque 125

feed ESI vital.

GON
telatLaid J seesAiteor
Pye ofeinweic
k
Fenality

Rotor surface

(e)
(d)
mechanical en-
Figure 3-2 Illustrating some construction features of an electro
a plan view; (c) a uniform airgap
ergy conversion device: (a) stator and rotor; (b)
h coil with axial length /;
flux density produced by coil 1-2; (d) a full-pitc
coil a-b with displac ement dx.
(e) changing flux linkage of

in the flux linking the coil

distance dx to the new position a’b’ causes a change
that is expressed by
db =2BdA G:1)
value of the flux density at the
The factor 2 accounts for the two coil sides; B is the
dA is the differe ntial area through which
points where conductors a and b lie, and
126 Fundamentals of Electromechanical Energy Conversion Chap. 3

the B-field penetrates. A little thought reveals this area to be equal to the axial
length of the rotor, /, times dx. Thus
dA = 1 dx (3-2)

However, if it is further assumed that dx results from a linear velocity v imparted

to the conductor for the differential time dr, it follows that

dx =vudt (3-3)
Insertion of Eqs. (3-2) and (3-3) into Eq. (3-1) then leads to
dod = 2Blv dt (3-4)

The corresponding emf induced in coil ab becomes, by Faraday’s law,

aia_ db _Phiag 2Blv (3-5)

Because the coil a-b is assumed driven from left to right, the 0 x B rule
indicates that the direction of the induced emf is into the paper for coil side a’ and
out of the paper for coil side b’. The magnetic field is considered to be directed
downward on side a’ and upward on side b’. Moreover, the application of the I x
B rule discloses that a developed force exists on sides a’ and b’ in association with
a coil current J that is directed from right to left. This produces a torque that
opposes the torque that drives the coil.
The factor 2 appears in Eq. (3-5) because the coil involves two conductors.
It then follows that, if the rotor is equipped with a total of Z conductors, which are
all series-connected, the total induced voltage can be expressed as
e = ZBlu (3-6)

When a current i flows through conductors a and b in the configuration of

Fig. 3-2(e), the direction of i and the direction of B are 90° apart. Accordingly, a
force exists on each conductor which is given by Eq. (1-2). Thus
Fe= Bli (3-7)
where F,. denotes the force developed on an individual conductor. However,
since there are now Z conductors on the rotor surface, it follows that the total
developed force ist
F = ZBli (3-8)
Moreover, if this force is made to act through a moment arm r, which is the radius
of the rotor, the resulting developed torque can be expressed as
T = Fr = ZBlri (3-9)

+This equation is valid provided that the conductors are paired to form series-connected coils
that have a span of one pole pitch. A pole pitch is equal to 180 electrical degrees.
Sec. 3-1 Electromagnetic Torque 127

A glance at Eqs. (3-6) and (3-9) shows that both the induced voltage and the
developed electromagnetic torque are dependent upon the coupling magnetic
field. The factor that determines whether the induced voltage e and the developed
torque T are actions or reactions depends upon whether generator or motor action
is being developed.
The two mechanical quantities T and w,, and the two electrical quantities e
and i, which are coupled through the medium of the magnetic field B, are related
further by the law of conservation of energy. This is readily demonstrated by
dividing Eq. (3-6) by Eq. (3-9). Thus
€ 27 Biv Wm
ie wane eal
where

Om = : mechanical angular velocity of the rotor (3-11)

Rearranging Eq. (3-10) leads to

ei = Tom (3-12)
me-
which states that the developed electrical power is equal to the developed
chanical power. This statement is valid for generator as well as motor action.
Finally, on the basis of the foregoing discussion, it should be apparent that a
com-
conventional electromechanical energy-conversion device must involve two
ponents. One is the field winding, which by definition is that part of the machine
which by
which produces the coupling field B. The other is the armature winding,
’’ emf e and the current
definition is that part of the machine in which the ‘‘working
L exist.

Example 3-1
of height 0.4 T. It
A two-pole electric machine has a rectangular flux-density wave
a radius of r= 0.1m. The rotor is
also has an axial length of / = 0.2m and a rotor with
turns. A current of 5 A is supplied to the
equipped with a single coil of 20 full-pitch
coil.
the coil.
(a) Find the maximum flux in webers that can be made to link
ed torque in N-m.
(b) When the coil carries 5 A, find the develop
a speed of 900 rpm,
(c) Assuming that the torque produced in part (b) results in
determine the coil induced emf.
re.
(d) Obtain the result in part (c) by an alternative procedu
(e) What is the mechanical power developed in watts?

is in the position relative the

Solution (a) The maximum flux occurs when the coil
Fig. 3-2(c). The area is that associated with one-half
flux density wave illustrated in
the rotor diameter. Thus

@ = BA = Barl = 0.47(0.1)(0.2) = 0.025133 Wb

 Chap. 3
128 Fundamentals of Electromechanical Energy Conversion

(b) By Eq. (3-9)

T = ZBiri = 2NBlri = 40(0.4)(0.2)(0.1)5 = 1.6 N-m

(c) From the principle of conservation of energy

ei = Tam

where

Om = at = ().10472(900) = 94.25 rad/s

at Tené wooo Berea y

(d) Invoking Eq. (3-9) we get

e = ZBlv = 2NBlv = 40(0.08)(0.1)a,r = 0.32(94.25)0.1 = 30.16 V

(e) Two alternative expressions provide the result.

To, Il 1.6(94.25) = 150.8 W

or

ei = 5(30.16) 150.8 W

Important note. The foregoing notions regarding the development of

electromagne tic torque can be applied in a general way to the actual field and
ampere-cond uctor distributions that are found in real ac and dc machines. This
treatment appears at the end of the book in the Appendix to Chapter 3. The
principal results are then invoked in Sec. 3-4 as the starting point in deriving more
readily useful forms of the torque formulas for these practical ac and de machines.

3-2 INDUCED VOLTAGES

The electromechanical energy-conversion process involves the interaction of two

fundamental quantities which are related through the law of conservation of en-
ergy and the coupling magnetic field. These are, of course, electromagnetic
torque and induced voltage. The manner in which electromagnetic torque is de-
veloped is described in the preceding section. Here we consider the generation of
voltage in the armature winding brought about by changing flux linkages. The
polarity of this voltage depends upon the direction of the relative motion of the
winding with respect to the magnetic flux and is independent of whether the
machine is a motor or a generator, i.e., the direction of energy flow.
Consider the configuration shown in Fig. 3-3. The flux-density curve is as-
sumed to be sinusoidally distributed along the air gap of the machine, and the
armature winding is assumed to consist of a full-pitch coil having N turns. A full-
pitch coil is one that spans 7 electrical radians. Just as Ampere’s law serves as
Sec. 3-2 Induced Voltages 129

Figure 3-3 Configuration for derivation

of the induced-emf equation.

the starting point in the development of the torque relationships, so Faraday’s law
serves as the starting point in establishing the induced-voltage relationships.
Equation (3-5) is not the result we are seeking, because it applies when the flux-
density curve is uniform over the pole pitch. It is more generally useful to deal
with a sinusoidal distribution of the flux density, as explained more fully later.
Since the energy-conversion process involves relative motion between the field
and the armature winding, we know that the flux linking the coil changes from a
positive maximum to zero to a negative maximum and back to zero again. This
cycle repeats every time the coil passes through a pair of flux-density poles. It
should be clear from Fig. 3-3 that maximum flux penetrates the coil when it is in a
position corresponding to 8 = 0° or 8 = 180° or multiples thereof. Minimum flux
linkage occurs when B = 90° or 270° or multiples thereof. It should be apparent
too that the maximum flux that links the coil is the same as the flux per pole, which
for sinusoidal distributions is given by Eq. (3-64) of Appendix to Chapter 3.
Moreover, as the coil moves relative to the field, it follows that the instantaneous
flux linking the coil may be expressed as

6 = @cos B = 5Bn cos B (3-13)

The cos f factor is included because of the sinusoidal distribution. Furthermore,

since 8 increases with time in accordance with
B= owt (3-14)
the
where @ is expressed in electrical radians per second, the expression for
instantaneous flux linkage becomes

d = :B,,Ircos wt = ® cos wt (3-15)

voltage
Then by Faraday’s law the expression for the instantaneous induced
becomes

e=
dd
rhb = wN® sin wt (3-16)
130 Fundamentals of Electromechanical Energy Conversion Chap. 3

It is important to keep in mind that ® represents the maximum flux, which is the
same as the flux per pole.
Equation (3-16) is applicable to dc as well as ac machines. Ina later section
the induced emf per phase in an ac machine and the induced voltage appearing
between brushes in a dc machine are derived starting with Eq. (3-16). The fact
that in the dc machine the flux-density curve is nonsinusoidal (as Fig. 3-16 of
Appendix to Chapter 3 shows) is relatively unimportant; what is of primary impor-
tance is the fundamental component of the nonsinusoidal distribution. It is for
this reason that Eq. (3-16) is derived for sinusoidal distributions. Recall that by
means of the Fourier series any nonsinusoidal periodic wave may be expressed
entirely in terms of sinusoids.
Example 3-2
The purpose of this example is to draw a distinction in the character of voltage that is
induced in a revolving coil when it finds itself under the influence of a vertically
directed uniform field as illustrated in Fig. 3-4(a) and a radially directed uniform field
as occurs in the configuration of Fig. 3-4(b). Assume that coil AA’ is square and
measures 30 cm on a side with a total of 40 turns. The coil is driven at a constant
speed of 200 rad/s. The value of the uniform magnetic field in both cases is 0.1 T.
(a) Obtain an expression for the flux linkage through the coil as a function of
time.
(b) Find the expression for the voltage induced in the coil as a function of time
using Faraday’s law.
(c) Determine the result found in part (b) by employing the Blu concept using
Eq. (3-5).
(d) Now assume that coil AA is put under the influence of a radially directed
field as shown in Fig. 3-4(b). Coil CC’ is assumed to carry a fixed unidirectional

B field silly,
(a) : (b)

Figure 3-4 Calculation of induced voltages: (a) coil AA’ operating in a vertically
directed field; (b) coil AA’ operating in a radially directed field.
Sec. 3-2 Induced Voltages 131

current, thereby producing a constant field distribution over the entire periphery of
the air gap. Find the coil emf in this case as a function of time.

Solution (a) Let time be counted starting with the coil in the horizontal position in
Fig. 3-4(a). As time elapses, the coil, which is rotating, takes a new position dis-
placed by @ = wf degrees from its reference position. In this new position, the
component of the vertical field that threads the coil in a quadrature fashion is
B cos 6 = B cos wt

The corresponding amount of flux that links with the coil is described by
BA cos 6 = BA cos ot
or
dm COS 0 = hm COS wt = b
linkage is
where ¢,, denotes the maximum flux that links with the coil. But flux
simply this flux multiplied by the number of turns in the coil. Thus

h = Nob = Nop, cos ot = 40(0.1)(0.3)(0.3) cos wt

= 0.36 cos 200 Wb-t
the rate of
(b) By Faraday’s law, we know that the induced emf is equal to
change of flux linkage with time. Hence

dy d
e= = ee as 7 AU. Sb 0s 01) (0.36)(200) sin 200

= 72 sin 200t Vv

This result states that the emf varies sinusoidally with time.
lines by the
(c) With the Blu approach it is the cutting action of the flux
is in the horizonta l position, the
conductors that receives the focus. When the coil
of the B field, so the instanta neous value of
conductors fail to be under the influence
coil sides are not cutting the flux lines. It is importan t to
the voltage is zero. The
r is collinear with the
note that at this position, the velocity vector of the conducto
the field lines is zero and so no
field. Hence the component that is in quadrature with
elapse and the coil takes a position 4 degrees
cutting occurs. As time is allowed to
e point, the quadratu re compone nt of the velocity vector
from its horizontal referenc
as can be seen from Fig.
is no longer zero. Rather, it is equal to v sin @ = v sin wt,
vector then permits the in-
3-4(a). Employing only the useful part of the velocity
duced emf to be written as
e = 2NBlv sin wt = 2NBlor sin wt

= 2(40)(0.1)(0.3)(200)(0.15) sin wt

= 72 sin wt = 72 sin 200t

of Fig. 3-4(a), the use ofa
(d) Unlike the situation that occurs in the geometry
is at all times in a position of
radially directed field means that the velocity vector
induced in this case is
quadrature with the B-field. Hence the voltage
= 72 V
e = 2NBlu = 2NBlor = 2(40)(0. 1)(0.3)(200)(0.15)
 132 Fundamentals of Electromechanical Energy Conversion Chap. 3

This quantity is constant in magnitude with time and specifically it is equal to the
peak value found in part (c). Another point of interest associated with the geometry
of Fig. 3-4(b) compared to that of Fig. 3-4(a) is that flux linkage in the case of Fig.
3-4(b) is not described by the cosine function but rather in the manner already
described in Sec. 3-1.

3-3 CONSTRUCTION FEATURES OF ELECTRIC MACHINES

The energy-conversion process usually involves the presence of two important

features in a given electromechanical device. These are the field winding, which
produces the flux density, and the armature winding, in which the working emf is
induced. In this section the salient construction features of the principal types of
electric machines are described to show the location of these windings, as well as
to demonstrate the general composition of such machines.

The Three-Phase Induction Motor

This is one of the most rugged and most widely used machines in industry. Its
stator is composed of laminations of high-grade sheet steel. The inner surface is
slotted to accommodate a three-phase winding. In Fig. 3-5(a) the three-phase
winding is represented by three coils, the axes of which are 120 electrical degrees
apart. Coil aa’ represents all the coils assigned to phase a for one pair of poles.
Similarly, coil bb’ represents phase b coils, and coil cc’ represents phase c coils.
When one end of each phase is commonly connected, as depicted in Fig. 3-5(b),
the three-phase stator winding is said to be Y-connected. Such a winding is called
a three-phase winding because the voltages induced in each of the three phases by
a revolving flux-density field are out of phase by 120 electrical degrees—a distin-
guishing characteristic of a symmetrical three-phase system.
The rotor also consists of laminations of slotted ferromagnetic material, but
the rotor winding may be either the squirrel-cage type or the wound-rotor type.
The latter is of a form similar to that of the stator winding. The winding terminals
are brought out to three slip rings as depicted in Fig. 3-6(a). This allows an

Figure 3-5 Three-phase induction

motor: (a) showing a stator with three-
phase winding and squirrel-cage rotor;
(b) schematic representation of a three-
phase Y-connected stator winding.
Sec. 3-3 Construction Features of Electric Machines 133

external three-phase resistor to be connected to the rotor winding for the purpose
of providing speed control. As a matter of fact it is the need for speed control
which in large measure accounts for the use of the wound-rotor type induction
motor. Otherwise, the squirrel-cage induction motor would be used. The squir-
rel-cage winding consists merely of a number of copper bars imbedded in the rotor
slots and connected at both ends by means of copper end rings as depicted in Fig.

(a)

(b)

(b) showing details

Figure 3-6 (a) Wound rotor for three-phase induction motor;
of rotor bars of squirrel-cage induction motor.
134 Fundamentals of Electromechanical Energy Conversion Chap. 3

3-6(b). (In some of the smaller sizes aluminum is used.) The squirrel-cage con-
struction is not only simpler and more economical than the wound-rotor type but
more rugged as well. There are no slip rings or carbon brushes to be bothered
with.
In normal operation a three-phase voltage is applied to the stator winding at
points a-b-c in Fig. 3-5. As described in Sec. 4-1, magnetizing currents flow in each
phase which together create a revolving magnetic field having two poles. The
speed of the field is fixed by the frequency of the magnetizing currents and the
number of poles for which the stator winding is designed. Figure 3-5 shows the
configuration for two poles. Ifthe pattern a-c'-b-a'-c-b' is made to span only 180
mechanical degrees and then is repeated over the remaining 180 mechanical de-
grees, a machine having a four-pole field distribution results. For a p-pole ma-
chine the basic winding pattern must be repeated p/2 times within the circumfer-
ence of the inner surface of the stator.
The revolving field produced by the stator winding cuts the rotor conduc-
tors, thereby inducing voltages. Since the rotor winding is short-circuited by the
end rings, the induced voltages cause currents to flow which in turn react with the
field to produce electromagnetic torque—and so motor action results.
Accordingly, on the basis of the foregoing description, it should be clear that
for the three-phase induction motor the field winding is located on the stator and
the armature winding on the rotor. Another point worth noting is that this ma-
chine is singly excited, i.e., electrical power is applied only to the stator winding.
Current flows through the rotor winding by induction. As a consequence both the
magnetizing current, which sets up the magnetic field, and the power current,
which allows energy to be delivered to the shaft load, flow through the stator
winding. For this reason, and in the interest of keeping the magnetizing current as
small as possible in order that the power component may be correspondingly
larger for a given rating, the air gap of induction motors is made as small as
mechanical clearance will allow. The air-gap lengths vary from about 0.02 in. for
smaller machines to 0.05 in. for machines of higher rating and speed.

Synchronous Machines

The essential construction features of the synchronous machine are depicted in

Fig. 3-7. The stator consists of a stator frame, a slotted stator core, which pro-
vides a low-reluctance path for the magnetic flux, and a three-phase winding
imbedded in the slots. Note that the basic two-pole pattern of Fig. 3-5S(a) is re-
peated twice, indicating that the three-phase winding is designed for four poles.
The rotor either is cylindrical and equipped with a distributed winding or else has
salient poles with a coil wound on each leg as can be seen in Fig. 3-7(b). The
cylindrical construction is used almost exclusively for turbogenerators which op-
erate at high speeds. On the other hand, the salient-pole construction is used
exclusively for synchronous motors operating at speeds of 1800 rpm or less.
When operated as a generator the synchronous machine receives mechani-
Sec. 3-3 Construction Features of Electric Machines 135

(b)

Figure 3-7 Salient-pole synchronous machine: (a) schematic diagram; (b) photo-
graph showing field coils wrapped around salient poles. Note too the squirrel-
cage winding embedded in the pole faces.

is driven at some fixed

cal energy from a prime mover such as a steam turbine and
ed from a de source, thereby furnishing a
speed. Also, the rotor winding is energiz
the rotor is at standsti ll and direct
field distribution along the air gap. When
no voltage is induced in the stator wind-
current flows through the rotor winding,
coils. Howeve r, when the rotor is
ing because the flux is not cutting the stator
in the stator windin g and upon
being driven at full speed, voltage is induced
can be delive red to it.
application of a suitable load electrical energy
136 Fundamentals of Electromechanical Energy Conversion Chap. 3

A little thought should make it apparent that for the synchronous machine
the field winding is located on the rotor; the armature winding is located on the
stator. This conclusion is valid even when the synchronous machine operates as a
motor. In this mode ac power is applied to the stator winding and dc power is
applied to the rotor winding for the purpose of energizing the field poles. Mechan-
ical energy is then taken from the shaft. Note, too, that unlike the induction
motor, the synchronous motor is a doubly fed machine. In fact it is this character-
istic which enables this machine to develop a nonzero torque at only one speed—
hence the name ‘‘synchronous.”’ This matter is discussed further in Chapter 6.
Because the magnetizing current for the synchronous machine originates
from a separate source (the dc supply or an exciter generator such as the one in
the leftmost part of Fig. 3-7(b)), the air-gap lengths are larger than those found in
induction motors of comparable size and rating. However, synchronous ma-
chines are more expensive and less rugged than induction motors in the smaller
horsepower ratings because the rotor must be equipped with slip rings and brushes
to allow the direct current to be conducted to the field winding.

Direct-Current Machines

Electromechanical energy-conversion devices that are characterized by direct

current are more complicated than the ac type. In addition to a field winding and
armature winding, a third component is needed to serve the function of converting
the induced ac armature voltage into a direct voltage. Basically the device is a
mechanical rectifier and is called a commutator.

Typical flux path

Field winding coil

section
(flat portion of f ; ,
each turn) Figure 3-8 Construction features of
the dc machine, showing the Gramme-
ring armature winding.
Sec. 3-3 Construction Features of Electric Machines 137

Appearing in Fig. 3-8 are the principal features of the dc machine. The
stator consists of a laminated ferromagnetic material equipped with a protruding
structure around which coils are wrapped. The flow of direct current through the
coils establishes a magnetic field distribution along the periphery of the air gap in
much the same manner as occurs in the rotor of the synchronous machine. Hence
in the dc machine the field winding is located on the stator. It follows then that the
armature winding is on the rotor. The rotor is composed of a laminated core,
which is slotted to accommodate the armature winding. It also contains the com-
mutator—a series of copper segments insulated from one another and arranged in
cylindrical fashion as shown in Fig. 3-9(a). Riding on the commutator are appro-
priately placed carbon brushes which serve to conduct direct current to or from
the armature winding depending upon whether motor or generator action is taking
place.
In Fig. 3-8 the armature winding is depicted as a coil wrapped around a
toroid. This is merely a schematic convenience. An actual winding is wound so
that no conductors are wasted by placing them on the inner surface of the rotor
core where no flux penetrates. In Fig. 3-8 those parts of the armature winding
which lie directly below the brush width are assumed to have the insulation
removed, i.e., the copper is exposed. This allows current to be conducted to and
from the armature winding through the brush as the rotor revolves. Ina practical
winding each coil is made accessible to the brushes by connecting the coils to
individual commutator segments as indicated in Fig. 3-9(a) and then placing the
brushes on the commutator.
For motor action direct current is made to flow through the field winding as
well as the armature winding. If current is assumed to flow into brush B, in Fig.
3-8, then note that on the left side of the rotor for the outside conductors current
flows into the paper while the opposite occurs for the conductors located on the
outside surface of the right side of the rotor. By Eq. (1-2) a force is produced on
each conductor, thereby producing a torque causing a clockwise rotation. Now
the function of the commutator is to ensure that, as a conductor such as | in Fig.
3-8 revolves and thus goes from the left side of brush B; to the right side, the
current flowing through it reverses, ensuring a continuous unidirectional torque
for the entire armature winding. Recall that a reversed conductor current in a flux
field of reversed polarity keeps the torque unidirectional. The reversal of current
in
comes about because the commutator always allows current to be conducted
it is
the same directions in either side of the armature winding whether or not
rotating.
Another point of interest in Fig. 3-8 concerns the location of the brushes. By
contrib-
placing the brushes on a line perpendicular to the field axis all conductors
a unidirect ional torque. If, on the other hand, the brushes were
ute in producing
rs would
placed on the same line as the field axis, then half of the conducto
and the other half counterc lockwise torque, yielding a
produce clockwise torque
zero net torque.
 138 Fundamentals of Electromechanical Energy Conversion Chap. 3

Figure 3-9(a) Armature winding of a dc machine, showing the coils connected to

the commutator. (Courtesy of General Electric Company.)

Figure 3-9(b) Complete stack of stator lamina- Figure 3-9(c) Complete stator structure of
tions for the two-pole dc machine equipped the two-pole dc machine of Fig. 3-9(b) with
with interpoles. : field and interpole windings in place.
Sec. 3-3 Construction Features of Electric Machines 139

with
Figure 3-9(d) Cutaway view of completely assembled two-pole de machine
notations for various constructio n features. See text for descriptions . (Courtesy
of General Electric Company.)

in the
Additional construction features of the de machine are displayed
stator stack of
remaining parts of Fig. 3-9. Shown in Fig. 3-9(b) is the laminated
are used to
the two-pole machine. The small protrusions are called interpoles and
current directio n in an armature coil as it passes from one side of
aid the change of
windings in
a brush to the other. Figure 3-9(c) depicts the field and interpole
ed machine.
place. Finally, Fig. 3-9(d) displays a cutaway view of the assembl
5 denotes the steel
Here item 2 locates the vacuum degassed steel ball bearings;
air passages for
enclosure; 11 is the commutator support which includes cooling
brush-access port
improved cooling of the armature conductors; 12 identifies the
insulation.
for maintenance; and 14 denotes contamination resistant armature
 140 Fundamentals of Electromechanical Energy Conversion Chap. 3

3-4 PRACTICAL FORMS OF TORQUE AND VOLTAGE FORMULAS

Our objective here is to modify the basic expressions for induced voltage and
electromagnetic torque to forms that are more meaningful for the specific design
of a particular electromechanical energy-conversion device, whether it is alternat-
ing or direct current.

Alternating-Current Machines

Voltage. If N denotes the total number of turns per phase of a three-phase

winding, the instantaneous value of the emf induced in any one phase can be
represented by Eq. (3-16), which is repeated here for convenience. Thus
e=oN® sin at (3-17)

Keep in mind that ® denotes the total flux per pole and w represents the relative
cutting speed in electrical radians per second of the winding with respect to the
flux-density wave. It is related to the frequency f of the ac device by

w = 2nf (3-18)
where f is expressed in cycles per second.
The maximum value of this ac voltage occurs when sin wt has a value of
unity. Hence

Emax = oN® V (3-19)

The corresponding rms value is

E= aS = V2 rfN® = 4.44fN® V (3-20)

A comparison of this expression with Eq. (2-18) reveals that the equations have an
identical form. There is a difference, however, and it lies in the meaning of ®. In
the transformer, ®,, is the maximum flux in time corresponding to the peak mag-
netizing current consistent with the magnitude of the applied voltage. In the ac
electromechanical energy-conversion device ® is the maximum flux per pole (a
space quantity) that links with a coil having N turns and spanning the full-pole
pitch.
In any practical machine the total turns per phase are not concentrated ina
single coil but rather are distributed over one-third of a pole pitch (or 60 electrical
degrees for each of the three phases). In addition the individual coils that make up
the total N turns are intentionally designed to span not the full-pole pitch but
rather only about 80% to 85% of a pole pitch. Such a coil is called a fractional-
pitch coil. The use of a distributed winding employing fractional-pitch coils has
the advantage of virtually eliminating the effects of all harmonics that may be
present in the flux-density wave while only slightly reducing the fundamental
Sec. 3-4 Practical Forms of Torque and Voltage Formulas 141

component. + The reduction in the fundamental component can be represented by

a winding factor denoted by K,,. Usually K,, has values ranging from 0.85 to
0.95. Accordingly, the final practical version of the induced rms voltage equation
for an ac machine is

E = 4.44fNK,,® V (3-21)

Torque. Analysis of the manner in which electromagnetic torque is pro-

duced by the interaction of the armature ampere-conductor distribution with the
field distribution that prevails in actual ac machines can be shown (see Appendix
to Chapter 3, Sec. 3A) to lead to the following basic expression:

T= gPe JIncosh N-m GoD

where p is the number of poles

® is the flux per pole in Wb
Jm is the equivalent current sheet that represents an idealized ampere-con-
ductor distribution expressed in A/rad
wis the phase displacement angle between the start of the current sheet
(i.e., ampere-conductor distribution) and the start of the flux density
wave beneath a pole.

A study of Eq. (3-22) stresses the three conditions that must be satisfied for
the development of torque in conventional electromechanical energy-conversion
be an
devices. There must be a field distribution represented by ®. There must
ampere-conductor distribution represented by J,,. Finally, there must exist a
it is
favorable space-displacement angle between the two distributions. Note that
and still the
possible to have both a field and an ampere-conductor distribution
torques are
torque can be zero if the field pattern is such that equal and opposite
developed on the conductors, i.e., y = 90°.
for
There are two useful and practical forms to which this basic expression
it is the am-
developed electromagnetic torque may be converted. In one form
ed. In the
pere-conductor distribution of the armature winding that is emphasiz
per pole, %,, is
other, it is the conservation of energy that is stressed. The mmf
related to J, by
Im = Fp (3-23)
given ac machine the
(see Appendix to Chapter 3, Eq. (3-50). Moreover, for any
as they appear in the
quantity #, is entirely determined from the design data
following expression:

+See Appendix B.
+See Appendix C.
142 Fundamentals of Electromechanical Energy Conversion Chap. 3

Fy = mee Pate Og (3-24)

7 Pp Pp
where g = number of phases of the armature winding
N> = number of turns per phase of the armature winding
K,,. = armature-winding factor
J, = armature-winding current per phase
p = number of poles

Hence the expression for electromagnetic torque as described by Eq. (3-22) be-
comes
7G
T= gP PF, cos Ww = 5 0:9PPN2Ky2ql2 COS
7
(3-25)
Also, gN> is the total number of turns on the armature surface. This can be
expressed in terms of the total number of conductors, Z,, by making use of the
fact that it requires two conductors to make one turn. Thus

NEMaar
2s (3-26)

Inserting this into Eq. (3-25) yields

T = 0.177 p®(Z,K,212) cos w N-m (3-27)

The quantity in parentheses emphasizes the role played by the ampere-conductor

distribution of the armature winding. Of course to compute the torque by this
equation we need information about the space-displacement angle in addition to
the design data, i.e., p, ®, Z2, K,2. Normally, & can be determined by the same
data that lead to the determination of J,, which is described later in the next
chapter.
A second approach to obtaining a useful form of the basic torque equation is
to replace ® with Eq. (3-21) and J, with Eqs. (3-23) and (3-24) in Eq. (3-22). Thus

N2 Ky2 )
” : Sacer I,} cosw
ee a leeseal =

Or

Pp
T= anf Gee cos W (3-28)

Moreover, since the mechanical angular velocity w,, is related to the electri-
cal angular velocity w by
Q 4nf
Om = ile —.
9 af
ar p)
.
=_-_—_
(3-29)
=

Vir
Sec. 3-4 Practical Forms of Torque and Voltage Formulas 143

it follows that Eq. (3-28) may be written as

1
T= ra gExI, cos wb N-m (3-30)

where E> and J, are the rms induced voltage and current per phase of the armature
winding.
It can be shown that the space-displacement angle w in this equation is
identical with the time-displacement angle @),+ which is the phase angle existing
between the two time phasors E and J,. Accordingly, the expression for the
electromagnetic torque may be written as

1
T= @.. qExlh COS 0, N-m (3-31)

Note the power balance, which states that the developed mechanical power is
equal to the developed electrical power.

Direct-Current Machines

Voltage. The useful version of the expression for the emf induced in the
armature winding of a dc machine and appearing at the brushes also derives from
Eq. (3-16). The presence of the commutator, however, makes a difference in the
manner in which this equation is to be treated to obtain the desired result. We are
interested in the voltage as it appears at the brushes. If we follow coil 1 in F ig. 3-8
as it leaves brush B> and advances to brush B,, we notice that the emf induced in
this coil is always directed out of the paper beneath the south-pole flux. In fact
this is true of any coil that moves from B, to B;. Having established the direction
of this voltage for the coils on the left side of the armature, we then note that upon
tracing through the winding starting in the direction called for by the sign of the
emf induced in coil 1 we terminate at brush B,. When coil | rotates to a position
of the
on the right side of the armature winding, such as position 11, the direction
induced emf is into the paper. If we start with this indicated direction and trace
the armature winding, we again terminate at brush B). Therefore it is reasonable
fixed
to conclude that brush B, is at a fixed positive polarity and brush B, is at a
along as observ-
negative polarity. In essence this means that, if we were to ride
B), the
ers on coil 1 as it traverses through one revolution beginning at brush
variation shown in
fundamental component of the induced emf would exhibit the
coil as it moves
Fig. 3-10. Note that relative to brush B, the voltage induced in the
in conjunc-
from B, to B> appears rectified because the action of the commutator
spite of its rota-
tion with the brushes is to fix the armature winding in space in
B, on the left side
tion. Accordingly, whether we look at a coil moving from B, to

+See page 175.

144 Fundamentals of Electromechanical Energy Conversion Chap. 3

Induced
coil emf

7 er wt neath Figure 3-10 Fundamental component

cies of the induced emf in a coil of the
Bo B, B2 Brush position armature winding of the dc machine.

or from B, to B, on the right side, the directions of the induced emf’s are such as to
make B, assume one fixed polarity, and B; the opposite fixed polarity.
For a single coil having N, turns the average value of the unidirectional
voltage appearing between the brushes is obtained by integrating Eq. (3-16) over 7
electrical radians and dividing by 7. Thus

Foe :i ON simardot= 4iNe@ (3-32)

It is customary to express the frequency in terms of the speed of the armature in
rpm which is given by the relationshipt

(3-33)

where p denotes the number of poles and n the speed in rpm. Inserting Eq. (3-33)
into Eq. (3-32) and rearranging leads to

E. = p®2N, hs
= = p®z bis.
60 60
where z denotes the number of conductors per coil.
This last expression is the average value of the induced emf for a single coil.
If many such coils are placed to cover the entire surface of the armature, the total
dc voltage appearing between the brushes can be considerably increased. If the
armature winding is assumed to have a total of Z conductors and a parallel paths,
then the induced emf of the armature winding appearing at the brushes becomes

(3-34)

tRotating a coil in a two-pole field at a rate of 1 rps results in a frequency of 1 Hz. For a four-
pole field each rps yields 2 Hz. For a p-pole distribution the frequency generated in hertz is

=e = DPhil
Pa TPS Bore iat
Sec. 3-4 Practical Forms of Torque and Voltage Formulas 145

where Kg is a winding constant defined as

Ke=eA
60a (3-35)

Whenever the speed is expressed in radians per second, the induced emf equation
can then be expressed as

E, = Kr®w (3-36)

where

hf
Kr; = oa (3-36a)

It is worth keeping in mind that the induced-emf equation for the ac machine
and that for the dc machine originate from the same starting point. To further
emphasize this common origin let it be said that we can derive Eq. (3-34) from
Eq. (3-21). It merely requires introducing the appropriate winding factor as it
applies to the dc armature winding. In fact, it is the quasi-annihilating effect of the
winding factor on the harmonics that allows the derivation of Eq. (3-34) to proceed
in the fundamental terms in spite of the nonsinusoidal field distribution character-
istic of dc machines. It is assumed throughout, however, that the fundamental
component of flux is virtually the same as the total flux per pole.

Torque. Analysis of the manner in which electromagnetic torque is devel-

oped by the interaction of the ampere-conductor distribution with the flux density
lead
wave in the dc machine can be shown (see Appendix to Chapter 3, Sec. 3B) to
to the following fundamental formula:

T=5J@ N-m (3-37)

where p is the number of poles
J is the current-sheet equivalent of the armature winding ampere-conduc-
tor distribution in A/rad
® is the flux per pole in Wb
nt upon
It is worthwhile to note again that electromagnetic torque is depende
tion repre-
a field distribution represented by ®, an ampere-conductor distribu
between the two distributions. Of
sented by J, and the angular displacement
ty the current sheet
course, # does not appear in Eq. (3-37) because for simplici
ensuring that all
was assumed to be in phase with the flux-density curve, thereby
parts of the current sheet experience a unidirectional torque.
s results
A practical version of this basic torque expression for dc machine
expression involving the
when the current sheet J is replaced by its equivalent
Applying Ampere’s
appropriate design data. In this connection refer to Fig. 3-11.
 146 Fundamentals of Electromechanical Energy Conversion Chap. 3

Current sheet

S N S

1 Figure 3-11 Direct-current machine:

(a) armature with current sheet; (b)
armature with finite ampere-conductor
(a (b) distribution.

circuital law to the typical flux path shown makes it apparent that the mmf per
pole pair is given by
=
F = (rad) J a
Te = be ences
(3-38)
-

Hence the mmf per pole is .

F wd
F,
Te Sa
5
—_>S>S
5
-
(3-39)
-

This expression is applicable whenever the armature winding is represented by a

current sheet. In a practical situation the armature winding is always represented
by a finite ampere-conductor distribution as depicted in Fig. 3-11(b). In such
cases the armature mmf per pole may be expressed in terms of the total current J,
entering or leaving a brush and the total conductors Z on the armature surface.
Thus
Ci LZa 4
Fy eeeaD mmf/pole, or A-t/pole
Bs (3-40)
5

Upon equating the last two expressions we obtain the equation for J. Thus
tba
eee (3-41)
mpa
Insertion of Eq. (3-41) into Eq. (3-37) then yields the desired expression for
the electromagnetic torque developed by the dc machine. Hence

i=Bape
5 Lo.=LOU ace
ah O/, (3-42)

or

T = K,Ol, N-m (3-43)

where

Kr =Peo
5 = torque constant (3-44)
Sec. 3-4 Practical Forms of Torque and Voltage Formulas 147

The foregoing form of the torque formula stresses the ampere-conductor distribu-
tion taken in conjunction with the flux field. It is analogous to Eq. (3-27) for the ac
machine. Of course in Eq. (3-43) & does not appear because it was made zero by
placing the brush axis in quadrature with the field axis. It is important to note,
however, that the fundamental quantities for the production of torque are there.
The equations differ only to the extent that the mechanical details of construction
differ.
An alternative expression for the electromagnetic torque results upon substi-
tuting in Eq. (3-42) the expression for ® obtained from Eq. (3-34). Thus, from Eq.
(3-34)
60a
® = Bin E, (3-45)

so that

_~ Qaa
PZ pZn
Oa, Eula,_= 60.
271 Ela ae

But
27Nn
Om = Eo (3-47)

Therefore

(3-48)

A glance at this result again points out the power-balance relationship that under-
lies the operation of electromechanical energy-conversion devices. Equation
(3-48) for the dc machine is analogous to Eq. (3-31) for the ac machine.
Example 3-3
An ac machine is equipped with four poles, three phases, and 36 armature winding
turns per phase. The field winding is designed to produce a flux per pole of 0.015
Wb. The armature winding is designed to be essentially resistive over its operating
range and the coil span is made equal 83.3% of a pole pitch.
in
(a) When the motor develops a torque of 115 N-m, find the armature current
amperes.
(b) Determine the armature mmf per pole.
(c) Find the developed torque when the armature current is 15 A.
(d) In part (a) the developed power is known to be equal to 21,677 W. Calcu-
late the armature induced voltage per phase.
of
(e) Compute the mechanical speed of the motor in rpm for the condition
part (a).
is the
Solution (a) We can use Eq. (3-27) directly here to get the current J,, which
first two matters need to be resolved. The first is the angle w.
armature current. But
 148 Fundamentals of Electromechanical Energy Conversion Chap. 3

By the problem statement this quantity is 0° because the inductive component of the
armature winding is considered to be negligible. Second, we need to determine
K,2. This quantity is readily calculated by invoking Eq. (B-2) of Appendix B. Thus

Ky2 = sin (0.833) 5 = sin 75° = 0.966

It therefore follows that the armature current is

ih 115
L,= 0.177pbZA0.966) — 0.177(4(0.015)(2 x 3 X 36)(0.966) a
(b) The armature mmf per pole is found from

F, = Jn = 0.94 = I, = 0.9(3) oeuee)

4 51.9 = 1218.3 A-t

(c) Using the second form of Eq. (3-25) we get

T = § (0.9)pdNoKwaqlr Os
= z (0.9)(4)(0.015)(26)(0.966)(3)(15) = 33.2 N-m

(d) From Eq. (3-31)

3 Ep = 21,677
21i6rt &
E, = 361.9). 139.2 V per phase

(e) By conservation of energy

Tom = 21,677

On = ere = 188.5 rad/s

115
or, expressed in rpm, we write

30
Nm = 7 em = 1800 rpm

where n,, denotes the mechanical speed in revolutions per minute.

3-5 FINAL REMARKS

For those readers who are interested in a detailed and rigorous treatment of the
derivation of the basic electromagnetic torque formulas for ac and dc machines,
the subject matter is continued at the end of the book in the section entitled
Appendix to Chapter 3. In the interest of continuity, both the equation numbers
and figure numbers pick up from the last notations used in Sec. 3-4. A full comple-
ment of problems is also included.
Chap. 3 Problems 149

PROBLEMS

3-1. A two-pole de motor is equipped with a single ten-turn full-pitch coil. It has a
rectangular field distribution of height 0.5 T. The radius of the rotor is 0.2 m and the
axial length is 0.3 m.
(a) Find the flux per pole in webers.
(b) Current is supplied to the coil and the motor is found to reach a speed of
625 rpm. Find the emf induced in the coil.
(c) What is the rate of change of flux with time?
3-2. The motor of Prob. 3-1 is used to supply a torque of 9 N-m to an attached load at a
speed of 500 rpm.
(a) Find the current drawn by the coil.
(b) Calculate the coil emf.
(c) What is the developed mechanical power?
3-3. At a particular load condition the motor of Prob. 3-1 delivers 400 W while drawing a
coil current of 8 A. All losses are negligible.
(a) Find the developed torque at the motor shaft.
(b) What is the coil emf?
(c) Determine the speed of rotation of the motor shaft in rpm.
3-4 At a given operating condition the machine of Prob. 3-1 runs at 350 rpm. Find the
coil induced emf.
3-5. The instantaneous voltage generated in a coil revolving in a magnetic field can be
calculated either by the flux-linkage concept or by the Blu concept. With this in mind
so
consider the following problem. A square coil 20 cm ona side has 60 turns and is
located that its axis of revolution is perpendicu lar to a vertically directed uniform
magnetic field in air of 0.06 Wb/m2. The coil is driven at a constant speed of 150
rpm. Compute:
(a) The maximum flux passing through the coil.
(b) The maximum flux linkage.
(c) The time variation of the flux linkage through the coil.
concepts
(d) The maximum instantaneous voltage generated in the coil using both
Indicate by a sketch the position of the coil at this
referred to in the foregoing.
instant.
cycle.
(e) The average value of the voltage induced in the coil over one
of the coil is 30° from the
(f) The voltage generated in the coil when the plane
vertical. Compute by both methods.
at a constant speed of
. A square coil of 100 turns is 10 cm on each side. It is driven
300 rpm.
cular to a vertically
(a) The coil is placed so that its axis of revolution is perpendi
uniform field of 0.1 Wb/m?. At time t = 0, the coil is in a position of
directed
the instanta neous voltage gen-
maximum flux linkages. Derive an expression for
erated. Sketch the voltage waveshape for one cycle.
remain the same as for part
(b) The coil is now placed ina radial field. All other data
numerical value of
(a). Sketch the voltage waveshape for one cycle showing the
the maximum voltage.
The coil is revolving at 1800 rpm
3-7. In Figs. P3-7(a) and (b) the flux per pole is 0.02 Wb.
and consists of 2 turns.
150 Fundamentals of Electromechanical Energy Conversion Chap. 3

(a) For the configuration of Fig. P3-7(a) derive an expression for the flux linking the
coil in terms of the flux per pole and ap.
(b) What is the instantaneous value of the coil voltage in Fig. P3-7(a)?
(c) Compute the maximum value of the voltage induced in the arrangement of
Fig. P3-7(b).
(d) Assuming that the coil of Fig. P3-7(b) is connected to a pair of commutator
segments, find the dc value of this voltage.

Uniform vertical field

vill
B &o

0° | | Rad
ao

(a) (b)
Figure P3-7

3-8. An electric machine has a field distribution as shown in Fig. P3-8. The coil spans a
full 180° and has N turns. The machine has two poles and an axial length / and rotor
radius r.
(a) Obtain the expression for the flux linkage per pole in terms of @ expressed in
electrical degrees.
(b) By using Faraday’s law find the expression for the induced emf in terms of the
flux per pole.
(c) Use the Blu form for the induced emf and verify the result of part (b).

cl
(oc + 77) Figure P3-8

3-9. A two-pole field distribution has the sinusoidal shape shown in Fig. P3-9. The maxi-
mum value of the flux density is 0.4 T. A coil having full pitch, 50 turns, an axial
length of 0.3 m, and a radius of 0.2 m is driven at 600 rpm.
(a) Derive the expression for the flux linkage of the coil when at position ag.
(b) Using the flux linkage approach, find the instantaneous value of the coil voltage
when ap = 120°.
(c) Repeat part (b) using the Blu approach.
(d) Find the maximum voltage induced in the coil and the position at which it
occurs.
(e) What is the average value of the induced coil voltage?
Chap. 3 Problems 151

B=B,, sina

Figure P3-9

3-10. A two-pole machine has the field distribution shown in Fig. P3-10. The coil is full
pitch and has N turns, radius r, and an axial length /.
(a) Show that the flux linkage (A) for the coil position is given by
NB, Ir ee
h=—7 (=5 - 2a3) where 0 S$ ay = .

(b) Assume that the coil is driven at wo rad/s. Find the expression for the induced
coil voltage using the change-of-flux-linkage approach. Express the result in
terms of the flux per pole and for the interval 0 = wot = 77/2.
(c) Repeat part (b) using the Blu approach.

em Electrical
degrees

Figure P3-10

[i.e., Eq. (3-31)].

3-11. Derive Eq. (3-27) by starting with the conservation of energy
rized by a sinusoida l rotor mmf and
3-12. A four-pole ac induction motor is characte
r, the flux per pole is 0.02 Wb. When operating at a
sinusoidal flux density. Moreove
conditio n, the space displace ment angle is found to be 53.3°. What is
specified load
a torque of 40 N-m?
the amplitude of the rotor mmf wave required to produce
Assume two poles.
V induced in its armature winding,
3-13. A 60-Hz, eight-pole ac motor has an emf of 440
which has 180 effective turns. When this motor develops 10 kW of power, the
amplitude of the sinusoidal field mmf is equal to 800 A-t.
(a) Find the value of the resultan t air-gap flux per pole.
wave.
(b) Find the angle between the flux wave and the mmf
of a 60-Hz, four-po le inducti on motor is 120 V. The
3-14. The induced emf per phase
motor develops a torque of
number of effective turns per phase is 100. When this
152 Fundamentals of Electromechanical Energy Conversion Chap. 3

60 N-m the space displacement angle is 30°. Find the value of the total armature
ampere-turns.
3-15. The three-phase armature winding of a 60-Hz, four-pole, ac induction motor is
equipped with 320 effective turns per phase. When the motor develops 10 kW it has
an armature current of 40 A, a space displacement angle of 37°, and a speed of
rotation of 1700 rpm.
(a) Find the amplitude of the armature mmf per pole.
(b) What is the value of the developed torque?
(c) Compute the induced emf in the armature winding per phase.
(d) Find the flux per pole.
3-16. Identify the physical location and the function(s) performed by each of the following
in the dc machine: shunt field winding, interpole winding, compensating winding,
and series field winding.
3-17. A machine is built with a two-pole single-phase winding and equipped with a dc
commutator winding as depicted in Fig. P3-17. The brush axis is displaced by 20°
from the coil axis. If a single-phase voltage.is applied to the stator winding, will a
continuous, unidirectional torque be developed? Explain. Assume negligible arma-
ture leakage flux.

Figure P3-17

3-18. Show that the basic equation for electromagnetic torque in the form
2
r=5() © sin A
T

is correct dimensionally.
 A

Three-Phase
Induction Motors

One distinguishing feature of the induction motor is that it is a singly excited

machine. Although such machines are equipped with both a field winding and an
armature winding, in normal use an energy source is connected to one winding
alone, the field winding. Currents are made to flow in the armature winding by
induction, which creates an ampere-conductor distribution that interacts with the
field distribution to produce a net unidirectional torque. The frequency of the
induced current in the conductor is affected by the speed of the rotor on which it is
located; however, the relationship between the rotor speed and the frequency of
the armature current is such as to yield a resulting ampere-conductor distribution
that is stationary in relation to the field distribution. As a result the singly excited
induction machine is capable of producing torque at any speed below synchronous
speed.* For this reason the induction machine is placed in the class of asynchro-
nous machines. In contrast, synchronous machines are electromechanical en-

applied to the field winding

+Synchronous speed is determined by the frequency of the source
machine is designed. These quantities are related by Eq.
and the number of poles for which the
(3-33). Thus
120f
synchronous speed = a =n,

153
 154 Three-Phase Induction Motors Chap. 4

ergy-conversion devices in which a net torque can be produced at only onet speed
of the rotor. The distinguishing characteristic of the synchronous machine is that
it is a doubly excited device except when it is being used as a reluctance motor.
The salient construction features of the three-phase induction motor are
described in Sec. 3-3. Because the induction machine is singly excited, it is neces-
sary that both the magnetizing current and the power component of the current
flow in the same lines. Moreover, because of the presence of an air gap in the
magnetic circuit of the induction machine, an appreciable amount of magnetizing
current is needed to establish the flux per pole demanded by the applied voltage.
Usually, the value of the magnetizing current for three-phase induction motors
lies between 25% and 40% of the rated current. Consequently, the induction
motor is found to operate at a low power factor at light loads and at less than unity
power factor in the vicinity of rated output.
We are concerned in this chapter with a description of the theory of opera-
tion and of the performance characteristics of the. three-phase induction motor.
Our discussion begins with an explanation of how a revolving magnetic field is
obtained with a three-phase winding. After all it is this field that is the driving
force behind induction motors.

4-1 THE REVOLVING MAGNETIC FIELD

The application of a three-phase voltage to the three-phase stator winding of the

induction motor creates a rotating magnetic field, which by transformer action
induces a working emf in the rotor winding. The rotor induced emf is called a
working emf because it causes a current to flow through the armature winding
conductors. This combines with the revolving flux-density wave to produce
torque in accordance with Eq. (3-27). Consequently, we can view the revolving
field as the key to the operation of the induction motor.
The rotating magnetic field is produced by the contributions of space-dis-
placed phase windings carrying appropriate time-displaced currents. To under-
stand this statement let us turn attention to Figs. 4-1 and 4-2. Appearing in Fig.
4-1 are the three-phase currents which are assumed to be flowing in phases a, b,
and c respectively. Note that these currents are time-displaced by the equivalent
of 120 electrical degrees. Depicted in Fig. 4-2 are the stator structure and the three-
phase winding. Note that each phase (normally distributed over 60 electrical
degrees) for convenience is represented by a single coil. Thus coil a-a’ represents
the entire phase a winding having its flux axis directed along the vertical. This
means that whenever phase a carries current it produces a flux field directed along
the vertical—up or down. The right-hand rule readily verifies this statement.
Similarly, the flux axis of phase b is 120 electrical degrees displaced from phase a,

}Theoretically, there are two rotor speeds at which a net torque different from zero can exist,
but at the second speed enormous currents flow, which makes operation impractical.
Sec. 4-1 The Revolving Magnetic Field 155

he : Figure 4-1 Balanced three-phase

N 21314 1S alternating currents.

and that of phase c is 120 electrical degrees displaced from phase b. The unprimed
letters refer to the beginning terminal of each phase.
Let us consider the determination of the magnitude and direction of the
resultant flux field corresponding to time instant f, in Fig. 4-1. At this instant the
current in phase a is at its positive maximum value while the currents in phases b
and c are at one-half their maximum negative values. In Fig. 4-2 it is arbitrarily
assumed that, when current in a given phase is positive, it flows out of the paper

Phase a
flux axis

Phase c Phase b
flux axis flux axis

Phase a flux axis

Phase b
flux axis flux axis

(c)
Figure 4-2 Representation of the rotating magnetic field at three different in-
stants of time: (a) time ¢; in Fig. 4-1; (b) time #5; (c) time fs.
 Three-Phase Induction Motors Chap. 4
156

with respect to the unprimed conductors. Thus, since at time ¢; i, is positive, a dot
for a’ because it
is used for conductor a. See Fig. 4-2(a). Of course a cross is used
that phase a
refers to the return connection. Then by the right-hand rule it follows
Moreover, the
produces a flux contribution directed upward along the vertical.
is a
magnitude of this contribution is the maximum value because the current
®,,, where ®,, is the maximu m flux per pole of phase a. It
maximum. Hence ¢, =
field with
is important to understand that phase a really produces a sinusoidal flux
the axis of phase a as depicted in Fig. 4-3. However ,
the amplitude located along
al distribu tion is convenie ntly represen ted by the vec-
in Fig. 4-2(a) this sinusoid
tor dy.
To determine the direction and magnitude of the field contribution of phase b
that
at time t;, we note first that the current in phase b is negative with respect to
in phase a. Hence the conducto r that stands for the beginning of phase b must be
assigned a cross while b’ is assigned a dot. Hence the instantan eous flux contribu-
tion of phase b is directed upward along its flux axis and the magnitude of phase b
flux is one-half the maximum because the current is at one-half its maximum
value. Similar reasoning leads to the result shown in Fig. 4-2(a) for phase c. A
glance at the space picture corresponding to time ¢,, as illustrated in Fig. 4-2(a),
should make it apparent that the resultant flux per pole is directed upward and has
a magnitude 3 times the maximum flux per pole of any one phase. Figure 4-3
depicts the same results as Fig. 4-2(a) but does so in terms of sinusoidal flux waves
rather than flux vectors. Keep in mind that the resultant flux vector in Fig. 4-2
shows the direction in which flux crosses the air gap. Once across the air gap, the
flux is confined to the iron in the usual fashion.
Next let us investigate how the situation of Fig. 4-2(a) changes as time
passes through 90 electrical degrees from f; to 5 in Fig. 4-1. Here phase a current

Axis of Axis of Axis of

phase a phase b phase c

Displacement
along air gap

Figure 4-3 Component and resultant field distributions corresponding to f, in

Fig. 4-1.
Sec. 4-1 The Revolving Magnetic Field 157

is zero, yielding no flux contribution. The current in phase b is positive and equal
to 3/2 its maximum value. Phase c has the same current magnitude but is
negative. Together phases b and c combine to produce a resultant flux having the
same magnitude as at time ¢;. See Fig. 4-2(b). It is important to note, too, that an
elapse of 90 electrical degrees in time results in a rotation of the magnetic flux field
of 90 electrical degrees.
A further elapse of time equivalent to an additional 90 electrical degrees
leads to the situation depicted in Fig. 4-2(c). Note that again the axis of the flux
field is revolved by an additional 90 electrical degrees.
On the basis of the foregoing discussion it should be apparent that the appli-
cation of three-phase currents through a balanced three-phase winding gives rise
to a rotating magnetic field that exhibits two characteristics: (1) it is of constant
amplitude and (2) it is of constant speed. The first characteristic has already been
demonstrated. The second follows from the fact that the resultant flux traverses
through 27 electrical radians in space for every 27 electrical radians of variation in
time for the phase currents. Hence, for a two-pole machine, where electrical and
mechanical degrees are identical, each cycle of variation of current produces one
complete revolution of the flux field. Hence this is a fixed relationship which is
dependent upon the frequency of the currents and the number of poles for which
the three-phase winding is designed. In the case where the winding is designed for
four poles, it requires two cycles of variation of the current to produce one
revolution of the flux field. Therefore it follows that for a p-pole machine the
relationship is

f=" x ps = 5% (4-1)
that
where fis in cycles per second and rps denotes revolutions per second. Note
Eq. (4-1) is identical with Eq. (3-33).
at
An inspection of the ampere-conductor distribution of the stator winding
e in such a
the various time instants reveals that the individual phases cooperat
in Fig. 4-2(a) the
fashion as to produce a solenoidal effect in the stator. Thus
the currents are such that they all enter into the page on the right side
directions of
the flux field is then
and leave on the left side. The right-hand rule indicates that
is similar except
directed upward along the vertical. In Fig. 4-2(b) the situation
resultant flux field is
that now the cross and dot distribution is such that the
be concluded that the
oriented horizontally toward the right. Therefore, it can
associated with the
rotating magnetic field is a consequence of the revolving mmf
stator winding.
d three-phase cur-
In the foregoing it is pointed out that the flow of balance
rotating field of constant
rents through a balanced three-phase winding yields a
satisfied, it is still
amplitude and speed. If neither of these conditions is exactly
not be of constant ampli-
possible to obtain a revolving magnetic field but it will
machine a rotating
tude nor of constant linear speed. In general, for a q-phase
when the following two
field of constant amplitude and constant speed results
 Three-Phase Induction Motors Chap. 4
158

conditions are satisfied: (1) there is a space displacement between balanced phase
windings of 27/q electrical degrees, and (2) the currents flowing through the phase
windings are balanced and time-displaced by 27/q electrical degrees. For the
three-phase machine g = 3 and so the now familiar 120° figure is obtained. The
only exception to the rule is the two-phase machine. Because the two-phase
situation is a special case of the four-phase system, a value of q equal to 4 must be
used.
One final point is now in order. The speed of rotation of the field as de-
scribed by Eq. (4-1) is always given relative to the phase windings carrying the
time-varying currents. Accordingly, if a situation arises where the winding is
itself revolving, then the speed of rotation of the field relative to inertial space is
different than it is relative to the winding.

Mathematical Analysis

The mmf distribution of the stator (field) winding of the three-phase induction
motor is usually trapezoidal in shape, but it contains a very prominent fundamen-
tal component. In fact it is this component that is depicted in Fig. 4-3. Although
some mmf harmonics do exist, we shall continue to confine our attention here to
the important fundamental term. .
The resultant stator mmf at any time instant is composed of the contributions
of each phase. It is important to keep in mind that each phase winding makes a
contribution that varies with time along a fixed space axis. An observer looking
only at a single phase thus sees an alternating field. If the axis of phase a in Fig.
4-3 is taken as the zero reference point for the displacement angle a along the air
gap, the alternating field created by this phase may be mathematically expressed
by
Faq = Fm COS wt COS a (4-2)
where #,, denotes the maximum value of the mmf per pole and is specified by Eq.
(3-24) when q is set equal to unity (one phase). It is worthwhile to pause a moment
to be sure that the meaning of Eq. (4-2) is understood. This equation states that an
observer stationed in space, say, at a = 0° (i.e., on the axis of phase a) sees at time
t = 0 the maximum value of phase a mmf. As time progresses the value of this
mmf changes. Thus at wt = 60° the observer standing at a = 0° now sees F,,/2 as
the phase contribution. At wt = 120° the contribution becomes —%,,/2, which
means that the pole reverses polarity. At wt = 180° the mmf contribution is found
to be —#,,. Accordingly, the observer placed at a = 0° sees the mmf contribution
of phase a alternate between a positive #,, and a negative %,, as time progresses.
If the point of observation is changed, Eq. (4-2) states that the observer continues
to see an alternating field but now it varies between different maxima. Thus, if the
observer is fixed at a = 30°, then as time elapses phase a mmf is noted to alternate
between +(V3/2)%,,. An observer stationed at this position can never see the
Sec. 4-1 The Revolving Magnetic Field 159

peak value of fundamental mmf of phase a. Of course an observer situated at

a = 90° sees a zero value of phase a mmf for all time instants.
The expression for the mmf of phase b can be written by introducing two
modifications in Eq. (4-2) relating to the time variable t and the space variable a.
Recalling that a balanced three-phase voltage is applied to the balanced three-
phase stator winding, it follows that the time-varying current in phase b winding is
displaced by 120 electrical degrees from that in phase a as depicted in Fig. 4-1.
Moreover, the axis of the winding of phase b is also displaced by 120 electrical
degrees in space. Accordingly, the expression for the instantaneous mmf of phase
b becomes
Fi =e COSiw@r — 120°) cos (a — 120°) (4-3)

where F,, carries the same meaning as for phase a. Proceeding in a similar fashion
the mmf of phase c can be written as
F. = Fm COS (wt — 240°) cos (a — 240°) (4-4)

As was demonstrated in the physical analysis, the resultant mmf of the total
three-phase winding is found by summing the instantaneous contributions of the
alternating mmf’s of each phase. Expressing it algebraically we have
Ff, = F, + FH, + F, = Fmlcos wt cos a + cos (wt — 120°) cos (a — 120°)

+ cos (wt — 240°) cos (a — 240°)] (4-5)

Upon introducing the trigonometric identity

cos x cos y = $.cos (x — y) + 3. cos (x + y) (4-6)

there results

ee [cos (wt — a) + cos (wt + a) + cos (wt — a)

Fe, = —

(4-7)
+ cos (wt + a — 240°) + cos (wt — a) + cos (wt + a — 120°)]
it should be clear
Keeping in mind that this last expression describes a space field,
terms, being equal in amplitu de and 120° apart,
that the second, fourth, and sixth
yield a net value of zero. Hence Eq. (4-7) simplifi es to

F, = 3F,, cS (wt — a) (4-8)

de. Note the corre-

This is the equation of a revolving field of constant amplitu
it is shown that the cooper-
spondence with the results depicted in Fig. 4-2. There
field having an amplitude
ation of the three alternating fields yields a resultant
ent with the factor 3¥,,
equal to 3 the amplitude of any phase. This is consist
ce we must now show that
appearing in Eq. (4-8). To complete the corresponden
a travel ing wave.
cos (wt — a) is the mathematical expression for
160 Three-Phase Induction Motors Chap. 4

In this connection let us consider an observer fixed in space at a position

a = 30° and investigate what happens as time is allowed to progress. By direct
application of Eq. (4-8) for a = 30° we obtain:

1. At wt; = 0°,

F, = 3F, COS (—30°) = $F» ae

2. At wt, = 30°,
I, = 35mn
3. At wt; = 120°,
Fe. = 3F,, cos (120° —,30°).=.0

Depicted in Fig. 4-4 are the various positions that the constant amplitude cosi-
nusoid must take to yield values consistent with those just computed. Note that,
as time progresses by 30°, the wave position at wt; must travel 30° in the positive
direction of a. Only in this way can the observer fixed at the space position
a = 30° see the peak value of the resultant wave as called for by Eq. (4-8). For an
additional elapse of time corresponding to 90° the observer sees a zero value for
the resultant wave. This happens when the wave shown in position 2 moves along
in the positive direction of a by 90° to position 3. Therefore, Eq. (4-8) does in fact
represent the equation for a traveling wave involving a trigonometric function.
The same reasoning employed in the foregoing can be used to illustrate that a
traveling wave of trigonometric character traveling in the negative a direction
bears the form cos (wt + a), where as before ¢ is a time variable and a is a space
variable.

© wt,=30°
® wt, =0°

7
6

|
|
Figure 4-4 Graphical demonstration that Eq. (4-8) is the expression for a travel-
ing wave.
 Sec. 4-2 Induction Motor Slip 161

4-2 INDUCTION MOTOR SLIP

The three-phase induction motor may be compared with the transformer because
it is a singly energized device which involves changing flux linkages with respect
to the stator and rotor windings. In this connection assume that the rotor is of the
wound type and Y-connected as illustrated in Fig. 4-5. With the rotor winding
open-circuited no torque can be developed. Hence the application of a three-
phase voltage to the three-phase stator winding gives rise to a rotating magnetic
field which cuts both the stator and rotor windings at the line frequency f;. The
rms value of the induced emf per phase of the rotor winding is given by Eq. (3-21)
as
Ey = 4.44 f; No Kyw2P (4-9)

where the subscript 2 denotes rotor winding quantities. Note that the stator fre-
quency f; is used here because the rotor is at standstill. Hence E, is a line-
frequency emf. Of course the flux ® is the flux per pole, which is mutual to the
stator and rotor windings.
A similar expression describes the rms value of the induced emf per phase
occurring in the stator winding. Thus
Kyi®
Nif,
E, = 4.44 (4-10)

From Eqs. (4-9) and (4-10) we can formulate the ratio

Ey __ MiKw
EN Ka se
ratio of the
Note the similarity of this expression to the voltage transformation
the winding factors of the
transformer. The difference lies in the inclusion of
for the motor as contrast ed
motor, necessitated by the use of a distributed winding
essence, then, the inductio n
to the concentrated coils used in the transformer. In
mer wherein the stator
motor at standstill exhibits the characteristics of a transfor
winding is the primary and rotor winding is the secondary.
under running
Next let us consider the behavior of the induction motor
similari ties to the trans-
conditions—again with the intention of pointing out
a running torque) it is
former. To produce a starting torque (and subsequently
winding . This is readily
necessary to have a current flowing through the rotor

Applied
3-phase
voltage
Figure 4-5 Schematic representation
Rotor winding of the three-phase wound-rotor induc-
Stator winding
tion motor. The dashed line indicates
short-circuit links for normal operation.
162 Three-Phase Induction Motors Chap. 4

accomplished by short-circuiting the winding in the manner indicated by the bro-

ken line in Fig. 4-5. Initially the induced emf EF; causes a rotor current per phase J,
to flow through the short-circuit, producing an ampere-conductor distribution
which acts with the flux field to produce the starting torque. The sense of this
torque is always to cause the rotor to travel in the same direction as the rotating
field. An examination of Fig. 4-6 makes this apparent. Assume that the flux field
is revolving clockwise at a speed corresponding to the applied stator frequency
and the number of poles of the stator winding. This speed is called the synchro-
nous speed and is described by Eq. (4-1). Thus

rpm (4-12)

By the 0 x B rule the induced emf in a typical conductor lying beneath a south-
pole flux is directed into the paper as indicated in Fig. 4-6. Then for this direction
of current the /, x B rule reveals the torque to be directed clockwise. Therefore
the rotor moves in a direction in which it tries to catch up with the stator field.
As the rotor increases its speed, the rate at which the stator field cuts the
rotor coils decreases. This reduces the resultant induced emf per phase, in turn
diminishing the magnitude of the ampere-conductor distribution and yielding less
torque. In fact this process continues until that rotor speed is reached which
yields enough emf to produce just the current needed to develop a torque equal to
the opposing torques. If there is no shaft load, the opposing torque consists
chiefly of frictional losses. It is important to understand that as long as there is an
opposing torque to overcome—however small or whatever its origin—the rotor
speed can never be equal to the synchronous speed. This is characteristic of
singly excited electromechanical energy-conversion devices. Since the rotor (or
secondary) winding current is produced by induction, there must always be a

Figure 4-6 Showing the direction of

induced voltage and developed torque
on a typical conductor.
Sec. 4-2. —_Induction Motor Slip 163

difference in speed between the stator field and the rotor. In other words, trans-
former action must always be allowed to take place between the stator (or pri-
mary) winding and the rotor (or secondary) winding.
This speed difference, or slip, is a very important variable for the induction
motor. In terms of an equation we may write
slip =n. 1 rpm (4-13)

where n denotes the actual rotor speed in rpm. The term slip is used because it
describes what an observer riding with the stator field sees looking at the rotor—it
appears to be slipping backward. A more useful form of the slip quantity results
when it is expressed on a per unit basis using synchronous speed as the refer-
ence. Thus the slip in per unit is

heaon
s= (4-14)
Ns

For the conventional induction motor the values of s lie between zero and unity.
It is customary in induction-motor analysis to express rotor quantities (such
as induced voltage, current, and impedance) in terms of line-frequency quantities
and the slip as expressed by Eq. (4-14). For example, if the rotor is assumed to be
operating at some speed n < n,, then the actual emf induced in the rotor winding
per phase may be expressed in terms of the line-frequency quantity E, as sE>.
This formulation has definite advantages, as described in the next section. Ina
similar fashion it is possible to express the rotor winding impedance per phase as
Zy = rp + jsx2 (4-15)

where z> denotes the rotor phase impedance, r is the rotor resistance per phase,
Of
and x, is the line-frequency leakage reactance per phase of the rotor winding.
n
course the effective value of this reactance when the rotor operates at a speed
(or slip s) is only s times as large. Keep in mind that the frequency of the currents
rotor
in the rotor is directly related to the relative speed of the stator field to the
winding. Accordingly, we may write
_ p(slip rpm) _ p(s = ”) .
fs 120 eal ol) (4-16)
By means of
where f; is the frequency of the emf and current in the rotor winding.
Eq. (4-14) it is possible to rewrite Eq. (4-16) as

bia s >wo
1200ek
120
(4-17)
merely multiplying the
which indicates that the rotor frequency f; is obtained by
slip. For this reason
stator line frequency by the appropriate per-unit value of the
fy is often called the slip frequency.
 164 Three-Phase Induction Motors Chap. 4

4-3 THE EQUIVALENT CIRCUIT

It is desirable to have an equivalent circuit of the three-phase induction motor in

order to direct the analysis of operation and to facilitate the computation of
performance. From the remarks made in the preceding section it should not come
as a surprise that the equivalent circuit assumes a form identical to that of the
exact equivalent circuit of the transformer. The derivation proceeds in a similar
fashion with necessary modifications introduced to account for the fact that the
secondary winding (the rotor) in this instance revolves and thereby develops
mechanical power.

The Magnetizing Branch of the Equivalent Circuit

All the parameters of the equivalent circuit are expressed on a per-phase basis.
This applies whether the stator winding is Y-.or A-connected. In the latter case
the values refer to the equivalent Y connection. Appearing in Fig. 4-7(a) is the
portion of the equivalent circuit that has reference to the stator (or primary)
winding. Note that it consists of a stator phase winding resistance r,, a stator
phase winding leakage reactance x; , and a magnetizing impedance made up of the
core-loss resistor r, and the magnetizing reactance x,. There is no difference in
form between this circuit and that of the transformer. The difference lies only in
the magnitude of the parameters. Thus the total magnetizing current J/,, is consid-
erably larger in the case of the induction motor because the magnetic circuit
necessarily includes an air gap. Whereas in the transformer this current is about
2% to 5% of the rated current, here it is approximately 25% to 40% of the rated
current depending upon the size of the motor. Moreover, the primary leakage
reactance for the induction motor also is larger because of the air gap as well as
because the stator and rotor windings are distributed along the periphery of the air
gap rather than concentrated on a core as in the transformer. The effects of the
actions that take place in the rotor (or secondary) winding must reflect themselves
at the proper equivalent-voltage level at terminals a-b in Fig. 4-7(a). We next
investigate the manner in which this comes about.

The Actual Rotor Circuit per Phase

For any specified load condition that calls for a particular value of slip s, the rotor
current per phase may be expressed as

= SE>
2 - ry + jsxz (4-18)

where EF) and x, are the standstill values. The circuit interpretation of Eq. (4-18) is
depicted in Fig. 4-7(b). It illustrates that J, is a slip-frequency current produced by
Sec. 4-3. The Equivalent Circuit 165

Ig i)

sE 2 jSXo

(a) (b)

Tome ne oS OP
a

E, < E>
— ig _—

b
(c) (d) (e)

T, n jx4 r2 iX2

(f)

Fe Re="4+lo Xe=i%1 + 5X2

(a) stator winding section;

Figure 4-7 Derivation of the equivalent circuit:
circuit; (d) modified equivalent rotor
(b) actual rotor circuit; (c) equivalent rotor
eferred equival ent rotor circuit; (f) exact equivalent circuit;
circuit; (e) stator-r
(g) approximate equivalent circuit.

a rotor circuit having an impedance

the slip-frequency induced emf sE) acting in
the current that would be seen by an
per phase of r, + jsx2. In other words, this is
ermore, the amount of real power
observer riding with the rotor winding. Furth
ed times the real part of the rotor
involved in this rotor circuit is the current squar
rotor copper loss per phase. Hence
impedance. In fact, this power represents the
as
the total rotor copper loss may be expressed
166 Three-Phase Induction Motors Chap. 4

a)
where q> denotes the number of rotor phases.

The Equivalent Rotor Circuit

By dividing both the numerator and the denominator of Eq. (4-18) by the slip s
we get
“{ E,
Gls)+ix es,
The corresponding circuit interpretation of this expression appears in Fig. 4-7(c).
Note that the magnitude and phase angle of J, remain unaltered by this operation.
However, there is a significant difference between Eqs. (4-18) and (4-20). In the
latter case I, is considered to be produced by a line-frequency voltage E, acting in
a rotor circuit having an impedance per phase of r2/s + jx). Hence the J, of Eq.
(4-20) is a line-frequency current, whereas the J; of Eq. (4-18) is a slip-frequency
current. It is important that this distinction be understood.
Manipulation of Eq. (4-18) by s has enabled us to go from an actual rotor
circuit characterized by constant resistance and variable leakage reactance [see
Fig. 4-7(b)] to one characterized by variable resistance and constant leakage reac-
tance [see Fig. 4-7(c)]. Moreover, the real power associated with the equivalent
rotor circuit of Fig. 4-7(c) is clearly

P= 132 (4-21)
Hence the total power for q2 phases is

A comparison of this expression with Eq. (4-19) indicates that the power associ-
ated with the equivalent circuit of Fig. 4-7(c) is considerably greater. For exam-
ple, in a large machine a typical value of s is 0.02. Hence P, is greater than the
actual rotor copper loss by a factor of 50.
What is the meaning of this power discrepancy? The answer lies in the fact
that by Eq. (4-20) Ip is a line-frequency current. This means that the point of
reference has changed from the rotor (where slip-frequency quantities exist) to the
stator (where line-frequency quantities exist). In accordance with the circuit rep-
resentation of Fig. 4-7(c) the observer changes his or her point of reference from
the rotor to the stator. This shift is significant because now, upon looking into the
rotor, the observer sees not only the rotor copper loss but the mechanical power
Sec. 4-3 The Equivalent Circuit 167

developed as well. The latter quantity is included because with respect to a

stator-based observer the rotor speed is no longer zero as it is relative to a rotor-
based observer. As a matter of fact Eq. (4-22) gives the total power input to the
rotor. It is the power transferred across the air gap from the stator to the rotor.
We can rewrite Eq. (4-22) in a manner that stresses this fact:
‘ r
(4-23)
me
Poe ali = = gol5 E i ma “a 5)|

In other words, the variable resistance of Fig. 4-7(c) may be replaced by the actual
the
rotor winding resistance r, and a variable resistance R,,, which represents
mechanical shaft load. That is,

(4-24)

load to be
This expression is useful in analysis because it allows any mechanical
Figure 4-7(d) depicts the modi-
represented in the equivalent circuit by a resistor.
Finally, it should be apparent on the
fied version of the rotor equivalent circuit.
equivale nt circuit is equivale nt only
basis of the foregoing remarks that the rotor
of the rotor current per phase are
insofar as the magnitude and phase angle
concerned.

The Stator-Referred Rotor Equivalent Circuit

is a line-frequency quan-
The voltage appearing across terminals a-b in Fig. 4-7(a)
ng across terminals a '-b' in
tity having N, K,,; effective turns. The voltage appeari
has N2Kw2 effective turns. In
Fig. 4-7(d) is also a line-frequency quantity but
be joined to a-b in Fig. 4-7(a) to
general E, # E, so that a '_b' in Fig. 4-7(d) cannot
this it is necessa ry to replace
yield a single-line equivalent circuit. To accomplish
having N, K,,; effecti ve turns
the actual rotor winding with an equivalent winding
all the rotor quantit ies must be
as was done with the transformer. In other words,
Fig. 4-7(e). The prime notatio n is
referred to the stator in the manner depicted in on
The derivat ion of the reducti
used to denote stator-referred rotor quantities. ix
to the stator appears in Append
factors to be used in referring a rotor quantity
wound rotor and squirre l-cage rotor machines are treated there.
D. Both the

The Complete Equivalent Circuit

in Fig. 4-7(e) is the same as that

The voltage appearing across terminals a-b
. Hence these terminals may be
appearing across terminals a-b in Fig. 4-7(a)
t as it appears in Fig. 4-7(f). Note
joined to yield the complete equivalent circui
inding transformer.
that the form is identical with that of the two-w
 168 Three-Phase Induction Motors Chap. 4

The Approximate Equivalent Circuit

Considerable simplification of computation with little loss of accuracy can be

achieved by moving the magnetizing branch to the machine terminals as illus-
trated in Fig. 4-7(g). This modification is essentially based on the assumption that
V, ~ E, = E5. All performance calculations will be carried out using the approxi-
mate equivalent circuit.

4-4 COMPUTATION OF PERFORMANCE

When the three-phase induction motor is running at no-load, the slip has a value
very close to zero. Hence the mechanical load resistor R,, has a very large value,
which in turn causes a small rotor current to flow. The corresponding electromag-
netic torque, as described by Eq. (3-27), merely assumes that value which is
needed to overcome the rotational losses consisting chiefly of friction and wind-
age. If a mechanical load is next applied to the motor shaft, the initial reaction is
for the shaft load to drop the motor speed slightly and thereby increase the slip.
The increased slip subsequently causes J, to increase to that value which, when
inserted into Eq. (3-27), yields sufficient torque to provide a balance of power to
the load. Thus equilibrium is established and operation proceeds at a particular
value of s. In fact for each value of load horsepower requirement there is a unique
value of slip. This can be inferred from the equivalent circuit, which shows that
once s is specified then the power input, the rotor current, the developed torque,
the power output, and the efficiency are all determined.
The use of a power-flow diagram in conjunction with the approximate equiv-
alent circuit makes the computation of the performance of a three-phase induction
motor a straightforward matter. Depicted in Fig. 4-8(a) is the power flow in state-
ment form. Note that the loss quantities are placed on the left side of a flow
point. Appearing in Fig. 4-8(b) is the same power-flow diagram but now ex-
pressed in terms of all the appropriate relationships needed to compute the perfor-
mance. It should be clear that to calculate performance one must first compute
the currents J, and J; from the equivalent circuit and then make use of the perti-
nent relationships depicted in Fig. 4-8(b).
Example 4-1
A three-phase, four-pole, 30-hp, 220-V, 60-Hz, Y-connected induction motor draws
a current of 77 A from the line source at a power factor of 0.88. At this operating
condition, the motor losses are known to be the following:

Stator copper losses = P.,; = 1033 W

Rotor copper losses = Pay = 1299 W

Stator core losses = P,. = 485 W

Rotational losses (friction, windage, and iron losses due to rotation)

= Pr = 540 W
Sec. 4-4 | Computation of Performance 169

Electrical power input

omen USEF] POWer
-" seme POWEr losses

Stator core loss Power transferred across the air-gap

Stator copper loss (i.e. rotor input power)

eee a

Rotor copper loss Mechanical power developed

s

eo ---""

| Rotational losses Mechanical power output

2 losses (a)

P,=qi\4l, cos

Ai 2i ls AP moet
Realy F = a2lg Ff= aE], cs 6, = q2Egl cos Op = Tay
eo
2 or"

QF =Q15t2=8P Pn=Ri1-8)
=Tp

= losses (b)

Figure 4-8 Power-flow diagram: (a) statement form; (b) equation form.

developed
Determine: (a) the power transferred across the air gap, (b) the internally
(c) the slip expressed in per unit and in rpm, (d) the
torque in newton-meters,
l power developed in watts, (e) the horsepow er output, (f) the motor speed
mechanica
(h) the torque
in rpm and in radians per second, (g) the torque at the output shaft,
(i) the efficiency of operation at the stated
needed to overcome the rotational losses,
condition.
the application of the
Solution The solution to this problem provides an exercise in
power flow diagram.
Thus
(a) Sufficient information is cited to permit finding the input power.
P, = V3 VI, (pf) = V3(220)77(0.88) = 25,820 W

Figure 4-8(a) then shows that the gap power is

P, = P; — P. — Pou = 25,820 — 485 — 1033 = 24,302 W

(b) Here use is made of the important relationship

Tw, = P,
170 Three-Phase Induction Motors Chap. 4

where in this case

=F p,= % ((1202) =-
Ws = 35 Ms = 3 soo)=
39 (1800) = 188.5 rad/s

Hence the internally developed torque is

12 24,302
T =
a
=a =
188.5
a =
128.93 N-m
-

(c) By Eq. (4-22), we have

Eig te i
= P. = 24302 ~ 0.0535 p-u_

Then

slip in rpm = sn, = (0.0535)(1800) = 96.2 rpm

(d) Again from the power flow diagram the mechanical power developed can be
expressed as

Pm = Pg — Pe = 24,302 — 1299 = 23,003 W

(e) For the output power we get
Po = Pm — Prot = 23,003 — 540 = 22,463 W
Accordingly

he (P28 463
= 30.1
Pa 46 546
(f) The actual motor speed is the synchronous speed less the slip in rpm. Thus

n =n, — sn, = 1800 — 96.2 = 1703.8 rpm

(g) The expression for the output torque is

where w,, is the actual motor speed expressed in radians per second, or

7 7
Om = 36 (n) = 30 (1703.8) = 178.42 rad/s

Therefore,

GAP? K22,403
T, = On 178.42 ~ 125.9 N-m

(h) The difference between the internally developed torque of 128.93 N-m and
the output torque of 125.9 N-m is the torque required to supply the rotational losses
associated with the operating speed of 1703.8 rpm. Thus

Trot = T — T, = 128.93 — 125.9 = 3.03 N-m

Sec. 4-4 Computation of Performance 171

It is instructive to observe that this same torque can be found from the expression
that relates torque, speed, and power. Accordingly, we can also write
Pci «DAY
Tot = ite = 178.42 = 3.03 N-m

(i) The efficiency is

IP Beil
heii Pane= DS 820
eis 0.87

Thus the motor operates at an efficiency of 87%.

Example 4-2
has the
A three-phase, four-pole, 50-hp, 480-V, 60-Hz, Y-connected induction motor
following parameters per phase:
rm = 0.10 0, x, = 0.35 0

r5 = 0.12 Q, x, = 0.40 0
rotational losses
It is known that the stator core losses amount to 1200 W and the
the motor draws a line current of 19.64 A at a
equal 950 W. Moreover, at no-load
power factor of 0.089 lagging.
and
When the motor operates at a slip of 2.5%, find: (a) the input line current
torque in newton-m eters, (Cc) the
power factor, (b) the developed electromagnetic
horsepower output, (d) the efficiency .
Hence the
Solution (a) The computations are carried out on a per-phase basis.
V, and the equivale nt circuit is depicted in
phase voltage is 480/\/3 or 277.13
Fig. 4-9. The stator-referred rotor current then follows from

is = ry + (r3/s)
V;
+ f(xy + x3)
6
4.9
2H AG
+ j0.75

=1155,97 (8.1 = 55,33 —=JhAd

taken equal to the no-load
For all practical purposes the magnetizing current may be
is negligib ly small. Thus
current because the corresponding rotor current
In = 19.64 /-85° = 1.75 — j19.58

f, ry =O1 —jlxy* XQ)=j0.75

19.64/-85° —-l2
Fadevo.ig? 2
Ves eC lalone eB cae Ggs entice

Figure 4-9 Equivalent circuit for Example 4-2.

172 Three-Phase Induction Motors Chap. 4

Hence, the input line current is

T, = I, + 15 = (55.33 + 1.75) — (19.58 + 8.47)

= 97-08 1 28.05)=63.6/—26.2° (4-25)
and
power factor = cos 0, = cos 26.2° = 0.895 lagging

(b) The developed torque is found from

jee (4-26)
Ws

Also

_ 2mns _ 27(1800)
60 60 = 607 rad/s

and

P, = qi? - = 3(55.97)24.8 = 45,110 W

Therefore

= eae = 239.32 N-m

(c) From the power-flow diagram the power output is

Po = Pm — Prt = PZ — 5) — Prot = 43,982.3 — 950 = 43,032.3 W

The horsepower output is therefore

43 032.3
horsepower = = 57.68
0.746
Note that this is slightly greater than the rated horsepower of 50. Rated horsepower
occurs at a slip somewhat less than 2.5%.
(d) It is more accurate to find the efficiency from the relationship

> losses
ys (4-27)
Pi

rather than from the output-to-input ratio. The tabulation of the losses is as follows:

P.. = core loss = 1200 W

stator copper loss = q,J?r; 3(63.6)70.1 = 1213.5 W

rotor copper loss = qlyrs = sP, = 0.025(45,110) = 1127.8 W

Prot = rotational loss = 950 W

2 losses = 4491.3 W
 Sec. 4-5 Correlation of Induction Motor Operation 173

Also, the input power is

P; = V3 480(63.6)0.895 = 47,324 W

Hence

_ , _ 44913 = | — 0.095 = 0.905

il 47,324
The efficiency is 90.5%.

4-5 CORRELATION OF INDUCTION MOTOR OPERATION

WITH THE BASIC TORQUE EQUATIONS
of the
The variation of torque with speed (or slip) is an important characteristic
shape of this curve can be identified in
three-phase induction motor. The general
(3-22)] and knowledg e of the performa nce
terms of the basic torque equation [Eq.
as at no-
computational procedure. When the motor operates at a very small slip,
current is very small—ju st enough to
load, Eq. (4-18) indicates that the rotor
losses. Moreover , the power-fac tor angle
develop torque to supply the rotational
of the rotor current is practically zero. That is,

6, = tan! i = 0° for s very small (4-28)

-cage induc-
Appearing in Fig. 4-10 is a cross-sectional view of a two-pole squirrel
the stator winding and the
tion motor showing the rotating flux field produced by
of the rotor. For the in-
corresponding induced ampere-conductor distribution
rotation of the field the emf
stant shown and the assumed counterclockwise
the plane of the paper. Why?
induced in bar 1 is a maximum and directed out of
m value and that
Note that the emf induced in bars 2 and 10 is less than the maximu
less still. If the value of the bar
induced in bars 3 and 9 at the same time instant is
ement along the rotor periph-
emf’s are plotted as a function of the spatial displac
See Fig. 3-12(b) of Appen-
ery, a discrete distribution of sinusoidal shape results.
actor angle is practica lly zero
dix to Chapter 3. Furthermore, because the power-f
is almost identica l with the emf
at low slips, the distribution of rotor currents
in Fig. 4-10 by dots and crosses
distribution. The current distribution is denoted
dot (or cross), the larger is the
placed outside the rotor surface. The thicker the
value of the current represented.
appearing in Eq. (3-22).
Let us now relate this picture to the quantities
omagne tic torque equation, two
Because there are two forms of the basic electr
form of Eq. (3-22) first, we note again
interpretations are possible. Taking the last r
field ®@, an ampere-conducto
that torque depends upon the presence of a flux
and a space displacement angle
distribution (or current sheet) as implied by J,
slips near zero the ampere-conduc-
w. Examination of Fig. 4-10 discloses that for
distribution, i.e., y ~ 0°. This
tor distribution is in space phase with the field
174 Three-Phase Induction Motors Chap. 4

Revolving sinusoidal flux Note: Dots and crosses inside

field produced by three-phase rotor bars denote instantaneous
stator winding. value of induced emf’s. Dots
and crosses outside rotor bars
denote corresponding current
values.

Figure 4-10 Cross-sectional view of two-pole squirrel-cage induction motor, de-

picting the flux field and corresponding induced rotor mmf distribution at very
small slips.

means that the best field pattern for the development of torque exists—each bar
produces a positive torque. However, in spite of this, very little net torque is
developed at such low slips because the value of rotor ampere-conductor distribu-
tion is also near zero.
In the second interpretation the rotor is treated like a solenoid. In Fig. 4-10
note that the rotor current distribution of dots and crosses creates an mmf directed
in quadrature with it (along the horizontal toward the right in this case). This mmf
is represented for one pole by the dashed-line sine wave and is assumed to have a
peak value of %,. Keep in mind that the magnitude of ¥, is directly dependent
upon the rotor current and is practically zero for very low values of slip. Since the
torque angle is the angle between the axis of the flux field distribution and the axis
of the armature (or rotor) mmf distribution, we see that its value here is almost
90°, again attesting to the existence of an optimum field pattern for the production
of torque.
What happens to the field pattern as mechanical load is applied to the motor
shaft? As already described in Sec. 4-4 the immediate reaction is for the motor
speed to decrease slightly, which increases the slip. The increased slip produces
an almost direct increase in rotor current because for values of slip below 10% the
sXx7 term in Eq. (4-18) is negligibly small. When rated horsepower is delivered the
slip lies between 3% and 5% for most induction motors and the value of the rotor
power-factor angle lies in the vicinity of 10° (compare with Example 4-1). In spite
of this, however, and for the convenience of pictorial representation we shall
assume that at rated load the slip is such that Eq. (4-28) yields a rotor power-factor
Sec. 4-5 Correlation of Induction Motor Operation 175

angle of 36°. Note that this is the number of degrees separating the rotor bars in
the machine configuration of Fig. 4-10. If the field distribution is assumed to take
the same instantaneous position as depicted in Fig. 4-10, then the emf distribution
must be the same. This is shown with the appropriate dot-cross notation inside
the rotor bars of Fig. 4-11(a). Note that the emf in bar 1 is a maximum because it is
under the influence of the maximum value of the flux density. As a result of the
counterclockwise rotation of the stator flux field we know that the emf was a
maximum in bar 2 by an amount of time equal to that which the flux field takes to
travel the distance between two bars, or more specifically of wt = 36°. At this
point it is helpful to recall the meaning of the power-factor angle @), which is the
phase angle between E, and I>. When 6; is a lag angle, it states that the peak value
the
of rotor current occurs 6> electrical degrees in time after the peak value of
rotor induced emf has occurred. Hence for the time instant illustrated in Fig.
4-11(a) the peak outward-directed rotor current exists in bar 2. By the same
and 4,
reasoning outward-directed currents can be found to exist in bars 1, 10, 3,
the cosine of the angle
but the magnitudes will be less than the peak current by
is confined just
that these bars are displaced from bar 2. As before, our attention
rotor ampere-con -
to fundamental components of all distributions. The resulting
4-11(a) and repeated for
ductor distribution then takes the position shown in Fig.
(or
clarity’s sake in Fig. 4-11(b). Note that bars 4 and 9 produce a negative
clockwise) torque. This is the reason that the field pattern is not optimum.
in Fig. 4-10
Nonetheless, the developed torque is much greater here than it is
at Eq. (3-27), which is
because the rotor current is many times greater. A glance
the quantities appear-
the practical form of Eq. (3-22), emphasizes this point. Of
voltage. Hence, if Ih
ing there, ® is essentially fixed by the applied terminal
decreases, the torque increases. In normal
increases at a greater rate than cos
operation the value of cos y rarely dips below 0.95.
4-10 and
These discussions revolving about the situations depicted in Figs.
n motor the space-
4-11 lead to a very useful conclusion: In the polyphase inductio
is equal to the power-factor angle of the armatur e winding,
displacement angle
i.é.,

(4-28a)

pment of torque in the

One final comment is in order concerning the develo
results of Eqs. (3-27) and (3-22)
polyphase induction motor. Keep in mind that the
-condu ctor distributions are
are valid provided that the flux field and the ampere
condit ion satisfi ed in the situa-
stationary in relation to one another. How is this
speed of the stator flux field
tion of Fig. 4-112 To begin with we know that the
relative to the stator structure is w,. It therefore remains to be shown that the
relative to the stator struc-
rotor ampere-conductor distribution has also a speed speed of
requency currents, the
ture of w,. Since the rotor winding carries slip-f
176 Three-Phase Induction Motors Chap. 4

=0,

ye Rotor ampere
t* conductor distribution
| (reversed )

Axis of rotor (armature)

mmf wave. Axis of
rotor N-pole lies
along here.

® \.— Rotor mmf wave

(a)

Rotor mmf wave

Rotor ampere
conductor distribution

@©-- emf
distribution
*~Current
distribution
“ in rotor

+ Rotor rotation

A=54° y=0,=36°
Field axis

(b)
Figure 4-11 Depicting the field distribution and associated ampere-conductor
and mmf waves of the rotor for 0, = tan~! (sx/r,) = 36°. Only the fundamental
components are shown. The field travels from right to left. (See p. 257 for defini-
tion of A.)
 Sec. 4-6 Torque-Speed Characteristic: Starting and Maximum Torques 177

the rotor ampere-conductor distribution relative to the rotor is at slip frequency,

i.e., sw,. However, because the rotor bars are imbedded in the rotor iron, which
in turn revolves at a speed of w, = w,(1 — s) with respect to the stator structure, it
follows that the total speed of rotation of the rotor ampere-conductor distribution
is the sum of sw, and w,(1 — s). Clearly this is w,. Hence the rotor ampere-
conductor distribution and the flux field distribution are stationary in relation to
each other. In fact this condition is fulfilled for all values of slip and is a conse-
quence of the singly energized nature of the induction motor.

4-6 TORQUE-SPEED CHARACTERISTIC: STARTING

AND MAXIMUM TORQUES

The torque-speed characteristic of the induction motor can be explained in light of

Eqs. (3-27) and (3-22). As slip is allowed to increase from nearly zero to about
10%, Eq. (4-18) shows that the rotor current increases almost linearly. Moreover,
for this same range of slip, w varies over a range of about zero to 15 degrees. This
means that cos & remains practically invariant over the specified slip range and so
torque increases almost linearly in this region.
As slip is allowed to increase still further, the rotor current continues to
increase but much less rapidly than at first. The reason lies in the increasing
importance of the sx, term of the rotor impedance. In addition, the space angle w
now begins to increase at a rapid rate which makes the cos diminish more
rapidly than the current increases. Since the torque equation now involves two
opposing factors, it is entirely reasonable to expect that a point is reached beyond
which further increases in slip culminate in decreased developed torque. In other
words, the rapidly decreasing cos factor predominat es over the slightly increas-
ing I> factor in Eq. (3-27). As # increases, the field pattern for producing torque
becomes less and less favorable because more and more conductors that produce
4-11
negative torque are included beneath a given pole flux. A glance at Fig.
. Accordingly , the composite torque-spee d
makes this statement self-evident
curve takes on a form similar to that shown in Fig. 4-12.
speed n
The starting torque is the torque developed when s is unity, i.e., the
s that for the case illustrat ed the starting torque is
is zero. Figure 4-12 indicate

200

rated)
(%
of
Torque

? Figure 4-12 Typical torque-speed

=
Sm Os S10 curve for a three-phase induction
0 n,<— Speed motor.
178 Three-Phase Induction Motors Chap. 4

somewhat in excess of rated torque, which is fairly typical of such machines. The
starting torque is computed in the same manner as torque is computed for any
value of slip. Here it merely requires using s = 1. Thus the magnitude of the rotor
current at standstill is

lS
Vi SS (4-29)
V(r, + 15)? + (xy + x3)?
The corresponding gap power is then
2a
qi Vir2
Ps = qi Pry =
(+ ri + Gey + xd? ae
It is interesting to note that higher starting torques can result from increased rotor
copper losses at standstill.
At unity slip the input impedance is very low so that large starting currents
flow. Equation (4-29) makes this apparent. In the interest of limiting this exces-
sive starting current, motors whose ratings exceed 3 hp are usually started at
reduced voltage by means of line starters. This matter is discussed further in
Sec. 4-10. Of course, starting with reduced voltage also means a reduction in the
starting torque. In fact if 50% of the rated voltage is used upon starting, then
clearly by Eq. (4-30) it follows that the starting torque is only one-quarter of its
full-voltage value.
Another important torque quantity of the three-phase induction motor is the
maximum developed torque. This quantity is so important that it is frequently the
starting point in the design of the induction motor. The maximum (or breakdown)
torque is a measure of the reserve capacity of the machine. It frequently has a
value of 200% to 300% of rated torque. It permits the motor to operate through
momentary peak loads. However, the maximum torque cannot be delivered con-
tinuously because the excessive currents that flow would destroy the insulation.
Since the developed torque is directly proportional to the gap power, it
follows that the torque is a maximum when P, is a maximum. Also, P, is a
maximum when there is a maximum transfer of power to the equivalent circuit
resistor r3/s. Applying the maximum-power-transfer theorem to the approximate
equivalent circuit leads to the result that

22 = WT + (x) + x4? (4-31)

That is, maximum power is transferred to the gap power resistor r3/s when this
resistor is equal to the impedance looking back into the source. Accordingly, the
slip 5 at which the maximum torque is developed is

r3
Vit +Gy SS
SS
Fai aa,
4-32
Sec. 4-6 Torque-Speed Characteristic: Starting and Maximum Torques 179

Note that the slip at which the maximum torque occurs may be increased by using
a larger rotor resistance. Some induction motors are in fact designed so that the
maximum torque is available as a starting torque, i.e., 5, = 1.
With the slip s,, known, the corresponding rotor current can be found and
then inserted into the torque equation to yield the final form for the breakdown
torque. Thus

a qvVi (4-33)
Ws 2[r; ae Vrt + (x + x47)

An examination of Eq. (4-33) reveals the interesting information that the maxi-
mum torque is independent of the rotor winding resistance. Thus increasing the
rotor winding resistance increases the slip at which the breakdown torque occurs,
but it leaves the magnitude of its torque unchanged. Figure 4-13 shows the effect
of increasing the rotor resistance on a typical torque-speed curve.
A word of caution is appropriate at this point. Keep in mind that the results
appearing in Eqs. (4-29), (4-30), and (4-33) are predicated on the approximate
equivalent circuit, which is used whenever the voltage drop across the primary
leakage impedance is small compared to the stator induced emf per phase. This
assumption is quite valid for normal operating conditions spanning the region from
no-load to rated load. However, when interest becomes focused on maximum
torque or starting torque at rated voltage, then the primary leakage voltage drop
begins to exert some noticeable influence. Therefore, if more accurate results are
required, we can base the analysis on the more exact equivalent circuit of Fig.
4-7(f) (see Example 4-4). Usually, errors of the order of 3% are incurred when
Be-
these calculations are based on the simpler, approximate equivalent circuit.
cause in practical situations the extent of these errors is often less than the effects
due to saturation and skin effects, which are not accounted for either, our policy
here is to stay with the approximate equivalent circuit.
Example 4-3
takes place
Refer to the induction motor of Example 4-2. Assuming that operation
(a) the slip at
at full-rated voltage with all saturation effects neglected, determine

High rotor Low rotor

:
resistance resistance
Torque

Figure 4-13 Showing the effect of

increased rotor resistance on torque-
Sm Sm 0 ~+—Slip speed curves.
1
180 Three-Phase Induction Motors Chap. 4

which the maximum torque is developed, (b) the current at maximum torque, and
(c) the value of the maximum torque. Compare these results with those found in
Example 4-2.

Solution The solution is obtained here employing the results obtained from use of
the approximate equivalent circuit.
(a) Application of Eq. (4-32) yields

Caras eed enya edit Zee BE 0.159

™ VrE
+ (xy + x42 =VO.1? + 0.75)?
(b) The rotor current referred to the stator is

pie 273 uf 277.13 nyne277 13

2 ry + (5/8) + 7G, + x4) 0.1 + (0.12/0.159) + 0.75. 1.137 /41.3°
243.74 [—41.3° A
This current is about 3.83 times as large as the current when the motor delivers a
little in excess of 57 hp as computed in Example 4-2 at near rated conditions.
(c) The maximum torque follows directly from Eq. (4-33). Thus

gab ean V9 stetty nseMinowtVtlor) SII Aine doefloeteiles

"Qo + Vet Gt xSe 20885) 7, + Ve + Ge FP
= (611.5) :
OT tev
= 611.15(1.181) = 721.77 N-m
which is approximately three times as large as the value found near rated conditions
in Example 4-2.

Example 4-4
Repeat Example 4-3 using the exact equivalent circuit in place of the approximate
one.

Solution In order to use the exact equivalent circuit, the value of the magnetizing
reactance, x4, is needed. This is readily found from the data appearing in Example
4-2 as

aati
eo" = ogee 42 0
The exact equivalent circuit appears in Fig. 4-14(a). Observe that the core-loss
resistor is omitted for simplicity; its absence causes negligible effects.
To obtain a result analogous to Eq. (4-31) by reducing the circuit in Fig. 4-14(a)
to an equivalent series circuit, we invoke Thévenin’s theorem. We replace the cir-
cuit to the left of terminals ab by an equivalent voltage source of value

Y=aw otarcce)
oegee Oe nl:
Sec. 4-6 Torque-Speed Characteristic: Starting and Maximum Torques 181

=
al
S-

(a) (b)

Figure 4-14 (a) Exact equivalent circuit with the core-loss resistor omitted;
(b) the Thévenin equivalent of (a).

and an internal impedance which is simply the parallel combination of (r; + jx;) and
jx». Thus
Sl Soe
& a ry + J(Xy =e x4) om Ri a x

The Thévenin equivalent circuit is shown in Fig. 4-14(b). Itis instructive to note that
this circuit is now equivalent in form to the approximate equivalent circuit which
upon analysis led to Eq. (4-31). The role of V; is now replaced by V;, and R; replaces
r,, and X; replaces x,;. Accordingly, Eqs. (4-32) and (4-33) are now applicable to the
exact equivalent circuit provided that the just-mentioned replacements are made.
Hence, before proceeding with the evaluation of the items specified in this example,
we first determine these replacement quantities.
— j14.2 Oss ll;ls 27003 : Y
Vi oA aaa?
-_ f14.20.1 + 0.35) = 0.09 + j0.343 = R; + JX;
; 0.1 + j14.55
Observe the relatively small change that occurs in the value of R; from r; and _X; from
x,;. In other words, the shunting effect of the magnetizing reactance is not very
great.
(a) The slip at which the maximum torque now occurs is
Ir? 0.12
So = eae oes) = 0.161
VRE CX a eS) V(0.09)? + (0.343 + 0.45)°
using the
This represents a change of slightly more than 1% from the value found
approximate equivalent circuit.
(b) The value of the current at maximum torque iS

, Vi {Ca 2 el °
Lh +j(K + xi) 0-09 + 0.7453 + j(0.343 + 0.4)
Rare Be Tehasies
 182 Three-Phase Induction Motors Chap. 4

Here again the difference is about 1%. Of course, there are motors where the differ-
ences can be of the order of 3% or 4%. In such cases the use of the more exact
equivalent circuit might appear to be justified. But it is important to note here that
this amounts to quibbling because there is a more important effect taking place at the
maximum torque condition which we have not accounted for. With currents as large
as three times the rated current flowing at the maximum torque condition, a consider-
able amount of saturation occurs in the teeth of the stator and rotor. A subsequent
decrease in the values of the stator and rotor leakage reactances of the order of 50%
and more can easily occur. The effect of this change is much more serious than
whether the approximate or the exact version of the equivalent circuit is used.
(c) Continuing with the original premise in the statement of the problem as set
forth in Example 4-2, we get for the maximum torque

ao qv; o 3(270.5)"
"IR: + VR? + (X; + x5] 2(188.5)[0.09 + V(0.09)? + (0.343 + 0.4)7]
= (582.26)(1.193) = 694.77 N-m

Here the difference is nearly 4%.

4-7 EQUIVALENT CIRCUIT PARAMETERS FROM NO-LOAD TESTS

The performance computations for a three-phase induction motor presuppose

knowledge of the parameters of the equivalent circuit. This information may be
available either from design data or from appropriate tests. When the design data
are not available information about the magnetizing branch can be obtained from a
no-load test that is performed by applying a balanced three-phase voltage to the
motor which is uncoupled from its load. Often the input power and line current
are measured at rated voltage and rated frequency. Sometimes this no-load test is
performed with a varying voltage, which leads to the plot depicted in Fig. 4-15.
Point a corresponds to operation at rated voltage. Experimental points below b
are not taken because the speed will no longer be close to synchronous speed.

input
No—load
power

Friction and windage Figure 4-15 Showing the variation of

loss no-load input power with applied volt-
Vv; age; ab is the curve obtained from
Phase voltage experimental data.
Sec. 4-7 Equivalent Circuit Parameters from No-Load Tests 183

Extrapolation of the curve to the ordinate axis gives a good indication of the
friction and windage loss at normal speeds.
At no-load and rated voltage the input power is used to supply three losses:
the stator copper loss, the stator core loss P., and the rotational losses P,.. In
equation form we have
Po = qilor ol + Prot (4-34)

where gq; denotes the number of stator phases and 1; is the effective stator resis-
tance per phase. The rotational losses include the friction and windage losses as
well as iron losses caused by the pulsations of flux in the stator teeth as the rotor
revolves. These iron losses due to rotation are larger in machines with open slots
than in those with semiopen slots. When specific information about P. or Prot 18
not known, it is customary to assume these quantities to be equal for they are
frequently found to be so in conventional machines. Based on this assumption,
the current through the core-loss resistor can then be found as follows:

ie Es _ (Po = alori) (4-35)

+) .aiVs 2q1Vi
Hence

=> (4-36)

To find the magnetizing reactance we must first obtain the reactive compo-
nent of the no-load current, J,. The power-factor angle at no-load is found from

6) = cos! = i (4-37)

Therefore
I4 = Ip sin 60 (4-38)

so that

fot (4-39)
The phasor diagram at no-load is depicted in Fig. 4-16. The angle 6 is
usually large (65° to 80°) because of the need for a large magnetizing current to
produce the required flux per pole in a magnetic circuit containing air gaps. The
‘total in-phase component is represented by three terms to take care of the three
losses appearing in Eq. (4-34). Note that the phasor sum of J. and J yields the
branch of the equivalent circuit.
total magnetizing current I, of the magnetizing
The quantity /49 is assumed to be equal to /, and represents the small rotor current
184 Three-Phase Induction Motors Chap. 4

Figure 4-16 Phasor diagram of the

Ig induction motor at no-load.

that flows at no-load to supply the rotational losses. Specifically, this quantity
may be written as :
_ 2 (4-40)
Is = Prot = Po qilon

The third part of the in-phase component of the no-load current, J,;, represents
the stator copper loss at no-load. This is,

Ir;
Fi 7 Vi (4-41)

This too can be readily computed from the measured no-load data.
Information about the winding resistances and leakage reactances is ob-
tained from a blocked-rotor test. This test is analogous to the short-circuit test of
the transformer. It requires that the rotor shall be blocked to prevent rotation and
that the rotor winding shall be short-circuited in the usual fashion. Furthermore,
since the slip is unity, the mechanical load resistor R,, is zero and so the input
impedance of the equivalent circuit is quite low. Hence, in order to limit the rotor
current in this test to reasonable values, a reduced voltage must be used—usually
about 10% to 25% of rated value. Moreover, operation at such reduced voltages
renders the core loss as well as the magnetizing current negligibly small. Accord-
ingly, the equivalent circuit in this test takes on the configuration shown in Fig.
4-17.
Assume now that the following instrument readings are taken in performing
a blocked-rotor test on a Y-connected three-phase wound-rotor induction motor:

Py total wattmeter reading, W

I, = line current of Y connection
V, = line voltage of Y connection
Sec. 4-7 Equivalent Circuit Parameters from No-Load Tests 185

T, Re=ntte iX.= ilxjtx)

Mp
v3 ‘
Figure 4-17 Induction motor equiva-
lent circuit for the rotor-blocked test.

Assume too that the dc phase winding resistances of the stator and rotor windings,
ride and r2q:, are made available from a simple dc test and that the ratio of
transformation, a, from stator to rotor is also available. Keep in mind that the
instrument readings are consistent with operation at the rated frequency of the
motor. Hence, if present, the influence of skin effect? is reflected in these read-
ings. From the foregoing measurements it follows that the effective value of the
equivalent winding resistance is given by

(4-42)

Here both r; and r3 denote effective winding resistances per phase. To separate
out the effective value of r; use is made of the following equation:

(4-43)

This last expression states that the ratio of the effective value of r; to the effective
value of the equivalent resistance bears the same relationship as the dc value of r;
to the dc value of the equivalent resistance. It is important at this point to under-
stand that, in the equivalent circuit of the induction motor as it is used to compute
performance at normal values of slip, the r; to be used is the effective value as
determined by Eq. (4-43) but the r, to be used is the de value. The latter quantity
is used because at normal slips (3 to 5%) the frequency of the rotor current is very
low (2 to 3 Hz).
The equivalent phase impedance is obtained from

(4-44)

the surface of a
+Skin effect refers to the tendency of time-varying current to crowd toward
the effective cross-secti onal area and thereby increases the resistance over
conductor which reduces
the de value.
186 Three-Phase Induction Motors Chap. 4

Finally, the equivalent leakage reactance is determined from

X= VLR Sy FX (4-45)

Note that, as long as computations are carried out using the approximate equiva-
lent circuit, it is sufficient to deal directly with X, without a further breakdown
into x, and x3.
In view of the importance of working with the dc value of r, in the equivalent
circuit for low values of slip, how is this value found for the squirrel-cage induc-
tion motor if the design data are not available? The use of the effective value of rz
leads to very serious error because the skin effect in the squirrel cage is very
large. This is caused by the use of solid copper bars imbedded in iron. The ratio
of effective to dc value at 60 Hz is often 3:1. Such large ratios do not occur in the
wound-rotor machine because of the use of stranded conductors. Since the phase
terminals of a squirrel cage are not available, an indirect means of determining the
dc value of the rotor phase resistance must be.employed. If the blocked-rotor test
is performed for various frequencies at reduced voltage, the effective equivalent
resistance can be found and plotted to yield the curve depicted in Fig. 4-18. Note

resistance
Equivalent

Figure 4-18 Illustrating how dc equiv-

alent resistance of a squirrel-cage
induction motor can be found from a
blocked-rotor test for varying fre-
Frequency quency.

that as the frequency is reduced the effective equivalent resistance assumes corre-
spondingly smaller values. By extrapolating the experimental curve to the ordi-
nate axis the dc equivalent resistance for the squirrel-cage motor is obtained.
Then the stator-referred dc resistance per phase of the rotor is found as

r) de a Rese =" Mysto (4-46)

Accordingly, in the equivalent circuit for the squirrel-cage induction motor at low
slips, the effective value of r; and the dc value of r; [Eq. (4-46)] must be used.
Example 4-5
The following no-load tests aré performed on a three-phase, four-pole, 60-Hz, 20-hp,
550-V, Y-connected squirrel-cage induction motor and the results are found to be:
Sec. 4-7 Equivalent Circuit Parameters from No-Load Tests 187

No-load test: 550 V, S5.8A, 754 W, 60 Hz

Blocked-rotor test at rated frequency: 123 V, 25 A, 2419W, 60 Hz
Blocked-rotor test at low frequency: SDV DON Aee ZOOS aWVe me tack,
De test on stator per phase: POV elo ek

The friction and windage losses for this motor are known to be 328 W. Determine the
parameters of the approximate equivalent circuit to be used at normal operating slips
(i.e., 3% to 5%).

Solution The dc stator winding resistance per phase is

lS)
ly de = 5 = 0.60

Although the stator winding is a wound winding using stranded wire, some small skin
effect is present. Therefore, the ac value of the winding resistance is needed. This
can be determined from the blocked-rotor test performed at rated frequency. Thus

2419
Inet “3056 29.0)

From the blocked rotor test which is performed at low frequency, the skin effect is
not considered to be important and so the equivalent ‘‘dc’’ resistance is found to be

Then, in accordance with Eq. (4-43), the ac stator winding resistance has the value

rae = NieR.FE = 0.6 (FF)

1.29
edc
= 0.70 °

This value is a bit on the pessimistic side for the reason that during the blocked-rotor
test at full frequency, the skin effect that takes place in the iron-embedded bars of the
rotor is greater than for the stranded wire of the stator winding. Nonetheless, this is
one of the six values of the equivalent circuit. A second parameter, the ‘“‘dc’’ value
of the rotor resistance referred to the stator, is determined by applying Eq. (4-46),
which here yields

Po de =s Rea =e VAs 1.1 — 0.6 = 0.50

The use of the approximate equivalent circuit means that the two parameters x;
leakage
and x; may be evaluated and used as a lumped quantity. Because these
the data of the rotor-bloc ked test at rated fre-
reactances are the standstill values,
quency are employed. According ly,

Abe eh Palins
Ze
Wate 325)
X, = V2 — R?,, = V2.84 = 1.29 = 2.53 O
Eq.
The two remaining parameters are found from the no-load test data. By
(4-37) the no-load power-factor angle is
 188 ; Three-Phase Induction Motors Chap. 4

Po eae Hi 754 .
0) = cos7! 82.2°
qaVilo 3(550/V3)5.8
so that the corresponding value of the reactive component of the no-load current
becomes

I, = Ip sin 09 = 5.8 sin 82.2° = 5.75 A

which in turn yields a magnetizing reactance of

V; 550
i, eee
Pine Ty DAS STS
Finally, the core-loss resistor is found by first calculating the value of core loss from

P, = Po — Pew — 31671 ac = 754 — 328 — 3(5.82)(0.7)

= 355.4 W
and then noting that

Hence the value of the core-loss resistor computes to be

_ gV¥t _ 3(550P _ (550)

= 851.20
“tg P, Msp 35.4

4-8 SPEED CONTROL

In most industrial applications the essentially constant speed characteristic of the

induction motor is a desirable one. However, there are some applications (e.g.,
conveyors, hoists, and elevators) where speed control capability is a transcendent
factor. Therefore in this section we explore the extent to which the induction
motor lends itself to adjustment of speeds under various load conditions.
The squirrel-cage induction motor is examined first because it is simpler
owing to the closed nature of its rotor winding, which eliminates all possibility of
control from that end. Any control that is possible must take place through ma-
nipulations at the stator. Here three possibilities present themselves, two of
which are evident from Eq. (4-12). The first is to change the line frequency. As
the line frequency is raised or lowered the synchronous speed is correspondingly
raised or lowered, which provides satisfactory control. However, the serious
shortcoming of this procedure is that a variable frequency source is not available
generally. Moreover, if such a source is to be provided locally, a motor-generator
set is needed equipped with appropriate controls and power rated comparably to
that of the induction motor it is to control. Furthermore, if maximum iron-flux
densities are not to be exceeded, as the frequency is dropped, corresponding
decreases in the applied voltage must be effected.
Sec. 4-8 Speed Control 189

A second method of control is to change the number of poles. Recall that

the number of poles is determined by the winding layout. It is possible to arrange
the winding in appropriately paired sections for each phase. When the two sec-
tions, for example, are connected in a way to ensure symmetrical orientations of
current flow, a four-pole machine results. However, if the current through the
second section is made to flow in the reversed direction by throwing a switch, a
two-pole machine results. Accordingly, the synchronous speed can be changed
by a factor of 2. Such a machine is called a multispeed motor. The motor can
operate at one of two speeds, but unfortunately once a particular speed is selected
there is no further control available. Speed control in this instance occurs in a
discrete fashion. For a line frequency of 60 Hz such a machine allows operation
near either 3600 or 1800 rpm. If speed control is desired between 1800 and 3600
rpm, it cannot be achieved with the multispeed squirrel-cage induction motor.
The third method of speed control involves reducing the applied line volt-
age. Operation at reduced voltages means that the ordinate values of torque on
the torque-speed curve assume new values in accordance with the square of the
reduced voltage ratio. Thus, if the squirrel-cage motor is operated at 70.7% of
rated voltage, all points on the new torque-speed curve have half the value of the
original curve. The maximum torque is reduced by one-half ; the starting torque is
also reduced to one-half its original value; and so on. Figure 4-19 depicts the
distinctions over the complete speed range. Also shown are the torque-speed
curves of a constant-torque load so characteristic of conveyors and the variable
load torque characteristic of fans and other propeller type loads. It is assumed
that at rated voltage the induction motor delivers rated torque to each load type.

V, (rated)

eet pi — torque
rated
of
Per
torque
cent _ 2 loa
Variable— torque -——~ ©
load— -_——
= <a
-——
—

s=O
s=1
n=0 n=n,
by reduc-
Figure 4-19 Illustrating how some degree of speed control is obtained
ing the applied voltage.
190 Three-Phase Induction Motors Chap. 4

The operating point is denoted by a, where the slip is in the vicinity of about 5%.
A reduction of 30% in the applied voltage causes the operating point for the
constant-torque load to move from a to b and from a to c for the variable-torque
load. Note that this scheme may allow the slip to double or perhaps triple. A little
thought indicates that this is not an appreciable amount of speed adjustment—and
certainly not for the price paid. Speed control by voltage control has serious
disadvantages which account for its infrequent use, especially for polyphase mo-
tors. First, it is expensive. Variable three-phase voltage is obtained efficiently
through the employment of a motor-generator set or a three-phase variac having
ratings at least as large as the motor being controlled. Second, when the motor is
operating at reduced voltage, its reserve capacity is dangerously reduced. For the
situation illustrated in Fig. 4-19, note that the reserve capacity for the constant-
torque load is 250% of rated torque at full voltage and only 125% at 70.7% of rated
voltage. Third, if an attempt were made to operate the motor over the full range at
the reduced voltage, a glance at Fig. 4-19 would make it clear that only the
variable-torque load can be so operated. At reduced voltage the motor’s starting
torque of 62.5% is insufficient to move the constant-torque load.
In conclusion therefore it can be stated that speed adjustment of the squirrel-
cage induction motor is difficult to achieve and expensive besides. This motor is
best suited for constant-speed loads.
Speed control is easier to achieve in the wound-rotor induction motor than in
the squirrel-cage motor because of the availability of the rotor terminals. Of
course the adjustment of speed by the three methods already outlined apply with
equal validity to the wound-rotor machine.
Three additional methods by which the speed can be controlled include the
following: increasing the total rotor reactance per phase, x2; increasing the total
rotor resistance per phase, r3; and injecting suitable voltages at the rotor termi-
nals. The use of an external three-phase reactor connected to the rotor terminals
is rarely, if ever, used because of its very adverse effect on the line power factor.
In addition the size of the reactor needs to be large because of the effect of the slip
term. By far the most common method of adjusting the speed of a three-phase
induction motor is to use an external three-phase resistor connected to the rotor
terminals. The larger the value of this external resistance per phase, the lower the
speed is. The effect is illustrated in Fig. 4-20. Recall from Eq. (4-32) that, as
resistance is added to the rotor winding, the slip at which the maximum torque is
developed increases. When this is combined with, say, the constant-torque char-
acteristic of the load, we see that the same torque can be delivered over a consid-
erable range of speed. It is customary with this method to obtain speed control to
about 50% below synchronous speed. Beyond this point the efficiency becomes
quite poor, being less than 50%. In fact this is the price paid for attaining speed
control. When rated torque is delivered to a load at a slip of 50% rather than the
usual 5% the power delivered to the load is less by about 45 per cent. This
difference of 45% of rated power is consumed as heat in the external resistors and
so is unrecoverable.
 Sec. 4-8 Speed Control 191

Some rotor
resistance added NO external
rotor resistance

rated
of
torque
Per
cent

s=1 s=O — Figure 4-20 Speed control by external

n=0 n =n, __ rotor resistance.

From the nature of the torque-speed curve we know that at a specified slip s
there occurs a developed torque T. How do we determine the amount of resis-
tance to be added to the rotor winding per phase so that this same developed
torque prevails at an increased slip s’? To preserve correspondence with the
discussion involving the situation illustrated in Fig. 4-20, we continue with the
condition of a load that calls for constant developed torque. Because Tw, = P, and
is a fixed quantity, it follows that constant developed torque by the motor implies
constant gap power. When no external rotor resistors are used, the expression for
gap power is given by Eq. (4-22), which is repeated here for convenience. Thus,
.
Py tpl (4-47)
If we now assume that a resistor of value R, /phase is added to each of the three
phase windings of the rotor, then Eq. (4-47) becomes

P, = gol 2 s!
(4-48)

as
where s’ is used to indicate that the insertion of R, changes the operating slip
note that no distinction is made in the
demonstrated in Fig. 4-20. It is important to
under
symbol for the rotor current in Eqs. (4-47) and (4-48). This is because
remains essentially fixed. Reference
constant torque conditions the rotor current
right side are
to Eq. (3-27) makes this statement self-evident since all terms on the
ever so slightly closer to unity.
virtually constant, including cos , which moves
to Eq. (4-47), the equation for R, is
Consequently, on equating the last expression
found. The result is

Q/phase (4-49)
192 Three-Phase Induction Motors Chap. 4

Equation (4-49) states that by adding R, to each phase of the rotor winding, a
given torque that is developed at a slip s with the rotor winding short-circuited can
now be developed at an arbitrarily selected slip s’ subject to the constraint s <
s' = 1. Observe that by choosing s = s,, = slip at maximum torque and s’ = 1, the
maximum torque can be made to be the starting torque.
To illustrate, let us find the amount of resistance to be added to the rotor
winding per phase to yield the maximum torque at starting. Assume that this
torque normally occurs at a 10% slip with the rotor short-circuited. In this case
s’ = 1 ands = 0.1; hence

11031
[ee or, (4-50)
Ba 0.1
Thus, by adding an external resistance equal to nine times the actual rotor winding
resistance per phase, the starting torque is equal to the maximum torque.
The achievement of speed control by injecting a three-phase voltage into the
rotor winding can be understood by studying the sequence depicted in Fig. 4-21.
For simplicity it is assumed that rated torque is to be delivered to a constant-
torque load at varying speeds. Moreover, the rotor leakage reactance is assumed
to be negligibly small. This means that the space-displacement angle w of Eq.
(4-28a) is zero so that the basic torque equation is T = K®J,. Since ® is fixed by
the constant applied voltage, it follows that to maintain constant torque at various
speeds it is necessary merely to keep J, constant. Depicted in Fig. 4-21(a) is the
situation existing in the rotor at normal operation, i.e., with the rotor short-
circuited and rated motor torque delivered to the load. Note that the resultant slip
voltage sE, need only be large enough to overcome the rotor winding resistance
drop. Now assume that a voltage of rms value V; per phase is injected into the
rotor in such a direction as to oppose the original sE voltage. The immediate
effect of V; is to cause the rotor current to drop, which in turn causes the devel-
oped torque to decrease. Then, since the motor can no longer supply rated torque
to the load, the speed decreases. As the speed decreases, the slip emf increases
and thereby gradually overcomes the influence of V; on the rotor current. Equilib-
rium is again established when the resultant rotor voltage per phase is once more
equal to nr). This happens when algebraically

S'E, = SE> a V; (4-51)

where s’ denotes the new increased slip. The condition is depicted in Fig.
4-21(b). The net effect of the injected voltage is to bring about an increase in slip
in a manner not unlike that achieved with external rotor resistance. Asa matter of
fact the corresponding value of external rotor resistance that yields the same slip
can be computed from

InRa = V; (4-52)
The basic difference between the two methods, however, lies in the fact that,
whereas with the resistance method power is wasted as heat, the injected voltage
Sec. 4-8 Speed Control 193

n=n,= rated speed

(a)

s' negative
n'>n,

(e)

Figure 4-21 Speed control by the rotor injected-voltage method: (a) normal oper-
ation, with rated motor torque delivered to constant-torque load; (b) V, injected
opposed to sE); (c) V, injected in-phase with sE, of part (a) but less in magnitude;
(d) V, injected in-phase with original sE, and equal to it; (e) V, injected in-phase
with original sE, but greater in magnitude.

speed or
source can either usefully consume the power associated with the drop in
else return it to the line source.
is
What happens to the induction motor speed when the injected voltage
but smaller in magnitu de? The
introduced in phase with the original slip emf
is to cause I) to
condition is depicted in Fig. 4-21(c). The immediate effect of V;
accelera ting torque that raises
increase beyond the rated value, which causes an
ium is then establis hed when
the motor speed beyond the normal value. Equilibr
this is a
s'E, + V; = bn. It is interesting to note by way of comparison that
The use of exter-
situation that cannot be realized with external rotor resistance.
d slips—n ever at reduced
nal rotor resistance allows speed control at increase
slips.
just large
An interesting situation occurs when the injected voltage is made
to Fig. 4-21(d). The
enough to establish the rotor current entirely by itself. Refer
 Three-Phase Induction Motors Chap. 4
194

motor is thus relieved of having to play any part in producing the rotor current.
Accordingly, the slip becomes zero and the motor operates at synchronous
speed. This should not cause too much surprise because in point of fact the motor
is no longer singly but, rather, doubly excited and so has every right to behave like
a synchronous motor. Depicted in Fig. 4-21(e) is the situation where V; is made
larger than is needed to produce the required rated rotor current. Consequently,
the motor will accelerate above synchronous speed and cause operation to take
place at negative slips.
What price is paid for speed control by the injected rotor voltage method?
To begin with, an auxiliary device is needed to generate the injected voltage. An
important characteristic of the unit is that it must make available at the rotor
terminals of the induction motor a voltage for injection that always has the correct
slip frequency; otherwise, addition to the motor slip-frequency emf cannot be
achieved. Moreover, the size and rating of the auxiliary unit must be consistent
with the degree of control to be affected. Thus, if the speed is to be controlled to
50% below synchronous speed, its rating must be at least equal to one-half that of
the motor being controlled. However, with the injected voltage source, speed
control to the extent of 50% is possible above as well as below synchronous speed
so that the total range over which control is possible is 3: 1 as compared with 2: 1
for the external rotor resistance method. The Schrage motor is an induction
motor that employs injected armature voltages to obtain speed control. This mo-
tor has been manufactured by the General Electric Company for many years to
supply up to 50 hp. GE calls it the “‘brush-shift’’ motor.
Example 4-6
A 15-hp, 60-Hz, four-pole, 480-V, three-phase, wound-rotor, Y-connected induction
motor has a line frequency rotor induced emf per phase of 141 V, a rotor leakage
reactance per phase of 0.27 ©, and a rotor winding resistance of 0.16 © per phase.
When the motor operates with the rotor winding short-circuited, it develops a torque
of 62.4 N-m at a slip of 3.5% and carries a rotor current of 31.1 A. Determine the
magnitude and phase of the injected voltage to be introduced at the terminals of the
rotor winding (with the short-circuit removed) so that the same torque is developed
at a slip which is 40% above synchronous speed. The rotor leakage reactance at this
high slip is not to be neglected.
Solution We know from the basic torque equation [see Eq. (3-27)] of ac machines
that the condition of constant torque requires that

Ih cos Wh = Ixcos Wz
where the * notation denotes operation at the new slip condition. The angle w of
course denotes the power factor angle at the rotor winding at the particular operating
slip. For the case at hand, we have

== SX po _, (0.035)(0.27)
w = tan = 3.4°
eas 0.16
and
Sec. 4-8 Speed Control 195

Em rca Vv
| s*E,= (0.4)(I41) =56.4 V

ER= SEo* Io io = 4.94 V

Er= 15 (re+jO4x0.27) =7.24V

I, V; = 63.64 V
(a) (b)

Figure 4-22 Diagram for Example 4-6: (a) showing relationships between slip-induced
the required injected voltage to
emf’s and currents at two values of slip; (b) determination of
achieve speed control above synchronous speed.

re ~,
*
S°%2 _ _, 0.4)(0.27) baa
LEE it UE RIa tre os
condition is
Accordingly the rotor current at the new slip

« — 7, £08 He _ ae st OY
34° i aes A
wy 31.1 cos
f th,cos

When the motor operates with the rotor short-circuited, the resultant emf that
produces the current of 31.1 A has a magnitude of
Ep = [y|r + jsxz| = 31.1|0.16 + j(0.035)(0.27)| = 4.94 V
in quadrature with
This quantity is shown in Fig. 4-22(a) placed along a vertical line
the flux phasor. However, when the motor operates at a high slip, the rotor winding
can be appreciab le as already demonstr ated (W7 = 34°). Although the
phase angle
relations hip to the flux that
resultant rotor emf at slip s* must still bear a quadrature
dependen t upon the winding resis-
produces it, the magnitude now is no longer just
makes a contribut ion. Thus the magnitud e of the
tance. The leakage reactance also
resultant rotor winding induced emf now computes to be

Ex = If |r. + js*x2| = 37.50.16 + j(0.4)(0.27)| = 7.24 V

phasor components.
This quantity is also depicted in Fig. 4-22(a) together with its
vertical is the same for both slip conditions,
Observe that the projection of J * on the
thus ensuring constant torque.
 196 Three-Phase Induction Motors Chap. 4

Keep in mind that by design the injected voltage is introduced always in

quadrature with the flux phasor for the achievement of speed control. Because in
this case we want speed control above synchronous speed, the injected voltage must
be introduced 90° lagging the flux phasor in the manner illustrated in Fig. 4-22(b).
The quantities Ez, s*E,, and V; are all collinear and the relationship between them
expressed arithmetically is simply

V; — s* E> — ER

V; = Ex + s*E, = 7.24 + (0.4)(141) = 63.64 V

4-9 ELECTRONIC METHOD OF SPEED CONTROL

Reference to Eq. (4-12) makes it obvious that an attractive means of speed control
is the availability of an adjustable frequency source. It is mentioned in the preced-
ing section that one method of obtaining such a source is to use a dc motor-ac
generator set. Unfortunately, this solution projects its own shortcomings such as
the need for a suitable dc supply, the need to purchase two additional machines, a
more restrictive limit on the range of speed control as well as reduced efficiency.
A better solution lies in the use of an inverter which can be made to change a de
source into a variable frequency source. Because dc power is not readily avail-
able, it becomes necessary to generate it through the use of suitable rectifier
circuits. A block diagram of this electronic scheme of speed control of a squirrel-
cage induction motor is shown in Fig. 4-23. It is significant to note that this is one
of the few methods that is available to achieve speed control for the highly rugged
squirrel-cage induction motor.
The three-phase rectifier in Fig. 4-23 serves the purpose of converting the
commonly available three-phase voltage to a dc source. Usually, the output of the
rectifier contains higher harmonics of the fundamental frequency of the ac source
and these are conveniently removed by an appropriate filter. Figure 4-24(a) illus-
trates a typical three-phase half-wave rectifier the output voltage of which is
drawn in Fig. 4-24(b). By passing this waveshape through the LC filter, the output
voltage becomes essentially dc. It then becomes the task of the inverter to gener-
ate a new three-phase voltage source which in general exhibits the properties of
variable frequency, adjustable voltage, and even adjustable phase. Simultaneous
adjustment of the output voltage of the inverter with frequency is needed to

Three Three-phase
phase squirrel —cage
rectifier motor

Figure 4-23 Block diagram to illustrate the principal equipment needed in the
electronic speed control of a squirrel-cage three-phase motor.
Sec. 4-9 Electronic Method of Speed Control 197

D1

D2

D3

+ Rectifier ———>+~__—_ Filter ——>

(a)

Vr
(V)

. ea aie Figure 4-24 (a) A three-phase half-

S mie Qi wt wave rectifier using semiconductor
diodes; (b) output waveshape of the
(b) rectifier.

prevent operation at values of flux per pole which deviate sharply from the rated
value of the controlled motor. This can be achieved by designing the speed con-
trol system so that it keeps the ratio of the inverter output voltage to the controlled
frequency a constant. By Eq. (4-9) it then becomes clear that by this technique
the flux per pole can be preserved over the operating range. If this precaution
were not to be observed, then at low frequency severe saturation of the magnetic
circuit could easily occur, while at high frequencies there would be a diminution of
the motor’s reserve capacity.
The circuitry of Fig. 4-24(a) shows the use of diodes in the rectifying cir-
cuit. It is useful to note that such diodes are available in voltage ratings up to
5000 V and current ratings of 7500 A. The output voltage can be made adjustable
by replacing the diodes with silicon-controlled rectifiers (SCRs). By placing a
suitable voltage pulse on the gate terminal of the SCR, the instant of firing of the
diode section can be delayed, thereby reducing the value of the rectified output.
Although the voltage ratings of the SCRs match those of the power diodes, the
current ratings are about half as much.
A schematic diagram of a typical electronic speed control system employing
are
the foregoing notions is depicted in Fig. 4-25. The components shown here
diagram of Fig. 4-23 but modified to include the
those called for by the block
ontrolled rectifiers. Moreover , the details of the
control circuit for the silicon-c
198
aj09 104jU09
4104 SYOS
psads
|O4JU0d
@BDIJOA

—aa1y]aspyd

BIQDIADA

304no0s
@HDIIOA
4Aa4lly

snjd
JOIJIJOOI

espyd—saiy]
| aspyd—seauy -|@auinbs
abps udlJONpuU!
40jOW

DANSSz-p
= INVWISYDSWeiseIp
1OJ d1U0.1D9]9
poeds JoUOD
Jo& sseyd-se1y} 28e9-jolinbs
uononput“10}0Uoy], pofjonuod-uosIIs
sIOYNIEI(YDS)
ore payxeU
J@AU|

[§ YSNOIY)
‘9S
Sec. 4-9 Electronic Method of Speed Control 199

inverter are also indicated. The switches marked S1 through S6 are the SCRs and
each is appropriately gated by signals obtained from the gate control section. It is
the purpose of the gate signal to sequence the ON and OFF times of each SCR so
that the output voltage appearing at terminals a, b, c will exhibit the characteris-
tics of a three-phase voltage set. The speed control voltage E, serves two func-
tions. First, it controls the rectifying action of the three-phase rectifier in a fash-
ion that allows the magnitude of the dc output to be directly proportional to E; ;
and second, the speed control voltage also determines the output frequency of a
voltage-controlled oscillator (VCO) in a direct manner. Expressed mathemati-
cally, we can write
Vac = KE; (4-53)

and
fo = kE, (4-54)

where V4, denotes the output voltage of the three-phase rectifier plus filter and f, is
the output frequency of the oscillator. The quantities k; and k, are appropriate
scale factors. In fact, it is this frequency which the inverter imposes at the termi-
nals of the squirrel-cage induction motor. The diodes that are placed in parallel
with the SCRs in a reversed orientation are needed to provide a path for reactive
currents when the corresponding SCR is gated off. They also are useful in allow-
ing energy to flow back to the source whenever the induction motor is made to
operate as a generator.
We now undertake a description of how the inverter takes the dc source and
manipulates it to produce a three-phase ac voltage set. The gate control section of
the system plays a crucial role in the process, for it provides the gating signals,
which in turn establish the proper gating sequence as well as the duration of the
gating (i.e., the portion of the frequency cycle each SCR is allowed to stay con-
ductive). The switching sequence used in this explanation is shown in Fig. 4-26.
One period of the frequency f, is shown and it is divided into six equal parts of a/3
radians. In the interest of realizing a maximum dc voltage each SCR is gated to be
in the ON mode for close to 7 radians. As a practical matter, the ON-time must
be kept just under 7 radians in order subsequently to allow S4 to be gated ON at
exactly 7 radians. A glance at Fig. 4-25 makes clear the importance of gating S1
to the OFF state before S4 is turned on. Failure to do so places a short-circuit
across the de supply.
The switching sequence appearing in Fig. 4-26 is determined as follows.
Start with S1 and draw a line over the first three 60° periods, thus indicating that
S1 is to be kept in a gated ON state for the first 180 electrical degrees. Next,
because we are dealing with three-phase, start the conduction period of S2 ata
position 120 electrical degrees later than the start of S1. Similarly, start the con-
duction period of S3 at a position 120 electrical degrees later than the start of S2.
This pattern is then repeated for the second set of SCRs. Accordingly, the ON-
time for S2 is made to begin at 7 radians and to endure for 180°. The conduction
200 Three-Phase Induction Motors Chap. 4

Si

S2

So

S4

$5

S6

Sed 27m Radians scale

Sa Time scale

Figure 4-26 Switching sequence diagram for the SCRs in Fig. 4-25.

period for SS is then made to begin 120° later than for S4 and to endure for 180°.
Finally, conduction for S6 is programmed to commence 120° after S4 is placed into
conduction and again allowed to stay conductive for almost 7 radians. By exam-
ining the first 60° column of Fig. 4-26, we can properly conclude that over this
period the SCRs that are gated on are S1, S3, and SS. The remaining three SCRs
are simultaneously and intentionally gated to the OFF state. During the period
that a specific SCR is programmed to be in the ON state, it is important to
maintain the gate signal so that the switch continues to stay closed whether or not
current is flowing at the moment. Upon repeating the procedure for each 60°
period, the following table is obtained that describes the switching sequence.

Interval SCRs in ON state

0-60° S1, S5, S3

60-120° S1, $5, S6
120-180° S1, S2, S6
180-—240° S4, $2, S6
240-300° S4, S2, $3
300-360° $4, S5, S3

Once the switching sequence is established, it is a routine matter to identify

the line-to-line voltages as they appear at the terminals of the squirrel-cage motor,
which is assumed to have a Y-connected stator winding. The inverter output
Sec. 4-9 Electronic Method of Speed Control 201

appears at terminals a, b, c. During the first 60° interval, switches S1, S3, and S5
are in the ON state. Because S1 and S3 are conductive, terminals a and b are
connected to the + side of the dc source and through the action of S5 terminal c is
joined to the negative side. This arrangement places phases a and c of the induc-
tion motor in parallel with each other and the combination in series with phase b.
Refer to the schematic diagram for the 60° interval in line (a) of Table 4-1, which
illustrates the distribution of the dc input voltage to the inverter during the first
three 60° intervals of the operating frequency determined by the voltage-con-
trolled oscillator in response to the speed control voltage E,. For balanced phase
windings it follows that the distribution of the de source voltage puts Vq./3 across
the two parallel-connected phases (i.e., coils an and cn) and 2/3 Vg. across phase
b. A simple application of Kirchhoff ’s voltage law then shows that the line volt-
ages are V,, = Vac, Vic = —Vac, and Vig = 0.
During the second 60° interval, the foregoing table shows that switches S1,
S5, and S6 are in the ON state. In this case, only terminal a is joined to the
positive side of the dc source while b and c are both connected to the negative
side. The effect is to put phases b and c in parallel and the combination in series

TABLE 4-1 DISTRIBUTION OF THE DC INPUT VOLTAGE DURING THE FIRST THREE
60° INTERVALS OF THE OPERATING FREQUENCY
eee
Voltage distribution Corresponding line
Interval ON-gated SCRs on induction motor phases voltages
_ Sei Le eee eee EEE ———— ee
(a) 0-60° $1, $3, $5 Vap = Vac
Voc oe —Vac
Veg = 0

(b) 60-120° Siz S5,156 Vap = Vac

Voe =0

Vea = Vac

(c) 120-180° S1, S2, S6 Vap = 9

Voc = Vae
Veneta cs
202 Three-Phase Induction Motors Chap. 4

with phase a in the manner illustrated in line (b) of Table 4-1. While reading these
diagrams, it is helpful to keep in mind that terminals marked with the same
polarity are joined together for the specified interval. Now the distribution of the
dc voltage puts two-thirds across the winding of phase a and one-third across each
of the coils b and c with the polarity indicated. The result is that the line voltages
during this second period are described by Va, = Vac, Vic = 0, and Veg = —Vac-
The situation that prevails in the third 60° interval leads to the results shown
in line (c) of Table 4-1. Here the coils of phase a and b are in parallel. The
corresponding line voltages become V,, = 0, Vic = Vac, and Veg = —Vac. A
continuation of this procedure through to the sixth 60° interval leads to the results
depicted in Fig. 4-27. An examination of these waveshapes reveals that the line
voltages do exhibit the characteristics of a balanced three-phase voltage system.
They are of equal magnitude and the individual line voltages are displaced in time
by the required 120 electrical degrees. Of course, these waveshapes are not sinus-
oidal but they do contain a large fundamental component which is the one that
serves to drive the induction motor.
The peak value of this fundamental component is readily found by applying

Figure 4-27 Variations of the inverter output line-to-line voltages corresponding

to the switching sequence of Fig. 4-26 for one full cycle of the controlled fre-
quency fo. ‘
Sec. 4-9 Electronic Method of Speed Control 203

Vit
Line
-to —line
voltage

Figure 4-28 Variation of each line-to-line voltage with zero time placed at the
center of the rectangular wave.

Fourier series analysis to the rectangular waveshapes that represent the line-to-
line voltages. For convenience, this basic waveform is redrawn in Fig. 4-28 with
the zero position relocated in the interest of taking advantage of the quarter-wave
symmetry contained therein. Then from Fourier series analysis the amplitude of
the fundamental component is

Vip =—1 |."

.f20
Vee cos wt dwt) = =~ [,Vac cos wt d(wt)
or

Vip =
2V3
= Var (4-55)

The corresponding rms value of this voltage is then

V V6
Vi= aye pds Vic = 0.78 Vac (4-56)

The rectangular nature of the line voltages means, of course, that harmonics do
exist in these voltages. However, the harmonic situation is not as bad as it may at
first appear. For example, there are no even harmonics in these line voltages
because of the mirror symmetry. Furthermore, although a third harmonic voltage
exists in the line-to-neutral voltages, they do not appear in the line-to-line voltages
because cancellation occurs when adding the phase voltages to obtain the line
voltages. Hence the first important harmonic becomes the fifth. But Fourier anal-
ysis tells us that the magnitude of the fifth harmonic is only one-fifth that of the
fundamental. When this is coupled with the fact that the frequency of the fifth
harmonic is five times that of the fundamental frequency and that the induction
the
motor presents an inductive load, which exerts a substantial filtering effect,
cause
total effect of this harmonic is not serious. The other higher harmonics
even less disturbance. The influence of the harmonic content of the line voltages
more
can be further reduced by employing a switching arrangement that yields a
 204 Three-Phase Induction Motors Chap. 4

nearly sinusoidal waveshape for the line voltages (see Prob. 4-62). Unfortunately,
this usually carries with it the penalty of reduced magnitude for the rms output
voltage.
As a final point, let us examine the variation, if any, that takes place in the
flux per pole of the induction motor over its operating range. Insertion of Eq.
(4-53) into Eq. (4-56) gives the rms value of the voltage applied to the induction
motor as
Vi = 0.78Vac = 0.78kK,Es = koEs (4-57)
where ky denotes a new constant equal to 0.78k,. Except at very low frequencies,
where machine inductance plays a reduced role in determining motor perform-
ance, Eq. (4-57), when divided by V3, also expresses the induced emf that
occurs per phase in the induction motor as it operates at various frequencies in
response to the speed control voltage E,. Hence we may write

—tVi = Ey = 4,44NK, fod = —-ko E,

V3 Sop V3
or

d = KES = ko Es (4-58)
V3 (4.44)NK, fo K fo
where K = (V3)4.44NKy, which is a constant. But the commanded frequency is
established by the voltage-controlled oscillator in accordance with the relation-
ship appearing in Eq. (4-54). Upon introducing this expression for fo into Eq.
(4-58), there results

o= KoEs _ ko = a constant (4-59)

"Kp Es 2 Kk;
Therefore, the speed control offered by the electronic system depicted in Fig. 4-25
is achieved in a manner that ensures operation at essentially constant flux.

4-10 RATINGS AND APPLICATIONS OF THREE-PHASE

INDUCTION MOTORS

Now that the theory of operation, the characteristics, and the performance of the
three-phase induction motor are understood, we can study their standard ratings
and typical applications. Of course, before it is possible to specify a particular
motor for a given application, the characteristics of the load must be known.
These include such items as horsepower requirement, starting torque, accelera-
tion capability, speed variation, duty cycle, and the environment in which the
motor is to operate. Once this information is available, it is often possible to
select a general-purpose motor to do the job satisfactorily. Table 4-2 is a list of
Sec. 4-10 Ratings and Applications of Three-Phase Induction Motors 205

such motors which are readily available and standardized in accordance with
generally accepted criteria established by the National Electrical Manufacturers
Association (NEMA). The table is essentially self-explanatory.
A brief description of the salient features of the common classes of squirrel-
cage induction motors follows. These distinctions arise from the differences in
the construction details of the rotor slots which accommodate the squirrel cage.

NEMA Class A: The General-Purpose Motor

The rotor stampings for this machine usually have semiclosed slots of medium
depth in order to limit the skin effect. See Fig. 4-29(a). The low-resistance cage
that is placed in these slots leads to a performance characterized by high efficiency
and high power factor under rated load conditions. The disadvantage of such a
construction is that at rated voltage the motor draws from the line source very
large starting currents in a ratio of five to eight times rated current. The limita-
tions imposed by the electric power companies in restricting the size of the motor
starting current often means that motors rated in excess of 7.5 hp cannot be
started directly across the supply lines. Instead, these motors must be energized
through a voltage-reducing device such as an autotransformer (also called a start-
ing compensator). When this procedure becomes necessary, the starting torque
as well as the accelerating torque available to a shaft-connected load will be
reduced. Recall that the starting torque is directly proportional to the square of

oon al
(c)

Figure 4-29 Construction details of the rotor slots of squirrel-cage induction

motors: (a) class A; (b) class B; (c) class C; (d) class D.
 OF POLYPHASE 60-CYCLE
TABLE 4-2. CHARACTERISTICS AND APPLICATIONS
AC MOTORS
Starting Maximum Starting
torque* torque* current®
Type hp (%)
classification range (%) (%)

HP Poles Torque Up to 225 but 500-1000

General-purpose, normal _0.5—-500
iS) 175 not less than
torque and starting
les 250 200
current, NEMA class A
5.0 150

—S o —i=]Nn

FN
PRNANAN

0.5-500 Same as above About the same About 500-550,

General-purpose, normal
as class A but less than aver-
torque, low starting
may be less age of class A
current, NEMA class B

200-250 190-225 About same as

High torque, low starting 3-200
class B
current, NEMA
class C

High torque, medium and 3-150 Medium slip Usually same as Medium slip
high slip, NEMA 350 standstill 400-800, high
class D High slip torque slip 300-500
275-315

——— eel
ee
Low starting torque, 40-200 Low, not less Low, but not Normal 500—
either normal starting than 50 less than 150 1000, low
current, NEMA class 300-500
E, or low starting
current, NEMA class F

Wound-rotor 0.5-5000 Up to 300 225-275 Depends upon

external rotor
resistance but
may be as low
as 150

a EEE aE
ee
aFigures are given in percent of rated full-load values.
Source: By permission from M. Liwschitz-Garik and C. C. Whipple, Electric Machinery, vol. II
(Princeton, N.J.: D. Van Nostrand Co., Inc.).

206
 Power
Slip factor Efficiency
(%) (%) (%) Typical applications

Low, 3-5 High, 87-89 High, 87-89 Constant-speed loads where excessive
starting torque is not needed and
where high starting current is toler-
ated. Fans, blowers, centrifugal
pumps, most machinery tools, wood-
working tools, line shafting. Lowest
in cost. May require reduced volt-
age starter. Not to be subjected to
sustained overloads, because of
heating. Has high maximum torque

3-5 A little lower 87-89 Same as class A—advantage over

than class A class A is lower starting current, but
power factor slightly less

3-7 Less than class A 82-84 Constant-speed loads requiring fairly

high starting torque and lower start-
ing current. Conveyors, compres-
sors, crushers, agitators, reciprocat-
ing pumps. Maximum torque at
standstill
————

Medium 7-11; Low Low Medium slip. Highest starting torque

high, 12-16 of all squirrel-cage motors. Used for
high-inertia loads such as shears,
punch presses, die stamping, bull-
dozers, boilers. Has very high
average accelerating torque. High
slip used for elevators, hoists, and
so on, on intermittent loads
ceo nee a clea a ae i EER EE Se ee
1 to 33 About same as About same as Direct-connected loads of low inertia
class A or class A or requiring low starting torque, such
class B class B as fans and centrifugal pumps. Has
high efficiency and low slip

I <<

3-50 High, with rotor High, with rotor For high-starting-torque loads where
shorted same shorted the very low starting current is required
as class A same as class or where torque must be applied
A, but low very gradually and where some
when used speed control (50%) is needed.
with rotor Fans, pumps, conveyors, hoists,
resistor for cranes, compressors. Motor with
speed control speed control more expensive and
may require more maintenance.
a a a

207
 Chap. 4
Three-Phase Induction Motors
208
the motor at
reduction in voltage applied to
the applied voltage. Hence a 20%
in the starting torque.
starting yields a 36% reduction ver, this
e line starting is tolerable, howe
In sizes and circumstances wher demanding
costly of the group and also the least
motor will be found to be the least gned to have a generous
class A motor is desi
in maintenance. Moreover, the m deve lope d torq ue, which
its value of maximu
reserve capacity, as indicated by position
torque. This puts the motor in a good
typically is close to 225% of rated y to be enco unte red in
overloads that are likel
to manage successfully temporary shap e of its torq ue-s peed
re 4-30 depicts the
general-purpose applications. Figu
curve.

ent,
NEMA Class B: Low Starting Curr
Normal Starting Torque
, as illus-
B motor may be either of two types
The rotor laminations in the class a singl e deep bar
ed to accommodate either
trated in Fig. 4-29(b). The slot is shap r cros s-se ctional
with the upper one of large
or two bars of slightly different sizes ing thro ugh the
flux produced by a current flow
area. At line frequency the leakage chan ges with
in Fig. 4-29(b). As this flux
deep bar assumes the pattern depicted end of the bar
ed emf induced at the deep
the changing bar current, the associat the bar (poin t 5).
at the upper end of
(point a) is greater than that produced bar than in the
in the upper portion of the
Accordingly, current tends to flow more

300

100

full—load
torque
of
percent
Torque,
50

100 Figure 4-30 Typical torque-slip curves

60° 80
10) 20 40 for various classes of squirrel-cage
induction motors.
Speed, percent of sychronous speed
 Sec. 4-10 Ratings and Applications of Three-Phase Induction Motors 209

3.0

'2ac
V2de ~a
2.0

ratio
effect
Skin

0.2 0.4 0.6 0.8 1.0

Figure 4-31 Illustrating the skin effect
Slip, p—-u in a deep rotor bar.

lower portion, thus reducing the effective cross-sectional area of the conductor for
current flow. Of course, for a nonvarying current flow (dc) the counter emf is
zero, so the entire cross-sectional area of the bar is effective in supporting current
flow. As the frequency of the bar current increases there is a greater squeezing of
the current toward the upper portion of the bar. Figure 4-31 depicts the effects of
this situation (the skin effect ratio) for a typical solid deep bar as a function of the
slip. Observe that at standstill the ac resistance is three times greater than the dc
resistance. In essence, then, the class B motor at starting exhibits a high rotor
resistance, which in turn yields ample starting torque at reduced starting current.
The latter is about 75% of the value of the class A motor used in application. On
the debit side for the class B motor, however, it is necessary to list a lower
maximum torque, slightly lower power factor, and a somewhat greater cost.

NEMA Class C: Low Starting Current, High Starting Torque

The rotor slot construction detail for the class C motor appears in Fig. 4-29(c). A
double squirrel cage is always used. At standstill the impedance of the lower cage
is much greater than the impedance of the upper cage. The upper cage is charac-
terized by high resistance and low leakage reactance while the lower cage exhibits
high leakage reactance and low resistance. Accordingly, at starting, the upper
the advan-
cage plays the dominant role; and because it has high rotor resistance,
Then, as the
tages of a high starting torque at low starting current are realized.
frequency is
motor reaches its normal operating speeds (where the cage current
a much lower
very low), control passes over to the lower cage, which displays
opera-.
resistance than the upper cage, thus improving efficiency during normal
is diminishe d in
tion. Of course, the high leakage reactance of the lower cage
power factor
importance by operation at low slips. However, both efficiency and
B. A glance at the
for the class C motor will be less than for either classes A or
 Chap. 4
Three-Phase Induction Motors
210
acceler-
motor shows that it can provide more
torque-speed curve for the class C acte ristic
motors. In fact, a distinguishing char
ating torque than either the A or B its pull -out
a starting torque that is larger than
of this motor is that it develops
torque. the
r has eliminated the need to specify
The availability of the class C moto
applications.
wound-rotor induction motor in some
ue, Lowest Starting
NEMA Class D: Highest Starting Torq
Current, High Slip
h and
class D motor. A slot of shallow dept
A single squirrel cage is used for the resis tivity
r to Fig. 4-29(d). Bars of higher
small cross-sectional area is used. Refe to
is to obtain the high resistance needed
are also often employed. The objective typic al
ng torques, as revealed by the
obtain very high starting and accelerati e
Fig. 4-30. Of course, the use of a singl
torque-speed characteristic shown in er
iency will be lower and the slip high
high-resistance cage means that the effic l
the high slip characteristic is a very usefu
than the preceding classes. However, s,
s, such as for punch presses, hoist
one in applications where this motor excel
a flywheel in these applications, the drop
cranes, and soon. When equipped with
on of the duty cycle enables the flywheel
in speed that occurs during the work porti of
permitting a decrease in the peak level
to release some of its stored energy, thus
e.
power the motor draws from the line sourc
limited by the allowable tempera-
The horsepower rating of an ac motor is
ugh copper, aluminum, and iron as
ture rise under operating conditions. Altho
temperature increases, this does not
individual elements can withstand very high
used for the stator winding in both the
apply to the insulating materials that are
motors as well as for the rotor wind-
wound-rotor and the squirrel-cage induction
ials deteriorate rapidly whenever
ing of the wound-rotor motor. Insulating mater
to withstand are exceeded. Conse-
the maximum temperatures they are designed
ial to withstand high temperature
quently, the capability of an insulating mater
cal dimensions of a motor with a
rises plays a crucial role in establishing the physi
is equipped with a superior insulating
specific horsepower rating. When a motor
for a given horsepower because it 1s
material, the motor size can be made smaller
ly (i.e., operate at higher flux densi-
possible to work the motor harder magnetical
operate at higher current densities in
ties in the iron) as well as electrically (i.e.,
the copper or aluminum conductors).
ting materials. Class B insula-
There are three classes of NEMA~-rated insula
fibers, and asbestos in a manner
tion is composed of such materials as mica, glass
when operation occurs with fan
that allows a maximum temperature rise of 80°C
nts in addition to some synthetic
cooling. Class F insulation includes similar eleme
temperature rise of 100°C. The
substances all designed to provide an allowable
which permits an even higher
class H category refers to a silicone-type insulation
temperature rise of 125°C.
 Sec. 4-11 Controllers for Three-Phase Induction Motors 211

180

O
: 160

=
ie
aEun lt0
&
oD
&
ec
-== |20

2
oO
|

100

90 1000
1 10 100

Life in percent, logarithmic scale

Figure 4-32 Illustrating the effect of winding temperature on motor life.

If an electric motor is operated at overload for a sustained period of time, the

life of the motor can be markedly reduced. This fact is graphically demonstrated
in Fig. 4-32 for an insulating material of the class B type. An examination of the
plot reveals an interesting bit of information: if a motor equipped with this insula-
tion is operated continuously at an overload condition that causes an increased
temperature rise of 8 to 10°C over the rated value of the insulation, the life of the
motor is approximately halved. Moreover, because of the linear character of this
semilogarithmic plot, the statement is true for each additional increase in operat-
ing temperature of 10°C.

S
4-11 CONTROLLERS FOR THREE-PHASE INDUCTION MOTOR
next step is to select
After the right motor is selected for a given application, the
functio ns of a controller are
the appropriate controller for the motor. The primary
ng without damage or inconven-
to furnish proper starting, stopping, and reversi
to the power system. However,
ience to the motor, to other connected loads, or
especia lly the followi ng:
the controller fulfills other useful purposes as well,

loads may be damaged if

1. It limits the starting torque. Some connected shaft
For exampl e, fan blades can be
excessive torque is applied upon starting.
can be strippe d. The controller
sheared off or gears with excessive backlash
speed picks up the voltage is
supplies reduced voltage at the start and as the
increased in steps to its full value.
 Three-Phase Induction Motors Chap. 4
212

2. It limits the starting current. Most motors above 7.5 hp cannot be started
of the excessive starting current
directly across the three-phase line because
t is limited only by the leakage
that flows. Recall that at unity slip the curren
quantity, especially in the larger
impedance, which is usually quite a small
annoying because it causes light
motor sizes. A large starting current can be
ted motors to stall. Reduced-
to flicker and may even cause other connec
nces.
voltage starting readily eliminates these annoya
al-pu rpose motors are designed to
3. It provides overload protection. All gener
overh eatin g. However, if for
deliver full-load power continuously without
say, 150% of its rated output
some reason the motor is made to deliver,
the deman d and burn itself up
continuously, it will proceed to accommodate
motor is based on the allowable
in the process. The horsepower rating of the
tion used for the field and
temperature rise that can be tolerated by the insula
heat that raises the tempera-
armature windings. The losses produce the
the rated values , there is no
ture. As long as these losses do not exceed
to becom e exces sive, damage
danger to the motor, but, if they are allowed r-
that will keep the tempe
will result. There is nothing inherent in the motor
the functi on of the con-
ature rise within safe limits. Accordingly, it is also
is achieved by the use
troller to provide this protection. Overload protection
ive to the heat produced by
of an appropriate time-delay relay which is sensit
the motor line currents.
at reduced voltage can be
4. It furnishes undervoltage protection. Operation
ds rated power. If the
harmful to the motor, especially when the load deman
automatically discon-
line voltage falls below some preset limit, the motor is
ler.
nected from the three-phase line source by the control

al and magnetic. We
Controllers for electric motors are of two types—manu
ages over the man-
shall consider only the magnetic type, which has many advant
protection. It can be
ual type. It is easier to operate. It provides undervoltage
er, the magnetic
remotely operated from one or several different places. Moreov
controller requires a
controller is automatic and reliable, whereas the manual
called for in a given
trained operator, especially when a sequence of operations is
antage is the greater initial cost of the magnetic type.
application. The one disadv
Appearing in Fig. 4-33 is the schemat ic diagram of a magneti c full-voltage
n motor. The operati on is simple. When the
starter for a three-phase inductio
coil M is energize d. This moves the relay arma-
start button is pressed, the relay
closing the main contact ors M, which in turn
ture to its closed position, thereby
When the relay armatur e moves to its energiz ed
apply full voltage to the motor.
contact or M, which serves as an electric al
position, it also closes an auxiliary
the start button without de-ener gizing
interlock, allowing the operator to release
than M,. The
the main relay. Of course contactors M are much larger in size
that enables it to handle the starting motor
former set must have a current rating
just the exciting current of the relay
current. The latter needs to accommodate
Sec. 4-11 Controllers for Three-Phase Induction Motors 213

OL heater element

3-—phase M
voltage

OL heater

Relay OL Control circuit

coil Figure 4-33 Full-voltage magnetic
Mg starter.

coil. Figure 4-33 also shows that the motor line current flows through two over-
load heater elements. If the temperature rise becomes excessive, the heater ele-
ment causes the overload contacts in the control circuit to open. In controller
diagrams it is important to remember that all contactors are shown in their de-
energized state. Thus the symbol —| |— means that the contactors M are open
when the coil is not energized. Similarly, the symbol —>{— means that these
contactors are closed in the de-energized state.
Undervoltage protection is inherent in the magnetic starter of Fig. 4-33.
This comes about as a result of designing the coil M so that, if the coil voltage
drops below a specified minimum, the relay armature can no longer be held in the
closed position.
A full-voltage magnetic starter equipped with the control circuitry to permit
reversing is illustrated in schematic form in Fig. 4-34. A three-phase induction
motor is reversed by crossing two of the three line leads going to the motor
terminals. In this connection note the criss-cross of two of the R contractors.
Pressing the forward (FwD) button energizes coil F which in turn closes the main
contactors F as well as the interlock F,. This allows the motor to reach its
forward operating speed. To reverse the motor the REV button is pushed. This

‘ Figure 4-34 Full-voltage starter

Relay coil equipped with reversing control.
 Three-Phase Induction Motors Chap. 4
214

Figure 4-35 Three-step reduced-

voltage starter for a three-phase induc-
tion motor.

F contactors. Two,
does two things. One, it de-energizes coil F, thus opening the
the R contact ors which apply a reversed-
it energizes the R relay coil, thus closing
it to attain full speed in the reverse direc-
phase sequence to the motor causing
interloc k circuit is a safety measur e which
tion. Putting the REV switch in the F,
ors closed at the same time.
prevents having both the R and F contact
limiting resis-
An illustration of a reduced-voltage magnetic controller using
4-35. This unit is frequent ly referred to as a
tors in the line circuit appears in Fig.
the line resistor s are remove d in three
three-step acceleration starter because
s coil M and closes contact ors M, thus
steps. Pushing the start button energize
through the full resistor s. In addition
applying a three-phase voltage to the motor
contactor
to contactors M and M, coil M is also equipped with a time-delayed
not close until a preset time after
Ty. This contactor is so designed that it does
is usually obtaine d through a mechani -
the armature of coil M is closed. The delay
d by the relay armatur e. Of course
cal escapement of some sort which is actuate
correspond-
the time delay is needed to permit the motor to accelerate to a speed
elapse of the preset time, the Ty
ing to the reduced applied voltage. After the
closes contact ors 1A, short-
contacts close, energizing coil 1A, which in turn
1A is also equippe d with a time-
circuiting the first part of the series resistor. Coil
delay contactor T;4, which is designed to allow the motor to accelera te to a higher
This
speed before it closes. When contactors 1A do close, coil 2A is energized.
section of the line
immediately closes contactors 2A, shorting out the second
Tp, closes, applyin g an
resistor. Then, after still another time delay, contactor
ors 3A the full line
excitation voltage to coil 3A. With the closing of contact
in
voltage is applied to the motor. In this manner the motor is brought up to speed
current or develop -
a ‘‘soft,”? smooth fashion without drawing excessive starting
ing large starting torques.
Chap. 4 Problems 215

PROBLEMS

4-1. A three-phase armature winding is shown in Fig. P4-1. Sinusoidal currents having an
amplitude of 100 A flow in the three phases. Each coil consists of three turns.
Carefully sketch to scale the actual mmf distribution along the air gap for a span of
two poles for the indicated time instants ¢, and f.

Figure P4-1

. A 60-Hz polyphase induction motor runs at a speed of 873 rpm at full load. What is
the synchronous speed? Find the frequency of the rotor currents.
. A 50-Hz polyphase induction motor runs at a speed of 1475 rpm at full load. What is
the synchronous speed of the motor? How many poles is it designed for?
. State the general conditions under which it is possible for m alternating fields (with
axes fixed in space) to yield a revolving field constant in amplitude and traveling at
constant speed.
. Determine the magnitude and direction of the resultant flux field in the machine
configuration of Fig. 4-2 corresponding to time instants and ft, in Fig. 4-1.
. In the time variation of the phase currents depicted in Fig. 4-1 assume that the
amplitude of phase b current is one-half that of phases a and c but that each is
displaced by 120° from the other. Find the magnitude and direction of the resultant
flux for the time instants f,, t;, and f;, and compare with the results shown in Fig.
4-2.
by
. An unbalanced three-phase mmf flows through three coils that are space displaced
120° as shown in Fig. 4-2. The nature of the unbalance is such that phase b lags
ampli-
behind phase a by 90° instead of 120°. However, each phase has the same
character of the resultant flux field correspond ing to various
tude. Investigate the
instants of time.
form
4-8. Demonstrate that a resultant mmf that is described by an equation of the
F, = F cos (wt + a)
that ris a
represents a wave that travels in the negative a direction. Keep in mind
time variable and a is a space variable.
are 90°
4-9, A balanced two-phase voltage is applied to identical coils, the axes of which
in terms of the mmf amplitude of
apart. Derive the expression for the resultant mmf
216 Three-Phase Induction Motors Chap. 4

either coil (¥,,), the angular frequency w of the exciting currents, and the space
variable a. Assume that each coil produces a sinusoidal field distribution.
4-10. A balanced, three-phase, 60-Hz voltage is applied to a three-phase, four-pole induc-
tion motor. When the motor delivers rated output horsepower, the slip is found to be
0.05. Determine the following:
(a) The speed of the revolving field relative to the stator structure, which accommo-
dates the exciting winding.
(b) The frequency of the rotor currents.
(c) The speed of the rotor mmf relative to the rotor structure.
(d) The speed of the rotor mmf relative to the stator structure.
(e) The speed of the rotor mmf relative to the stator field distribution.
(f) Are the conditions right for the development of a net unidirectional torque?
Explain.
4-11. Repeat Prob. 4-10 for the case where the rotor structure is blocked, thus preventing
rotation in spite of the application of a balanced three-phase voltage to the stator.
4-12. Repeat Prob. 4-10 for the case where the slip has a value of 0.03.
4-13. Determine the no-load speed of a six-pole, wound-rotor, three-phase induction mo-
tor, the stator of which is connected to a 60-Hz line and the rotor of which is
connected to a 25-Hz line, when:
(a) The stator field and the rotor field revolve in the same direction.
(b) The stator field and the rotor field revolve in opposite directions.
4-14. Determine the speed of operation of a four-pole, wound-rotor, three-phase induction
motor the stator of which is connected to a 60-Hz line and the rotor of which is
connected to a 50-Hz line when:
(a) The stator field and the rotor field revolve in the same direction.
(b) The stator field and the rotor field revolve in opposite directions.
4-15. Answer whether or not the following statements are true or false. When the state-
ment is false, provide the correct answer.
(a) In a three-phase induction motor the rotor ampere-conductor distribution is
stationary with respect to the stator magnetic field distribution at all speeds of
the rotor.
(b) The load on the shaft of a three-phase induction motor is increased. This causes
the speed of the rotor mmf relative to the stator mmf to change, thus permitting
an increased input current to provide power balance.
4-16. The shaft output of a three-phase, 60-Hz induction motor is 75 kW. The friction and
windage losses are 900 W, the stator core loss is 4200 W, and the stator copper loss is
2700 W. If the slip is 3.75%, what is the percent efficiency at this output?
4-17. Refer to Prob. 4-16. The rotor winding resistance referred to the stator is known to
be 0.1 . Determine the value of the stator-referred rotor current for the specified
operating conditions.
4-18. A 15-hp, 220-V, three-phase, 60-Hz, six-pole, Y-connected induction motor has the
following parameters per phase: r; = 0.128 Q, r; = 0.0935 0, x; + x3 = 0.496 O,
re = 183 O, xg = 8 QO. The rotational losses are equal to the stator hysteresis and
eddy-current losses. For a slip of 3%, find:
(a) The line current and power factor.
(b) The horsepower output.
(c) The starting torque.
Chap. 4 Problems 217

4-19. A 10-hp, 220-V, Y-connected, four-pole, 60-Hz, three-phase induction motor draws
a line current of 26.2 A at a pf of 0.78 lag when it operates at a slip of 5%. Rotational
losses amount to 250 W. The motor is known to have the following parameters
expressed in ohms per phase:

ry = 0.3 Xx, = x» = 1225 r, = 150 xg = 18

(a) Calculate the rotor current per phase referred to the stator.
(b) Find the value of output horsepower.
(c) Determine the efficiency.
(d) Find the developed torque in N-m.
(e) Calculate the starting line current at rated line voltage.

4-20. A three-phase, 60-Hz, six-pole, wound-rotor induction motor draws 10 kW when

driving its normal load. It draws 700 W when the load is disconnected. The rotor
and stator copper losses under normal load are 295 W and 310 W, respectively.
Calculate the efficiency, the speed, and the shaft torque of this motor under normal
load conditions. Assume equal rotational and core losses and negligible no-load
copper losses.

4-21. A three-phase induction motor has a Y-connected rotor winding. At standstill the
rotor induced emf per phase is 100 Vrms. The resistance per phase is 0.3 , and the
leakage reactance is 1.0 © per phase.
(a) With the rotor blocked, what is the rms value of the rotor current? What is the
power factor of the rotor circuit?
(b) When the motor is running at a slip of 0.06, what is the rms value of the rotor
current? What is the power factor of the rotor circuit?
(c) Compute the value of the developed power in part (b).

4-22. A three-phase, 12-pole, 60-Hz, 2200-V induction motor runs at no-load with rated
voltage and frequency impressed and draws a line current of 20 A and an input power
of 14kW. The stator is Y-connected, and its resistance per phase is 0.4. The rotor
resistance r is 0.2 0 per phase. Also, x; + x5 = 2.0 per phase. The motor runs at
a slip of 2% when it s delivering power to a load. For this condition compute:
(a) The developed torque.
(b) The input line current and power factor.

4-23. A three-phase, 440-V, 60-Hz, Y-connected, eight-pole, 100-hp induction motor has
the following parameters expressed per phase:

r; = 0.06 0, xp = xe = 0.26 0
ry = 0.048 0 r= 107.5
a = S4T' 0.
deter-
The rotational losses are 1600 W. Using the approximate equivalent circuit
mine for s = 0.03:
(a) The input line current and power factor.
(b) The efficiency.
 Three-Phase Induction Motors Chap. 4
218

induction
4-24. A three-phase, 335-hp, 2000-V, six-pole, 60-Hz, Y-connected squirrel-cage
has the following parameter s per phase that are applicable at normal slips:
motor
r= 0.20 x, = x} = 0.7070
rs = 0.203 2 r. = 450 0
x= 710
The rotational losses are 4100 W. Using the approximate equivalent circuit, com-
pute for a slip of 1.5%:
(a) Line power factor and current.
(b) Developed torque.
(c) Efficiency.
4-25. A six-pole, three-phase, 40-hp, 60-Hz, induction motor has an input when loaded of
35 kW, 51 A, 440 V, and a speed of 1152 rpm. When uncoupled from the load, the
readings are found to be: 440 V, 21.3 A, 2.3 kW, and 1199 rpm. The resistance
measured between terminals for the stator winding is 0.25 for a Y-connection. The
stator core losses and the rotational losses are known to be equal. Determine:
(a) The power factor of the motor when loaded.
(b) The motor efficiency when loaded.
(c) The horsepower rating of the load.
4-26. Answer each part briefly:
(a) What is the effect of doubling the air gap of an induction motor on the magnitude
of the magnetizing current and of the maximum value of the flux per pole?
Neglect the effect of the leakage impedance.
(b) Describe the effect of a reduced leakage reactance on the maximum developed
torque, the power factor at full load, and the starting current of a three-phase
induction motor.
4-27. The manufacturers’ specifications concerning the ratings of induction motors invari-
ably list the maximum (or breakdown) torque as a percentage of rated torque. Its
value is usually 200%. Explain the significance of choosing so high a value.
4-28. In a given situation only 220-V, 30-Hz, three-phase service is available. A plant
manager has an opportunity to buy cheaply a 10-hp, 440-V, three-phase, 60-Hz, four-
pole squirrel-cage motor, which is to be used to supply power to a constant torque
load. The load torque corresponds to that of the 10-hp motor.
(a) Can the motor be used? Justify your answer.
(b) If so, what would be its rating?
(c) How would the efficiency of operation be affected?
(d) Find the approximate change in starting torque.
4-29. Would you recommend use of a 115-V, 60-Hz, three-phase induction motor on a
460-V, 400-Hz system to deliver power to a constant torque load? Explain.
4-30. When a 10-hp, 220-V, 60-Hz, three-phase induction motor operates at 220 V, 50 Hz
delivering the same constant rated torque, what change occurs in the rotor current?
4-31. In what way, if any, does a reduced applied voltage affect the breakdown torque of a
three-phase induction motor? In what way, if any, does an increased frequency
affect the breakdown torque?
Chap. 4 Problems 219

4-32. A 115-V, three-phase, 60-Hz induction motor is used on a 115-V, 50-Hz source.
Assuming that the delivered torque is to be the same when operating at either
frequency, determine:
(a) The change in operating flux.
(b) The change in rotor current.
(c) The change in synchronous speed.
(d) Would you permit this motor to be used continuously on the 50-Hz source?
Explain.
4-33. Refer to the motor of Prob. 4-19.
(a) Compute the slip at which maximum torque occurs.
(b) Determine the value of the maximum torque.
4-34. Refer to the motor of Prob. 4-19.
(a) For the operating condition described determine the missing parameter of the
exact equivalent circuit: namely, the stator-referred rotor resistance per phase.
(b) Using the exact equivalent circuit obtain the speed at which maximum torque
occurs.
(c) Find the stator-referred rotor current at maximum torque. How does this differ
from the input stator current?
(d) Compute the value of the maximum torque.
4-35. A three-phase, Y-connected, 440-V, 200-hp induction motor has the following
blocked-rotor data: P, = 10 kW, J, = 250 A, V, = 65 V, and r; = 0.02 ©. Find the
value of the rotor resistance referred to the stator.
4-36. A six-pole, 60-Hz, three-phase, Y-connected wound-rotor (three phases also) induc-
tion motor has a standstill induced rotor voltage of 130 V per phase. At short-circuit
with the rotor blocked, this voltage produces a current of 80 A at a power factor of
0.3 lagging. At full-load the motor runs at a slip of 9%. Find the full-load developed
torque.
4-37. A 500-hp, three-phase, 2200-V, 25-Hz, 12-pole, Y-connected wound-rotor induction
motor has the following parameters: r; = 0.225 0, r3 = 0.235 Q, x; + x3 = 1.43 Q,
Xp = 31.8 0, re = 780 Q. A no-load and a blocked-rotor test are performed on this
machine.
(a) With rated voltage applied in the no-load test, compute the readings of the line
ammeters as well as the total wattmeter reading.
(b) In the blocked-rotor test the applied voltage is adjusted so that 228-A line current
is made to flow in each phase. Calculate the reading of the line voltmeter and the
total wattmeter reading.
4-38. A three-phase, 2000-V, Y-connected wound-rotor induction motor has the following
no-load and blocked-rotor test data:
No-load: 2000 Vo" al :3 An LOK W.
Blocked-rotor: 440 V, 170.0 A, 36.4 kW

The resistance of the stator winding is 0.22 © per phase. The rotational losses are
equal to 2 kW. Calculate all the necessary data for the approximate equivalent
circuit at a slip of 2%, and draw the circuit showing all parameter values.
220 Three-Phase Induction Motors Chap. 4

4-39. For the machine of Prob. 4-37 compute the following:

(a) the slip at which maximum torque occurs.
(b) The input line current and power factor at the condition of maximum torque.
(c) The value of the maximum torque.
4-40. Refer to the motor of Prob. 4-37, and find the value of resistance that must be
externally connected per phase to the rotor winding in order that the maximum
torque be developed at starting. What is the value of this torque?
4-41. A three-phase, 440-V, 50-hp, 60-Hz, Y-connected induction motor yields the follow-
ing no-load test data:
Dc test on stator per phase: Save 50 A
Blocked-rotor test at rated frequency: 106 V, 65 A, 3423 W, 60 Hz
Blocked-rotor test at low frequency: 42 V, 65 A, 2910W, 15 Hz

No-load test: 440 V, 16.3 A, 2157 W, 60 Hz

The friction and windage losses for this motor are 834 W. Compute the parameters
of the approximate equivalent circuit to be used at normal operating slips.
4-42. From the no-load test on a 10-hp, four-pole, 230-V, 60-Hz, three-phase, Y-connected
induction motor with rated voltage applied, the no-load current was found to be 9.2 A
and the corresponding input power 670 W. Also, with 57 volts applied in the
blocked-rotor test, it was found that the motor took 30 A and 950 W from the line.
The stator winding resistance was measured to be 0.15 © per phase. When this
motor is coupled to its mechanical load, the input to the motor is found to be 9150 W
at 28 A and a power factor of 0.82. The stator core losses are equal to the rotational
losses.
(a) Compute the rotor current referred to the stator.
(b) Find the developed torque.
(c) What is the value of the slip?
(d) At what efficiency is the motor operating?
4-43. A Y-connected, wound-rotor induction motor is found to have the following parame-
ters per phase:
r, = 0.26 Q, x, = 0.60, ro = 143 0
r, = 0.40 0, Xo Sara, Xo = 22.2.0

(a) If rated voltage of 220 V is applied to the motor at no-load, compute the line
current and power factor. Assume the stator core losses are equal to the rota-
tional losses.
(b) If the rotor is short-circuited and blocked and a line voltage of 60 V is applied,
find the power input in watts.
4-44, A three-phase induction motor has the following data: 20 hp rated, 220 V, Y-con-
nected, four-poles, 60 Hz. Stator effective resistance is 0.118 ( per phase; rotor
effective resistance referred to the stator is 0.102 per phase. Equivalent stator
referred reactance is 0.208 © per phase. Total core loss is 408 W, of which one-half
is for the rotor. Friction and windage is 400 W.
(a) What is the starting torque at full voltage?
(b) If the motor is driven at 1800 rpm against the revolving field, what is the brake
effect obtained in watts at the shaft? Assume that rotor resistance and core loss
remain unchanged.
Chap. 4 Problems 221

4-45. A three-phase wound-rotor induction motor delivers fixed torque to a load. The
rotor leakage reactance is negligible. For this load condition the rotor resistance
drop is found to be 20 V. Determine the rotor injected voltage needed to produce
twice this torque at 10 times the slip.
4-46. A 625-V (line voltage), three-phase, Y-connected wound-rotor induction motor has a
ratio of transformation from stator to rotor of 3.0. At a specified load condition with
the rotor short-circuited, the motor operates with a slip emf (i.e., sEy) of 6 V per
phase.
Find the value of the injected voltage needed to increase the slip to a value of
40%. Assume that the torque varies in a manner proportional to the square root of
the slip. Neglect rotor leakage reactance and rotational losses.
4-47. A four-pole, three-phase, wound-rotor induction motor is running at 1700 rpm with
the rotor short-circuited and rated voltage at rated frequency applied to the stator
winding. The rotor slip emf (sE) is 10 V per phase and the rotor current is 40 A per
phase. An injected voltage of slip frequency and constant rms value of 80 V per
phase is then introduced. This voltage is in phase opposition to the normal slip emf.
Rotor leakage reactance is negligible.
(a) What is the new motor speed if the torque required by the load is constant as the
speed varies?
(b) Repeat part (a), assuming that the torque varies directly with the speed.
4-48. A three-phase wound-rotor induction motor delivers rated torque and draws a line
current of 100 A at a pf of 0.707. The effective turns ratio from stator to rotor is 1.41
and the rotor resistance per phase is 0.1 . The rotor is wound for three phases.
Assume that the stator core loss, stator winding resistance, and stator and rotor
leakage reactances are all negligible.
(a) Calculate the value of rotor injected voltage needed to increase the line pf to 90%
lagging.
(b) Find the power supplied by the source of injected voltage.
4-49, A 60-Hz, six-pole, three-phase induction motor has its speed adjusted by a rotor
injected voltage. When it develops rated torque, the motor slip is 5%, and the
normal slip frequency voltage is 20 V. The rotor resistance per phase is 0.1 ohm and
the rotor leakage reactance is negligible. An injected voltage of 100 V per phase is
inserted into the rotor in phase with the original sE, voltage. If the torque demand
remains unchanged, determine:
(a) The new value of slip.
(b) The new operating speed in rpm.
(c) The power supplied by the injected voltage source.
4-50. A three-phase, 1000-hp, 2200-V, 25-Hz, 12-pole, wound-rotor induction motor has a
ratio of transformation from stator to rotor of 22 to 15. Both stator and rotor wind-
ings are Y-connected. Ata slip of 2% the motor develops its rated torque and draws
from the line a current of 180 A at a pf of 0.866. Consider the quantities r,, x,, x2,
and r, negligible. The motor is to develop the same torque at a slip of 10%, while
simultaneously operating at unity pf, by means of a rotor injected voltage. Deter-
mine the magnitude and phase of the injected voltage relative to the original slip emf.
4-51. A three-phase, Y-connected, wound-rotor induction motor delivers rated torque at a
slip of 4%, and draws a line current of 100 A at a pf of 80%. The effective turns ratio
from stator to rotor is unity, and the rotor is wound for three phases. The rotor
 Three-Phase Induction Motors Chap. 4
222

200 V per phase. For

resistance per phase is 0.1 ©, and the rotor standstill emf is
simplicity assume that ry; = x = x2 = 0.
torque at a slip of 20%
(a) Find the rotor injected voltage needed to give the same
below synchro nous speed.
pf unity.
(b) Calculate the rotor injected voltage required to make the line
to a constant torque
4-52. A three-phase induction motor with slip rings delivers power
The rotor current is 40 A per phase; the rotor resistanc e is 0.40 per phase. To
load.
whose magnitud e in this case
bring the pf to unity, a voltage is injected into the rotor
in the rotor at this particula r slip. The speed
is equal to the normal slip emf induced
than before.
of the rotor is now to be decreased so that the slip is five times greater
phase of the in-
The torque remains fixed. What must now be the magnitude and
jected voltage to produce unity pf?
at a slip of 2%. At
4-53. A three-phase, wound-rotor induction motor delivers rated torque
The rotor resistanc e per phase is
this slip the value of the rotor induced emf is 10 V.
produce twice rated torque at the same slip of
0.2 © per phase. If it is desirable to
2%, determine:
injected into
(a) The magnitude, phase, and frequency of the voltage that must be
the rotor.
(b) The amount of power that the auxiliary source must supply to the induction
motor.
Note: At the low slip of 2% the rotor leakage reactance may be neglected.
motor has
4-54. A 1000-hp, three-phase, 2200-V, 25-Hz, 12-pole, wound-rotor induction
the following parameters per phase:

r, = 0.093 QO, x, = 0.295 0

ry = 0.072 Q, xp 0S 1
stator to rotor effective turns ratio: 22/15.

When operating at a slip of 1.72% this machine develops a torque of 15,000 lb-ft,
corresponding to a rotor current per phase of 205 A. The friction and windage losses
total 6 kW.
(a) Determine the magnitude and phase of the minimum value of injected rotor
voltage needed to change the slip to 50% above synchronous speed at constant
torque. Do not neglect rotor leakage reactance.
(b) Determine the mechanical power output of the rotor in part (a).
(c) Determine the rotor electrical power “‘output”’ in part (a).
(d) Find the total rotor copper loss. How is this loss supplied?
4-55. In the configuration of Fig. P4-55 balanced three-phase currents of 20 Hz variation
are fed to the brushes of the de winding. The rotor rotates at a frequency of 35 Hz.
What is the frequency of the emf induced at the brushes? Why?
4-56. The resistance measured between each pair of slip rings of a three-phase, 60-Hz, 300-
hp, 16-pole induction motor is 0.04 Q. With the rings short-circuited, the full-load
slip is 0.025, and it may be assumed that the slip-torque curve is a straight line from
no-load to full-load. The motor drives a fan which requires 300 hp at the full-load
speed of the motor. The torque required to drive the fan varies as the square of the
speed. What value of resistance, should be connected in series with each slip ring so
that the fan will run at 300 rpm?
Chap. 4 Problems 223

From
balanced
three-phase
source

Figure P4-55

4-57. A three-phase, 60-Hz, four-pole, wound-rotor induction motor has a rated speed of
1746 rpm. The actual rotor resistance per phase is 0.8 . What starting resistance
must be inserted in the rotor circuit so that rated torque will be developed at start-
ing? Neglect any change of rotor resistance with frequency.
4-58. For the polyphase induction motor, sketch the variation of torque as a function of
horsepower output over the speed range from zero to synchronous speed. Hint:
Apply the circle diagram.
4-59. Continue the analysis displayed in Table 4-1 by completing the tabulation for the
remaining three 60° intervals of the cycle.
4-60. Draw the line-to-neutral voltage variations for each phase over one complete cycle
for the speed control system of Fig. 4-25 corresponding to the switching sequence
displayed in Fig. 4-26.
4-61. In the interest of generating a more nearly sinusoidal waveform for the line voltages
in the speed control system of Fig. 4-25, the gated ON interval for each SCR is
specified as 120°. Draw the switching sequence diagram for the inverter.
4-62. (a) Use the results of Prob. 4-61 and sketch the variation of each line-to-neutral
voltage over one cycle of the control frequency.
(b) Sketch the variations of the line-to-line voltages over one complete cycle of the
control frequency.
(c) Determine the expression for the rms value of the line-to-line voltages and com-
pare it with the result that appears in Eq. (4-56).
 a
Three-Phase
Synchronous Generators

for
The synchronous generator is universally used by the electric power industry
supplying three-phase as well as single-phase power to its customers . The single-
one
phase power that is brought to homes, shops, and offices originates from
phase of the three-phase system. Moreover, the assignment of commercial load
circuits to each phase is made in an effort to keep the phases balanced.
The basic construction features of these machines are described in Sec. 3-3
and illustrated in Fig. 3-17. It is worth noting here that synchronous generators
are classified in two types. The first is the low-speed (engine- or water-driven)
type, which is characterized physically by having salient poles, a large diameter,
and small axial length. The second is the turbogenerator, which uses the steam
turbine as the prime mover. The usual salient-pole rotor construction is aban-
doned in these high-speed generators in favor of the cylindrical (or smooth) rotor
because the protruding-pole construction gives rise to dangerously high mechani-
cal stresses. For 60-Hz three-phase power the two-pole generator must operate at
3600 rpm. Moreover, since turbogenerators are invariably designed for two poles,
it follows that these machines are further characterized by a small diameter and
long axial length.

224
 Sec. 5-1 Generation of a Three-Phase Voltage 225

5-1 GENERATION OF A THREE-PHASE VOLTAGE

How is a three-phase voltage generated? This can be readily understood from a

study of Fig. 5-1(a). Depicted here is a two-pole rotor, the field winding of which
is assumed to be energized from a dc source to create the pole flux. It is further
assumed that the pole pieces are shaped to produce a sinusoidal flux field. Ap-
pearing in the stator is a balanced three-phase winding with the axis of each phase
displaced by 120°. The complete winding of each phase is represented in Fig. 5-1
by a single coil. Now consider that the rotor is driven by the prime mover in a
counterclockwise direction at synchronous speed. Applying the v x B rule for the
field direction shown reveals that the instantaneous voltage induced in coil sides
a, b',c’ is directed out of the paper while in coil sides a’, b, c itis directed into the
paper. Furthermore, since coil sides a and a’ are located beneath the maximum
value of the flux-density wave, the induced emf in phase a is at its maximum
value. The corresponding induced emf in phase b, as identified by coil side b, is
seen to be of opposite sign (it is under a south-pole flux) and of smaller magni-
tude. In fact, since coil sides b and b’ are both displaced from the maximum flux-
(cos
density position by 60°, it follows that this instantaneous voltage is at one-half
result for phase c.
60°) its maximum value. A similar reasoning yields the same
Next consider that the rotor has advanced by 60° in the counterc lockwise
b’,
direction. This puts the amplitude of the north-pole flux directly beneath
in phase b. Also, coil
indicating that the induced emf is now a negative maximum
indicated in
c-c’ now finds itself under the influence of flux of reversed polarity as
of the paper and into the
Fig. 5-1(b). Hence the direction of the emf in c is now out
voltages for the first and
paper for c’. The instantaneous values of the three-phase
identifie d as ft; and f2,
second time instants are depicted in Fig. 5-2 and are
respectively.
results for
By repeating the foregoing procedure many times and plotting the
5-2. Note that the
each instant of time we obtain the complete curves of Fig.
other two except for
induced-emf variation for each phase is identical with the
of fact this varia-
time displacements of 120° and 240°, respectively. As a matter
phase space-displaced
tion is a direct consequence of having the beginning of each

Figure 5-1 Induced-voltage distribu-

tion in the stator winding of a synchro-
nous generator: (a) phase a at a posi-
tive maximum (time ¢, in Fig. 5-2);
(b) phase b at a negative maximum
(a) (b) (time 12).
 Three-Phase Synchronous Generators Chap. 5
226

-_—
E
©
33)=)
5©)
£

Figure 5-2 Time history of three-phase

induced emf’s for the generator of
Fig. 5-1.

by 120 electrical degrees. Physical connection of the ends of each phase (points
a’, b', c') gives a Y-connected stator winding.

5-2 LINEAR ANALYSIS BY THE GENERAL METHOD

The basic principles and characteristics of the cylindrical-pole synchronous gener-

ator can be readily described by developing the phasor diagram as was done for
the transformer. However, there exists an important distinction that must be
understood right at the start. The transformer (as also the three-phase induction
motor) is essentially a constant-flux device, i.e., the amplitude of the resultant
operating air-gap flux is fixed by the applied ac voltage. In contrast the resultant
air-gap flux in the synchronous machine depends upon the magnitude of the
exciting current in the field winding (located on the rotor structure) as well as on
the value of the armature (or load) current.+ Recall that for the synchronous
machine the armature winding is the stator winding. Consequently, when analyz-
ing the synchronous generator, it becomes necessary to introduce the magnetiza-
tion curve that applies for its magnetic circuit. If the design data are available, the
magnetization curve can be computed by following the procedure outlined in
Chapter 1. Otherwise, if the machine is available in the laboratory, the magnetiza-
tion curve can be found by performing a simple open-circuit test. The wiring
diagram for such a test is shown in Fig. 5-3. The rotor is driven at synchronous
speed throughout the test by the prime mover. The dec field current is varied by
means of the rheostat and at each setting the corresponding line-to-neutral voltage
is measured. A plot of these data points is depicted as curve (a) in Fig. 5-4 and is
often called the no-load characteristic. The straight-line portion of this curve is
called the air-gap characteristic. In this section our interest is confined to linear

+Here we are treating the synchronous machine as an isolated unit. If it were assumed to be
connected to an infinite bus system (i.e., in parallel with other synchronous generators many times its
size), the air-gap flux would be invariant.
Sec. 5-2 Linear Analysis by the General Method 227

Stator phase
winding

Figure 5-3 Wiring diagram for determining the no-load characteristic of the syn-
chronous generator.

used
analysis so that for our immediate purposes the magnetization curve to be
will be curve (b) of Fig. 5-4.
generator
Another useful characteristic in connection with synchronous
by short-ci rcuiting termi-
analysis is the short-circuit characteristic. This is found
synchro nous speed,
nals a, b, and c in the diagram of Fig. 5-3, driving the rotor at
of field current. A
and measuring the short-circuit line currents for various values
to curve (c) of Fig.
plot of the short-circuit line current versus field current leads
air-gap flux needed to
5-4. This characteristic is linear because the amount of
small as to cause little
produce even as much as twice rated armature current is so
flux is needed to pro-
or no saturation of the magnetic-circuit iron. Only enough
duce an emf which overcomes the leakage impedance drop.
, a few remarks
Before proceeding with a description of the phasor diagram
phasor diagram of both
are in order at this point concerning the inclusion in one
and the mmf space
the time-varying quantities (such as voltage and current)
The field winding is lo-
sinusoids. Figure 5-5 depicts a two-pole turbogenerator.
slots. When a direct cur-
cated on the rotor and is distributed in suitably placed
-condu ctor distribu tion is
rent is made to flow through this winding, an ampere
into the conduct ors on the
produced. The illustration shows the current moving
left side. The corresp onding
right side, and flowing out of the conductors on the
by this field winding has a sinusoid al distribution in space with its
mmf produced
m value of the flux density
axis directed along the vertical. Because the maximu
induced in coil ey-ey is a
lies along this vertical too, it follows that the voltage
rule indicates that this emf is
maximum value. Moreover, application of the v x B
into the paper for the bottom
directed out of the paper for the top conductor and
228 Three-Phase Synchronous Generators Chap. 5

voltage
Line-to-neutral

Fig
Stator
circuit
short
current,
amperes

Pyeae 10,000kVA,3-phase, Y- connected

13,800 volts, 2-pole, 60: Hz
turbogenerator

O 50 100 I50 200 250 300 350 400 450

Field amperes

Figure 5-4 No-load characteristics: (a) nonlinear open-circuit characteristic;

(b) linear air-gap characteristic; (c) short-circuit characteristic.

conductor. Although it is assumed that the stator is equipped with a three-phase

winding, essentially coil e-ey can be considered to represent the emf induced in
this entire winding for the instant shown. Accordingly, if one were to look at the
emf induced in each coil side of the entire three-phase winding, a sinusoidal emf
distribution would result, with its peak lying along the vertical. As a matter of
convenience, therefore, it can be said that insofar as the armature induced emf is
concerned, the entire three-phase winding can be represented by the single coil
er-e. Note that the axis of this coil lies along the horizontal. In general, the axis
of the coil that represents the emf distribution in the three-phase winding is always
90 electrical degrees displaced from the axis of the field mmf #-. As a result of
this quadrature relationship it will be customary for us in the phasor diagrams
which follow to locate the field mmf #- at a position 90 degrees leading the
direction of the excitation emf phasor. The excitation emf is the voltage induced
by the field winding flux alone. It is helpful to note in Fig. 5-5 that when the emf in
coil er-e is a maximum, the flux linkage is zero.
A similar convention is adopted concerning the phasor representation of the
armature current and the corresponding mmf associated with this current. For
example, in Fig. 5-5, coil i,-i, is used to denote the sinusoidal ampere-conductor
Sec. 5-2 Linear Analysis by the General Method 229

Axis of ans

oir Wiis &Axis of coil e¢ —e;

Axis of Fy

Figure 5-5 Diagram illustrating the relationship between mmf space sinusoids
and the associated time-varying quantities of excitation voltage and armature
current. The rotor is assumed to be driven counterclockwise at w radians per
second.

distribution of the entire three-phase armature winding for a phase lag relative to
the excitation emf of & degrees. Note that the axis of this coil coincides with the
location of the peak value of the corresponding sinusoidal armature mmf #, and is
space displaced from the field mmf axis ¥- by (90° + &). Another point worthy of
note is that as the phase angle between armature current and excitation emf
changes, a corresponding change occurs in the axis of #4. This fact is noted in
subsequent phasor diagrams by placing the armature mmf coincident with the
phasor for armature current.
In synchronous machine analysis it is important to note that the two space
sinusoids ¥ and #4, having the same frequency of rotation and the same number
of poles, can be added to yield a resultant space sinusoid denoted by ¥g, which in
wind-
turn is responsible for generating the resultant induced emf in the armature
represent ed in the
ing E,. As a convenience, the result of this addition is also
sinusoids and not
phasor diagram in spite of the fact that these mmf’s are space
time-varying sinusoids. This procedure is permissible because space sinusoids
can also be summed by phasor methods.
One final point. Application of the force rule (I x B) discloses that an
counter-
average torque is produced on the armature (stator) winding to cause
there occurs in-
clockwise rotation. However, because the stator is immobile,
orque to
stead a clockwise torque on the rotor. The latter represents a counter-t
the prime mover of the turbogenerator.
e basis.
The development of the phasor diagram is done on a per-phas
e x, per phase
Assume that the armature winding resistance r, and leakage reactanc
 Three-Phase Synchronous Generators Chap. 5
230

ous
are known along with the air-gap characteristic. Assume too that the synchron
armature current to a lagging power-fac tor load at
generator is delivering rated
rated terminal voltage per phase V, and the load
rated terminal voltage. Call the
power-factor angle 6. Let it be desired that the value of the field current corre-
sponding to this load condition should be found. Before proceeding, it is helpful
to keep in mind the factors that must be reckoned with in establishing the rated
terminal voltage when rated current flows. It is shown in Sec. 4-1 that the exis-
tence of three-phase currents in a three-phase winding leads to a revolving mmf
the
F,. It follows then that the resultant mmf Fr is made up of the vector sum of
field mmf ¥;- and the armature mmf #4. In equation form we have
Fee oh Fp ae Fr (5-1)

In turn, this resultant mmf is responsible for the resultant air-gap flux ®, that
induces the actual armature winding voltage E,. The terminal voltge is then ob-
tained by subtracting the armature leakage impedance drop per phase from E,.
We shall now develop the phasor diagram by starting with the known termi-
nal conditions: rated armature current J, lagging the rated terminal voltage V, by @
degrees. Refer to Fig. 5-6. The induced armature winding voltage per phase is
obtained as the phasor sum of V, and the armature leakage impedance drop.
Thus, employing the horizontal axis as the zero reference line for angle measure-
ments, we can compute ;

de = V; a: Tire eM PT

or

E, /y = Vi {0 + In[—9 (ra + JX) (5-2)

The quantity R that produces E, is found by entering the air-gap characteristic.
Note that R is expressed in field amperes rather than field ampere-turns Fr. This
is irrelevant here in view of our objective, which is to find the required field
current for the specified load condition.
The complete expression for the field current representing the resultant flux
is then
Riz Rk [90D 2 (5-3)

where y is determined from the right side of Eq. (5-2). It is convenient now to
write the counterpart of Eq. (5-1) in field amperes, which after all is the quantity
that usually appears as the abscissa quantity in the experimentally established
curves of Fig. 5-4. Thus
F+A=R (5-4)
where A denotes the armature mmf expressed in equivalent field amperes.
i A glance at Eq. (5-4) shows that with R available F can be determined once
A is known. There are two way’ to determine A, either from the design data or
from appropriate no-load tests. The latter approach is discussed in Sec. 5-4.
Sec. 5-2 Linear Analysis by the General Method 231

r
Figure 5-6 Phasor diagram of a synchronous generator for a lagging power-facto
load. The general method of analysis is indicated.

represent the
When the design data are available, Eq. (3-24) may be used to
in mind that this equation is derived (see Appendix
armature mmf per pole. Keep
distribution of the
C) as the amplitude of the fundamental component of the mmf
the distribu tion of the field winding on the cylindrical-
armature winding. Also,
having a
rotor surface is such that it leads practically to a sinusoidal mmf wave
where Ny is the number of turns per pole. Accord-
peak mmf per pole of Nyl;,
and the field mmf waves are both of the same shape,
ingly, since the armature mmf
mmf as describe d by Eq. (3-24) can be replaced
it follows that the actual armature
by an equivalent field winding current A. Thus
1 N2Kw2 ps p
nt field amperes (3-3)
5-5
N;
Av===0.9.:9 3 q
equivale

where g = number of phases

p ll
= number of poles
N>K,y2 = effective armature winding turns per phase
[, = armature current per phase.
nship as already
Equation (5-5) indicates that A and J, differ by a fixed relatio
lent field current repre-
described. Hence the phasor expression for the equiva
load is
senting the armature mmf for lagging power-factor
A=A/-8 (5-6)

diagram of Fig. 5-6 note that A

where A is described by Eq. (5-5). In the phasor
of convenience.
and J, are shown on the same line as a matter
232 Three-Phase Synchronous Generators Chap. 5

With R and A determined, the required field current is found to be

F=R-A=R/90+y-A/0 (5-7)
The quantity Ey appearing in Fig. 5-6 and shown 90° behind F is the voltage
produced by F at the armature terminals with the load removed (i.e., J, = 0) and
no saturation of the iron. It is usually called the excitation voltage. The foregoing
procedure for finding the field current is referred to as the general method.
The positive direction of rotation for the phasor quantities and the space
sinusoids depicted in Fig. 5-6 is counterclockwise and is indicated by the direction
of the arrow marked w. It was also previously pointed out that for generator
action a counter torque exists on the rotor. This is indicated by the arrow marked
T in Fig. 5-6. It is useful to note here that in generator action the torque on the
field structure is always such as to bring the field axis in alignment with the
resultant flux.

Example 5-1
A 10,000-kVA, three-phase, Y-connected, two-pole, 60-Hz, 13,800-V, line-to-line
turbogenerator has the air-gap characteristic depicted in Fig. 5-4. The armature
winding resistance is 0.07 per phase and the armature winding leakage reactance
per phase is 1.9. Also, the armature mmf at rated current is known to be equal to
155 equivalent field amperes. Find the field current that produces rated terminal
voltage when rated armature current is delivered to a 0.8 lagging power-factor three-
phase balanced load.

Solution The rated phase voltage is

13,800
V,=—
t V3 = 7967.4 V (5-8 )

and the rated armature current is

10,000
= : = 418.4A 5-9
V3 13.8 i
The actual armature induced emf is found from Eq. (5-2) to be

= E, /y = 7967.4 /0° + 418.4 /=36,8" (OL07 sg l.9) (5-10)

Note that r, is small enough to be neglected. Thus,

E, = 7967.4 + 418.4(0.8 — j0.6)j1.9

= 7967.4 + 477 + j636

= 8444.4 /4.3° (5-11)

The corresponding value of R from the air-gap characteristic is 157 A. Hence by
Eq. (5-3)
R= 157 (90° 14.37 = 1577 94:3° (5-12)
 Sec. 5-3 Linear Analysis by the Synchronous Reactance Method 233

Equation (5-7) then yields the desired result:

Pe dT re 15) 90,0.

= 115747 156.6 — 124 +-j92:8

= —135.79 + j249.1 = 284 /119° A (5-13)

The excitation voltage produced by this field current is found on the air-gap line to
be
Ey = 15,300 /29° V (5-14)
This quantity is almost twice the rated terminal voltage. It is a highly exaggerated
result because of the assumption of no saturation. Practical and useful results are
obtained when this assumption is dropped as described in Sec. 5-4.

5-3 LINEAR ANALYSIS BY THE SYNCHRONOUS REACTANCE

METHOD. THE EQUIVALENT CIRCUIT

In any treatment of the behavior of synchronous machines it is necessary to take

due account of the effects produced by the current that flows in the armature
winding—frequently referred to as armature reaction. In the preceding section
this was accomplished by working directly with the armature winding mmf ex-
pressed in equivalent field amperes. A second approach is possible by treating the
armature reaction in terms of a suitable flux and its associated induced voltage.
Since the assumption of no saturation has been made, it is to be expected that
either viewpoint will lead to the same result. As a matter of fact this is precisely
the reason for imposing the linearity condition. In this way the introduction of
new concepts can readily be correlated with preceding ideas with consistency.
By means of superposition the effect of the rotating armature mmf can be
treated individually. Accordingly, the set of revolving mmf poles created by the
three-phase armature current in the three-phase winding can be considered to
relation-
produce an armature reaction flux ®,,. Of course this flux bears a fixed
ship to its mmf A. Moreover, it revolves with respect to the stationary armature
a
winding at a speed corresponding with 60 Hz. Hence ®,, can be said to induce
in Fig.
voltage (rise) of value E,, which can be placed 90° behind ®,, as depicted
of this kind it is customary to replace such an armature flux
5-7(a). In situations
drop that involves the associated current (J, in this case)
voltage by a reactance
of leakage
directly. Recall that this same technique was used to replace the effect
as well as armature leakage flux occurring around the slots
flux in the transformer
of the stator windings. Expressed mathemati cally this is
234 Three-Phase Synchronous Generators Chap. 5

where x, is the armature reaction reactance. The —j factor accounts for the 90°
lag existing between I, and E,,. Furthermore, the linearity condition enables a
constant value of x, to replace the effect of armature flux. Thus a 50% increase in
armature reaction flux brought about by a 50% increase in armature current causes
E,, to increase to 1.5 times its previous value. But the presence of J, on the right
side of Eq. (5-15) automatically accounts for this. If J, and ®,, were not assumed
to be linearly related, this correspondence would not prevail. Finally, it should be
apparent that, when the synchronous generator operates at a field current F, the
associated excitation voltage must contain a component equal and opposite to that
specified by Eq. (5-15).
Based on this background the complete phasor diagram of the effect of
armature mmf treated in terms of its equivalent reactance drop can be developed.
We begin by assuming rated armature current flows to a balanced three-phase
lagging power-factor load at rated terminal voltage. As with the general method,
the armature leakage impedance drop can be added to V, to yield E,. Refer to Fig.
5-7(b). However, at this point, instead of dealing with mmf’s, the analysis con-
tinues in voltages. Thus to E, is added an armature reaction reactance drop that is
equal and opposite to E,,. The resultant is the excitation voltage. An important

Se, = ~jTQx¢
(b)
Figure 5-7 Illustrating the replacement of the effect of armature mmf by an
equivalent induced emf E,,.
Sec. 5-3 Linear Analysis by the Synchronous Reactance Method 235

observation to make in Fig. 5-7(b) is that the leakage reactance drop and the
armature reaction reactance drop add along the same line, both being perpendicu-
lar to I,. This allows the two quantities to be added algebraically, which leads to a
total reactance called the synchronous reactance and denoted by x;. In equation
form

where x; is the armature winding leakage reactance per phase and x, is defined by
Eq. (5-15).
A study of Fig. 5-7(b) discloses that with the use of the synchronous reac-
tance the required field excitation for any specified load condition can be found in
two steps. The first step involves finding E; directly from the equation
Ep = V, + [(ra + jxs) ~ VilO? + La /—8 jas (5-17)
In step two the air-gap characteristic is entered at the level of E, and the corre-
reac-
sponding value of F is recorded. This procedure is called the synchronous
mind that the synchrono us reactance is a
tance method. It is useful to keep in
of the armature winding leakage flux and
fictitious quantity that replaces the effect
the armature winding rotating mmf.
circuit
A circuit interpretation of Eq. (5-17) leads directly to the equivalent
depicted in Fig. 5-8. This is an immedia te
of the synchronous generator and is
consequence of introducing the concept of synchro nous reactanc e.
are avail-
How is the synchronous reactance determined? If the design data
stator as well as the reluctan ce of the
able so that the effective turns of the
expressi on like Eq. (3-84) of Appen-
magnetic circuit can be computed, then an
air-gap charac-
dix to Chapter 3 can be used. Otherwise x, can be found from the
1.e., through the use of curves (b) and
teristic and the short-circuit characteristic,
examine the phasor
(c) of Fig. 5-4. To understand how this comes about let us
ns with the field
diagram of the synchronous generator under short-circuit conditio
rated armature current to flow. The
current adjusted to the value that causes
but the terminal voltage is zero. Ac-
armature winding mmf remains the same,
to that shown in Fig. 5-9(a).
cordingly, the phasor diagram of Fig. 5-7(b) reduces
to neutrali ze the effect of the
Note that the field current is used almost entirely
only large enough to overcom e the
armature winding mmf. The remainder, R,-, is
purposes is equal to the leakage
leakage impedance drop, which for all practical
short-circuit conditions lags
reactance drop. Note too that the armature current in

Vt Load Figure 5-8 The per-phase equivalent

circuit of the synchronous generator
lies to the left of the load terminal.
236 Three-Phase Synchronous Generators Chap. 5

Air-gap

voltage
Line-to-neutral current
phase
Armature

(e)

Field current, amps

(a) (b)

Figure 5-9 Determining the synchronous reactance from no-load tests: (a) phasor
diagram at short-circuit; (b) no-load characteristics.

the excitation voltage by nearly 90°. For obvious reasons at such low power
factors the quantities R, A, and F are essentially collinear. However, the impor-
tant thing to note in Fig. 5-9(a) is that the excitation voltage at short-circuit is
practically equal to the synchronous impedance drop. This observation provides
the key for the determination of x,. If F,. is maintained and the short-circuit is
removed, the open-circuit voltage produced by this field current is obviously
Eysc. Therefore, with F,. known, the air-gap characteristic can be used in the
manner illustrated in Fig. 5-9(b) to obtain E;,-. Similarly, by entering the short-
circuit characteristic the corresponding current produced by F,. can be found.
The synchronous reactance is then the ratio of these two quantities. Thus

Xs — Efe (5-18)
Ta
Keep in mind that this is the unsaturated value of synchronous reactance because
we are still working with the air-gap line.
Example 5-2
Determine the following information for the machine of Example 5-1: (a) the value of
the unsaturated synchronous reactance and (b) the field current needed to establish
rated terminal voltage when rated current is delivered to a balanced three-phase load
of 0.8 power factor lagging, using the synchronous reactance method.

Solution (a) A glance at Fig. 5-4 indicates that a field current of F,. = 170 A is
needed to force rated current of 418.4 A through the armature winding at short-
circuit. Furthermore, the excitation voltage corresponding to 170 field amperes is
Eysc,= 9200 V. Hence
 Sec. 5-4 Nonlinear Machine Analysis 237

E 9200
Xs =
Rpm
ems
ee
—eOoro-,-—sas =
220 (5-19)
-

(b) By Eq. (5-17) we have

Ep = V, [0° + flax, = 7967.4

/0° + 418.4(0.8 — j0.6)j22
ll 7967.4 + 5523 + 77364 = 13,490 = j7364

= 15,369 /28.6° (5-20)

Then from the air-gap characteristic .
F=285A (5-21)

A comparison of these results with Eqs. (5-13) and (5-14), which are obtained
by the general method, shows them to be almost identical. Such correspondence
is entirely in keeping with the linearity assumption.

5-4 NONLINEAR MACHINE ANALYSIS

The power output capability of electric machines is directly dependent upon the
degree to which the material is worked magnetically and electrically. The higher
the operating flux density in the iron parts of the machine and the current density
in the copper parts are, the greater is the output. Consequently, in most electric
machines the level of flux density chosen puts operation in the saturated or nonlin-
ear region. If analytical procedures are to be used to predict machine perfor-
mance, then the effect of the nonlinearity must be appropriately accounted for.
Otherwise realistic results will be almost impossible to attain. This is a point to
keep well in mind when sophisticated procedures are employed to predict ma-
chine behavior, formulated in terms of inductance parameters based on linearity.
The obvious way to account for operation of the machine in its nonlinear
region is to work directly with the nonlinear characteristic usually represented in a
graphic form such as curve (a) in Fig. 5-4. This curve is the actual open-circuit
characteristic and may be either computed from the design data by employing the
procedures discussed in Chapter 1 or found experimentally as described in Sec.
5-2. It should not be inferred that the mere use of the actual nonlinear characteris-
tic guarantees realistic results. That this need not be so will be illustrated pres-
ently. Attention is now turned to the application of the general method and the
synchronous reactance method to find the necessary excitation currents for syn-
chronous generators by utilizing the actual, nonlinear open-circuit characteristic
instead of the linear air-gap characteristic. It is worthwhile to note that informa-
tion about the range of the excitation currents is necessary for the proper design of
a voltage regulator which will act to maintain constant terminal voltage in the
presence of changing load demands.
 Three-Phase Synchronous Generators Chap. 5
238

The General Method

voltage when
Again, let us find the excitation needed to produce rated terminal
diagram
rated current flows to a balanced lagging power-factor load. The phasor
e winding emf
of Fig. 5-6 applies with undiminished validity. The actual armatur
induced by the resultant flux is computed as in the linear case. Thus
E= V, [0° + I,(ra + jx1) = 7966.4 + 418.4/—36.8° (0.07 + JAD)
= 8444.4 /4.3° V (5-22)

The value of the resultant mmf expressed in field amperes that produces this result
is found by entering the nonlinear open-circuit characteristic (Fig. 5-4) at 8444.4
volts and reading the associated abscissa value. This yields
R=272.4
so that the complete phasor expression becomes
R = 222/94.3° A (5-23)
A comparison with the result obtained through linear analysis indicates a wide
discrepancy (222 versus 157 A), which attests to the fact that the machine is
operating at a fairly high level of saturation.
The armature reaction mmf is again treated in terms of an equivalent field
current of 155 A. Application of Eq. (5-7) then provides the required value of the
field current:
F = R — A = 222/94.3° — 155/-36.8°
= -140.6 + j314
= 344/114° A (5-24)
Eq. (5-13) shows that the result obtained by linear analysis is considerably lower
and, if used for the design of a voltage regulator, would obviously lead to an
inadequate design.
The field excitation voltage that corresponds to the field current of 344
amperes is found on the curve of the open-circuit characteristic to be
Er = 9560 V (5-25)

This value stands in sharp contrast to the figure of 15,300 found in the linear case.
The latter quantity is highly unrealistic, and it is easy to appreciate why by noting
in Fig. 5-4 the wide divergence existing between the open-circuit and the air-gap
characteristics for field current in the vicinity of 300 A.
Equation (5-25) identifies the value of the terminal voltage that occurs if the
excitation current is maintained at 344 A and the load is disconnected. In other
words the no-load terminal voltage is 9560 V. Accordingly, the voltage regulation
of this machine in percent is found to be
Sec. 5-4 Nonlinear Machine Analysis 239

nr OY Een, _ 9560 — 7960 Mg

voltage regulation = Vann 100 = 7960 100 = 20.1% (5-26)

where Eris the excitation voltage found on the magnetization curve. This figure is
not only realistic but typical of such machines. Calculation of the voltage regula-
tion for the linear case yields a meaningless result.

The Saturated Synchronous Reactance Method

It is shown in Sec. 5-3 that the use of the unsaturated synchronous reactance
method in finding excitation currents leads to unrealistic solutions for reasons
therein described. However, it is possible to alter the method of computation in a
way that is consistent with the state of saturation that the machine is in and
thereby obtain results that are comparable with those of the general method.
Interested readers are referred to Appendix E for details of a method that is based
on the flux density conditions associated with the resultant mmf R.
The generally accepted way of accounting for a saturated value of synchro-
nous reactance is to deal with the flux density condition that is associated with the
mmf needed to produce rated voltage at no load. To illustrate, let the field current
that is needed to generate rated terminal voltage (Vr) be represented by Oa in
Fig. 5-10. Now, if this same field current is used in a short-circuit test, there
results a short-circuit armature current (J,..) that is often in the vicinity of the rated
value (Iz). In Fig. 5-10 this current is shown to be slightly less than the rated
current. The conditions that prevail in the short-circuit case are such that there
exists a synchronous reactance drop, J,-x;, that is virtually equal in value to the
rated terminal voltage because the field current Oa is common. Thus, we can
write
LX = Ve (3-27)
or
Li
Vr
sT om saturated synchronous reactance (5-28)
SC

® ®

S
oe
8
a =
& o
2 E
Te} =)

: :E
E
oO

z < Graphical data used to

Figure 5-10
find the saturated value of synchronous
reactance; also the short-circuit ratio
(SCR).
240 Three-Phase Synchronous Generators Chap. 5

By way of comparison it is instructive to note that the unsaturated value of

synchronous reactance is, of course, a larger quantity than that displayed in
Eq. (5-28) since in the calculation use is made of the voltage on the air-gap line
rather than on the open-circuit characteristic for excitation Oa.

Short-Circuit Ratio. A useful synchronous generator parameter results

when Eq. (5-27) is expressed in terms of per-unit values. Accordingly, we have
(p-u I,.)(p-u x5) = p-u Vr
Since in machine analysis, the rated voltage is taken as the base quantity, the
corresponding per-unit value is unity. Therefore, it follows from the last expres-
sion that the per-unit value of the saturated synchronous reactance can be given
simply by
1
p-u Xs; = ints (5-29)

A modified form of this last expression is obtained by recalling the definition of the
per-unit value of the armature current, namely,
_ I,- (in amperes) _
ao Ip (in amperes) — Hoy G30)
which is also defined as the short-circuit ratio (SCR). Hence,
I 1
p-u x, =
p-u Isc - Son (5-31)

Another view of the short-circuit ratio is possible by noting from the similar
triangles found in Fig. 5-10 that

pS dl Da 5-32
Ob bd. Tp oe
Accordingly, the short-circuit ratio can also be identified as the ratio of field
current (Oa) that generates rated voltage at no load to the field current (Ob) that
produces rated current at short-circuit. The short-circuit ratio is useful because
its reciprocal denotes the per-unit value of the saturated synchronous reactance,
one of the important parameters of the synchronous machine.
Example 5-3
For the machine of Example 5-1 find the excitation emf corresponding to rated
terminal voltage when rated armature current is delivered to a 0.8 lagging pf three-
phase balanced load. Use the saturated synchronous reactance method. Also find
the short-circuit ratio and the per-unit value of the saturated synchronous reactance.

Solution The field current needed to generate the no-load rated phase voltage of
7967.4 V is found from Fig. 5-4 to be
Oa = 188A
Sec. 5-4 Nonlinear Machine Analysis 241

The armature current on the short-circuit characteristic produced by this same field
current is
Ij. = 462 A

Hence, the value of the saturated synchronous reactance becomes by Eq. (5-28)

IOLA
Ra ap oy 17.25 O/phase

The per-unit value of the saturated synchronous reactance can be found by two
procedures. The first uses the short-circuit ratio which by Eq. (5-30) is found in this
case to have a value of

SCR be
= p-u I,. =a Aa Ais 1.105

Then invoking Eq. (5-31) the per-unit saturated synchronous reactance is

SEU eee

The second procedure makes use of the base impedance which for the example at
hand is given by
tt Virgen 01-460
Lene = mee AiR ae 19.04 0

Accordingly,
pee NPs
p=-Ux, = yee = 19.04 = 0.906

The excitation voltage is found by applying Kirchhoff’s voltage law to the

equivalent circuit of Fig. 5-8. Thus,

E, = V, + Ifa + ix) ~ Vi + 1,/-36.8° Gix;)

7967.4 /0° + (418.4)(0.8 — j0.6)j17.25

13,585.84 /25.15° V

Comparing this result with that found for the unsaturated case displayed in Eq. (5-14)
by em-
indicates a marked improvement. Still more improvement can be realized
described in Appen-
ploying the modified saturated synchronous reactance method
dix E.

The field excitation current needed to maintain constant terminal voltage at

fixed lagging load power factor rises with increasing kVA demand at the load
terminals. This is obvious from the results of the preceding examples. In those
circumstances where the load draws a leading current, the need for field excitation
actually diminishes with increasing load kVA because of the magnetizing effect
caused by the armature mmf. These results are graphically portrayed in Fig. 5-11.
 Three-Phase Synchronous Generators Chap. 5
242

0.8 pf lead

1.0 pf

0.8 pf lag

for
Field
current
operation
voltage
terminal
constant
at Figure 5-11 Change in field current
needed to maintain constant terminal
voltage as load kVA varies for different
RatedkVA Load kVA load power factors.

5-5 EXPERIMENTAL DETERMINATION OF LEAKAGE REACTANCE

x, AND ARMATURE REACTION MMF A

Nonlinear machine analysis by the general method presupposes knowledge of the

armature winding leakage reactance as well as the armature winding mmf ex-
pressed in equivalent field amperes. When the complete design data of a machine
are available, these quantities can be readily calculated. However, in the absence
of such data other means must be sought to furnish the information.
As a review, let us return to Fig. 5-9, which depicts the situation in the
synchronous generator at short-circuit with rated current flowing. Recall that the
armature current flows at almost zero power factor and so the quantities E.R
-and A are collinear. In Fig. 5-9(b), F,, is a measured quantity and is accordingly
available. If the winding data are known, the quantity A can be computed from
Eq. (5-5) and then algebraically subtracted from F,, to yield R,-. Then on the air-
gap line corresponding to R,., the leakage reactance drop is found. Strictly speak-
ing, this ordinate quantity is E,,,.), but because x; > r, little error is made by taking
Exsc) = Tax. Since Exs5:) and I, are then both known, x; can be found. Unfortu-
nately, this determination of x; is dependent upon knowledge of A which is gener-
ally not available and must also be found.
On the basis of the foregoing it appears worthwhile to explore further the
operation of the synchronous generator at zero power factor. The wiring diagram
for such a test is shown in Fig. 5-12(a). The zero power-factor load consists of
purely inductive coils arranged in a Y-connection. Initially the coils are set for
zero reactance and the field current is adjusted to yield rated current. This corre-
sponds with point F,. in Fig. 5-9(b). The coil reactance is then increased, which
causes a drop in current. The field current is then increased to reestablish the
rated value of the armature current. Now the terminal voltage is no longer zero
but, rather, equal to J,X,, where.X, is the load reactance. By continuing this
procedure, rated terminal voltage can be made to appear across the zero power-
 Sec. 5-5 Leakage Reactance x, and Armature Reaction mmf A 243

Figure 5-12(a) Wiring diagram for obtaining the zero power-factor character-
istic.

aa
re)
>
£

8 Ghee eee ae a,
Sy
88
<=
a. Air-gap char.

|
Zero pf char. |
191 for rated I, Figure 5-12(b) Illustrating how the
zero power-factor characteristic can be
F, derived from the short-circuit data for
Field current, amp conditions of no saturation.

factor load with rated armature current flowing. A plot of terminal voltage versus
field current yields the zero power-factor characteristic.
the
If the machine were entirely linear (i.e., with no saturation occurring),
knowledg e of F;,
zero power-factor characteristic could be readily drawn from a
of Fig. 5-12(b).
alone. The construction should be obvious from an inspection
line Oc to
The zero power-factor characteristic is generated merely by displacing
at
higher and higher voltage levels. Note that to produce rated terminal voltage
A from F,, R
zero power factor, the total field current must be F,. By subtracting
 244 Three-Phase Synchronous Generators Chap. 5

voltage,
Phase
volts
c

Triangles Oac and Oo'c’

are identical.

Figure 5-13. Graphical determination

of leakage reactance x, and armature
mmf expressed in field amperes.
Triangle a'b'c' is known as the Potier
Field current, amp. triangle.

is obtained corresponding to which is the armature induced voltage, E, on the air-

gap line. The vertical distance between E, and V, (line a’b’ in Fig. 5-12(b)) con-
tinues to be the leakage reactance drop because at zero power factor E, and V, are
also essentially collinear in the phasor diagram.
Let us next consider what happens to the situation depicted in Fig. 5-12(b)
when the actual nonlinear characteristics are used. Refer to Fig. 5-13. Curve 1
denotes the open-circuit characteristic and curve 2 denotes the zero power-factor
characteristic for rated armature current. A little thought should make it clear
that the zero power-factor curve cannot be derived in the manner just described
for the linear case because of saturation. Hence this characteristic must be found
experimentally (or at least points c and c’). A useful thing happens if we attempt
to construct triangle O’a'c' of Fig. 5-12(b) on the characteristics of Fig. 5-13. To
accomplish this, it is necessary to mark off line O’c’ equal to Oc. Then at O’ a line
is drawn parallel to the air-gap line until it intersects the open-circuit characteristic
at point a’. Thus the triangle Oac is duplicated and, what is more, as a result of
this construction the two items being sought are immediately available. Line a’b'
represents the leakage reactance drop and can easily be measured in volts. Then

(5-33)

This quantity is called the Potier reactance. Moreover, the armature winding
mmf expressed in equivalent field amperes is immediately available upon measur-
ing line b’c' against the abscissa axis. Thus

A = b'c’ | in field amperes (5-34)

 Sec. 5-6 Theory for Salient-Pole Machines 245

The rationale for drawing a line at O' that is parallel to the air-gap line lies in
the fact that triangle Oac involves quantities in which saturation bears no influ-
ence, namely, stator leakage reactance, x, (which involves chiefly air paths) and
the armature winding mmf A (which is dependent upon solely winding data and
current). This is not to say that saturation is not present; it is and a comparison of
F, corresponding to c’ in Fig. 5-12b with F, in Fig. 5-13 makes this obvious.
However, by using c’ in Fig. 5-13 as a starting point and then constructing the
Potier triangle, O’a'c’, in a manner that ignores saturation, valid information
about x, and A thereby results.

5-6 THEORY FOR SALIENT-POLE MACHINES

In contrast to the cylindrical-pole synchronous machine, which has a uniform air

gap, the salient-pole machine has a highly nonuniform air gap because of the use
of a protruding-pole structure. A glance at Fig. 3-7 makes this obvious. A flux
path that includes two protruding poles has two small gaps to cross. This path is
indicated in Fig. 5-14 by ¢, and is called the direct axis. It is the path of minimum
reluctance. On the other hand the path denoted by @, in Fig. 5-14 involves two
large air gaps and is the path of maximum reluctance. This is called the quadra-
ture axis. In normal operation the armature (or stator) winding mmf is distributed
with its peak value located somewhere between the direct and quadrature axes.
Accordingly it produces a significant effect in each axis—but in a different way
because of the considerable difference in the reluctance variation of each axis.
Consequently, the cylindrical-rotor theory developed in the preceding sections
cannot be used for the salient-pole machine with satisfactory results. A modified
theory is required that takes due account of this difference. In this section such a
theory is developed for the salient-pole machine in both a two-reactance method
(analogous to the synchronous reactance method for smooth-rotor machines) and
a method based on mmf’s.

Two-Reactance Method for Salient-Pole Machines

To provide background, let us consider first the manner of locating the position of
the armature winding mmf relative to the rotor field distribution when current
flows. Call the angle between the excitation voltage and the armature current w.
Again refer to Fig. 5-14(a). For the indicated direction of the rotor flux and speed
the induced emf in those conductors under the influence of the north-pole flux is
directed outward; and in those inductors under the influence of the south-pole
flux, it is directed into the plane of the paper. Coil e;-e; is a single coil that
corresponds to the location of the amplitude of the fundamental sinusoidal emf
distribution that takes place in the entire three-phase winding for the instant
shown. The resultant distribution of the currents in the same three-phase winding
for a lagging power-factor load is represented by coil i,-i,, Which is located behind
 Three-Phase Synchronous Generators Chap. 5
246

Direct-axis
flux path

Quadrature -axis
flux path

Armature ampere-conductor
distribution

(b)

Figure 5-14 (a) Identification of direct- and quadrature-axis flux paths; () illus-
trating how the armature mmf, AN,, is resolved into two components: one acting
in the direct axis, A,N;; the other acting in the quadrature axis, A,N,.

the emf distribution by the specified phase angle . Appearing in Fig. 5-14(b) is
the developed representation of Fig. 5-14(a). The armature ampere-conductor
distribution is shown dotted and lagging behind the field flux distribution by yw. On
the other hand the armature mmf, which in synchronous machines is being de-
noted in terms of equivalent field mmf, is represented by AN;, where A is the
Sec. 5-6 Theory for Salient-Pole Machines 247

equivalent field current as described by Eq. (5-5). It is always desirable to work in

equivalent field current so that the abscissa axis of the open-circuit characteristic
may be used directly in accounting for armature reaction mmf. Again note that
the armature mmf is 90 degrees behind the ampere-conductor distribution. Ac-
cordingly, the armature mmf curve AN; takes the position shown in Fig. 5-14(b).
Clearly, the location of this wave is dependent upon the phase angle between Ey
and [,, which in turn is dependent upon the power factor of the load and the
amount of power delivered (see Sec. 5-7).
It is helpful to consider for a moment the general effects caused by various
locations of the armature mmf wave. If the load conditions are such that w be-
comes equal to 90°, the armature mmf wave lies entirely in the direct axis and
directly opposes the field flux. In other words, the armature mmf has a demagne-
tizing effect on the field flux. When W assumes a value of 0 degrees (as nearly
happens for a unity power-factor load), then clearly the armature wave lies en-
tirely in the cross axis and its effect is to cause a distortion of the field flux. That
is, the armature mmf wave now has a cross-magnetizing effect on the field flux.
When the armature current leads the excitation voltage (as happens with leading
power-factor loads) by, say, 90°, the effect of the armature mmf is to produce
magnetization of the field flux. Because of the nature of industrial and domestic
loads, the armature current usually lags behind the excitation voltage so that the
situation depicted in Fig. 5-14 is the customary one. This means that generally the
armature mmf wave is so located that it produces simultaneously an effect on both
the direct and the cross axes.
In view of the fact that the direct axis is influenced by iron saturation as mmf
is added or subtracted in this axis whereas iron saturation is of little consequence
in the cross axis, it is imperative that for any specified / the armature mmf wave
be composed of two components—one that acts in the direct axis and the other
that acts in the quadrature axis. The amplitude of the direct-axis component is
denoted by A,N; in Fig. 5-14. A study of Fig. 5-14(b) reveals that

A,=Asinw <A (5-35)

where A, is the direct-axis armature mmf expressed in equivalent field amperes.
Similarly, the component of the quadrature-axis thmf can be expressed as

A,=Avcos A (5-36)

Appearing in Fig. 5-15 are the respective flux-density wave forms produced
by the direct-axis and quadrature-axis armature mmf components. The direct-
axis mmf produces a quasi-truncated sinusoid, whereas the quadrature-axis mmf
produces a saddle-shape flux-density wave because of the high reluctance of the
interpolar space. The effects are indeed quite different and accordingly must be
treated separately.
The key to the reactance method of treating the salient-pole synchronous
machine lies in Eqs. (5-35) and (5-36). In accordance with these equations the
 248 Three-Phase Synchronous Generators Chap. 5

Quadrature axis, mmf Direct-axis, mmf

Quad. axis flux Direct-axis flux

density Bg v4 density By

} Figure 5-15 Flux-density waveforms

N S I produced by the direct- and quadrature-
¢ axis mmf’s.

armature current can be considered composed of a direct-axis and a quadrature-

axis current. That is,
I, =1, sin (5-37)
I, = Iq cos (5-38)
Hence, if we assume for the moment that E;, [,, and are known, then J, and J,,
as well as their associated mmfs A, and A,, will be represented as depicted in Fig.
5-16. Note that the direct axis is that along which the field mmf F acts. By
superposition each mmf is then considered to. produce its own induced emf in the
armature winding. Thus by following the same procedure outlined in the synchro-
nous reactance method for the cylindrical rotor machine, through superposition
we can associate with A, an emf which is replaced by the reactance drop in the
direct axis, i.e., j[yXaqa. The j term provides the necessary 90-degree lead that
must exist between a current and its related reactance drop. Here X,q is the
armature reaction reactance in the direct axis. Ina similar fashion the effect of A,
is replaced by a reactance drop +j/,Xaq leading A, (or I,) by 90°. The quantity xa,
F

Figure 5-16 Illustrating the composi-

tion of the armature current and its
associated mmf’s into direct- and
quadrature-axis components.
Sec. 5-6 Theory for Salient-Pole Machines 249

Figure 5-17 Development of a phasor diagram for the salient-pole synchronous

machine, by the armature-reaction reactances.

is called the armature reaction reactance in the quadrature axis. A single arma-
ture reaction cannot be used here because of the wide difference in the flux-
density curves associated with the two axes. The value of xq, is always less than
Xaa since the emf induced by a given mmf acting on each axis is always smaller for
the quadrature axis owing to its higher reluctance.
We next describe the development of the phasor diagram for the salient-pole
synchronous machine, assuming the following data to be available: rated terminal
voltage per phase, V;; rated armature current per phase, J,; load power-factor
angle, @; armature winding leakage reactance per phase, x7; and the armature
reaction reactance in each axis expressed per phase, x,4 and x,,. For simplicity
the armature winding resistance is assumed negligible. Refer to Fig. 5-17. Add
jI,x; to V, to get the induced emf in the armature winding caused by the resultant
gap flux. This is quantity Ob. At b mark off the quantity + jlgXaq leading I, by 90°
and terminating at point d. The excitation voltage E, lies along line Od. Finally,
the magnitude of Eis found by adding the armature reaction reactance drop in the
direct axis, i.e., jlaXaq, along line Od starting at point d. The quantity Of mea-
sured on a suitable voltage scale yields the value of the excitation voltage.
Generally, unless the design data are furnished, the quantities x,, and Xqq are
not available. Even an appropriate no-load test does not provide information
com-
about the armature reactances directly. Rather, it is information about the
of the leakage reactance with the armature reaction reactances that
bined effects
dia-
becomes directly available. In this connection then let us study the phasor
gram of Fig. 5-17 in more detail. First note that
x chd =
250 Three-Phase Synchronous Generators Chap. 5

It then follows that

ae bd _ Wgkaq _ Ia COS W Xag _ Pe (5-39)

cosw cos w cos

In addition,
ac = ab + bc = I,(x1 + Xaq) (5-40)

Therefore we can define

where x, is the quadrature-axis synchronous reactance. Note the correspon-

dence with Eq. (5-16) for the cylindrical rotor machine. Equation (5-40) furnishes
an alternative way of locating the direction along which the E,phasor lies. It merely
requires that the quantity j/,x, be added to V, to locate point c in Fig. 5-17.
By proceeding in a similar fashion a synchronous reactance associated with
the direct axis can be defined. From the geometry of Fig. 5-17 we can write that
df os TgXad ve Iq sin w Xad
Conn isin vw sinw sine = LaXad
Also, .
ae = ab + be = I,(x, + Xaa)

Therefore

where xq denotes the direct-axis synchronous reactance. This quantity is compa-

rable to x, for the smooth-rotor machine. Fairly accurate values for x, and xg can
be obtained from a no-load test called the slip test, which is described in Appendix
F. Our discussions hereinafter proceed on the assumption that information about
xq and xq is available either from the design data or from an appropriate no-load
test such as the slip test.
The determination of Ey (and thence the corresponding field current F’) by
the use of x, and x, has caused the method to be called the two-reactance
method. Depicted in Fig. 5-18 is the phasor diagram expressed solely in terms of
synchronous reactance drops. The solution procedure then involves just three
steps as outlined next:

1. Locate the direction of Ey from

eh sin 0 + Ix,
w = tan
V, cos 0 (5-43)
Sec. 5-6 Theory for Salient-Pole Machines 251

Figure 5-18 Phasor diagram of the salient-pole machine in terms of the synchro-
nous reactances in the direct and quadrature axes.

2. Compute the phasor expressions for direct and quadrature currents.

I, = UI, cos y)/6 (5-44)
Iq = (I, sin y)/5 — 90° (5-45)

where
b= -—0 (5-46)

and the angles are all positive for lagging power factor.
3. Finally, as can be seen from Fig. 5-18,
Ey — V, aF shbes an jlaxa (5-47)

The field current that produces this excitation voltage is then found from the
open-circuit characteristic.

The same limitations that apply to the synchronous reactance method apply
to the two-reactance method. To obtain reliable results the direct-axis synchro-
the
nous reactance must be found corresponding to the proper flux conditions in
such problem because it is associate d with
machine. Of course, x, presents no
whenever
large air paths where saturation is of little or no consequence. Surely,
of dealing with the effect of armature reaction
there is a choice, the preferred way
with the open-circ uit characteri stic. This is
in the direct axis is to work directly
method for salient-po le machines which is
possible by employing the general
described next.
252 Three-Phase Synchronous Generators Chap. 5

General Method for Salient-Pole Machines

The phasor diagram to be used is the one depicted in Fig. 5-17. The procedure is
as follows:

1. Find w by means of Eq. (5-43).

2. Compute the emf induced by the resultant flux in the direct axis. This is the
quantity Od and shall now be called E,. Its magnitude is given by
E, = V,; cos (w —-0) + I,x; sin w (5-48)

3. Corresponding to Ez on the open-circuit characteristic, find the resultant

mmf expressed in field amperes. Call this Ry.
4. Find A by means of Eq. (5-34) and compute
Av=Asinw
5. Then for w lagging the required field current is found from the algebraic sum
of steps 3 and 4. Thus
F=R,+Asinw (5-49)

It is interesting to note that by computing E, in the manner described by Eq.

(5-48), there is no need to have knowledge of.x,,. Of course the magnitude of Ez is
the same as that which would be obtained from the expression
Eg = Vi > jl te + jlghag (5-50)
Finally, note that steps 4 and 5 of the foregoing procedure are the mmf equivalent
of adding the reactance drop Ijxaq to Ey to obtain the excitation voltage. The
advantage of the mmf version is that one automatically accounts for the correct
level of saturation by working directly with the nonlinear open-circuit characteris-
tic. This method yields highly realistic results.

Example 5-4
A three-phase, 15-kVA, 220-V, Y-connected, 60-Hz, six-pole salient-pole synchro-
nous generator has the open-circuit, short-circuit, and zero power factor characteris-
tics depicted in Fig. PS-10. The armature winding resistance is negligible. It is also
known that for this type machine the synchronous reactance in the quadrature axis is
60% of the value associated with the direct axis. Moreover, the leakage reactance is
known to have a value of 0.9 ( per phase.
(a) Identify the base quantities to be used in calculating the performance em-
ploying the per-unit notation.
(b) Determine the per-unit values of the leakage reactance and the direct-axis
synchronous reactance per phase as derived from the graphical data.
(c) Determine the excitation emf when this machine delivers rated kVA to a 0.8
pf lagging load using the two reactance method. Make all calculations in terms of
per-unit quantities.
Sec. 5-6 Theory for Salient-Pole Machines 253

(d) Find the excitation emf for the conditions of part (c) using the general
method for the salient-pole machine.

Solution (a) It is customary to use the rated phase voltage as the base voltage and
the rated phase current as the base current. Thus in this case

V, 220
Voese = Vonase = Vi tay 127 V/phase

15,000
Tiegce = 1 hace = =
— = 394 A/phase
Se 1951( 320) .
Correspondingly, the base impedance becomes

Vise 27s
Diese = ie Oe 3.22 (/phase

A useful, alternative form for the calculation of the base impedance can be obtained
by writing

V phase ep Ne Von
Loaseie cat
pe VAIV3 Vi VA

where V, is the line-to-line voltage (irrespective of whether or not the stator winding
is Y- or A-connected) and VA denotes the volt-ampere rating of the generator.
Observe that with this formulation the V3 is not a factor in the evaluation. Applied
to the case at hand, we get

Zoase = tenRAX = 3:23 0/phase

(b) The per-unit value of the leakage reactance per phase readily follows from
the definition of a per-unit value. Thus
x(Q) 0.9 _

A convenient way to find the per-unit value of the direct-axis synchronous

reactance when graphical data is available is to find the short-circuit ratio (SCR). By
definition the SCR is the armature current, expressed in per unit, which flows in a
short-circuit test corresponding to that value of field excitation that produces rated
voltage in an open-circuit test. Algebraically stated, we have
SCR = p-u J,.

where the right side is the short-circuit armature current caused by that field current
that on open circuit yields rated voltage. Applying this definition to the graphical
data, we see that to produce a rated voltage of 127 V on the open-circuit characteris-
tic a field current of 4.15 A is needed. However, this same field current causes a
short-circuit current of 41 A, which expressed as a per-unit quantity is p-u J,, = 41/
39.4 = 1.041 = SCR. Consequently, the per-unit value of xz becomes simply
p-u Vr ay 1 1 1
p-u xq = Dali
tee GeDeVidsc SC 1.041
254 Three-Phase Synchronous Generators Chap. 5

at their
(c) Delivering rated kVA means that both the voltage and current are
rated values which on a per-unit basis means values of unity. The solution procedure
factor
for this part is outlined by Eqs. (5-43) through (5-47). For a 0.8 pf the power
angle is
6 = cos! 0.8 = 36.9°
Also, the p-u value of the synchronous reactance in the quadrature axis is
p-u Xq = (0.6)(p-u xq) = 0.577

Hence
weston (1) Se Bey ARS 0.6 ~een ~ 55 8°

/8
p-u I, = (p-u I, cos p) /8 = cos 55.8°/6 = 0.562
where
8 = b — 0= 55.8° — 36.9° = 18.9°
p-u I, = (p-u J, sin p)/—90° + 6 = 0.827/-71.1°
Therefore,

p-u Ey = V, + jlyxq + jluxa = 1/0° + (0.562) /18.9°0.577) /90° + (0.827) /=71.1°(0.961) /90°
= 1/0° + (0.562)/18.9°(0.577) /90° + (0.827) /=71.1°(0.961) /90°
= 1 /0° — 0.105 + j0.307 + 0.752 + j0.257 = 1.647 + j0.564
= 1.74/18.9°
In accordance with this result it follows that the voltage regulation of this machine is
ahigh 74%. This value is highly pessimistic, which we already know is characteristic
of the synchronous reactance method. A more realistic result can be obtained by
using the general method, which is done in part (d).
(d) The general method also begins with the evaluation of p. The emf induced
by the resultant flux in the air gap of the direct axis is then found with Eq. (5-48).
Thus

p-u Ey = V, cos (Ww — 6) + I,x; sin

= (1) cos (18.9°) + (1)(0.28) sin 18.9° = 0.946 + 0.091 = 1.037

The corresponding voltage value is

Eq = 1.037(127) = 131.7 V

and the resultant field current, Rz, associated with this voltage as obtained from the
open-circuit characteristic is found to be

Ry =4.6A
Also, by construction of the Potier triangle the armature reaction mmf expressed in
terms of equivalent field amperes is found to be A = 2.96 A. Accordingly, the
 Sec. 5-7 Power 255

resultant field current that produces the excitation emf is determined by Eq. (5-49) to
be
F=R,+ Asin w = 4.6 + 2.96 sin 55.8° = 7.05 A

Entering the open-circuit characteristic for this value of field current yields an excita-
tion voltage of
Ey = 166 V/phase
The per-unit value is
166
p-u Ef ae 7 ihe3ilk

Thus a much more realistic value of 31% voltage regulation is now obtained.

5-7 POWER

The expression for power in an electromechanical energy-conversion device is

always a matter of primary importance. It was shown in Chapter 3 that, once the
developed torque is computed, the developed power can be found from
Pm = Tom (5-51)

where w,, denotes the mechanical speed of rotation expressed in radians per
second. However, it is frequently useful to have an equation for power that is
expressed in terms of appropriate machine parameters, voltages, and phase angles
because of the insight it gives regarding machine performance. Such an equation
can be derived from the phasor diagram for the cylindrical and also the salient-
pole machine.

Cylindrical-Rotor Machine

Fig. 5-19.
Use is made of the phasor diagram of Fig. 5-7(b), which is repeated in
e is very small, it is neglecte d here.
Since the effect of armature winding resistanc
d power can be written in one of
Consequently, the expression for the develope
two ways, 1.e.,
Pm = QVila cos 0 = GEfl, cos (5-52)
of E; on the
This statement follows from Fig. 5-19 upon noting that the projection
Moreove r, from the geometry
[, phasor is identical with the projection of V,onJ,.
be expresse d in two ways. Thus
of Fig. 5-19 we see that the quantity ab may also
ab = Ey sin 6 = I,x5 cos 6 (5-53)

from which it follows that

L, cos 0 = i sin 8 (5-54)

s
256 Three-Phase Synchronous Generators Chap. 5

Figure 5-19 Phasor diagram of the

cylindrical-rotor machine for lagging
power-factor.

Insertion of Eq. (5-54) into Eq. (5-52) yields the desired result. Hence the power
developed in the synchronous generator where armature winding resistance is
negligible is given by

(5-55)

where q denotes the number of phases.

Because of the dependence of the power upon the angle 6, this angle has
come to be called the power angle. Equation (5-55) indicates that, if the power
angle is zero, the synchronous machine cannot develop useful power. It further
reveals that the developed power is a sinusoidal function of the power angle and
that it has its maximum value at 6 equal to 90°. Figure 5-20 depicts the variation
P,, as a function of positive and negative values of 6. As described in Chapter 6,
negative values of power angle refer to the operation of the synchronous machine

Fn

|
Generator

90°

6, power angle,in
electrical degrees
Figure 5-20 Graphical representation
of developed power as a function of the
power angle for a cylindrical-rotor
synchronous machine.
Sec. 5-7 Power 257

as a motor. It is also worthwhile to note the correspondence of Eq. (5-55) with

Eq. (3-22) and Eq. (3-97) of Appendix to Chapter 3. Both the developed electro-
magnetic torque and the developed power are sinusoidal functions of an appropri-
ate power angle.

Salient-Pole Machine

The derivation of the expression for the developed power of the salient-pole
machine proceeds in a similar fashion. However, in this case use is made of the
phasor diagram based on the two-reactance theory. The diagram is repeated for
convenience in Fig. 5-21. Again, the phasor diagram as it applies to the synchro-
nous generator (i.e., positive power angle) is used, but it should be understood
that the results apply equally for operation as a synchronous motor.

Figure 5-21 Synchronous-generator

phasor diagram based on the two-
reactance theory.

Because of the assumption of negligible armature resistance, the power out-

put and the power developed are equal. Accordingly we may write
Pm = GVilq COs 0 (5-56)

is in
A study of Fig. 5-21 discloses that the component of armature current that
phase with the terminal voltage is related to J, and I, by
I, cos 9 = I, cos 6 + Iy sin 6 (5-57)
geometry of
Moreover, an expression for J, can be found by noting that by the
Fig. 5-21
IjXq = V; sin 6 (5-58)

power angle 6 of Eq.

+Take care not to confuse the torque angle A of Eq. (3-97) with the
of the flux-density wave produced by
(5-55). In the former, A is defined as the angle between the axis
90° displaced from the armature
the field winding and the axis of the armature winding mmf # (which is
ampere-conductor distribution). In the latter case, 6 is defined as the time angle between the excitation
by the resultant flux minus the
voltage (produced by the field flux) and the terminal voltage (produced
leakage flux). See Fig. 6-18 for further clarificati on.
258 Three-Phase Synchronous Generators Chap. 5

from which

I, = ba sin 6 (5-59)
Xq

Ina similar fashion an expression for J, can be found expressed in terms of voltage
and the machine parameters. Thus,

Er = V, cos 6 + LaXa (5-60)

Therefore,

Bs
Is eee 2 (5-61)
Xd

Upon substituting Eqs. (5-59) and (5-61) in Eq. (5-57), there results
: V; .
I cos 6 =! siPeiecso a sin 6 si cans (5-62)
Xq Xd Xd

The expression for the developed power then becomes

2
Pm =4q ML sin 6 + q at (xq — Xq) sin 6 cos 6 (5-63)
Xd XaXq

Introducing the trigonometric identity

sin 26 = 2 sin 6 cos 6 (5-64)

into the last equation yields the desired result. Hence,

2
sin 5 + q a (xq — Xq) sin 26 (5-65)
VE,

Arata Xd S68

A plot of this equation appears in Fig. 5-22 for motor as well as generator
operation. Curve (a) is a plot of the first term of Eq. (5-65) and is identical with the
result obtained for the cylindrical-rotor machine. It represents the power associ-
ated with the electromagnetic torque produced by the mutual coupling between
the rotor field winding (indicated by E;) and the stator armature winding (indicated
by V,). Curve (b) is a graphical representation of the second term of Eq. (5-65). It
is independent of excitation voltage which discloses that this term exists even
when the field current is zero. Of course it is assumed that the machine is con-
nected to an infinite bus system, otherwise the second term would be meaning-
less. When connected to the infinite bus, an armature current can flow without
field excitation. Accordingly, a revolving mmf exists and it attempts to align itself
with the minimum reluctance path as explained in Chapter 3. Since it arises
because of the difference in reluctance between the direct and quadrature axes, it
is called the reluctance power and has a value usually in the vicinity of 20% to 25%
Sec. 5-7 Power 259

30

Generator

(a)

180
6, power angle,
degrees

Figure 5-22 Developed power as a function of the power angle for the synchro-
nous salient-pole machine.

of the rating of the machine. Moreover, it involves double the power angle and
thereby causes a peaking of the resultant developed power curve as indicated by
curve (c) in Fig. 5-22. The effect of the reluctance power term in the expression
for developed power is not only to increase the maximum value of developed
power but also to cause it to occur at a value of 6 less than 90°. Finally, it is
interesting to note that, when Eq. (5-65) is applied to the cylindrical-rotor ma-
chine, the expression reduces to that of Eq. (5-55) because x, and x, are equal.
Example 5-5
(a) Calculate the electrical power delivered to the specified load by the three-phase
turbogenerator of Example 5-1.
(b) Using the results found in the solution to Example 5-2, find the value of the
mechanical power developed by the turbogenerator assuming negligible armature
winding losses.
(c) Repeat part (b) for Example 5-3.

Solution (a) The power output using phase quantities is given by

P, = 3V1, cos 6 = 3(7.9674)103(0.4184)103(0.8) = 8 x 10° W = 8 MW

(b) Invoking Eq. (5-55) yields

Pp, =3 Ms reef sone 736TH INT 381 Pe 28.6°

= 7.993 x 10° W = 7.993 MW

 260 Three-Phase Synchronous Generators Chap. 5

(c) Again by Eq. (5-55) the result is

pas V Er cin of 3 U: 9674) 103(13.586) 10° : /25.15 O = 8 x 10° Dine

W = 8 MW
e 17.25 sin

It is instructive to observe that the result for power is the same whether the calcula-
tion is based on the results of the excitation emf and the associated power angle
found using the unsaturated or the saturated synchronous reactances.

5-8 PARALLEL OPERATION OF SYNCHRONOUS GENERATORS

The electrical needs of industry, commercial establishments, and individual con-

sumers are supplied almost exclusively by synchronous generators. In many
large regions of the country, there are virtually hundreds of such generators
operating in parallel to meet these needs. It is the purpose of this section to
describe some of the principles that govern theRastedy: state behavior of alterna-
tors that operate in parallel.

The Need to Synchronize

Initially we focus on the parallel operation of two isolated synchronous generators

of equal ratings. Assume that the first generator G; is already connected to a
three-phase resistive load as indicated in Fig. 5-23 by the closed load switch and
that it is operating at rated frequency and rated voltage. It is now desirable to
place the second alternator in parallel with the first in order that it might help share
the current load as well as to prepare for the future increased power requirements
of this load.
If severe shocks to both generators are to be avoided, a definite procedure,
called the synchronizing procedure, must be followed before the switch (or circuit
breaker) that parallels the two machines can be closed. The process is started by
adjusting the speed of the prime mover of generator G) to correspond exactly to

Load
switch

Figure 5-23 Synchronizing two synchronous generators (alternators) for parallel

operation.
Sec. 5-8 Parallel Operation of Synchronous Generators 261

that of generator G,. The field current of G, is then adjusted to yield the same
nominal voltage between lines a’, b', and cc’. A suitable bank of lampst is placed
in series with lines a'-a, b'-b, c'-c, as shown in Fig. 5-23. These lamps will light
up if either the voltage or the frequency of G, differs slightly from those of G.
When the two frequencies are identical but the magnitudes of the line voltages are
slightly different, the lamps will be on steadily. A subsequent adjustment of the
field current of G, can then be used to cause the lamps to darken, at which time the
synchronizing switch can be closed, thereby placing both machines in parallel
without incident. On the other hand, if the two sets of three-phase voltages are
equal in magnitude but a bit different in frequency, the lamps will again register
light but now they will flicker at a rate equal to the difference frequency. At one
point in the difference cycle, the voltage across the lamps will reach twice the line
voltage; a half-cycle later it will be zero. An adjustment in the prime mover speed
can serve to reduce this difference frequency to a very small value. Then, as a
dark period is approached, the line switch can be closed with little or no distur-
bance.

Effect of Field Excitation (Control of Reactive AC Power)

Once G3 is placed in parallel with G,, in the manner described, and no further
changes are made, the second machine assumes a state of simply idling on the
line. If the assumed fixed load demand is to be equally shared by G, and Gy, it is
necessary to make appropriate adjustments on both prime movers in a fashion to
be described presently. For the moment, however, let us assume that each gener-
ator is delivering half the load power at rated frequency and rated load. Now we
investigate the effect of varying the field excitation of each machine that is operat-
ing in parallel. This initial state of affairs is depicted in F ig. 5-24, corresponding to
the unprimed quantities. This diagram is drawn for the case of negligible armature
(stator) winding resistance. Moreover, the excitation current is initially assumed
to be set at that value which yields unity power factor for each machine. This is
illustrated by placing J, and J,, the respective G; and G stator currents, in phase
with the voltage phasor. It is instructive to note here that synchronous generators
operating in parallel have a unique field excitation that yields unity power factor at
a fixed load share.
To better understand the sequence of events that unfolds with variation of
excitation, we begin by first increasing the field current of G,. In turn, this causes
an increase in the field excitation phasor Ey; to a larger value E;,. The new
position of phasor E;, in Fig. 5-24 must be such that its tip lies on the horizontal
locus line to ensure that the vertical distance between this line and the line along
which V lies remains a constant. This condition is demanded by Eq. (5-55), which
for constant power requires that Ey; sin 6; = E;, sin 5;. In addition, the operating

pe.
+The equivalent commercial version of these lamps is called a synchrosco
262 Three-Phase Synchronous Generators Chap. 5

Ete Eg=E¢e En Locus

jlyx, = jl2x,

Figure 5-24 Phasor diagram drawn to illustrate the effect of field excitation on
two synchronous generators operating in parallel. Real power remains fixed.

power angle of G; must assume a reduced value in the presence of the augmented
Ep.
What happens to this power system if no subsequent adjustment is made on
the second generator? The phasor diagram for the new condition in G; immedi-
ately reveals that G; can no longer operate at unity power factor. A reaction is,
therefore, set in motion to reduce the effect of the increased excitation. In a
generator overexcitation calls for a lagging reactive current component because it
produces a demagnetizing effect. But because the load is assumed to be resistive,
this generator lagging current component cannot be accommodated by the load.
Therefore, it must exist as a circulating current between the two stator windings
and as such must necessarily take on the character of a leading current in Gp.
Since a leading current in a generator produces a magnetizing effect on the air-gap
flux, it follows that the net effect of increasing only one field current is to cause the
terminal voltage of the parallel set to increase to V’ located midway between the
horizontal projections of Ey, and E;/; onto the V line.
On the basis of the foregoing analysis, it should be apparent that to maintain
operation at the original value of terminal voltage (after the field current of G; is
increased) it becomes necessary to decrease the field current of G» by a corre-
sponding amount. The required condition is depicted in Fig. 5-24 by the prime
notation on the quantities of G,. Here again note that the phasor voltage differ-
ences between the new excitation voltages and the terminal voltage call for a
reactive current flow—lagging in the overexcited machine G, and leading in the
underexcited machine G,. Accordingly, we are led to the conclusion that the
effect of changing the field excitation of generators operating in parallel is merely
to change the amount of reactive volt-amperes associated with each machine.
There is no change in the amount of real power delivered.
Sec. 5-8 Parallel Operation of Synchronous Generators 263

Effect of Prime Mover Power (Control of Real Power)

Attention is next directed to a study of altering the power delivered to the shaft of
each of two synchronous generators operating in parallel with fixed field excita-
tions. The starting point is again taken to be that corresponding to which each
generator delivers the same electrical power to the load at that excitation that puts
the power factor of each machine at unity. It is important to understand at this
juncture that the electrical energy developed by each generator is mechanically
derived from the individual prime movers. These prime movers may be steam
turbines, gasoline engines, waterwheels, or any practical primary energy source.
The speed-power characteristics of these prime movers are decisive in establish-
ing the manner in which generators operating in parallel will share a common
load. Moreover, these characteristics assume a drooping bias in the fashion illus-
trated in Fig. 5-25. Because G; and G; are assumed to be identical machines, their
speed-torque characteristics are also assumed to be the same. They are depicted
as mirror images in Fig. 5-25 for reasons of clarity. At the initial operating point,
the prime movers are revolving at a speed that yields the frequency required at the
electrical load. At this frequency the prime mover of G, delivers power P; while
the prime mover of G, delivers power P; = P}.
Suppose now that by suitable adjustment of the prime mover of G,; the power
delivered to G; is reduced. Simultaneously, in the interest of maintaining a con-
stant frequency, let the input power to G, from its prime mover be increased in
order to continue to operate at constant frequency. These prime mover adjust-
ments are represented in Fig. 5-25 by a displacement of the speed-power charac-
teristics as illustrated. Observe that to reduce the share of load supplied by gener-
ator G, its speed-power characteristic is displaced downward from position G, to
Gj. Similarly, the characteristic of the prime mover of machine two is raised in
reference to its initial position. The result is that, at the maintained frequency, G,
is now assuming a smaller share of the load P| < P;, while G, takes on a greater
share P} > P,. The sum of the primed and unprimed powers is constant, how-
ever.
Freguency

Power (PM,)
Power (PM2)

of generators G,
Figure 5-25 Speed-power characteristics of the prime movers
and G, of Fig. 5-23.
264 Three-Phase Synchronous Generators Chap. 5

Figure 5-26 Phasor diagrams illustrating the effect of changing the prime mover
power at constant excitation for two generators operating in parallel.

The effects of these prime mover adjustments on the electrical quantities of

the individual generators are shown in Fig. 5-26. The reduced share of the electri-
cal load supplied by G, is accompanied by a decrease in its power angle from 6; to
5{ as demanded by Eq. (5-55). The phasor voltage difference between Ep; and V is
correspondingly diminished and this in turn produces a smaller stator current Ii
for G;. Two points are worth noting about this new value of G; current. First, it
has a smaller in-phase component, which is consistent with the fact that it is
delivering a smaller share of the electrical power taken by the load. Second, the
power factor of this current is no longer unity. In fact, it is a lagging power
factor. Corresponding changes take place in generator G). Here the power angle
increases from 8 to 64, thus producing a larger phasor voltage difference which
acts to establish an increased value of stator current in G) as represented by J}.
This current exhibits a greater in-phase component because it delivers a greater
load share. In addition, it too now possesses a reactive current which is a leading
component. Since the load is resistive, the reactive current of both generators
constitute a common circulating current. Hence, if it is of a lagging nature in one
generator, it must be leading in the other. Once again, we have an illustration of
the fact that at each load condition there is a unique excitation for operation at
unity power factor.

Connection to an Infinite Bus

For the sake of argument, let G, in Fig. 5-23 represent many large synchronous
generators connected to a common power grid. This interconnection of genera-
tors frequently involves the use of solid copper bars and has come to be called a
bus. Appearing at such a bus are characteristics of the electric power system that
exhibit virtually constant voltage and frequency. It has come to be called an
infinite bus because no single generator on the interconnected system can by itself
Chap. 5 Problems 265

in normal circumstances influence either of these quantities to any noticeable

extent.
The synchronizing procedure needed to connect G; to the infinite bus is the
same as for the interconnection of two isolated generators as described previ-
ously. Assume that this has been done and that G; is now idling on the power grid
line. Any increase in the excitation of G) produces a lagging current with a power
factor very close to zero. This reactive current serves to neutralize the effect of
the increased field mmf in order that the terminal voltage stay invariant as de-
manded by the infinite bus. Actually, this reactive generator current serves to
relieve the other generators on the power grid of some of their reactive volt-
ampere burden. Correspondingly, a decrease in excitation causes G, to draw a
leading current from the bus system. This current serves to produce a magnetiz-
ing effect of sufficient quantity to ensure an air-gap flux that is consistent with the
bus voltage. A return of the excitation to a value of 100% (i.e., Ey = V), puts the
machine once again in the idling state.
Let the prime mover of G, now be adjusted to raise its speed-power charac-
teristic. This has the effect of increasing the input power to G), which in turn
responds by delivering a corresponding amount to the power grid. As the speed of
the prime mover of G, is momentarily increased to raise its speed-power charac-
teristic, it does make an effort to raise the frequency of the entire power system.
Of course, it fails to do so because it is no match for the enormous total capacity
already on the grid to which G) would need to supply synchronizing power in
order to bring about a frequency change. Instead, G2 remains content with simply
relieving the system of a tiny fraction of the delivered electrical power.

PROBLEMS

5-1. A synchronous generator has a short-circuit characteristic such that rated armature
current is obtained by 0.75 per-unit excitation. Rated voltage on the air-gap line is
obtained by 1.25 per-unit excitation. What per-unit excitation is required to supply a
load at rated voltage, rated kVA, and 80% pf lagging?
5-2. Repeat Prob. 5-1 for the case where the load power factor is 80% leading.
5-3. A synchronous generator has the following rating:
180-kVA, 440-V, 300-rpm, 60-Hz, three-phase, Y-connection,

r,~0, and x; = 0.296 0

The
The air-gap line is described by the equation E = 17I;, expressed per phase.
short-circuit characteristic is described by /,. = 10.75J. Compute the field current
pf of 0.8
needed to provide rated voltage when rated current is delivered to a load at a
leading. Use the general method linear analysis. Assume that J, ~ 17.9 equivalent
field amperes.
5-4. Repeat Prob. 5-3 for the case where the pf of the load is 0.8 lagging.
266 Three-Phase Synchronous Generators Chap. 5

5-5. When the machine of Prob. 5-3 delivers rated armature current to a unity power
factor load, the field winding current is measured to be 30 A. Moreover, the arma-
ture current is found to be lagging the excitation voltage by 50 electrical degrees.
(a) Compute the resultant voltage induced in the armature winding by the air-gap
flux.
(b) What is the value of the terminal voltage per phase?
(c) Find the power delivered to the load.
5-6. Repeat Prob. 5-5 for the case where the pf of the load is 0.8 lagging.
5-7. A 5-kVA, 220-V, 60-Hz, six-pole, Y-connected synchronous generator has a three-
phase winding with 135 effective turns per phase. Moreover, its field winding is
equipped with 200 turns. The leakage reactance per phase is 0.78 1 and the armature
winding resistance is negligible.
(a) Compute the field current required to establish rated voltage across the terminals
of a unity power factor load that draws rated generator armature current.
(b) Determine the field current needed to provide rated terminal voltage to a load
that draws 125% of rated current at 0.8 lagging pf.
5-8. A 5-kVA, 220-V, six-pole, 60-Hz, Y-connected synchronous generator has a three-
phase winding with 135 effective turns per phase and an armature winding leakage
reactance of 0.8 © per phase. The equation for the air-gap line is E = 25I;, ex-
pressed in induced volts per phase. The generator has 200 field turns. When this
machine delivers rated armature current to its connected load, like poles of the
armature mmf are known to lag behind its field poles by 150 electrical degrees.
Assume negligible saturation. :
(a) Find the value of the terminal voltage per phase.
(b) What is the pf of the load?
(c) How much power does the load take?
5-9. The machine of Prob. 5-3 has a cylindrical rotor.
(a) Find the value of the unsaturated synchronous reactance.
(b) What is the value of the armature reaction reactance?
(c) By means of the synchronous reactance method find the field current needed to
yield rated terminal voltage at rated current for a 0.8 lagging pf load.
(d) Prepare a carefully drawn phasor diagram.
5-10. A three-phase, 15-kVA, 220-V, Y-connected, six-pole synchronous generator has a
cylindrical rotor and negligible armature resistance. The open-circuit, short-circuit,
and full-load zero pf characteristics are shown in Fig. P5-10.
(a) Determine the armature leakage reactance per phase.
(b) Compute the voltage regulation at unity pf by the general method.
(c) Find the voltage regulation at pf = | by the use of the saturated synchronous
reactance.
(d) Compare the results of parts (b) and (c) and comment.
5-11. The number of field turns per pole for the machine of Prob. 5-10 is 400. Determine
the number of effective armature turns per pole.
5-12. At no-load, the field current of the machine of Prob. 5-10 is adjusted to 5.8 A and kept
at this value. A resistive load is then applied to the generator until rated current
flows through the armature winding. At this condition the angle between the no-load
excitation voltage E, and the, full-load terminal voltage is measured to be 46 electrical
degrees. Find the value of the terminal voltage per phase by the general method.
Chap. 5 Problems 267

200
Fd a
0 SA
eee eT
Side eeSala)7a Fall al
ia S-C characteristic
80

scale,
Current
amperes
volts
phase,
Voltage
per

60

Field amperes
Figure P5-10

rating:
5-13. A cylindrical pole synchronous generator has the following
200-kVA, 440-V, 300-rpm, 60-Hz, three-phase, Y-connection, rz ~ 0
in Fig. P5-13.
The open-circuit and rated current zero pf characteristics are shown
when rated current is
Compute the field current needed to provide rated voltage
delivered to a load at a pf of 0.8 lagging. Use the general method.
268 Three-Phase Synchronous Generators Chap. 5

w 400
oe
g
ov
Ss
$ 350
ig
5
2
2
£
=I
300

250

200

150

100

50

Field amperes

Figure P5-13

5-14. A 40,000-kVA, 14,000-V, three-phase, Y-connected alternator has negligible arma-

ture resistance and a leakage reactance of 1 (1 per phase. Other pertinent data are as
follows:

Short-circuit characteristic: I, = Tlf

Air-gap line, volts per phase: E = 331;

Leet Sete 21,3001

Open-circuit characteristic volts per phase: E = B04

The equation for the open-circuit characteristic is not valid for values of J; about the
origin. For a constant terminal voltage of 14,000 V (line) compute the variation of
Chap. 5 Problems 269

field current from no-load to full-load kVA at 80% lagging pf. Use the general
method for nonlinear analysis.
5-15. A 2100-kVA, 6600-V, Y-connected, 25-Hz, three-phase alternator is operated at
synchronous speed with the armature short-circuited. The excitation is adjusted to
produce full-load current in the armature. Without changing the field current, the
armature circuit is opened and the terminal voltage is found to rise to 2900 V.
Armature resistance is negligible.
(a) Determine the voltage regulation for a 0.8 lagging pf by the synchronous reac-
tance method.
(b) Repeat for a pf of 0.8 leading.
5-16 A three-phase, Y-connected synchronous generator is rated at 200 kVA and 2300 V.
It is supplying rated current at rated terminal voltage to a 0.8 lagging pf load. When
the load is completely disconnected (without any change of field excitation), the
terminal voltage rises to 3800 V.
(a) Draw a phasor diagram, illustrating the foregoing load condition. Mark each
phasor by its symbol and its numerical value.
(b) Calculate the synchronous reactance of this generator. Assume r, negligible.
5-17. A 50-kVA, 550-V, 25-Hz, single-phase alternator has an effective resistance of
0.25 © and a synchronous reactance of 3.2 0. Determine the percent regulation for
rated load and 0.8 lagging pf.
5-18. A cylindrical-pole synchronous generator supplies rated kVA at a leading pf of 80%.
Its per-unit synchronous reactance is 1.0, and the armature resistance may be as-
sumed negligible. The base impedance for such a machine is defined as the ratio of
the rated phase voltage to rated phase current. Compute the per-unit value of the
excitation voltage and the associated power angle.
5-19. A synchronous generator operates at rated voltage and rated current supplying
power to a 0.866 leading pf load. The per-unit value of excitation voltage is 1.732.
The base voltage is the rated value of the phase voltage. Armature resistance is
negligible. Find the per-unit value of the synchronous reactance.
5-20. A 1800-rpm, 60-Hz, 50,000-kVA, 13,800-V, synchronous generator has the following
test data available:

Open-circuit test:
I; (A): 175 200 225 250 PAS
V (line to line) kV: 2 13.0 13.8 14.5 (Sot

Short-circuit test:
I, = 193 A, I, = 2090 A

Zero pf characteristic:
I; = 468 A, I, = 2090 A, V = 13.8 kV
kVA at
Compute the field current needed to maintain rated terminal voltage for rated
80% pf.
generator is
5-21. A 5-kVA, 220-V, Y-connected, three-phase, salient-pole synchronous
means of a slip test, the direct-axis
used to supply power to a unity pf load. By
to be 12 Q and the quadratur e-axis synchron ous
synchronous reactance is found
that rated current is delivered to the load at rated voltage and
reactance 79. Assume
that armature resistance is negligible .
 Three-Phase Synchronous Generators Chap. 5
270

(a) Compute the corresponding value of the excitation voltage.

(b) What is the value of the internal phase displacement angle yp?
(c) What is the value of the power angle?
(d) Draw a carefully labeled phasor diagram.
5-22. Repeat Prob. 5-21 for the case where the load pf is 0.8 lagging.
5-23. Repeat Prob. 5-21 for the case where the load pf is 0.8 leading.
5-24. Assume that the characteristics depicted in Fig. P5-10 apply to the machine of Prob.
5-21. Moreover, assume that rated kVA is delivered to a unity pf load at rated
voltage.
(a) By the general method for salient-pole machines compute the value of excitation
voltage Ey and compare with the result of Prob. 5-21(a).
(b) What is the field current?
(c) Draw a carefully labeled phasor diagram for this load condition.
5-25. Repeat Prob. 5-24 for the case where the load pf is 0.8 leading.
5-26. Repeat Prob. 5-24 for the case where the load pf is 0.8 lagging.
5-27. A six-pole, 60-Hz, Y-connected synchronous generator is rated at 15-kVA, 220-V,
three-phase. A slip test is performed at a fixed reduced voltage of 96 V (line to line).
The maximum and minimum readings of the ammeter used in the test are 20 A and
12 A, respectively. Compute the value of the excitation voltage when this machine
delivers rated current to a 0.8 lagging pf load. Neglect armature resistance.
5-28. What physical significance can be attached to the synchronous reactances Xq and x,
as they relate to the salient-pole synchronous machine? Which reactance is larger
and why?
5-29. A three-phase, four-pole, 60-Hz cylindrical-rotor synchronous machine is rated at
10,000 kW, 5000 V, and 0.8 pf lead. The synchronous reactance is 0.8 p-u. The
normal open-circuit field current is 200 A. Neglect saturation and armature resis-
tance. Assume a constant field excitation. Calculate each of the following when the
machine is taking rated power from the line at rated current and voltage: (a) reactive
volt-amperes, (b) field current, and (c) power angle.

Figure P5-32
Chap. 5 Problems 271

5-30. Refer to the schematic diagram of Fig. 5-23. Describe the reactions that occur during
the synchronizing procedure when the speeds of the two generators are identical and
the switch is closed at an instant when the voltages of the two generators are not
exactly equal.
5-31. Repeat Prob. 5-30 for the case where the voltages are exactly the same but the
frequencies of the two generators are slightly different.
5-32. Two generators of equal rating are operating in parallel supplying power to a lagging
pf load for a condition as depicted in Fig. P5-32. Describe the adjustments to be
made so that each generator is made to operate at the same power factor of the load.
Is there any advantage to the new operating condition? Explain.
5-33. For the situation depicted in Fig. PS-32, describe the adjustments that must be made
in order to permit both generators to operate at unity power factor when the load is
unequally shared. What happens to the power angles of the individual generators?
 6
Three-Phase
Synchronous Motors

motors in contrast to
The significant and distinguishing feature of synchronous
cal energy is supplied to
induction motors is that they are doubly excited. Electri
torque can be devel-
both the field and the armature windings. When this is done
speed the average
oped at only onet speed—the synchronous speed. At any other
speed at which the
torque is zero. The synchronous speed refers to that rotor
(or the armature
rotor flux field and the armature ampere-conductor distribution
te this point con-
flux field) are stationary with respect to each other. To illustra
It has a de
sider the conventional synchronous motor equipped ith two poles.
applied to the
voltage applied to the rotor winding and a three-phase ac voltage
stator winding which is usually Hz.60 The three-phase voltage produces in the
of 60 Hz. The de
stator an ampere-conductor distribution which revolves at arate
field which is
(or zero-frequency) currents in the rotor set up a two-pole flux
a situation in
stationary so long as the rotor is not turning. Accordingly, we have
ary rotor
which there exists a pair of revolving armature poles and a pair of station
a cycle
poles. A little thought should make it clear that for a period of half
flux and the
(or zo S) a positive torque is developed by the cooperation of the field
ution. Howeve r, in the next half-cy cle this
revolving ampere-conductor distrib
torque is zero. This is the reason that a
torque is reversed so that the average
synchronous motor per se has no starting torque.

of the excessive
+There is a second speed, too, but it is only of academic interest because
currents that flow.

272
 Sec. 6-1 Phasor Diagram and Equivalent Circuit 273

To develop a continuous torque, then, it is necessary not only to have a flux

field and an appropriately displaced ampere-conductor distribution, but the two
quantities must be stationary with respect to each other. In singly fed machines
such as the induction motor this comes about automatically. The relative speed
between the stator field and the rotor winding induces rotor currents at those slip
frequencies which yield a revolving rotor mmf. Superposing this upon the rotor
speed gives a revolving rotor mmf, which in turn is stationary relative to the stator
field. In the synchronous motor this condition cannot take place automatically
because of the separate excitation used for the field and armature windings. In
light of these comments it is reasonable to expect that, if an induction motor is
doubly energized, it too will behave as a synchronous motor. Refer to Sec. 4-8.
On the basis of the foregoing comments it follows that to produce a continu-
ous nonzero torque it is first necessary to bring the de excited rotor to synchro-
nous speed by means of an auxiliary device. Sometimes the auxiliary device takes
the form of a small dc motor mounted on the rotor shaft. Most often, however,
use is made of a squirrel-cage winding similar to that used in induction motors and
imbedded in the pole faces. By means of this winding (also called the amortisseur
winding) the synchronous motor is brought up to almost synchronous speed.
Then, if the field winding is energized at the right moment, a positive torque will
be developed for a sufficiently long period to allow the armature poles to pull the
rotor poles into synchronism.
A general expression may be written to identify that frequency of rotation
required of the rotor so that the stator and rotor fields are stationary with respect
to each other. Let
f; = frequency of the currents through the stator winding
fy = frequency of the currents through the rotor winding
f, = frequency of rotation of the rotor structure
Then the defining equation is

feral: (6-1)
In the conventional synchronous motor the frequency of the rotor current fr is
at
zero, since it is direct current. Hence for an f, of 60 Hz the rotor must revolve
. On the other hand, note that,
60 Hz in order for nonzero torque to be developed
if the rotor winding were energized with 20-Hz current in place of direct current,
then for an f, of 60 Hz the synchronous speed would be 40 Hz.

6-1 PHASOR DIAGRAM AND EQUIVALENT CIRCUIT

generator to motor
When the synchronous machine makes the transition from
action, there occurs a reversal of power flow. Thus, instead of current flowing out
the current flows into the
of the armature winding to a suitable electrical load,
 Three-Phase Synchronous Motors Chap. 6
274

armature winding originating from a suitable electrical source. Since this transi-
tion occurs with the direction of rotation and the direction of field current flow
unaltered from the generator case, it follows that the relationship between F and
its associated excitation voltage Eyremains invariant. An observer looking only at
this relationship certainly could not distinguish whether operation is as a motor or
generator. One way to make this distinction, however, is to determine whether
the armature current flows in a direction that is essentially opposed to that of the
excitation voltage. If it is, the machine is behaving as a motor.—In-this condition
the phasor diagram of the synchronous motor can then be shown to differ from
that of the generator in two respects. The first involves the voltage equation for
the stator circuit. In the generator, E; (the excitation voltage) played the role of a
source voltage and the terminal voltage V, was dependent upon the magnitude of
E; and the synchronous impedance drop. In the motor the roles are inter-
changed. Now V, is the source voltage applied to the synchronous motor arma-
ture winding and Eyis a reaction or counter emf which is internally generated. It is
assumed that the terminal voltage originates from an infinite bus system? and so
remains invariant. Applying Kirchhoff’s voltage law to the synchronous motor
for negligible armature resistance leads to

V,/0° = Ey + jloxs (6-2)

Clearly, this equation states that the applied stator voltage is equal to the sum of
the drops. The excitation voltage is treated as a reaction voltage drop in much the
same way as E, was treated in the case of the transformer.
The circuit interpretation of Eq. (6-2) leads to the equivalent circuit as it
applies to the synchronous motor. This appears in Fig. 6-1.
The second point of view in which the phasor diagram of the motor differs
from that of the generator involves the angle 5. Physically, for the generator, the
phase of F; is ahead of V, in time because of the driving action of the prime
mover. Keep in mind that Ey may be considered associated with the rotor field
axis. A similar line of reasoning pertains for the motor. At no-load, 6 is zero and
so the axis of a field pole and the resultant flux axis associated with the terminal
voltage are coincident. As shaft load is applied, however, the rotor falls slightly
behind its no-load position and in this way causes 6 to increase but in a sense
opposite to that for the generator.
On the basis of these modifications the phasor diagram of the synchronous
motor becomes that shown in Fig. 6-2. Note that E; now lags V, and that it is the
phasor sum of Ey and j/,x, which equals V, as called for by Eq. (6-2). Another
interesting point about this diagram is revealed upon comparison with Fig. 5-19.
In both cases the magnitude of the excitation voltage Ey exceeds that of the
terminal voltage V,. This is referred to as a condition of overexcitation (i.e.,

+A power system of tremendous capacity compared with the rating of the synchronous motor.
 Sec. 6-2 Performance Computation 275

Figure 6-1 The equivalent circuit of the Figure 6-2. Phasor diagram of the over-
synchronous motor appears to the right excited synchronous motor.
of terminals ab.

|E,| > |V,|). Note, however, that overexcitation in the synchronous generator is
accompanied by a current of lagging power factor, whereas in the synchronous
motor overexcitation is accompanied by a current of leading power factor. When
the magnitude of E; is equal to V, the condition is referred to as 100% excitation.
A situation where |E;| < |V,| is described as underexcitation.

6-2 PERFORMANCE COMPUTATION

Cylindrical-Pole Motor

by
The analysis of the performance of a synchronous motor is readily determined
the phasor diagram, the power-flow diagram, and the expressio n for the mechani-
been
cal power developed. The usefulness of the power-flow diagram has already
induction motor. Hence it is presented in
demonstrated in connection with the
There is one fewer step in this flow
Fig. 6-3 without much additional comment.
loss
graph because the ac source does not have to supply the rotor winding copper

Electrical power input Pots cos 8 — = Useful power

Pot
-
ist ==<_=— Power losses

e
Pye ¢
eye
' ’
\ 1
VEt sind=lw
Mechanical ! P_=qEsl, cosw=q-——
Stator
copper power alata Soci ES Lad obit :
loss developed

Pe e
o* =
ef
ave oo
' H
\
Core loss Mechanical
i
R, Po=Pmi- Tos
Pe Prot=
rotational losses power Prot
output
only.
Figure 6-3 Power-flow diagram of the synchronous motor, ac power
276 Three-Phase Synchronous Motors Chap. 6

since the rotor is traveling at synchronous speed. This is not to say that the rotor
winding copper losses do not exist. They do, of course, but the power is supplied
from the separate de source. Figure 6-3 represents the flow of power from the ac
source to the shaft of the synchronous motor.
What is the mechanism by which the synchronous motor becomes aware of
the presence of a shaft load, and how is the electric energy source made aware of
this so that it proceeds to provide energy balance? To answer this question we
start first with a study of the conditions prevailing at no-load. As the power-flow
diagram indicates, the only mechanical power needed is that which is required to
supply the rotational and core losses. This calls for a very small value of 6, as
depicted in Fig. 6-4(a). The machine is assumed to be overexcited since E;> Ve
Note that 5 is just large enough so that the in-phase component of I, is sufficient
to supply the losses. Consider next that a large mechanical load is suddenly
applied to the motor shaft. The first reaction is to cause a momentary drop in
speed. In turn this appreciably increases the power angle and thereby causes a
phasor voltage difference to exist between V, and the excitation voltage E;. The
result is the flow of an increased armature current at a very much improved power
factor compared with no-load. As a matter of fact the speed changes momentarily
by a sufficient amount to allow the power angle to assume that value which
enables the armature current and input power factor to assume those values which
permit the input power to balance the power required at the load plus the losses.
Figure 6-4(b) is typical of what the final phasor diagram looks like. Note that the
power component of the current is considerably increased over the no-load case.
The increase in power angle from no-load to the load case can be measured by
means of any stroboscopic instrument. The stroboscope furnishes a convincing
demonstration of the physical character of the power angle.
The elements of the phasor diagram of the synchronous machine are the
same whether it is used as a motor or generator. The geometry of the diagrams

(a) (b)

Figure 6-4 (a) Phasor diagram of synchronous motor at no-load with overexcita-
tion; (b) under load at fixed excitation.
 Sec. 6-2 Performance Computation 277

differs in two details as already indicated: The signs of the power angles are
reversed and Eq. (6-2) is used for the motor, whereas Eq. (5-22) applies for the
generator. In either case the expression for the mechanical power developed is
given by Eq. (5-55) and is repeated here for convenience. :
V,E¢
Pm=q sin 6 (6-3)
s

Power Angle and Static Stability

When the synchronous motor is connected to a line source that behaves as an

infinite bus, it follows that the maximum power developed corresponds to a power
angle of 90° so that Eq. (6-3) becomes
VE
(P.mak =q W (6-4)
Xs

It is useful to keep in mind, however, that during normal operation the power
angle is considerably less than 90° even at rated conditions. The differential
serves as a reserve capacity to help preserve static stability (i.e., maintain syn-
chronism) during load changes.
Unfortunately, this stability limit is seriously compromised when the syn-
chronous motor is supplied by a local generator which is of comparable rating. To
illustrate the situation, refer to Fig. 6-5. For simplicity, armature resistances and
other losses are assumed negligible. The generator is assumed electrically cou-
pled to the synchronous motor through the synchronous reactances of both ma-
chines as depicted in Fig. 6-5(a). Moreover, the motor field excitation is adjusted

Generator Motor

(a) (b)
of comparable size.
Figure 6-5 Synchronous motor supplied by a synchronous generator
an overexcited synchro-
(a) Line diagram using per-phase quantities; (b) phasor diagram for
nous motor and underexcited synchronous generator.
278 Three-Phase Synchronous Motors Chap. 6

for overexcitation and, correspondingly, the excitation in the generator is set for
underexcitation. The power developed by the generator can be expressed as
P = qEfgla COS a (6-5)

where Ey, is the generator excitation voltage and a is the phase angle between Er,
and J,. Also, in triangle Oab of Fig. 6-5(b), the distance ab can be identified as
ab = Ejm sin (8m + 5g) = Em sin 6 (6-6)

In triangle abc this quantity is alternatively given by

ab = Iq(Xsg + Xsm) COS @ (6-7)
From the last two equations it then follows that

I, COS a a= ence
~ k riia sinhes 6
hha (6-8)

Inserting Eq. (6-8) into Eq. (6-5) yields the result

P=q Elie Gs (6-9)

Xsg + Xsm

where

5 = 8, + 8m (6-10)
Here 6, is the power angle of the synchronous generator and 6, is the power angle
of the synchronous motor.
The maximum power represented by this motor-generator combination is
obtained from Eq. (6-9) by again setting 6 equal to 90°. But a comparison of this
result with the case where the synchronous motor is supplied from an infinite bus
reveals two important differences. (1) In the motor-generator case the angle 6
includes both the generator as well as the motor power angles. Thus, in the case
where the generator has the same nominal rating as the motor the individual
power angles are each equal to 45°. Hence the maximum allowable motor power
angle needed to maintain synchronism has been reduced to one-half the value
allowed when the motor is supplied from an infinite bus. (2) The coefficient of sin 6
in Eq. (6-9) is smaller than the corresponding coefficient of Eq. (6-3) because the
denominator reactance term is now greater by a factor of two and, furthermore,
because E;, is less than V,. The conclusion then follows that the static stability of
the synchronous motor is much less when it is supplied by a comparably rated
synchronous generator than when it is supplied from an infinite bus.

Example 6-1
A 2300-V, three-phase, 60-Hz, Y-connected, cylindrical-rotor synchronous motor
has a synchronous reactance of 11 0 per phase. When it delivers 200 hp the effi-
ciency is found to be 90% exclusive of field loss, and the power angle is 15 electrical
degrees as measured by a stroboscope. Neglect ohmic resistance and determine (a)
Sec. 6-2 Performance Computation 279

the induced excitation voltage per phase, E,, (b) the line current J, ,and (c) the power
factor. Assume the motor is supplied from an infinite bus.
Solution (a) The power input and the mechanical power developed are the same.
Hence

3 V.Er ging. 200(746)

xe 0.9

Inserting V, = 2300/V3, 5 = 15°, and x, = 11 yields

Ey = 1793 V/phase
(b) The armature current follows directly from Eq. (6-2). Thus
PIM327 OIA 1793)/ 15°
Le 1327 /0)"1793 15°" = 55.9 [41°

11 [90°
Thus since the angle of J, is positive, the power factor is leading.
(c) The line power factor is merely

power factor = cos 41° = 0.755 leading

Salient-Pole Motor

The phasor diagram for an overexcited salient-pole synchronous motor is depicted

in Fig. 6-6. The motor is assumed to be connected to an infinite bus and operating
at a leading power factor angle 6. A comparison of this diagram with Fig. 5-18,
which illustrates the phasor diagram of an overexcited synchronous generator,
reveals a considerable similarity. There are differences, however. One is that
overexcitation produces opposite effects regarding the power factor; it is leading
for the motor and lagging for the generator. Another is that for the generator the
terminal voltage plus the reactance drops yields the excitation voltage, whereas

Figure 6-6 Phasor diagram for an overexcited salient-pole synchronous motor.

280 Three-Phase Synchronous Motors Chap. 6

for the motor it is the excitation voltage summed with the reactance drops that
equates to the terminal voltage.
Because the geometry of the phasor diagram is the same for Fig. 6-6 as for
Fig. 5-18, the expression for the determination of the angle &, which locates the
direction along which the excitation phasor lies, is identical to that of Eq. (5-43),
which is repeated here for convenience. Thus
V,t sin 6 + I,x
ee = a“q
(6-11)
A selgsh V, cos 0

The angle 6 is to be used as an absolute quantity because the expression derives

from geometrical considerations. Also, although the expressions for the magni-
tudes of the direct-axis and quadrature-axis currents are the same in both cases,
the angles of these complex quantities are reversed in sign. Accordingly, we have
I, = Ua cos )/—6
(6-12)
Ty = Ua sin Ws)/90° — 6

where in absolute terms the power angle takes on the value

5=y- 0
Finally, with V, taking on the role of the applied line voltage, Kirchhoff ’s voltage
law applied on a per phase basis to the stator winding yields
V, = E; +P jlaXa = ghixe (6-13)

from which the expression for the excitation voltage readily follows. Hence
Es = V, = jlaxa = jlgxq (6-14)

The expression for the mechanical power developed by a salient-pole synchro-

nous machine is derived in Sec. 5-7 specifically for the synchronous generator.
The result is displayed in Eq. (5-65) and graphically portrayed in Fig. 5-22.
Because the result applies with equal validity to the salient-pole synchronous
motor, it is repeated here for ease of reference. Thus,
2
t
Poa 7 a sind + q (X¢ —ete) Si 25 (6-15)
2XaXq
Several points are worth noting in connection with Eq. (6-15). First, it is helpful
to keep in mind that the derivation is based on the assumption that the energy
source for the motor comes from an infinite-bus system. Also, the second term on
the right side is called the reluctance power of the salient-pole synchronous motor
because it owes its presence to a difference in reluctance between the direct axis
and the quadrature axis of the machine. This term is independent of the excitation
voltage and so does not require a field current for it to exist. As long as an
armature winding current is permitted to flow, there exists a rotating armature
mmf which produces a revolving air-gap flux that produces a force that acts to
Sec. 6-2 Performance Computation 281

keep the rotor locked in a minimum reluctance position as the rotor revolves in
synchronism with the revolving armature flux field. This reluctance power can be
as large as 25% of the motor’s rating. Finally, because the reluctance power isa
function of double the power angle associated with the excitation component of
developed power, the superposition of the two components plotted versus power
angle causes the maximum power developed to occur at a power angle that is less
than 90°. See Fig. 5-22.

Example 6-2
Consider that the synchronous machine of Example 5-4 is now operated as an over-
excited synchronous motor drawing rated current at rated voltage from the three-
phase line source at a leading pf of 0.8.
(a) Determine the excitation voltage of the synchronous motor at this operating
point.
(b) Obtain the value of the mechanical power developed in watts as it derives
from the presence of the field excitation.
(c) Find the reluctance power in watts.
(d) Compute the total mechanical power developed and compare it with the
input power drawn from the line.
(e) Calculate the developed motor torque.

d
Solution (a) Angle w is found to be the same as the value found for the overexcite
generator. Thus

Peet at ate page gen 3690.8 OA a

V, cos 0

Also,
I, = Ua sin p) /90°— 6 = 0.827/71.1°
I, = UI, cos p)/—8 = 0.562 /—18.9°
where 6 is found from
5=~-0

Finally, from Eq. (6-14) we get

p-u E; = p-u V, — j(p-u Ia)xa — (PU 1)Xq

WA (0.827 /71.1°)(0.961 /90°) = (0.562 /—18.9°)(0.577 /90°)
ll 1 — (—0.752 + j0.257) — (0.105 + 0.307)
1.74/—18.9°
in Example 5-4 for
It should not be surprising that this result is identical to that found
of the power angle.
the overexcited synchronous generator except for the sign
(b) The developed mechanical power that originate s from the excitation of the
term on the right side of Eq. (6-15). Thus,
field winding is described by the first
evaluated in terms of per-unit quantitie s, we get
282 Three-Phase Synchronous Motors Chap. 6

f a) (p-u V,)(p-u Ey) . te LCR Taye ee

p-u Pu uae sin 6 0.961 sin 18.9 0.5865

To obtain this result in watts it is simply necessary to multiply by the base kVA
which is the rating of the motor, namely, 15 kVA. Thus,
Pm = (0.5865)(15,000) = 8797 W
(c) The expression for the reluctance power is the second term of Eq. (6-15).
Calling this part P,,,, we evaluate

p-u Po, ee eSX,) (p-u xg ‘— p-u xz) sin 26

= Apu x4 \(p-U

1? ; 5

= 0.2122
and expressed in watts this becomes
Por = (0.2122)(15,000) = 3183 W
(d) The total mechanical power developed by this motor is the sum of the
excitation power and the reluctance power. On a per-unit basis this value is
p-u P,, = p-U Ppp + p-u Pp, = 0.5865 + 0.2122 = 0.7987

This compares favorably with the per-unit input power, which is

p-u P; = (p-u kVA) cos @ = 1(cos 36.9) = 0.8

Expressed in watts this input power is

P; = (0.8)(15,000) = 12,000 W
The corresponding expression derived from the total mechanical power developed is
Pm = (0.7987)(15,000) = 11,985 W
(e) As a six-pole machine operating on a 60-Hz power supply, the synchronous
motor shaft speed is

_ 120(60)
= 1200 rpm
Ss
oe 4
Or, expressed in radians per second,

T
w, = 1200 (
z) = 125.7 rad/s

Hence the torque expressed in newton-meters is

bab ake Los

1257 77 Nm
which in English units is

T = (0.7376)95.4 = 70.35 lb-ft

 Sec. 6-3 Power-Factor Control 283

6-3 POWER-FACTOR CONTROL

For a fixed mechanical power developed (or load) it is possible to adjust the
reactive component of the current drawn from the line by varying the dc field
current. This feature is achievable in the synchronous motor precisely because it
is a doubly excited machine. Thus, although operation at a constant applied volt-
age demands a fixed resultant flux, both the dc source and the ac source may
cooperate in establishing this resultant flux. If the field winding current is made
excessively large, then clearly the resultant air-gap voltage in the motor tends to
be larger than that demanded by the applied voltage. Accordingly, a reaction
occurs which causes the armature current to assume such a power-factor angle
that the armature mmf exerts that amount of demagnetizing effect which is needed
to restore the required resultant flux. Similarly, if the field is underexcited, then
the resultant gap flux tends to be too small. This also creates a phasor voltage
difference between the line voltage and the motor excitation voltage which acts to
cause the armature current to flow at that power-factor angle which enables it to
magnetize the air gap to the extent needed to provide the necessary resultant flux.
Whether the reactive component of current must be leading or lagging read-
ily follows from an investigation of the phasor diagrams depicted in Fig. 6-7 for
various values of excitation voltage. The diagram applies to an assumed constant
mechanical power developed. Hence as FE; is changed, the sin 5 must change
correspondingl y to keep E; sin 6 invariant. In Fig. 6-7 this means that, as the
excitation is varied, the locus of the tip of the Ey phasor is the broken line.
Furthermore, the in-phase component of the current in each case must be the
same. In the first case, where the excitation voltage is Epi , the field current is
producing too much flux. This creates a reaction between the motor and the
source which calls for a leading current of such a magnitude that it provides an
amount and direction of synchronous reactance drop that when added to Ep yields
the fixed terminal voltage V,. Physically, what is happening is that a leading
reactive current is made to flow, which acts to demagnetize the flux field to the
extent needed. When the excitation is reduced to Ep, there is no excess flux
produced by the field winding. Consequently, the ac line current contains no
reactive component. It merely has the value of the in-phase component needed to

Figure 6-7 Showing the effect of

varying excitation on power-factor.
284 Three-Phase Synchronous Motors Chap. 6

supply power to the load. At the excitation corresponding to E;3 the machine is
greatly underexcited. To compensate for this a reaction occurs which allows the
line to deliver a large lagging reactive current, which helps to establish the value of
air-gap flux demanded by the terminal voltage. Note, too, that as the excitation is
decreased the power angle must increase for a fixed mechanical power developed.
The foregoing description can be neatly summarized by observing that at a
given load condition, 5, the synchronous motor has no need to draw reactive
current whenever
E,cos 6 = V,
However, it draws a leading reactive current whenever

Es COs6>.V,

and a lagging reactive current whenever

E, cos 6 < V,

If a plot is made of the armature current as the excitation is varied for fixed
mechanical power developed, the current is observed to be large for underexcita-
tion and overexcitation and passing through a minimum at some intermediate
point. The plot in fact resembles a V shape as demonstrated in Fig. 6-8.

Synchronous Motor as a Synchronous Capacitor

The ability of the synchronous motor to draw leading current when overexcited
can be used to improve the power factor at the input lines to an industrial estab-
lishment that makes heavy use of induction motors and other equipment drawing
power at a lagging power factor. Many electric power companies charge in-
creased power rates when power is bought at a poor lagging power factor. Over
the years such increased power rates can result in an appreciable expenditure of
money. In such instances the installation of a synchronous motor operated at
overexcitation can more than pay for itself by improving the overall input power
factor to the point that the penalty clause no longer applies.
When the synchronous motor is operated solely for the purpose of furnishing
power factor correction, it is called a synchronous capacitor. In such a situation

I
: Unity pf
Lag pf Lead pf

Full load
} full load
No load
Figure 6-8 Synchronous-motor V
e) Ir curves.
 Sec. 6-4 Synchronous Motor Applications 285

the synchronous motor serves the function of supplying the reactive volt-amperes
required by the singly excited induction motors that are operating on the same
power lines and thereby relieves the power source of the need to do so. In
addition to the favorable power rates that can result, the overall efficiency of the
plant is improved because of the reduced line losses that accompany the improved
line power factor.
Example 6-3
An industrial plant is supplied from a 2300-V, three-phase infinite-bus transmission
line. The plant uses several large three-phase induction motors drawing a total of
2600 kW at a lagging power factor of 0.87 when fully loaded.
(a) Find the line current drawn from the transmission line.
(b) Determine the kVA rating of a synchronous motor used as a synchronous
capacitor that is needed to improve the power factor of the plant input to unity.
(c) What is the new value of the input current to the plant?
(d) If the selected synchronous capacitor is found to have a synchronous
reactance of 1.6 0, find the required excitation voltage per phase needed for part (c).

Solution (a) From the expression for power in a three-phase system

_ __ 2600 x 10
L = 750.2 A
/3(2.3)103 (0.87)
(b) The current rating of the synchronous capacitor is the total reactive compo-
nent of the input current found in part (a). Thus,

lem = I, sin 6 = 750.2 sin (cos! 0.87) = 750.2 sin 29.54° = 369.9 A
kVA rating = V3(2.3)(369.9) = 1473

(c) The new value of the input current from the transmission line is simply the
in-phase component of the condition described in part (a). Hence,

new I, = 750.2 cos 29.54° = 750.2(0.87) = 652.7 A

(d) From Eq. (6-2)

Ey = V, — jlsmXs = 2300 — 369.9 [90° (1.6)

V3
current.
where I, = 369.9 [90° because the synchronous capacitor draws leading
Accordingly,

Ey = ae [0° — j1.6(j369.9) = 1327.9 + 591.8 = 1919.7 V

V3

6-4 SYNCHRONOUS MOTOR APPLICATIONS

-speed range
Synchronous motors are rarely used below 50 hp in the medium
with inductio n motors. In
because of their much higher initial cost compared
and control devices
addition they require a dc excitation source, and the starting
 286 Three-Phase Synchronous Motors Chap. 6

are usually more expensive—especially where automatic operation is required.

However, synchronous motors do offer some very definite advantages. These
include constant-speed operation, power-factor control, and high operating effi-
ciency. Furthermore, there is a horsepower and speed range at which the disad-
vantage of higher initial cost vanishes, even putting the synchronous motor at
advantage. This is demonstrated in Fig. 6-9. Where low speeds and high horse-
power are involved, the induction motor is no longer cheaper because it must use
large amounts of iron in order not to exceed a gap flux density of 0.7 T. In the
synchronous machine, on the other hand, a value twice this figure is permissible
because of the separate excitation.

Synchronous
Horsepower motors

Synchronous or
induction motors

Induction motor
3600 1200 720. 514 400 and lower Figure 6-9 The general areas of appli-
1g00 900 600 450 360 cation of synchronous and induction
Rpm motors.

Appearing in Table 6-1 are some of the more important characteristics of the
synchronous motor along with some typical applications. In addition to the area
of application listed there, these motors are also quite prominently found in the
following applications, which are characterized by operation at low speeds and
high horsepower: large, low-head pumps; flour-mill line shafts; rubber mills and
mixers; crushers; shippers; pulp grinders and jordans and refiners used in the
papermaking industry.

6-5 FIELD EXCITATION BY RECTIFIED SOURCES

A distinguishing feature of the conventional synchronous motor is its need for dc

as well as an ac electrical supply. Because the ac source is the one that is fre-
quently available, special provision must be made for a de source. In years past,
one solution was to equip the synchronous motor with an auxiliary self-excited dc
generator attached to the motor shaft. Once the synchronous motor reached
nearly synchronous speed through the induction-motor action of its squirrel-cage
winding, the de generator field rhedstat was adjusted to furnish the proper excita-
Sec. 6-5 Field Excitation by Rectified Sources 287

TABLE 6-1 SYNCHRONOUS MOTOR CHARACTERISTICS AND APPLICATIONS

i

Type Synchronous, high speed, Synchronous, low speed,

designation above 500 rpm below 500 rpm

Starting torque Up to 120 Low 40

(% of normal)
Pull-in torque 100-125 30
(% of normal)
Pull-out torque (%) - Up to 200 Up to 180
Starting current (%) 500-700 200-350
Slip Zero Zero
Power factor High, but varies with load and High, but varies with excita-
with excitation tion
Efficiency, % Highest of all motors, 92-96 Highest of all motors, 92-96
Typical applications Fans, blowers, dc generators, line Lower-speed direct-connected
shafts, centrifugal pumps and loads such as reciprocating
compressors, reciprocating compressors when started
pumps and compressors. unloaded, dc generators,
Useful for power-factor correc- rolling mills, band mills,
tion. Constant speed. Fre- ball mills, pumps. Useful
quency changers for power-factor control.
Constant speed. Flywheel
used for pulsating loads.
tu di ee ES ee oS a
hi a
vol. I
Source: By permission from M. Liwschitz-Garik and C. C. Whipple, Electric Machinery,
(Princeton, N.J.: D. Van Nostrand Co., Inc., 1946).

into
tion for the synchronous motor, which in turn pulled the synchronous motor
In applicati ons of large synchron ous motors
synchronism with the line voltage.
by
today, the exciter generator is conveniently replaced by a dc source obtained
and silicon-c ontrolled rectifiers (SCRs),
electronic methods employing diodes
a wide range of
which are fully capable of handling a large range of currents over
voltages.
This
A typical controlled electronic rectifier circuit is shown in Fig. 6-10.
controll ed rectifier which can be used to adjust
circuit is a single-phase, half-wave
field winding is
the field current in the field winding of a synchronous motor. The
inductan ce L and is called the load. Two
represented by the resistance R and the
Dyy that serves to shunt the load and the
electronic elements are used: the diode
as a controll able switch. Of course
silicon-controlled rectifier S1 which behaves
e ac supply. The manner in which this
the source voltage is the frequently availabl
ac supply to an adjustab le dc source
circuit performs the conversion from a fixed
to recall that the SCR
can be understood by referring to Fig. 6-11. It is helpful
at its three terminal s. The first is that
conducts when two conditions are satisfied
positive. The second is that the gate-
the voltage from plate (P) to cathode (K) be
take place during the time the plate-
cathode voltage be set to allow conduction to
of this gate voltage it is possi-
cathode voltage is positive. By proper adjustment
time the plate- cathod e voltage is
ble to permit conduction to occur over the entire
288 Three-Phase Synchronous Motors Chap. 6

Figure 6-10 Single-phase, half-wave

controlled rectifier circuit.

positive or during no part of this time. Figure 6-11(a) shows two cycles of the ac
source voltage. In Fig. 6-11(b) the rectified version of the ac source is shown.
Observe that a positive potential appears across S1 (the SCR) only during the
positive portion of the ac input voltage variation. During the negative portion of
the cycle, a reversed (negative) voltage appears across the SCR and so conduction
is not possible. This accounts for the zero level of load voltage from wt = 7 to
wt = 27. In the first half-cycle, however, S1 is made to conduct at the angle 6,
which is determined by the voltage on the gate terminal. This angle is often
referred to as the firing angle of the SCR and is adjustable between 0 and z radians
in this case. When firing of the thyristor (SCR) is made to occur at 6 = 0°, the
voltage developed across the load is a maximum. The associated average value of
this maximum voltage can easily be shown to be V,,/7 where V,, denotes the peak
value of the ac supply voltage.
The variation of the load current produced by the rectified voltage of Fig.
6-11(b) is depicted in Fig. 6-11(c). It is particularly interesting here to note that
action of the inductance causes the current to exist beyond the time when the
thyristor ceases to conduct. Since the current at wt = 7 is not zero and there
exists an amount of energy stored in the magnetic field of the inductor consistent
with this nonzero current, a path must be provided to permit this current to
continue to flow. As soon as the SCR cuts off at wt = 7, the load current tries to
change instantaneously. Then, by Lenz’s law, an internally generated emf is
induced in the inductor of a polarity that acts to sustain the original direction of
the current. This action puts a positive potential across the shunt diode, Dw,
which makes it conductive, thus providing a path for the load current. Current
decay subsequently proceeds at a rate determined by the field winding time con-
stant. The shunt diode is called a free-wheeling diode for the reason that it freely
responds to conditions that call for an auxiliary current path whenever inductive
elements are involved.
Sec. 6-5 Field Excitation by Rectified Sources 289

(a)

Mi

Ofene 1 On em7+@ 3ir 41r wtirad

(b)

IL

O @ 1 om 217+ om 41 wt,rad

(c)

is

O ] 1 em 27+ Shep 41 wt, rad

(d)

itw

O wt,rad

(e)

Figure 6-11 Time variations of the electrical quantities appearing in Fig. 6-10:
(c) the R-L load
(a) source voltage; (b) rectified voltage across the R-L load;
variation; (d) source current; (€) current in the free-whee ling shunt diode.
current
The firing angle of the SCR (S1) is 0.

in the free-
The variations with time of the source current and the current
Figs. 6-11(d) and (e), respect ively. The abrupt
wheeling diode are depicted in
= 7 is allowed on the assumpt ion that the ac
change in the source current at wf
nce, which is reasona ble. The magnit ude of
source has negligible internal inducta
course matches the cutoff value that is found
the step current that appears in ify of
in the variation of i,.
290 Three-Phase Synchronous Motors Chap. 6

The half-wave controlled rectifier has the disadvantage of relatively low

currents because of the absence of any contributions from the source when its
potential is undergoing a negative excursion. This situation, however, is nicely
resolved by employing the circuitry illustrated in Fig. 6-12, which is called a full-
wave controlled rectifier. By adding an additional SCR and diode along with
appropriate modifications in the gate control circuitry, conduction to the load can
be achieved in a controlled fashion during each successive half-cycle of the source
frequency. For an explanation of the operation of this circuit refer to Fig. 6-13.
During the first positive half-cycle of the source voltage, a positive potential
appears across the plate-cathode terminals of thyristor S1. In Fig. 6-12 the left
side of the source is placed positive while the cathode of D2 is placed negative.
When the gate terminal is pulsed at 6, S1 conducts and thereby applies the source
voltage to the R-L load by completing the circuit through diode D2. In other
words, when the source voltage is positive, there exists a series path connecting
the source to the load through thyristor S1 and diode D2. As soon as the source
goes negative just beyond 7 radians in Fig. 6-13(a), S1 goes off as does D2. But
now the right side of the source in Fig. 6-12 is positive and the left side is negative,
which means that a positive potential appears across both S2 and D1. Again a
closed path becomes established between the source and the load upon the arrival
of the gate pulse at wf = 7 + 0, but now it is thyristor $2 and diode D1 that are
effective in providing the connection. Observe too that the sequencing of the
switching operation is such as to maintain the polarity of the voltage across the
load the same as it was in the preceding half-cycle. See Fig. 6-13(b). This ensures
a unidirectional current flow. Because a contribution occurs for each half-cycle of
the source voltage, the average voltage that is available to the load doubles over
that of the half-wave rectifier to a maximum value of (2/77) V,,._ Moreover, control

Gate control
Circuitry

is
Vs
Figure 6-12 Single-phase, full-wave controlled rectifier circuit.
Vs

wt, rad

(a)

VL

Cai? 7 21 37 4 wt, rad

(b)
in

Oni? 1 27 37r 4tr wt, rad

(c)

i tw

1 wt, rad
O ree
(d)

I,

Tt wt, rad
Oo @
(e)

Figure 6-13 Time variations of the electrical quantities found in Fig. 6-12:
(a) source voltage; (b) rectified load voltage; (c) resultant load current variation in
steady state; (d) current in the free-wheeling diode; (e) the source current. The
SCR firing angle.is set to 6.
291
292 Three-Phase Synchronous Motors Chap. 6

is available from zero to this maximum through the action of the gate control
circuit, which varies @ from 7 radians to zero.
The corresponding variation of the current in the load is depicted in Fig.
6-13(c). This graph shows the variation after steady state is reached following a
suitable transient response. Note the appreciable improvement over the half-
wave circuit. A glance at Fig. 6-13(d) reveals the essential contribution of the
free-wheeling diodes which is to provide a current path over the time interval
between the start of any half-cycle of the source voltage and the point of firing of
the thyristors. Figure 6-13(e) illustrates the variations of the source current during
the time that the SCRs are in the ON state. These current pulses are supplied
alternately through thyristors S1 and S2 to the load from the source.
The foregoing analysis illustrates the improvement that occurs when a full-
wave rectifier circuit is used in place of a half-wave rectifier circuit. However, a
further improvement is possible when a three-phase full-wave arrangement is
used in place of the single-phase full-wave circuit. Such a circuit is depicted in
Fig. 6-14. It is useful to note that a total of six thyristors (SCRs) is needed with
this arrangement. The availability of a rectifier circuit that uses a three-phase
source is attractive for the reason that the three-phase source is already needed to
drive the three-phase synchronous motor and, accordingly, is assumed to be
available. By the proper sequential pulsing of the gate terminals of the thyristors
in the circuitry of Fig. 6-14 it is possible to vary the average voltage across the
field winding over a range that varies from zero to (3/77) V,,, where V,, denotes the
maximum value of the ac voltage between lines.
The success of this circuit depends upon a biasing scheme that permits only
two thyristors to be in a conducting state at any time while simultaneously ensur-
ing that the remaining SCRs are in a state of reversed bias. A typical switching
sequence for this circuit is illustrated in Fig. 6-15 as it applies to a line voltage
sequence of a, b, c. Each thyristor is scheduled for a maximum ON period of 120

Three - phase
source

Load
Figure 6-14 Three-phase, full-wave controlled rectifier circuit.
Sec. 6-5 Field Excitation by Rectified Sources 293

Figure 6-15 Switching sequence dia-

gram for the thyristors of Fig. 6-14.
The heavy lines indicate the specific
portions of a cycle when each thyristor
eT, wt is forward biased for conduction to
occur subject to the action of the gate
signal.

electrical degrees for each cycle of the source voltage. The gating pulse is de-
signed to allow a full variation over this interval. Starting with the beginning of a
cycle, S1 is forward-biased to be in the ON state potentially over the first 120
electrical degrees while S2 and S3 are scheduled potentially to be in the ON state
for each of the next 120° intervals respectively. Then, beginning with the start of
the next half-cycle at 7 radians, a similar schedule is established for thyristors S4,
S5, and S6. Accordingly, S4 is arranged to have a total possible ON period of 120°
extending from 7 to 3a radians. Similarly, S5 is forward-biased for a possible ON
period extending from 37 radians right on into the first 7/3 radians of the next
cycle. As a matter of convenience this second 60° portion is shown reflected into
the first cycle in Fig. 6-15. Thyristor S6 is subsequently scheduled for a potential
ON period covering the next 120 electrical degrees. Reflected back into the first
cycle, S6 is scheduled to occupy the period from 7/3 to 7 radians.
As a result of the switching sequence schedule depicted in Fig. 6-15 for the
line voltage sequence a, b, c, it follows that both thyristors S1 and S2 are forward-
vg, at
biased for conduction during the first 7/3 radians of a cycle that begins with
passing through zero
(3/2) V_ and moving to its peak while v,, is simultaneously
at which conducti on
and moving positively. The specific point in this interval
Similarly, in the next 60°
actually occurs is determined by the gate pulsing signals.
conducti on process. In
interval, it is thyristors S1 and S6 that are involved in the
over to thyristors S2
the third 60° interval the responsibility for conduction passes
and then repeats
and S6. This process continues for the remainder of the cycle
winding dc
itself for each cycle. The result is an effective control of the field
current at entirely acceptable levels.
 294 Three-Phase Synchronous Motors Chap. 6

Brushless Excitation for Synchronous Generators

Synchronous generators employed by the public utility companies can have rat-
ings of 1000 MVA and more (where MVA denotes megavolt-amperes). To equip
such machines with de power to be conducted to the rotating field winding through

To balanced three-phase
Revolving rectifier bus system
attached to shaft

L1 2 eS

Synchronous
Exciter
generator generator
(large rating)
WMA, over

Auxiliary
dc source
(static type) (a)

(b)
Figure 6-16 (a) Schematic diagram illustrating how brushless excitation of a large synchro-
nous generator is obtained. (b) Rotor of synchronous generator showing the exciter genera-
tor as well as the revolving rectifier.
 Sec. 6-6 Synchronous Motor Operation with the Basic Torque Equation 295

slip rings presents a significant maintenance problem as well as the loss of a

substantial amount of power consumed as a heat loss in the carbon brushes used
to conduct the large de currents to the slip rings. An innovative application of the
use of the circuitry of Fig. 6-14 can be used to eliminate these problems. The
arrangement is illustrated in Fig. 6-16(a) and a photograph of an actual machine
appears in Fig. 6-16(b).
The object of this scheme is to make available a dc source which is revolving
at the same speed as the field winding of the main generator. A direct connection
to the terminals of the field winding is then possible thereby eliminating the need
for brushes. Attached to the shaft of the main generator is an exciter generator
the size of which is sufficient to provide the power required by the field winding to
maintain the level of dc current needed by the synchronous generator. Also at-
tached to the shaft is the rectifier of Fig. 6-14. A small de current (in comparison
to that required by the synchronous generator) is supplied to the field winding of
the exciter generator. An auxiliary dc source derived from a static rectifier can be
used because the currents involved here are manageable. When the prime mover
of the synchronous generator reaches synchronous speed, the ac exciter generator
to the
develops a balanced three-phase voltage. Application of this voltage set
shaft produces a dc volt-
three-phase full-wave rectifier attached to the generator
same velocity as the genera-
age source at terminals F1—F2 which revolves at the
and slip
tor field winding itself. Hence, a firm, terminal connection is possible
rings and brushes for the main generator are no longer necessary.

6-6 CORRELATION OF SYNCHRONOUS MOTOR OPERATION

WITH THE BASIC TORQUE EQUATION
the manner in which the
It is the purpose of this section to outline in some detail
of Appendix to Chapter
factors appearing in the basic torque equation [Eq. (3-97)
power factor condi-
3] are manifested in synchronous motor operation for several
motor compli cates the situ-
tions. The doubly excited feature of the synchronous
of overexc itation or underexci-
ation somewhat because field winding conditions
re current to become leading
tation influence the reactive component of the armatu
directe d first to the case where
or lagging, respectively. See Sec. 6-3. Attention is
with the excitation voltage (and
w equals 0°, i.e., the armature current is in phase
Our concern is with the cylindri-
treated as a voltage drop). Refer to Fig. 6-17(a). is
-pole structure; the latter
cal-rotor machine in spite of the use of a salient
. For the indicated po-
employed merely to facilitate the graphical representation
the location of the amplitude
larity of the rotor field flux and direction of rotation,
winding is represented by coil
of the emf distribution of the entire three-phase
When the machine is operating
er-e; having the instantaneous directions shown.
d to the polarity of the er shown
as a motor, the armature current flows oppose
with respect to the field flux that
because ef is represented here as a voltage rise
296 Three-Phase Synchronous Motors Chap. 6

(a)

¢,, produced by armature

@;, flux field mmf associated with
distribution amp-—cond. distribution

Displacement along
inner periphery of
stator

Armature amp-cond. |
distribution

(c)
Figure 6-17 Space and time relationships in the synchronous motor for y = 0°:
' (a) space diagram depicting the axis of ampere-conductor distribution; (b) corre-
sponding phasor diagram; (c) detailed representation of (a). Field motion is to-
ward the left.

produces it. However, when viewed with respect to the current,+ the E; phasor is
treated as a voltage drop and is so represented in Fig. 6-17(b). The coil igi, in
Fig. 6-17(a) represents the location of the amplitude of the assumed sinusoidal
ampere-conductor distribution. Note that the armature current flows inward for
the upper conductor and outward for the lower conductor because i, is specified
as in-phase with the excitation voltage expressed as a voltage drop and hence
must be opposite to the direction of e, treated as a voltage rise. The armature mmf
(represented by A) is therefore directed from right to left and cooperates with F to

+See Sec. 2-1.

Sec. 6-6 Synchronous Motor Operation with the Basic Torque Equation 297

produce the resultant R. Note that the R in Fig. 6-17(a) is consistent with that
shown in Fig. 6-17(b). For simplicity the armature leakage reactance is assumed
to be negligible so that the armature induced emf E, may be taken equal to the
terminal voltage V,. It is for this reason that R is drawn perpendicular to V, in Fig.
of the
6-17(b). Then to be consistent with this the amplitude of the distribution
terminal voltage in the armature winding can be represented in the space picture
of
of Fig. 6-17(a) by coil v,-v;. It is worthwhile to note at this point that by means
the power angle 6 is obtained.
such a representation a physical interpretation of
Also worthy of note is the fact that application of the I x B rule indicates a
the torque on
clockwise torque produced on the armature (stator) winding, while
, the direction of torque
the rotor is counterclockwise. For the motor, therefore
6-17(b) note again that the
and the direction of rotation are the same. In Fig.
to align the field axis with
direction of the torque on the rotor is such that it tends
the axis of the resultant flux, as was the case for generato r action.
represented in
The information contained in Figs. 6-17(a) and (b) can also be
The funda-
more detail by employing the developed form depicted in Fig. 6-17(c).
denoted by ¢;. The ampere-
mental component of the field flux produced by F is
The armatur e flux produce d
conductor distribution is denoted by the dashed line.
or distribu tion is repre-
by the armature mmf associated with the ampere-conduct
conductor
sented by ¢,; this @, wave is properly displaced 90° behind the ampere-
wave. Of course the resultant
distribution and in-phase with the armature mmf
and is permitt ed by the linearit y assump-
flux is produced by summing @¢; and
a constant terminal voltage the
tion. This is denoted by ¢,. Keep in mind that for
of these distribu tions makes
amplitude of the ¢, wave remains invariant. A study
dent. Because p = 0°,
the correspondence with the basic torque equation self-evi
in the same direction and
each conductor beneath a given field pole carries current
This viewpoint is consis-
thereby makes a contribution to the developed torque.
describ ed by
tent with that version of the basic torque equation

T = 3 PPIn cos w = 5 P?OIn (6-16)

and J,, denotes the amplitude of

where ® is the flux produced by the field winding
the ampere-conductor distribution.
angle A. This angle is
The second viewpoint involves the use of the torque
en the axis of the field pole and the
defined in Fig. 3-18 (p 574) as the angle betwe angle
always the complement of
axis of the armature mmf pole. Moreover, A is second
is 90° for this case so that the
w. A glance at Fig. 6-17(c) discloses that A
form of the basic torque equation becomes

T= q POF, sin A = 5 POF, (6-17)

where ¥, denotes the armature mmf per pole.

s in the synchronous motor when
Figure 6-18 depicts the situation that exist
rn for producing torque 1s not optimum
wy = 60° leading. Note that the field patte
 298 Three-Phase Synchronous Motors Chap. 6

(b)

Arm. amp-cond.
distribution

Displacement along
._ Stator periphery

(c)

Figure 6-18 Space and time relationships in the synchronous motor for b = 60°
leading: (a) space diagram; (b) phasor diagram; (c) detailed form of (a).

because now some of the armature conductors find themselves

under a pole flux of
the wrong polarity. This is accounted for in E
q. (6-16) by the need to use cos y =
cos 60° rather than unity as in the preceding case. When the situatio
n is described
in terms of the formulation of Eq. (6-17), the required value
of A is obviously 30°.
It is worthwhile to note in Fig. 6-18(c) that the orientation
of the armature mmf (or
its associated flux) with the field flux is such as to produce a net
demagnetization.
This is entirely consistent with the vector relationshi p shown
between F and A in
Fig. 6-18(a).
Chap. 6 Problems 299

ae PROBLEMS

6-1. A three-phase, 1732-V (line to line), Y-connected, cylindrical-pole synchronous mo-

tor has r, = 0 and x, = 100 per phase. The friction and windage plus the core losses
amount to 9 kW. The motor delivers an output of 390 hp. The greatest excitation
voltage that may be obtained is 2500 V per phase.
(a) Calculate the magnitude and pf of the armature current for maximum excitation
at the specified load.
sm
(b) Compute the smallest excitation for which the motor will remain in synchroni
for the given power output.
us motor has
. A2300-V, three-phase, 60-Hz, Y-connected, cylindrical-rotor synchrono
phase. When it delivers 200 hp, the efficiency is
a synchronous reactance of 11 © per
angle is 15 electrical degrees. Neglect r, and determine : (a) the
90% and the power
excitation emf per phase; (b) the line current; (c) the power factor.
current causes
. Asynchronous motor is operated at half-load. An increase in its field
in armature current. Does the armature current lead or lag the terminal
a decrease
voltage? Explain.
80% leading pf. The
6-4 A three-phase, Y-connected synchronous motor is operating at
armature winding resistance is
synchronous reactance is 2.9 0 per phase and the
20 A per phase. The applied line voltage is
negligible. The armature current is
440 V.
(a) Find the excitation voltage and the power angle.
phase and maintain the
(b) It is desired to increase the armature current to 40 A per
in a phasor diagram how the field excitation
pf at 80% leading. Show clearly
Can this be accompli shed if the shaft load remains fixed?
must be changed.
Explain.
onous motor has an armature
6-5. A Y-connected, three-phase, 60-Hz, 13,500-V synchr
ce of 37.4 © per phase.
resistance of 1.52 © per phase and a synchronous reactan
is 96% and the field current is SO
When the motor delivers 2000 hp, the efficiency
current of 85 A.
adjusted that the motor takes a leading
(a) At what pf is the motor operating?
(b) Calculate the excitation emf.
(c) Find the mechanical power developed.
how the magnitude and pf of
(d) If the load is removed, describe (do not calculate)
current compar e with the original .
the resulting armature
synchr onous motor receives constant
. A 1250-hp, three-phase, cylindrical-pole,
V. Armatu re windin g resista nce is negligible and the
power of 800 kW at 11,000
e is Y-conn ected. The motor
synchronous reactance per phase is 50 2. The armatur
current shall not exceed 135% of
has a rated full-load current of 52 A. Ifthe armature throug h
the excitat ion emf can be varied
this value, determine the range over which
adjustment of the field current.
nt-voltage source. At no-load the
6-7. A synchronous machine is supplied from a consta
negligi ble when the excitation is 1.0 per unit.
motor armature current is found to be
nts are xg = 1.0 and x, = 0.6.
The per-unit motor consta
angle between the quadrature axis
(a) If the machine loses synchronism when the
300 Three-Phase Synchronous Motors Chap. 6

and the terminal voltage phasor direction is 60 electrical degrees, what is the per-
unit excitation at pullout?
(b) What is the load on the machine at pullout? Assume the same excitations as
part (a).
6-8. A synchronous machine operating at rated voltage and rated current shows a leading
pf of 86.6%. The per-unit excitation voltage is 1.732. The per-unit synchronous
reactance is 1.00.
(a) Does this condition represent motor or generator operation? Explain.
(b) Compute the power angle.
6-9. As a synchronous motor is loaded from zero to full-load, can the pf ever become
leading if the field excitation is maintained at 75%? Explain.
6-10. A synchronous motor delivers rated power to a load. Is there a lower limit on the
excitation current beyond which it may not be reduced? Explain.
6-11. A 440-V, three-phase, Y-connected synchronous motor has a synchronous reactance
of 6.06 © per phase. The armature resistance is negligible, and the induced excita-
tion emf per phase is 200 V. Moreover, the power angle between V, and E; is 36.4
electrical degrees.
(a) Calculate the line current and pf.
(b) What values of excitation emf and power angle are necessary to make the pf
unity for the same input?
6-12. A synchronous motor at no-load is connected to an infinite bus system. The field
circuit is accidentally opened. Explain what happens to the armature current.
6-13. An eight-pole synchronous motor draws 45 kW from the 208-V, 60-Hz, three-phase
power system at a pf of 0.8 lagging. The motor is Y-connected and has a synchro-
nous reactance of 0.6 ( per phase. Armature resistance is negligible. Without any
further manipulations on the motor, what is the highest possible value of its steady-
state torque?
6-14. Answer whether the following statements are true or false. When the statement is
false, describe why.
(a) Increasing the air-gap of a machine reduces the synchronous reactance and
raises the steady-state power limit of the machine.
(b) For a synchronous motor, the developed electromagnetic torque is in a direction
opposite to the direction of rotation.
(c) Ina synchronous machine, the space-phase relation between the field mmf wave
and armature-reaction-mmf wave is determined by the pf of the load.
(d) The sum of the armature mmfs in the three phases of the stator winding is zero
at any given instant.
(e) The sum of the main field flux ¢y and the armature reaction flux ®,, always
add
to give the resultant flux ®,.
(f) A synchronous motor operating at leading pf is underexcited.
(g) The high currents flowing during a short-circuit cause a synchronous
machine to
operate under saturated conditions.
(h) The load on a synchronous motor is increased. This causes the main
field mmf
axis to drop further behind the air-gap mmf axis, so as to increase the torque
angle. 1
(i) An overexcited synchronous motor is a generator of lagging kvars.
This means
that it supplies lagging kvars to the system to which it is connected.
Chap. 6 Problems 301

about
(j) For large machines the armature resistance and leakage reactance are
10% on the machine rating and the synchrono us reactance is about 100%.
that
(k) The equivalent circuit developed for the synchronous machine assumes
saturation is negligible and that the air gap of the machine is uniform.
the mag-
(1) The electromagnetic torque developed in a machine is proportional to
mf waves and the sine of the angle be-
nitudes of the main-field and air-gap-m
tween them.
waves are no
(m) By loss of synchronism we mean that the field and air-gap mmf
longer stationary with respect to each other.
voltage, through a
6-15. A synchronous motor is fed from an infinite bus with 1.00 per-unit
no resistanc e. With a per-unit excitatio n of 0.9, the
line containing inductance and
is negligibl e at no-load. However , when the motor is loaded without
motor current
of rated output. Full-
changing excitation, it falls out of step ata load of 0.81 per unit
n during a short-cir cuit test.
load current is produced by 0.75 per-unit excitatio
(a) What is the per-unit inductiv e reactanc e of the line?
(b) What is the current under the preceding conditions?
per-unit synchronous reac-
6-16. A six-pole synchronous motor (cylindrical rotor) has a
Its excitati on emf is adjusted to a
tance of 0.75, the resistance being negligibly small.
is loaded, the rotor is observe d to drop back 10
per-unit value of 1.50. As the motor
position corresp onding to zero load. Losses are negligi-
degrees from its initial space
:
ble. For rated voltage impressed on the motor, compute
(a) Line current in per-unit values.
(b) Power factor, lagging or leading.
or false. When the state-
6-17. Answer whether or not the following statements are true
a Y-connection and 0.8 pf lead.
ment is false, provide the correct answer. Assume
n of the field mmf axis and the
(a) In a synchronous machine the relative positio
armature mmf axis is solely depend ent upon the power factor of the load.
onous motor is increas ed. This causes the speed of the
(b) The load on a synchr
so that more input current
armature mmf to drop to a lower value at steady state
can flow to supply the need for increased shaft power.
ng at no-load with the per-unit
6-18. A synchronous motor with salient poles is operati
rated terminal voltage V,; = 1 and per-uni t excitat ion voltage of Ey = 0.65. The motor
Assuming that the ratio xg/x, =
is found to draw an armature current of 0.5 per unit.
axis and quadrature-axis synchro-
1.75, determine the per-unit values of the direct-
nous reactances.
reactances of a salient-pole synchro-
6-19. The direct- and quadrature-axis synchronous resistance
0.50 per unit. The stator
nous generator are xy = 1.00 per unit, and x, =
generator is connected to an infinite
and the rotational losses are both negligible. The
genera tor field excitation is adjusted so that
bus of rated voltage and frequency. The
zero. The power input to the generator from its prime
at no-load the line current is
tor is pulled out of synchronism.
mover is then slowly increased until the genera
the machine is on the verge of
What is the per-unit line current and the pf when
m illustr ating your calculations.
pulling out? Prepare a neat, clear phasor diagra
ine the following.
6-20. For the machine described in Example 6-2 determ
is developed.
(a) The power angle at which maximum power
of the develo ped power expressed in per unit.
(b) The maximum value
302 Three-Phase Synchronous Motors Chap. 6

6-21. The synchronous motor of Example 6-2 is operated with a change in field current
that causes the power factor to change to 0.8 lagging from 0.8 leading. Determine the
new value of the excitation voltage.
6-22. A cylindrical-pole synchronous motor is to be used to improve the power factor of a
load of 1500 kW at a power factor of 0.85 lagging. The synchronous motor is to
operate at a 0.75 leading power factor. What must be the kVA rating of the synchro-
nous motor if it is required to raise the total power factor to 0.95 lagging?
6-23. An industrial plant uses two 3-phase induction motors rated at 460 V. At full load
one draws 232 A at 0.85 pf lagging; the other draws 169 A at 0.82 pf lagging.
Find the kVA rating of a synchronous capacitor that can serve to bring the
plant power factor to 0.9 lagging.
6-24. The net input to an industrial plant supplied from a 2300 V infinite bus system is
measured to be 765 A at a 0.92 pf lagging. Although most of the loads are inductive,
the input pf was improved by the installation of a synchronous capacitor operating at
its rating of 1000 kVA.
Determine the original power factor of the plant.
6-25. The source voltage appearing in Fig. 6-12 is assumed to be 120 V rms and 60 Hz.
(a) Determine the value in volts of the average load voltage when the firing angle of
the SCR is set for 7/4 radian.
(b) Let the resistor in this circuit be 10 © and the inductor 5 H. Compute the
average value of the load current for the firing angle of part (a).
 7
Direct-Current
Generators

starting torques
The dc machine is a highly versatile machine. It can provide high
and decelera ting torques. It is capable of quick rever-
as well as high accelerating
a range of 4:1 is achieve d with relative ease in
sals, and speed control over
ion devices. These
comparison with all other electromechanical energy-convers
for its use in the really tough jobs in industry,
are features that are responsible
Unfortu nately, the need for a mechani cal rectifier
such as are found in steel mills.
to convert the ac emf that is induced in each
(in the form of a commutator)
voltage makes it one of the least rugged of electric
armature coil to a unidirectional
machines as well as more expensi ve.
induced voltages in
The principles underlying basic torque production and
we investigate the opera-
the dc machine are outlined in detail in Chapter 3. Here
types of dc machin es, beginning with a de-
tional characteristics of the various
influence of the armature
scription of how the direct voltage is obtained. The
particular emphasis on its
winding mmf on machine behavior is explored with
l charact eristic s. The subject of commuta-
effect on commutation and its externa
in the study of dc machines because
tion occupies a position of preeminence
is render ed almost useless. The motor
without good commutation the machine
the various method s of speed control.
speed-torque curves are analyzed and also
e perfor mance throug h the use of the
Finally, the procedure for finding machin

303
 304 Direct-Current Generators Chap. 7

governing equations, the equivalent circuit, the power-flow diagram, and the mag-
netization curve is outlined and illustrated.

7-1 GENERATION OF UNIDIRECTIONAL VOLTAGES

It is helpful at the start of our study of dc machines to understand how a direct

voltage can be produced at the armature output terminals. Appearing in Fig.
7-1(a) is a plot of the flux density produced by the field winding as a function of

Bt

27r
of, displacement
along air —gap

(a)

Commutator segment

E,,induced
emf
armature

Figure 7-1 Generation of unidirectional voltage: (a) shape of air-gap flux density
produced by field winding; (b) emf generated in a single-coil armature for four
instants of time; (c) plot of E, as a function of time, with specific instants of part
(b) shown.
Sec. 7-1 Generation of Unidirectional Voltages 305

displacement along the periphery of the rotor. The flat-topped portion of the plot
is attributable to the constant air gap between the rotor surface and the faces of the
pole pieces. A sharp falloff occurs in the interpolar space because of the effect of
the large air gaps there. Figure 7-1(b) depicts a cross-sectional view of a two-pole
machine having an armature winding consisting of a simple coil, a—a’. Coil side a
is joined to coil side a’ by a back connection that is not shown. The front ends of
each coil side are joined permanently to copper segments as indicated. Note that
both coil sides are effective in voltage production with this arrangement, unlike
the Gramme-ring winding. The copper segments are part of the rotor structure,
thereby rotating with it. Any emf that is induced in the coil a—a’ appears at these
copper segments. Placed in contact with the copper segments and fixed in space
are two carbon brushes B, and B. These are used to collect the voltage induced
in the armature winding and to make it available to the external circuit. The coil in
Fig. 7-1(b) is shown in four different positions relative to the flux field. The rotor
is assumed revolving at speed n in the counterclockwise direction. At time instant
1, the emf induced in the coil is zero because each coil side finds itself at a point of
zero flux density. At instant 2, there is an emf induced in each coil side and its
magnitude is proportional to the value of flux density, as obtained from Fig. 7-1(a),
as well as to the velocity and length of the coil sides. The direction of the induced
voltage in coil side a is such that it makes the polarity of brush B, positive. As the
rotor revolves an additional 90 degrees, the situation illustrated in 3 of Fig. 7-1(b)
is found to prevail. Again note the zero voltage value. At instant 4, coil side a is
under the influence of south-pole flux and so has an emf of reversed polarity
induced. However, note that its attached commutator segment is now in contact
with brush B), which keeps this brush at a negative polarity. A complete plot of
the voltage appearing at the armature terminals is depicted in Fig. 7-1(c) with the
four time instants of Fig. 7-1(b) specifically indicated. It is important to note that,
although the voltage in each coil side alternates for each revolution, the voltage
appearing at the brushes is unidirectional because of the effect of the segmented
commutator. t
Although the use of a single coil in conjunction with the commutator fur-
nishes a unidirectional voltage, the resulting wave shape of Fig. 7-1(c) is unsatis-
ble
factory because the magnitude is not constant over the full period. A considera
number of arma-
improvement can be achieved by increasing and distributing the
having
ture coils. Figure 7-2(a) depicts in developed form a two-pole machine
yielding a total of
eight slots on the rotor equipped with two coil sides per slot,
in practical ma-
eight coils. This representation is typical of the situation found
the various
chines. The lower part of the figure illustrates the manner in which
shown is
coil sides are joined to give a summed quantity. The winding layout

making. The arrange-

+For better understanding the distinction with the ac generator is worth
becomes an ac generator when each coil side is connected to a closed copper ring
ment of Fig. 7-1(b)
the positive and negative variations of induced emfs
and each brush is made to ride on one ring. Then
at the brushes and therefore the armature terminals.
are made to appear
 ~ ~“
x ~ ~ ey
3 SISAL Ps, S25 2 ig

volts
E,,

Total induced voltage

|| between brushes
Coil 5-14.
Coil 7-16, Lp KK OOK
Coil 1-10
Coil 3-12 TNL
\
to ty a= w,t
(c)
Figure 7-2 Generation of a dc voltage: (a) winding layout for a simple machine;
(b) schematic representation of two armature paths existing between the brushes
at the time / illustrated in (a); (c) total induced armature voltage as a function of
time; complete cycle is shown for each of four coils.

306
Sec. 7-1 Generation of Unidirectional Voltages 307

referred to as a Wave winding for obvious reasons.t As a general rule the sides of
a given coil are made to span one pole pitch (i.e., 180 electrical degrees). More-
over, in the double layer winding, such as the one illustrated here, each coil is so
placed that one side occupies a lower position and the other side an upper posi-
tion. For example, coil sides 1 and 10 are joined by the back-end connection to
constitute one coil. Note that the span is 180° and that coil side 10 occupies the
lower position in slot 5, while coil side 1 occupies the upper position in slot 1.
A study of Fig. 7-2(a) discloses that the armature winding completely closes
on itself and that in particular the wave winding has two armature paths with
respect to the brushes. The arrowheads attached to each coil side indicate the
direction in which the induced emf causes current to flow when a load is placed
across the brushes. In the plan view of Fig. 7-2(a) the direction of the emf for the
direction of rotation shown is downward beneath a north pole and upward beneath
a south pole. For clarity’s sake the south pole is repeated once; on the left side
the upper coil sides are specifically illustrated, while on the right side the coil sides
located in the lower portions of the slots are shown. Also brush B) is assumed to
be resting on commutator segment 4, and simultaneously brush B, is assumed to
be in contact with commutator segment 8. Examination of the circuitry at B,
discloses that current converges at this brush from two directions—from coil side
2 and coil side 7. Further investigation reveals that coil side 2 is associated with a
path that consists of four coils: 8-15, 6-13, 4-11, and 2-9. These coils are drawn
with solid lines in Fig. 7-2(a) and are denoted by small series-connected dc sources
7—
in Fig. 7-2(b). Similarly coil side 7 is associated with the remaining four coils:
with broken
16, 5-14, 3-12, and 1-10. These coils are represented in Fig. 7-2(a)
lines and in Fig. 7-2(b) by series-connected de sources.
flux-
If attention is directed solely at the fundamental component of the
time instant being
density curve of Fig. 7-1(a), it should be clear that for the
example, the
considered the values of the coil voltages are not the same. For
d from their
instantaneous values of coils 3-12, 4-11, 5-14, and 6-13 are displace
by 67.5°.
positions of maximum value by 22.5° while the remaining coils are away
s of the armature
A partial time history of these coil voltages for various position
field distribu tion is depicted in Fig. 7-2(c). The instant represented
relative to the
over any full cycle
in Fig. 7-2(a) is identified as to. Note that each coil voltage
tor. Further-
appears as a rectified wave because of the action of the commuta
are depicted . The
more, only the four coils on one side of the armature winding
B; at time f is the sum
total induced voltage appearing between the brushes B, and
COS 22.) + 2E,, COS
of the instantaneous coil voltages (Eq0), i-e., Eao = 2EL
conditions leading to
67.5°, where E,, is the peak voltage induced in the coil. If the
hen coils 5-14 and
Fig. 7-2(b) are analyzed for a time instant t;—22.5° later—t
their peak values, while the emf induced in coils 2-9 and
6-13 will experience
emfs in the remaining
1-10 will be zero. The instantaneous values of the induced
¢, is therefore Eg; = Em +
coils will be E, cos 45°. The total contribution at time

+Another frequently used arrangement is the lap winding.

 308 Direct-Current Generators Chap. 7

2E», cos 45° + 0. This total is identified in Fig. 7-2(c) and is less than the value
occuring at time fo.
A glance at Fig. 7-2(c) makes it obvious that an almost constant voltage now
appears between the brushes in contrast to the situation of Fig. 7-1(c). The small
ripple is attributable to the small number of slots per pole. In practical machines
this number is very large so that the ripple is hardly detectable even with sensitive
instruments.

7-2 DIRECT-CURRENT GENERATOR TYPES

The dc machine functions as a generator when mechanical energy is supplied to

the rotor and an electrical load is connected across the armature terminals. In
order to supply electrical energy to the load, however, a magnetic field must first
be established in the air gap. The field is necessary because it serves as the
coupling device permitting the transfer of energy from the mechanical to the
electrical system. There are two ways in which the field winding may be ener-
gized to produce the magnetic field. One method is to excite the field separately
from an auxiliary source as depicted in the schematic diagram of Fig. 7-3. But
clearly this scheme is disadvantageous because of the need of another dc source.
After all, the purpose of the dc generator is to make available such a source.
Therefore, invariably dc generators are excited by the second method which
involves a process of self-excitation. The wiring diagram appears in Fig. 7-4, and
the arrangement is called the self-excited shunt generator. The word shunt is used
because the field winding appears in parallel with the armature winding. That is,
the two windings form a shunt connection.
To understand how the self-excitation process takes place we must start
with the magnetization curve of the machine. Sometimes this is called the satura-
tion curve. Strictly speaking, the magnetization curve represents a plot of air-gap
flux versus field winding mmf. However, in the dc generator where the winding

Separate
dc source 7
Field
winding

Prime
mover

Figure 7-3 Schematic diagram ofa | , Figure 7-4 Schematic diagram of a self-ex-
separately excited dc generator. cited shunt generator.
Sec. 7-2 Direct-Current Generator Types 309

Voltage
Magnetization curve
(constant speed)

Field resistance
line

Figure 7-5 Illustrating the build-up

procedure of a self-excited shunt gener-
bd Field winding current ator.
O

constant Kz is known and the speed n is fixed, the magnetization curve has come
to represent a plot of the open-circuit induced armature voltage as a function of
the field winding current. With Kz and n fixed, Eq. (3-33) shows that ¢ and E,
differ only by a constant factor. Figure 7-5 depicts a typical magnetization curve,
valid for a constant speed of rotation of the armature. It is especially important to
note in this plot that even with zero field current an emf is induced in the armature
of value Oa. This voltage is due entirely to residual magnetism, which is present
because of the previous excitation history of the magnetic-circuit iron. The linear
curve appearing on the same set of axes is the field-resistance line. It is a plot of
the current caused by the voltage applied to the series combination of the field
winding and the active portion of the field rheostat. Clearly, then, the slope of the
linear curve is equal to the sum of the field-winding resistance Ry and the active
rheostat resistance R,,. The voltage Oa due to residual magnetism appears across
the field circuit and causes a field current Ob to flow. But in accordance with the
a
magnetization curve this field current aids the residual flux and thereby produces
larger induced emf of value bc. In turn this increased emf causes an even larger
of
field current, which creates more flux for a larger emf, and so on. This process
the induced emf produces just enough field cur-
voltage build-up continues until
order for
rent to sustain it. This corresponds to point f in Fig. 7-5. Note that in
three conditions must be satisfied: (1) There
the build-up process to take place
field winding mmf must act to aid this residual
must be a residual flux.t (2) The
value. The
flux. (3) The total field-circuit resistance must be less than the critical
which makes the resistance line coincide with
critical field resistance is that value
the linear portion of the saturation curve.
In addition to the shunt generator there are the other generator types—the
compound generator and the series generator. A compound generator is a shunt
generator equipped with a series winding. The series winding is a coil of compara-
In such cases the field
+When the iron is in its original virgin state, the residual flux is negligible.
machine a usable level of residual flux. This
winding must receive an initial excitation to give the
process is called flashing the field.
310 Direct-Current Generators Chap. 7

Output
terminals

Series field

Cie soot Shunt field

kA ofA)

© 2 POS il
(ata weelei
YI PSs DAS |

F Field
Sa rheostat
=O
Output
terminals
©

(b)
Figure 7-6 Cumulatively compounded generator: (a) schematic diagram; (b) de-
tailed diagram depicting the location of series and shunt windings on the pole
structure.

tively few turns wound on the same magnetic axis as the field winding and con-
nected in series with the armature winding. Consult Fig. 7-6. Because the series
field winding must be capable of carrying the full armature current, its cross-
sectional area is much greater than that used in the shunt field winding. The
purpose of the series field is to provide additional air-gap flux as increased arma-
ture current flows, in order to neutralize the armature winding resistance drop as
well as the voltage drops occurring in the feeder wires leading to the load. In such
cases the generator is usually referred to as a cumulatively compounded generator
because the series field aids the shunt field flux. If the series field connection were
 Sec. 7-3 Demagnetizing Effect of the Armature Winding mmf 311

reversed, its flux would oppose the shunt field flux in which case the configuration
is referred to as a differentially compounded generator.
By imposing the appropriate constraint on the connection diagram depicted
in Fig. 7-6(a), we can identify any one of the three modes of operation of the dc
generator. Thus in addition to the armature winding we have the following:

Compound generator: shunt and series field windings

Shunt generator: shunt field winding

Series generator: series field winding

Since the series generator is rarely used except for special applications, all further
treatment of generators is confined to the shunt and compound modes.

7-3 DEMAGNETIZING EFFECT OF THE ARMATURE WINDING MMF

The armature winding mmf produces two adverse effects: it causes a net reduction
in the field flux, and it makes it more difficult for the armature current in the coils
to commutate. We turn our attention to the first of these here; the problem of
commutation is treated in Sec. 7-6.
A glance at Fig. 7-6(b) shows that the flow of current in the armature winding
a
produces an ampere-conductor distribution that makes the armature behave like
brush axis and in a
solenoid. The associated mmf is directed downward along the
n of the iron
position of quadrature to the field axis. If there were no saturatio
present, the effect of this cross-armature mmf would be merely to cause a distor-
around
tion of the flux. However, in most practical machines operation occurs
ization takes
the knee of the magnetization curve and as a result a net demagnet
illustrate d in Fig.
place. To understand why this happens, consider the situation
(the latter not
7-7 which is that of a shunt generator delivering current to a load
This is a valid
shown). The armature mmf wave is assumed to be triangular.
large; other-
assumption whenever the number of surface armature conductors is
produced by this
wise it would be a stepped trapezoid. The flux-density wave
of the interpola r
armature mmf is saddle-shaped because of the high reluctance
the B, curves. It
space. The resultant air-gap flux density is the sum of the By and
in the trailing half
should be noted that the armature mmf acts to increase the flux
it causes a diminutio n of flux in the leading half. If the increase
of the pole, while
leading half, no net
of flux in the trailing half is equal to the decrease in the
half is less than the
reduction in field flux takes place. If the increase in the trailing
In Fig. 7-7 the cross-
decrease in the leading half, demagnetization occurs.
in the leading
hatched area represents the amount of demagnetization that occurs
amount of magnetiza-
half of the pole, the double cross-hatched area denotes the
a net demagnetization
tion caused by the cross-magnetizing armature mmf. When
former area exceeds the latter. Because machines are
of the field flux results, the
312 Direct-Current Generators Chap. 7

B,, flux density produced

by armature mmf

) © © © OWD OO GROOM Rotor direction

of rotation
——_

By, field flux density

Figure 7-7 Effect of armature mmf on the air-gap flux density. Demagnetization
occurs if the cross-hatched area to the left of the pole axis is greater than the
double-cross-hatched area to the right of the pole axis.

operated in a partially saturated state, there often exists some degree of magneti-
zation.
A measure of the amount of this demagnetization expressed in equivalent
field amperes can be obtained from a knowledge of the magnetization curve and
appropriate design data. The procedure is graphically illustrated in Fig. 7-8. The
air-gap flux density per pole produced by the field winding acting alone is denoted
by Br. Let p be defined as the ratio of the pole face width, b,, to the width of a
pole pitch, 7,. That is,

cas>»
Then the amount of armature mmf acting at either pole tip is
Lola
armature mmf at pole tip = p Pet (7-1)

where Z denotes the total armature conductors, p is the number of poles and a
denotes the number of armature paths. The factor 2 converts conductors to
turns. To express Eq. (7-1) in equivalent field amperes, it is necessary to divide
by the number of turns per pole of the field winding, N;. This quantity is shown
on the abscissa axis of Fig. 7-8 with some exaggerations for the sake of clarity.
The value of the resultant air-gap flux density at the center of the pole is Br.
However, at the trailing pole tip the magnetizing action of the armature mmf raises
this quantity to B, (as is also indicated in Fig. 7-7), whereas at the leading pole tip
it drops to B,.. Area acd is a measure of the amount of demagnetization that
occurs in the leading half of a pole; area abe is a measure of the amount of
magnetization that occurs in the trailing half of a pole. For machines that exhibit a
measurable degree of saturation area acd is greater than area abe. Hence demag-
 Sec. 7-4 External Generator Characteristics 313

B,

By

density:
flux

Figure 7-8 Illustrating the demagnetiz-

ing mmf by means of the magnetizing
O I; Field, amp curve.

netization occurs. The value of this demagnetization is found by displacing line

dae to that position d'a’e’ such that area a'd'c equals area a'be'. The quantity
aa' projected on the field current axis yields the desired result. Thus
Tg = a'a" measured in field amperes (7-2)

where If, is used to denote the demagnetizing effect of the armature mmf in
equivalent field amperes.

7-4 EXTERNAL GENERATOR CHARACTERISTICS

voltage
The external characteristic of a generator is described by a plot of terminal
speed held fixed. A generato r that is equippe d with
versus load current with the
a single field winding can be operated either in the
an armature winding and just
the shunt excited mode. The external characte ristic
separately excited mode or
of compari-
for each mode is depicted in Fig. 7-9 with solid lines. For the purpose
adjusted so that the terminal voltage at no-load is
son, each mode is assumed to be
then allowed to increase to its rated value, there
the same. As the load current is
with the situation being more pronoun ced for
occurs a fall-off of terminal voltage
separate ly excited mode. Of course, in both
the shunt connection than for the
flow of the increasing
cases there is a drop in voltage that is associated with the
circuit resistanc e. This is indicated in Fig.
armature current through the armature
r, for the separate ly excited mode of opera-
7-9 by the broken-line curve. Moreove
from the broken-l ine characte ristic that
tion there is an additional drop in voltage
effect of the armature current as de-
comes about because of the demagnetizing
two effects —the armature circuit resis-
scribed in the preceding section. These
e the resultant external charac-
tance voltage drop and armature reaction—defin
teristic of the separately excited generator.
 314 Direct-Current Generators Chap. 7

SS |
Caused by /,A, drop
os
Separate excited
Caused by armature reaction

Caused by reduced field winding voltage

voltage
Terminal

Load current

Figure 7-9 External characteristics of dc generator operated in shunt and separately ex-
cited modes. Speed is constant.

When the dc generator is operated as a self-excited shunt generator, then in

addition to the two factors already identified, there is a third factor that is respon-
sible for a further reduction of terminal voltage at a given load current. Keep in
mind that with the shunt connection the supply voltage for the field current is in
fact the very terminal voltage of the generator which is undergoing a decrease with
increasing load demand. Consequently, this reduced voltage causes a corre-
sponding reduction in field current and thereby in the air-gap flux which in turn
causes a further reduction in terminal voltage. Accordingly, the external charac-
teristic of the shunt generator assumes a position that is below the one obtained
for the separately excited mode of operation.
A dc generator that is equipped with a series field winding in addition to the
shunt field can be designed to counter the effects that cause the drooping charac-
teristics of Fig. 7-9. By supplying the series field winding with the right number of
turns, it is possible to exactly neutralize all three effects that cause the terminal
voltage of the shunt motor to decrease with load current. When this happens the
compound generator is said to be flat compounded. This condition is illustrated in
Fig. 7-10. If the machine is designed to have a number-of turns fewer ‘han the
number required for flat compounding, .the generator is said to be undercom-
pounded. In such a case only a partial compensation of the voltage drop of the
shunt mode is achieved. However, more commonly, the number of series turns
designed into the series winding exceeds that required for flat compounding. This
gives rise to an external characteristic that is called overcompounded. It is helpful
to note that an overcompounded generator offers the additional feature of permit-
ting the neutralization of the effect of the voltage drop incurred in the feeder lines
Sec. 7-4 — External Generator Characteristics 315

Overcompounded

100
Flat compounded
Undercompounded
~ Separately excited
Shunt

rated
%
voltage
Terminal
as

0 100
Load current as % rated

Figure 7-10 External characteristics of various dc generator types at constant

speed.

that are used to connect the generator to the load. This is an especially desirable
feature when long feeder lines are necessary. The voltage available for the feeder
line drop at full load in Fig. 7-10 is represented by the distance ab.
In the interest of providing some flexibility in dealing with situations where
different feeder lengths are encountered, a machine designed with a large number
parallel
of series field turns can be used in conjunction with a resistor connected in
with the series winding. By proper adjustmen t of this resistor the contribut ion of
be controlle d by diverting more or less
the series field mmf to the air-gap flux can
resistor. This resistor is known as the di-
of the armature current through the
verter. ;
various
Analternative way of representing the external characteristics of the
7-11. Here adjustme nts are made in each
types of dc generators is depicted in Fig.
voltage when rated load current is
mode of operation to establish rated terminal
to diminish to zero at constant speed,
delivered. As the load current is allowed
of the no-load
the external characteristics take on the positions shown. The value
r is the smallest while that of the
terminal voltage of the overcompounded generato
shunt generator is the largest (why?).

Voltage Control. In typical applications it is desirable to keep the termi-

regardless of the load
nal voltage at the load terminals fixed at the rated value
result: (1) adjust the driving speed
condition. There are two ways to achieve this
rheosta t that is placed in series with
of the generator prime mover; (2) adjust the
a flat-co mpounde d generat or and
the field winding (see Fig. 7-6). Of course, for
316 Direct-Current Generators Chap. 7

Separately excited

= oOoO

Overcompounded

rated
%
voltage
Terminal
as

Figure 7-11 External generator charac-

teristics drawn with full-load current at
0 100 rated terminal voltage as the common
Load current as % rated operating point. Speed is constant.

negligible feeder line voltage drop, no adjustment is needed if operation is con-

fined to no-load or to full load. However, for operation at half-load from no-load,
it is obvious from the external characteristic appearing in Fig. 7-10 that either a
drop in speed or a reduced field current adjustment is required to keep the terminal
at the rated value. On the other hand, if the operating generator mode is that of
undercompounding, separate excitation, or shunt, maintenance of constant termi-
nal voltage calls for an increase in speed or an increase in field current. A combi-
nation of the two can also be employed. Of course, the biggest adjustments must
be made for the shunt mode of operation.
Example 7-1
A 50-Kw, 500-V cumulatively compounded generator equipped with a diverter deliv-
ers rated load current to a load at rated voltage. This condition is achieved by proper
adjustment of the diverter resistance and field rheostat. Refer to Fig. 7-12 which
shows the shunt field placed directly across the load terminals (called the long-shunt

50 kW load

Figure 7-12 Diagram for Example 7-1.

Sec. 7-4 — External Generator Characteristics 317

connection).t The measured currents through the field winding and diverter are 4 A
and 21 A, respectively. The armature and series field winding resistances are 0.1 and
0.02, respectively.

(a) What is the series field current?

(b) Find the diverter resistance.
(c) Compute the generated armature voltage.

Solution (a) The armature current is

I, =I, += 0.5
50.
~et4= 104A
I, =z — 14 = 104 — 21 = 83.A

where J, denotes the diverter current.

(b) The voltage drop across the diverter is the same as that across the series
field winding. Hence,

IgRq = 83(0.02) = 1.66 V

P06. 166”
doting Oat i = 0.0790

(c) By Kirchhoff’s voltage law in the circuit of Fig. 7-12

E, = V, + IgRa + TaRa = 500 + 104(0.1) + 1.66 = 512 V

Derivation of External Characteristic of Self-Excited

Shunt Generator

The precise way in which the factors shown in Fig. 7-9 combine to yield a stable
the
operating point can be understood in the graphical representation appearing on
left side of Fig. 7-13. Assume that the following informatio n is available: armature
circuit resistance, field circuit resistance, magnetization curve, and the equivalent
mmf
field current that represents the demagnetization caused by the armature
from the design data or an appropria te
(i.e., Iya). The last item is available either
line with the nonlinear magnetiza tion
test. The intersection of the field resistance
ding to the
curve shows that at no-load the equilibrium field current is [yo correspon
condition is such that the armature
no-load terminal voltage Vy. When the load
that the terminal voltage has the
current is J,;, the external characteristic shows
value V;. Extension of a horizontal line at the V; level results in a point of
projection of O'
intersection with the field resistance line identified as O’. The
is now Ip1.
onto the field axis discloses that the current flowing in the field circuit
reduction occurs
A comparison with the value at no-load makes it clear that a
equal to
Tro ae Tr (7-3)
Aly >

in Fig. 7-6 instead of S2.

+There is also a short-shunt connection where F1 is joined to S1
318 Direct-Current Generators Chap. 7

Field resistance

Magnetization
curve

fll
Tio Tt Tenet Oa To
Field current Armature current

Figure 7-13 Derivation of the external characteristic of a self-excited shunt gen-

erator.

To find the net field current, which in turn is responsible for producing the air-gap
flux that provides the induced armature voltage E,, it is necessary to subtract
from I; the demagnetization effect Ij of the armature mmf. Thus

Tpnet = [pi — Tye (7-4)

Or expressing this more generally in terms of the field current existing at no-load,
we can write

Tpnet = Ivo — Aly — La (7-5)

The induced armature voltage for the specified load condition E,; is then
found on the nonlinear open-circuit characteristic corresponding to Ifner. This is
represented in Fig. 7-13 by point b. Upon subtracting from E,; the armature
circuit resistance drop /,;R,, the terminal voltage V; results. The J,;R, drop is
denoted by line ab in Fig. 7-13. It should be noted that there is only one position
in the vicinity of the upper portion of the magnetization curve where triangle O'ab
will probably fit between the field resistance line and the magnetization curve. In
fact it is this position that defines the equilibrium point that is being described.
On the basis of the foregoing discussion it can be seen that the external
characteristic can be derived from the magnetization curve, the field resistance
line, and knowledge of J;; and R,. To find the terminal voltage corresponding to
any specified value of J,, first construct the triangle Oab at the origin as shown in
Fig. 7-13. Side ab is equal to J, R, and side Oa is equal to the amount of demagne-
tization associated with the specified J,. Then draw a line through point b parallel
Sec. 7-4 External Generator Characteristics 319

to the field resistance line and intersecting the magnetization curve in two places.
Triangle Oab may then be translated to position O'ab taking care to keep O' on
the field resistance line and b on the magnetization curve. The terminal voltage is
the ordinate value corresponding to point a. By repeating this procedure for
various values of armature current the complete characteristic is obtained.

Derivation of External Characteristic of Cumulatively

Compounded Generator

A similar procedure may be used to describe the external characteristic of the self-
excited cumulatively compounded generator. Very often the contribution of the
series field is responsible for a rising external characteristic such as the one
depicted on the right side of Fig. 7-14. This characteristic can also be explained in
terms of the same information that is used for the shunt generator, but in addition
information about the series field contribution is required, usually expressed in
terms of an equivalent field current. The last quantity is readily written as

Uys = Hy
Ns
(7-6)
the
where /;, is the field current equivalent of the series field mmf, N, denotes
number of turns of the series field per pole, N; denotes the number of shunt field
turns per pole, and J, is the current flowing through the series field winding. Often
e
I, is equal to the armature current, but sometimes a portion of the armatur

Field resistance

a x

Magnetization
curve j

Ip, linet ly lio

Armature current

ively com-
Figure 7-14 Derivation of the external characteristic of a cumulat
pounded generator.
 320 Direct-Current Generators Chap. 7

current is diverted through a low resistance shunt (called a diverter) placed across
the series field winding in order to limit its contribution.
At no-load the operating point is on the magnetization curve. The no-load
terminal voltage is identified as Vy and its corresponding field current as Jj). It is
helpful to think of this point as identical with the no-load point of the self-excited
shunt generator. Next consider that load current is allowed to flow so that the
armature current becomes J/,,, corresponding to which the terminal voltage be-
comes V, as shown in Fig. 7-14. With the assumption that (V,, J.) is the new
equilibrium point, it follows that the field current is greater by the amount Al;
which yields a total current in the field winding of value J;; = Ij + Aly. The plus
sign for Al; always appears whenever the external characteristic is a rising one.
The current J, is associated with point O’ on the field resistance line.
To the shunt field current must now be added the effect of the series field as
expressed by Eq. (7-6). This yields a total main axis excitation expressed by
N,
Tn = Ip ar a= Ile = Ip ae Ie (7-7)

The net field current, which is responsible for the armature induced emf, is found
by subtracting the demagnetizing effect of the armature mmf. Thus
Dene =p ada (7-8)
Corresponding to /;pe is the induced armature voltage E,,. The terminal voltage
is then found by removing the armature circuit resistance drop from E,,, which
yields V;. The /,,(R, + R,) drop is represented by line ab in Fig. 7-14.
The external characteristic can be derived by following a procedure similar
to that described for the shunt generator. Construct triangle Oab at the origin.
Line ab is equal to the armature circuit resistance drop for the specified armature
current. Line Oa is equal to the resultant of the magnetizing effect of the series
field and the demagnetizing effect of the armature mnf, i.e.,
ON tia (7-9)
Then through b draw a line parallel to the field resistance line intersecting the
magnetization curve at point b. Dropping down from b by the /,,(R, + R,) drop
gives the desired terminal voltage V,;. A comparison with the procedure em-
ployed for the shunt generator discloses that the triangle Oab is placed on the left
side of the origin rather than the right side. This is dictated by the fact that the
base of the triangle for the cumulatively compounded generator represents a net
magnetization rather than a net demagnetization.

7-5 COMPUTATION OF GENERATOR PERFORMANCE

Appearing in Fig. 7-15 is the equivalent circuit of the compound generator. The
armature winding is replaced by a source voltage having the induced emf E, and a
resistance R,, which represents the armature circuit resistance. The armature
Sec. 7-5 Computation of Generator Performance 321

E>
[teed Figure 7-15 Equivalent circuit of the
compound dc generator.

circuit resistance includes the armature winding resistance, interpole winding

resistance, compensating winding resistance, plus the effect of the voltage drop in
the carbon brushes. The series field is replaced by its resistance R,, and the same
for the shunt field.
The governing equations for determining the performance are the following:

(7-10)

PZ
5 wy =a K7Pl,
ie Ol, (7-11)

V, = Ea i. T(Ra x R;) (7-12)

I, = I, =f Ir (7-13)

These include the two basic relationships for induced voltage and electromagnetic
torque as developed in Sec. 3-4. Equations (7-12) and (7-13) are merely state-
ments of Kirchhoff’s voltage and current laws as they apply to the equivalent
circuit of Fig. 7-15. For the compound generator the expression for the air-gap
flux must include the effect of the series field as well as the shunt field. Thus on
the assumption of negligible saturation,

® = D5), ar ®, (7-14)

where ®,, denotes the flux produced by the shunt field winding and ®, denotes the
flux caused by the series field winding. For the shunt generator, of course, ®, iS
zero in Eq. (7-14) and so too is R, in Eq. (7-12).
The power-flow diagram for the dc generator is depicted in Fig. 7-16. Note
the similarity it bears to that of the synchronous generator. The field winding loss
is included in the power flow directly because it is assumed that the generator iS
self-excited. Of course, if the field winding is separately excited, the field losses
are not supplied from the prime mover and so must be handled separately.
Moreover, note that the field losses are represented in terms of the product of the
field terminal voltage and the field current. This ensures that the field winding
rheostat losses are included as well.
The quantity P,., that appears in the power flow diagram represents the iron
losses that arise from rotation of the armature (rotor) structure as well as friction
and windage losses. The iron losses associated with rotation include hysteresis
322 Direct-Current Generators Chap. 7

P; (mechanical) Useful power

- Se =<=<= Power losses
2
Pot Pn = Egla
=Tw = TSH
es ad ad
o* es

12(R, +R,) y= VI (electrical)

V, 1
i Figure 7-16 Power-flow diagram of the
> losses dc generator.

and eddy current losses. Although the slotted rotor structure, which accommo-
dates the armature winding, is composed of laminations, eddy currents are caused
to flow in the individual laminations as the rotor iron is driven alternately beneath
poles of N polarity and S polarity. This very same alternation of the flux in the
iron is responsible for the hysteresis losses represented by the generated hystere-
sis loops. Both losses are functions of the maximum flux density and the speed.
Moreover, for machines with open slots, there is an additional component of Pro
that accounts for the losses incurred in the pole faces because of the ripple that is
made to appear in the field distribution curve by the slot effect. The major part of
the frictional losses is contributed by the friction of the brushes on the commuta-
tor segments. Also included but to a lesser degree are the machine bearing
losses. Typically, the total of these losses represents 3% to 5% of the machine
rating. However, it is useful to keep in mind that the per-unit value of all machine
losses diminishes as the machine rating increases.
The armature winding copper losses are essentially losses that occur under
load conditions. As such these losses account for all armature winding copper
losses, including the skin effect that takes place in the individual coils of the
armature winding, the electrical loss in the carbon brushes, (which is frequently
taken to be 2/, on the assumption that the voltage in the brushes is about 2 V), and
the short-circuit loss in those coils that undergo commutation. For convenience
all of these are assumed to be represented in the armature winding circuit resis-
tance. Thus the parameter R, is not to be treated as merely the dc armature
winding resistance.
The per-unit value of the armature circuit resistance indicates the ratio of the
armature circuit losses expressed in units of power to the rated power output of
the generator. If we denote the rated armature current as the base current, J;, the
rated armature voltage as the base voltage Vg, and the rated output power as the
base power Pz, then

IpRa pat IpRa = K; en Ra

p-u Ra =
Vale Ve 7 Va/Ip Bu Rp

where Rz is the base resistance and is determined as the ratio of the base voltage
to the base current. Thus a value of p-u R, = 0.05 means that 5% of the rated
output of the generator is consumed as losses in the armature circuit when rated
Sec. 7-5 Computation of Generator Performance 323

current is being delivered. This per-unit value takes on larger values for genera-
tors of small ratings. Similarly, for large generators the p-u R, diminishes.
Example 7-2
The magnetization curve of a 10-kW, 250-V, dc self-excited shunt generator driven at
1000 rpm is shown in Fig. 7-17. Each vertical division represents 20 V and each
horizontal unit represents 0.2 A. The armature circuit resistance is 0.15 Q and the
field current is 1.64 A when the terminal voltage is 250 V at rated load. Also, the
rotational losses are known to be equal to 540 W. Find at rated load (a) the armature
induced emf, (b) the developed torque, (c) the efficiency. Assume constant speed
operation.

Solution (a) With the generator delivering rated load current it follows that

l= 10,000.
say _oA

Hence from Eq. (7-13) the armature current is

I, = I, + I = 40 + 1.64 = 41.64A

Papa dl:32a) olde a a

/

HE Ped baglessESHEE
pee
EWE .
=

af NEE)
NES
ea Rielle Pa:
=

aoe Magnetization curve

BPsBAAN
Bites
Hy
ESied
=eEs
See
i Hires

Sy nm b re) o N @ ico) ro) iS) rer)

ie}

Figure 7-17 Magnetization curve for Example 7-2.

324 Direct-Current Generators Chap. 7

The induced armature voltage then follows from Eq. (7-12):

E, = V, + I,Rq = 250 + 41.64(0.15) = 256.25 V (7-15)

(b) To determine the developed torque it is first necessary to compute the
electromagnetic power E,/,. Thus

Eyl = 256.25(41.64) = 10,670 W

The power-flow diagram reveals that this power may also be found from

Eyl, = Po + 12Ry + Vilp = 10 KW + (41.64)70.15 + 250(1.64)

10 kW + 260 + 410 = 10,670 W (7-16)
which checks with the preceding calculation. Hence the developed torque is

mad
(7-17)
3 10,670
ee,
Bala
ie Eoeeh 27100060
(c) The efficiency is found from Eq. (4-27), which for the dc generator takes the
form

> losses
isos rte et pe (7-18)
In this case
Y losses = Pro + 12Ra + Vilr
= 540 + 260 + 410 = 1210 W
Hence

Lage © 1210 oy els ee ¥

mee joe10 + 340 | Fina are Bee
The efficiency is therefore 89.2%.

Voltage Regulation

The voltage regulation of a dc generator is defined as the change in terminal

voltage from full load to no-load expressed as a percentage of the full-load volt-
age. In mathematical form we can write

VR _ no-load voltage — full-load voltage . Via vn

100 = 100
full-load voltage Vo V;

(7-19)
Here the full-load voltage, Vj, is the rated terminal voltage (V,) and Vai is the
voltage measured when the load is disconnected from the generator. The latter
quantity is identified as V, in Figs. 7-13 and 7-14.
 Sec. 7-6 Commutation 325

Example 7-3
Find the voltage regulation for the load condition described in Example 7-2. Assume
that the equivalent field current that represents the demagnetizing effect of the arma-
ture mmf at rated load is 0.2 A.

Solution We resort to a graphical solution employing Fig. 7-17. On the magnetiza-

tion curve of Fig. 7-17 locate point b (see Fig. 7-13) at a level of E, = 256.25 V. Along
the ordinate line from b mark off the armature circuit resistance drop of 6.25 V.
Next, along the horizontal from a mark off the value of the equivalent field current of
the demagnetizing effect of the armature mmf of value 0.2 A. This locates point o’
(as identified in Fig. 7-13). Finally, draw the field resistance line that passes through
o' from the origin. The intersection of this line with the magnetization curve locates
the no-load terminal voltage of V, = 268 V. Hence,

268 — 250 |
VRo= 750 00 = 7.2%

7-6 COMMUTATION

Good commutation is indispensable for the satisfactory operation of dc ma-

chines. This matter is so important that it takes precedence in design consider-
ations over the problem of heat dissipation. Without proper commutation, deteri-
oration of the carbon brushes and the copper commutator segments can rapidly
set in. This causes severe sparking and thereby renders the machine useless.
Even such an important characteristic as the maximum torque is limited by com-
mutation rather than by heating.
Commutation is concerned with providing a suitable transition of the arma-
beneath a
ture current in a coil from a value +/,/a to a value of —J,/a as it passes
brush from one pole to an adjacent pole. A glance at Fig. 7-18(a) makes it evident
beneath the brushes and occupies the position now
that when coil 1-10 passes
being occupied by coil 15-8, the current in the coil will have reversed completely
1-10
as indicated by the cross-dot notation. Commutation of the current from coil
B, makes contact with commutato r segment I as
begins to take place when brush
that when this happens coil 1-10 is actually short-
shown in Fig. 7-18(b). Note
are
circuited by brush B, through segments 8 and I. All commutator segments
by a suitable material. A similar situation occurs at
insulated from one another
short-circu ited. Because of this and in the inter-
brush B> where coil 2—9 is being
in turn can cause
est of preventing large circulating currents in these coils, which
currents are interrupted upon leaving the brush,
severe sparking as the circulating
should be free of voltage sources. The time
the coils undergoing commutation
in Fig. 7-18(b) is 22.5° later than that shown in Fig. 7-18(a).
instant depicted
space where the flux
Hence coils 1-10 and 2-9 find themselves in the interpolar
is zero. Accordingl y, no Blu voltage is induced
density produced by the field mmf
326 Direct-Current Generators Chap. 7

cad

ee

Figure 7-18 Illustrating the commutation problem: (a) coil 1-10 must undergo a
complete current reversal as it advances counterclockwise to the left side of the
brush axis; (b) for the position shown, coil 1-10 is short-circuited through seg-
ments 8 and | by B,, and coil 2-9 is short-circuited through segments 4 and 5
by B, .

in these coils from this source. However, the presence of the armature mmf
causes the resultant flux density to have a finite value in the interpolar space as is
evident from Fig. 7-7. Since the polarity ofthis flux is the same as that of the main
field pole which the short-circuited coils are leaving, an emf is induced that acts to
maintain current flow in the same direction as exists before commutation begins.
Therefore this induced emf due to the armature mmf acts to hinder commutation:
 Sec. 7-6 Commutation 327

it becomes more difficult for the coil current to reverse by the time it leaves the
brush.

Types of Commutation

There are two types of commutation possible: resistance commutation and volt-
age commutation. Moreover, the latter may be divided into the two classes of
undercommutation and overcommutation. The distinction between these types is
depicted by the various commutation curves of Fig. 7-19.
Curve a represents linear commutation, and it occurs only when the brush
contact resistance is the exclusive factor influencing commutation. All induced
emf’s from whatever sources must add to zero and the coil resistance must be
negligible compared with the brush contact resistance. Under these conditions
the current in the coil changes uniformly during the commutation period T,. The
mechanism of commutation by this method is illustrated in Fig. 7-20. Figure
is
7-20(a) depicts the situation just before commutation of the current in coil 2—9
the total
about to occur. Brush B> is resting on commutator segment 4, and of
path
armature current J, flowing through this brush half comes from one armature
involving coil 16-7 and the remaining half comes from the second path involving
brush is
coil 2-9. The direction of rotation of the armature with respect to the
of the brush
towards the left. Depicted in Fig. 7-20(b) is the situation where 25%
of that part of
is in contact with commutator segment 5. The contact resistance
as the resistance
the brush surface in contact with segment 5 is three times as large
coming from
of the brush in contact with segment 4. Hence division of the current
the current
the two armature paths takes place in accordance with this ratio. Thus
of 3: 1 so that the
I,/2 coming from the right side divides at junction a in the ratio
$(/,/2) or I,/8. The
current passing to B, from segment 5 from this source is
the current
portion passing to B, through segment 4 is #(/,/2) or sla. Similarly,
so that 3(/,/2) or la
[,/2 originating from the armature path on the left side splits,
b to a to segment 4 and
flows to B, through segment 4 and 4(/,/2) or 17 flows from

2-9
Current
coil
in
Time

d
Time somewhere >3e.
between brushes. Figure 7-19 Commutation curve of
coil 2-9 in Fig. 7-18. During the com-
mutation period the curve is a plot of
Leading Trailing the circulating current in the short-
brush tip brush tip circuited coil undergoing commutation.
328 Direct-Current Generators Chap. 7

I,/2 I,/2
ee b eats a Tav2

Wi
(a) (b)
Figure 7-20 Illustrating linear commutation in coil 2-9: (a) conditions just before commuta-
tion begins; (b) conditions at a point one-quarter through the commutation period; (c) condi-
tions at a point halfway through the commutation period.

thence to B,. The net current that flows from segment 4 to B> is therefore 3/, + 21,
or 3/,, whereas the net current to B from segment 5 is (/,/8) + (/,/8) or I,/4. The
resultant current in coil 2—9 flows from a to b and is equal to 3/, — $J, or I,/4.
Consequently, one-quarter through the period of commutation the current in coil
2-9 has changed from /,/2 to I,/4. Keep in mind that the commutation curve is a
plot of the current in the short-circuited coil as a function of time. Appearing in
Fig. 7-20(c) is the distribution of currents that prevail at a time midway through
the commutation period. Note that now the current in the short-circuited coil is
zero.
Curve b in Fig. 7-19 represents resistance commutation. It is the curve that
results when the resistance of the short-circuited coil is not negligible compared
with the brush contact resistance. Note that at the leading brush tip the slope of
curve b is steeper than that of curve a, which indicates that the coil resistance
enhances the current reversal. This is apparent from an examination of Fig.
7-20(b) where it is seen that at junction a the presence of coil resistance means that
more current flows to segment 5. However, at the trailing brush tip (or during the
second half of the commutation period) the coil resistance serves to retard the
current reversal. The current density, which is represented by the slope of the
commutation curve, is therefore higher at both the leading and trailing brush edges
for resistance commutation.
The commutation curve denoted as c in Fig. 7-19 is described as under
commutation. It is a form of voltage commutation in which current reversal has
been delayed by the action of the emf of self-induction in the short-circuited coil
or of the Blu voltage induced in the short-circuited coil by the flux produced by the
armature mmf or of a combination of both. A decelerated commutation of this
kind is characterized by high current density at the trailing brush tip, which can
promote a deterioration of the brush material and also bring about undesirable
chemical changes in the copper segments of the commutator.
A scheme that is frequently used to neutralize this emf of self-induction and
Sec. 7-6 Commutation 329

the effect of the armature mmf is the inclusion of small poles located on the brush
axis and energized by a coil carrying the armature current. Because of their
location and function, these poles are called interpoles or commutating poles.
The use of armature-current excitation is dictated by the fact that both the emf of
self-induction in the short-circuited coil and the flux in the brush axis produced by
the armature mmf are proportional to armature current. For generator action the
polarity of the interpole must be the same as that of the field pole into which the
coil is moving. Figure 7-21 depicts the situation for a two-pole machine. It is
helpful to compare the resultant flux-density curve here with that appearing in Fig.
7-7. Note that the value of the flux density on the interpolar axis in Fig. 7-21 is
opposite to that occurring in Fig. 7-7. If the magnitude of the flux density pro-
duced by the interpole is adjusted properly, it can be made to neutralize the emf of
self-induction completely. Of course the effect of the armature mmf has already
been neutralized in the representation of Fig. 7-21 because the resultant flux
density is shown to be of opposite polarity. With such an adjustment the type of
commutation that prevails is linear commutation, provided that the coil resistance
compared to the brush contact resistance is inconsequential. Otherwise resis-
tance commutation takes place.
A considerable improvement in the commutation process can be achieved by
increasing the interpole flux to the point at which, in addition to canceling the
effects of the emf of self-induction and the armature mmf, there is induced a
voltage equal to twice the brush voltage drop, 2/,R,. Under these conditions that
special condition of overcommutation results that leads to a value of zero current
density at the trailing edge of the brush. This situation is denoted by curve d in
Fig. 7-19. Note that the current in the coil undergoing commutation has already
reached its value of —J,/2 before the coil leaves the brush. This results in no
sparking at the trailing edge of the brush and represents the most favorable
condition for good commutation.

Interpole Main field pole

Direction of armature rotation

—\—P

Resultant air gap flux density

a de
Figure 7-21 Influence of interpoles on the resultant flux-density wave of
machine.
330 Direct-Current Generators Chap. 7

Bek
10 fava

Figure 7-22 Illustrating the arc-type

drop
Brush
voltage
contact
O (00 amp/in® —_conduction characteristics of carbon
Current density brushes.

The relationship between the contact voltage drop at a brush and the current
density is shown in Fig. 7-22. For average carbon brushes the contact drop tends
to remain essentially constant at about | volt. This is consistent with the fact that
the conduction in the space between the copper segments and the carbon brushes
occurs in accordance with arc phenomenon.
When dc machines are used in applications that involve severe duty cycles
(e.g., steel mills), swiftly changing loads and high overloads can cause breakdown
of the air space between commutator segments. This breakdown can spread rap-
idly over the entire commutator, thus causing damage to the commutator as well
as a short-circuit on the dc supply lines. Unfortunately, this situation can occur in
spite of the presence of interpoles.
The most effective way to prevent this flashover condition is to neutralize
the armature reaction that causes the severe peaking of the air-gap flux density
that leads to breakdown. This is best achieved by the use of a compensating
winding embedded in slots distributed along the pole faces. By allowing armature
current to flow through this pole-face winding with a polarity opposite to that of

() -F) «—— Interpole winding

(e)
oe Field winding

OOM

Figure 7-23 Schematic diagram of a two-pole dc machine equipped with inter-

pole and compensating windings.
Chap. 7 Problems 331

the armature winding and a suitable number of ampere-turns, there can result
complete cancellation of the armature reaction. The chief disadvantage of this
method of eliminating or reducing the commutation problem is the high expense of
pole-face windings.
Figure 7-23 shows the schematic diagram of a two-pole dc machine equipped
with a compensating winding located in the pole faces. Beneath each pole face the
current carried by the compensating winding flows in the opposite direction to that
which flows in the armature winding.

PROBLEMS

7-1. The rotor of a two-pole de machine has 12 slots. The wave-wound, full-pitch arma-
a
ture winding has two coil sides per slot. There are 12 commutator segments. Show
the
plan view of the winding layout. Clearly identify the position of the brushes on
in
commutator and relative to the poles. Also show the direction of the induced emf
the coil sides for a counterclockwise rotation of the rotor.
having
7-2. The rotor of a two-pole dc machine carries a wave-wound armature winding
two coil sides per slot in its 12 slots. The winding is full-pitch.
between
(a) Draw the schematic diagram that shows the two armature paths existing
dc
the two brushes. Represent each suitably marked coil voltage by small
that instant when slots |
sources with properly identified polarities. Consider
and 7 are located exactly between the poles.
density in the
(b) Directing attention solely to the fundamental component of the flux
air gap, compute the voltage value appearing at the brushes.
r axis by 15°.
7-3. Repeat Prob. 7-2 for the case where slot 1 lies to the left of the interpola
the variation of the armature induced voltage
7-4. For the machine of Prob. 7-2 draw
as a function of w,t where w, is the angular rotor velocity.
measured at the brushes
curve, show the
Dealing only with the fundamental component of the flux density
total. Compute the maximum and
variation of each coil voltage as well as that of the
minimum values of the coil and total voltages .
ations: two poles, six
. Repeat Prob. 7-1 for a de machine with the following specific
winding having two coil sides per slot.
slots, full-pitch, wave-wound armature
of the emf induced in each coil
7-6. For the machine of Prob. 7-1 sketch the variation
where w, is the rotor angular
voltage as well as between brushes as a function of w,f,
nt of the air-gap flux
velocity. Direct attention solely to the fundamental compone
the maximu m and minimu m values of coil and brush voltages.
density. Indicate
7-7. Explain your answer to each part:
the knee of its magnetiza-
(a) Can a separately excited de generator operate below
tion curve?
of its magnetization curve?
(b) Can a de shunt generator operate below the knee
at a constant speed of 1000
7-8. The magnetization curve for a de shunt generator driven
represe nts 20 V; each horizontal
rpm is shown in Fig. 7-17. Each vertical division
division represents 0.2 A.
(a) Compute the critical field resistance.
332 Direct-Current Generators Chap. 7

(b) What voltage is induced by the residual flux of this machine?

(c) What must be the resistance of the field circuit in order that the no-load terminal
voltage be 240 V at a speed of 1000 rpm?
(d) Determine the field current produced by the residual flux voltage when the field
circuit resistance has the value found in part (c).
(e) At what speed must the generator be driven in order that it would fail to build up
when operating with the field circuit resistance of part (c)?
. A de shunt generator has a magnetization curve given by Fig. 7-17, where each
ordinate unit is made equal to 20 V and each abscissa unit is set equal to 0.2 A and the
speed of rotation is 1000 rpm. The field circuit resistance is 156 ©. Determine the
voltage induced between brushes when the generator is operated at a reduced speed
of 800 rpm.
7-10. It is desired to reverse the terminal voltage polarity of a generator that has been
operating properly as a cumulative compound generator. The machine is stopped.
Residual magnetism is reversed by temporarily disconnecting the shunt field and
separately exciting it with reversed current. The connections are then restored ex-
actly as they were before.
(a) Does the terminal voltage build up? Explain.
(b) If answer to part (a) is yes, will the compounding be cumulative or differential?
Explain.
7-11. A two-pole dc machine is equipped with a wave winding having 120 armature con-
ductors. The pole faces span 70% of a pole pitch and the field winding has 1050
turns. This machine operates with a fixed field current of 1.6 A. The magnetization
curve is depicted in Fig. 7-17 when each ordinate unit represents 0.1 Wb and each
abscissa unit represents 0.2 A. At rated load the armature current is 40 A.
(a) Determine the amount of demagnetization of the field flux caused by the arma-
ture mmf.
(b) What is the value of the air-gap flux under load?
(c) What is the value of flux at the leading pole tip?
7-12. In a lap-wound armature winding there are as many brushes as there are poles.
Hence, for a four-pole dc machine it is customary to use four brushes, with each pair
of brushes being responsible for conducting one-half of the total armature current.
Show the distribution of current in the coils undergoing commutation for two in-
stances: just before undergoing commutation and five-eighths through the commuta-
tion period. Assume resistance commutation throughout.
7-13. The no-load characteristic of a 10-kW, 250-V, dc self-excited shunt generator driven
at 1000 rpm is shown in Fig. 7-17. Each vertical unit denotes 20 V, and each horizon-
tal unit represents 0.2 A. This armature circuit resistance is 0.3 ©, and the field
circuit resistance is set at 157 . Determine:
(a) The critical resistance of the shunt field circuit.
(b) The change in voltage from no-load to full-load. Neglect armature
demagnetiza-
tion.
(c) The shunt field current under rated load conditions.
7-14, Repeat Prob. 7-13(b) and (c) for the case where the armature demagnetizing effect Ij,
is known to be 0.2 A expressed in terms of equivalent field current.
7-15. The machine of Prob. 7-13 operates with a field winding resistance of
143 Q.
(a) What is the no-load voltage?
Chap. 7 Problems 333

(b) When the generator delivers power to a specified load, the terminal voltage is
found to drop to 220 V and the corresponding demagnetization effect of the
armature mmf is known to be 0.25 equivalent field ampere. Determine the arma-
ture winding current.
(c) What is the value of the field
current at the load condition of part (b)?
(d) Find the armature winding emf under load.
7-16. The no-load characteristic of a S-kW, 125-V shunt generator driven at 1000 rpm is
shown in Fig. 7-17. Each ordinate unit represents 10 V, and each abscissa unit
represents 0.1 A. The armature circuit resistance is 0.2 ©, and the field circuit
resistance is 157 ©. When the generator delivers rated armature current, the arma-
ture induced emf is 116 V.
(a) Find the terminal voltage.
(b) What is the value of the demagnetization of the armature mmf expressed in
equivalent field amperes?
(c) What is the field current?
(d) How much power is delivered to the load?
7-17. The no-load characteristic of a dc shunt generator driven at a speed of 800 rpm is
approximated in its useful range by the equation

E=
3001;
2+Ir
where E is armature induced emf and /; is the field current. The armature circuit
resistance is 0.1. The field winding resistance is 20. The demagnetizing effect of
armature reaction may be neglected.
(a) The terminal voltage is to be 225 V when the armature current is 150 A. The
generator is driven at 800 rpm. Find the resistance of the field rheostat for this
condition.
(b) Change the field rheostat setting to 10 2. Let the load be disconnected and the
speed reduced to 720 rpm. Find the terminal voltage.
(c) Return to the conditions of part (a). Assume the rotational losses to be 2 kW.
Find the output power, the efficiency, and the shaft input torque from the prime
mover of the generator.
7-18. The no-load characteristic of a dc machine in its useful operating range and driven at
1500 rpm may be approximated by

a
sn?
34+;
ap lle tf
aC Ze— E,
400
is
This machine is operated as a dc shunt generator whose field circuit resistance
50 ©.
rpm,
(a) Compute the no-load terminal voltage when the generator is driven at 1500
indicating any necessary assumptio ns.
to 750
(b) What will be the terminal voltage at no-load when the speed is reduced
rpm? Account for the large difference.
provided
7-19. The magnetization curve of adc shunt generator is represented in Fig. 7-17,
by 0.5 A. The
that each ordinate unit is multiplied by 40 V and each abscissa unit
is 80 (.
armature circuit resistance is 0.4 Q, and the field winding resistance
(a) This machine is made to deliver an armature current of 60 A. Find the resistance
334 Direct-Current Generators Chap. 7

of the field rheostat (in series with the field winding) so that the terminal voltage
will be 520 V at a speed of 1000 rpm.
(b) With the field rheostat set as in part (a) find the terminal voltage at no-load and
1000 rpm.
7-20. A cumulative-compound generator driven at 1000 rpm has the magnetization curve
shown in Fig. 7-17, provided that each unit of the ordinate axis is multiplied by 20 V
and each abscissa unit is multiplied by | A. The armature circuit resistance is
0.06 ©. In addition:
Series turns per pole (N,) = 20
Shunt turns per pole (Ny) = 100
Armature-reaction demagnetizing effect at
I, of 100 A = 600 A-t per pole

The field circuit resistance is 45 Q.

(a) Determine the no-load terminal voltage.
(b) Find the terminal voltage when the load is such that it causes 100 A to flow
through the armature winding.
(c) Find the new value of the field current.
(d) What is the value of the armature induced emf at the specified load.
7-21. Repeat Prob. 7-20 for the case where the field circuit resistance is reduced to 31.5
and the number of series field turns per pole is increased to 40. The armature current
remains at 100 A.
7-22. Refer to the external characteristics depicted in Fig. 7-11.
(a) The shunt generator is operating at rated voltage, rated load current, and rated
speed. Assume the load demand is reduced to one-half the rated value. De-
scribe the adjustments you would make to maintain constant terminal voltage.
(b) Repeat part (a) for the overcompounded operating mode.
7-23. A 10-kW, 250-V generator is equipped with a ratio of shunt-to-series turns of 40: 1.
When operated as a self-excited generator, it takes an increase in field current of
0.3 A to maintain the terminal voltage at rated value at full-load current. The total
field current for this field condition is 1.6 A. Determine whether or not a diverter is
needed when this generator is operated cumulatively compounded. Assume a long-
shunt connection.
7-24. A cumulative-compound generator has the magnetization curve shown in Fig. 7-17,
where each ordinate unit is 20 V and each abscissa unit is 1 A. Determine the
required field circuit resistance that allows the generator to operate at a terminal
voltage of 240 V when the armature current is 100 A. In addition:

Armature circuit resistance drop = 10 V

Series field turns per pole = 25

Shunt field turns per pole = 1000

Armature demagnetizing A-t per pole = 800

 8
Direct-Current
Motors

A dec motor is a dc generator with the power flow reversed. In the de motor
electrical energy is converted to mechanical form. Also, as is the case with the
generator, there are three types of dc motors: the shunt motor, the cumulatively
compounded motor, and the series motor. The compound motor is prefixed with
the word cumulative in order to stress that the connections to the series field
winding are such as to ensure that the series field flux aids the shunt field flux.
The series motor, unlike the series generator, finds wide application, especially
for traction-type loads. Hence due attention is given to this machine in the treat-
ment that follows.

8-1 MOTOR SPEED-TORQUE CHARACTERISTICS

is the
How does the dc motor react to the application of a shaft load? What
it
mechanism by which the dc motor adapts itself to supply to the load the power
in terms
demands? The answers to these questions can be obtained by reasoning
of a pair of simple equations . Initially our remarks are confined to the shunt
line of reasoning applies for the others. For our purposes the
motor, but a similar
two pertinent equations are those for torque and current. Thus

335
336 Direct-Current Motors Chap. 8

(8-1)
and

(8-2)

Note that the last expression results from Kirchhoff’s voltage law and replacing
E,, with Eq. (3-34). With no shaft load applied, the only torque needed is that
which overcomes the rotational losses. Since the shunt motor operates at essen-
tially constant flux, Eq. (8-1) indicates that only a small armature current is
required compared with its rated value to furnish these losses. Equation (8-2)
reveals the manner in which the armature current is made to assume just the right
value. In this expression V,, R,, Kz, and ® are fixed in value. Therefore the
speed is the critical variable. If, for the moment, it is assumed that the speed has
too low a value, then the numerator of Eq. (8-2) takes on an excessive value and in
turn makes /, larger than required. At this point the motor reacts to correct the
situation. The excessive armature current produces a developed torque which
exceeds the opposing torques of friction and windage. In fact this excess serves
as an accelerating torque, which then proceeds to increase the speed to that level
which corresponds to the equilibrium value of armature current. In other words,
the acceleration torque becomes zero only when the speed is at that value which
by Eq. (8-2) yields just the right J, needed to overcome the rotational losses.
Consider next that a load demanding rated torque is suddenly applied to the
motor shaft. Because the developed torque at this instant is only sufficient to
overcome friction and windage and not the load torque, the first reaction is for the
motor to lose speed. In this way, as Eq. (8-2) reveals, the armature current can be
increased so that in turn the electromagnetic torque can increase. As a matter of
fact the applied load torque causes the motor to assume that value of speed which
yields a current sufficient to produce a developed torque to overcome the applied
shaft torque and the frictional torque. Power balance is thereby achieved, be-
cause an equilibrium condition is reached where the electromagnetic power E, I, is
equal to the mechanical power developed, Tw,,.
A comparison of the dc motor with the three-phase induction motor indi-
cates that both are speed-sensitive devices in response to applied shaft loads. An
essential difference, however, is that for the three-phase induction motor devel-
oped torque is adversely influenced by the power-factor angle of the armature
current. Of course no analogous situation prevails in the case of the dc motor.

Shunt Motor

From the foregoing discussion it should be apparent that the speed-torque rela-
tionship of the de motor is an important characteristic. Accordingly, it is useful
to
obtain an expression for speed as a function of the developed torque. We
begin
with Eq. (8-2) by solving for the speed. Thus,
Sec. 8-1 Motor Speed-Torque Characteristics 337

(8-3)

The subscript on ¢ emphasizes that consideration for the moment is confined to

the shunt motor. The torque variable can next be introduced by Eq. (8-1). Hence,
— V, me Ka
n
= Kp@a) KeKr®), | (ey
If it is now assumed that the shunt motor is equipped with a compensating winding
of negligible resistance so that the demagnetizing effect of the armature winding is
essentially neutralized (i.e., Jj = 0), ®,, stays constant and so Eq. (8-4) may be
more conveniently written as
n= kV, — kT = no — koT (8-5)

where

k, = ! rpm/V; kn = ucctiant arpm/N-m (8-6)

KE® 5; KeKr®5n

and
No = kV, = no-load speed (8-7)

Clearly, Eq. (8-5) is a straight line having a negative slope of k2 rpm/N-m and an
intercept of n,. The solid line of Fig. 8-1 shows the plot. The broken-line curve
that also appears is the speed-torque characteristic with the demagnetizing effect
of the armature mmf present; in other words, when the machine is not equipped
with a compensating winding. The reduction in ®,, with armature current causes
some compensation for the drop in the numerator term of Eq. (8-3).

n
E : payee te dsp (reduced by demagnetization
me of armature mmf)
3° den fixed
o
a
”

0 Torque, N-m

with and without a

Figure 8-1 Speed-torque characteristics of a shunt motor
compensating winding.
338 Direct-Current Motors Chap. 8

Example 8-1
A 20-hp, 230-V, 1150-rpm shunt motor has design parameters which lead to values
for k,; and ky in Eqs. (8-6) and (8-7) which are respectively 5.32 rpm/V and 0.4
rpm/N-m.
(a) Find the no-load speed.
(b) Determine the speed when full-load torque of 124 N-m is delivered. As-
sume the motor is equipped with a compensating winding.

Solution (a) From Eq. (8-7) the no-load speed is

Ng = 5.32(230) = 1223.6 rpm

(b) Prom Eq. (8-5)

n = 1223.6 — 0.4(124) = 1150.4 rpm

Speed determination at various load conditions. Speed and load condi-

tions are related through the induced armature voltage equation, E, = Ke@n.
Because E, is determinable from the terminal voltage and the armature-circuit
resistance drop, it follows that at each load condition E, is known. Letting sub-
script | denote conditions at some original operating point where speed, input line
current, field current, and so on are known and subscript 2 denote conditions at a
new operating point, the speed at the changed load condition can be found from
the relationship

oo = te (8-8)
This relationship reduces to a strikingly simple one for motors where the flux per
pole remains essentially invariant with load changes. Such is surely the case with
a shunt motor equipped with a compensating winding or one in which the demag-
netizing effect of the armature mmf is negligible (ja = 0). Then the new speed is
related to the original speed simply by

hy hy (8-9)

The change in motor speed that occurs as the load is changed from rated
value to no-load is an important performance indicator and is called the speed
regulation (SR). Expressed in percent of the full-load value it is written as

no-load speed — full-load speed

Sk = 100 (8-10)
full-load speed

Example 8-2
A 20-hp, 230-V, 1150-rpm shunt motor equipped with a compensating
winding has a
total armature-circuit resistance of 0.188 Q. At rated output
the motor draws a line
current of 74.6 A and a field current of 1.6 A.
Sec. 8-1 Motor Speed-Torque Characteristics 339

(a) Find the speed when the input line current is 38.1 A.
(b) What is the no-load speed if J, = 1.9 A?
(c) Determine the speed regulation.

Solution (a) The armature induced emf at rated load is

Eq = V; — IgRq = 230 — (74.6 — 1.6)0.188 = 216.3 V

n; = 1150 rpm at rated output

At I, = 38.1 A
E,2 = 230 — (38.1'— 1.6)0.188 = 223.14 V

Hence by Eq. (8-10)

nm = E:;en, = 223.14
a 1150 = 1186.4 rpm

(b) Now
Ego = 230 — 0.3(0.188) = 229.9 V

229.3
ny == 716.3 1150 = 1223.3 rpm

(c) From Eq. (8-10)

122375 — li50
SR = DOs ee, ne 100 = 6.29%

Effect of demagnetization on speed. When the shunt motor is not pro-

vided with a compensa ting winding, the demagnet izing effect of the armature mmf
causes a change in the flux per pole so that Eq. (8-9) is not directly applicable.
One way to account for this effect is to find J; in the manner described in Sec. 7-3
mag-
corresponding to specific armature currents. This requires working with the
of the motor. The following example illustrate s the procedure .
netization curve
Example 8-3
The machine of Example 7-3 is used as a shunt motor. The motor’s magnetization
a specific load
curve, which is obtained at 1000 rpm, is displayed in Fig. 7-17. At
A and an armature current of 40 A.
condition this motor draws a field current of 1.6
current on the pole flux is known in terms
The demagnetizing effect of this armature
current to be 0.3 A. The armature-circuit resistance is 0.15 ©.
of an equivalent field
Find the motor speed at the given load condition.
curve to find
Solution The solution strategy in this case is to use the magnetization
the magnetiz ation
the state of the pole flux at the given armature current. Since
curve is found at 1000 rpm, we have
E, = K,®(1000)
could be found by enter-
If there were no demagnetization effect, the quantity, K,;-®,
But to account for the
ing the magnetization curve at a field current of 1.6 A.
340 Direct-Current Motors Chap. 8

demagnetizing effect of the 40-A armature current we must use a value of Ey that
corresponds to a net field current of 1.6-0.3 = 1.3 A. From Fig. 7-17 this yields

E, = 11.8(20) = 236 = K,®(1000)

Hence, the state of pole flux at 40 A armature current is represented by

236
Kg® = 7799 = 0-236
oe (8-11)
8-11

Then at the specified operating point the actual induced armature voltage is

E, = 250 — IR, = 250 — 40(0.15) = 244 V

This value of E, derives from a circuit evaluation. A field evaluation, which must
yield the same result, is based on the result in Eq. (8-11). Hence

E, = K-@®n

244 = (0.236)n
n = 244/0.236 = 1033.9 rpm

Compound Motor

The expression for the speed of a cumulatively compounded motor differs from
Eq. (8-3) by the fact that the net air-gap flux contains the contribution of the series
field. Moreover, the armature circuit also contains the resistance of the series
field for a long-shunt connection. Thus, the analogous expression for the speed at
a given armature and field excitation is

"KG +B) Sri

V, Ta T(Ra at R;)

The summation of fluxes is used on the assumption of linearity for the magnetic
circuit. Although the total resistance drop is larger for the compound motor than
it is for the shunt mode, the strong contribution usually associated with ®, is the
dominant factor. Consequently, for the same armature current as for the shunt
mode of operation, the speed is lower and the speed-torque curve displays a more
pronounced droop as illustrated in Fig. 8-2. A better appreciation of the interplay
of these factors can be gained from a study of the following example.

Example 8-4
The motor of Example 8-2 is also supplied with a series field winding having a
resistance of 0.06 . When operating as a shunt motor the rated output is obtained
with an armature current of 73 A. The same constant torque is to be supplied when
this machine is arranged to operate as a cumulatively compounded motor. At this
load condition the series field winding contributes a 25% increase in pole flux.
(a) Find the armature current of the compound mode of operation.
(b) What is the speed of the compound motor?
Sec. 8-1 Motor Speed-Torque Characteristics 341

Shunt

Speed,
rpm
Cumulative
compound

Series ‘
Figure 8-2 Speed-torque characteris-
tics of various dc motor types using
common no-load speed (except for
Torque, N-m series motor).

(c) Find the speed regulation (SR) assuming the series field effect is negligible
at no-load.

Solution (a) At constant torque the increased flux field calls for a smaller armature
current. Thus,

daee
tat gg é

(b) As a shunt motor at rated load and speed the induced armature emf is
E, = 230 — 73(0.188) = 216.3 V

As a compound motor
El, = 230 — 58.4(0.188 + 0.06) = 215.5 V

Employing Eq. (8-8) then yields

D459 3916.7 rpm

Fa Ben, 0 = ANA
gig Ha 26311,.05
(c) The no-load speed from Example 8-2 is 1222.3 rpm. Hence,

_ 1222.3916.7
— 916.7 100 = 33.3%
SK =

a shunt motor.
Note how much larger this quantity is compared to operation as

Series Motor

from those of
The speed-torque characteristic of the series motor differs sharply
a surpris ing expecta tion in view of
the shunt and compound motors. This is not
constan t flux supplie d by a shunt field
the absence of a substantial and virtually
flux must be produc ed by the
winding. In the series motor all of the working
derive the express ion that relates
series field winding. It is instructive now to
speed and torque.
342 Direct-Current Motors Chap. 8

For simplicity we assume no saturation in the iron paths of the motor. This
permits the series winding flux to be related to the armature winding by

®, = Ksla (8-13)
where K, is an appropriate proportionality factor expressed in Wb/A. Inserting
Eq. (8-13) into the basic torque equation of Eq. (8-1) gives

je CN ke JGA he (8-14)
Furthermore, in the armature circuit of the series motor, Kirchhoff’s voltage law
states that

V, a Ey ad T(Ra us R;) a K;®,@ a5 I,R (8-15)7

where for convenience R = R, + R,. Then by employing Eq. (8-13) in the last
expression we get more simply

V, = [,(KrKso@ + R)
or .
V,
(8-16)
Lee
The desired speed-torque relation follows on inserting Eq. (8-16) into Eq. (8-14).
Hence,

bee K7K,
(K7K,@ = R)

or

Vie eee ti
Kiko thet VEO. (8-17)
In Eq. (8-17) w is the motor speed and T is the developed electromagnetic torque;
all other quantities are known. A plot of this equation takes the form depicted in
Fig. 8-2.
A more manageable expression results from Eq. (8-17) by assuming R small
compared to K;K,n. (See Prob. 8-7.) This step offers the advantage of making it
easier to see the manner in which these two variables are related. Accordingly,
we get

Kr K, VT

Next, upon introducing

V, Kr V,
Qa —S =
8-19
Kr K, V K7K,

TSee Eq. (8-30) and the footnote on p 350.

Sec. 8-1 Motor Speed-Torque Characteristics 343

we get finally,

(8-20)

Equation (8-20) shows that an increase in torque by a factor of 4 causes the speed
to halve. Similarly, a torque output that is nine times larger causes the motor to
reduce its speed to one-third its former value. A complete plot of Eq. (8-20) takes
the form depicted in Fig. 8-2.
A matter that is especially noteworthy about the series motor is the serious
consequence that can result if it is ever allowed to operate at no-load (i.e.,
T — 0). Equation (8-20) clearly indicates that the series motor speed tends to
infinity. Of course such a speed will never be attained. What will happen instead
is that the motor armature coils will fly out of their slots and be ripped apart unless
appropriate circuit-breaker protection is provided soon enough. A cardinal rule
that applies to the safe operation of the series motor is that it must never be
uncoupled from its load. Belt couplings must not be used in order to avoid the
possibility of their slipping and thereby causing damage.
Example 8-5
85 A at rated
A 250-V, 25-hp, 600-rpm series motor draws an armature current of
circuit resistance is 0.12 © and the series field
load torque of 314 N-m. The armature
winding resistance is 0.09 Q. Rotationa l losses are negligible .
is re-
Find the motor speed when the torque requirement at the motor shaft
duced to 20 N-m.

Solution The induced armature voltage at rated load is

E, = 250 — 85(0.21) = 232.15 V
first need to find the
To find the value of E, at the reduced load condition, we
This is readily determi ned by invoking Eq. (8-14). Thus
new armature current.
i ne. elie
7 \5 = Jay = 0.2524

i | = 0.2524(85) = 21.45 A
where primes denote the reduced load condition. Therefore,

E! = 250 — 21.45(0.21) = 245.5 V

specified loads, we get
Finally, forming the ratio of induced voltages at the
(a Ay KeK[qn' ~s ne
Eh RL = (0.2524) Fe

245.5 1
n= 537.15 0.2524 0” = 251329 rpm

value and can mean trouble.

Observe that this speed is more than four times the rated
344 Direct-Current Motors Chap. 8

\Series

Cumultatively “
140 compounded

0 100 Figure 8-3 Typical speed-torque

Torque (% rated) curves of dc motors.

A more useful representation of the graphs of Fig. 8-2 is shown in Fig. 8-3.
These curves are drawn with a common point which is rated speed at rated
torque. In this situation the series motor can share with the others the point of
commonality because the flow of rated armature current no longer poses a threat
to the series motor.

Torque versus Armature Current

The characteristic curves that describe the variation of developed torque as a

function of the motor’s drawn armature current for the various motor types read-
ily follows from the basic torque equation, Eq. (8-1). A shunt motor equipped
with a compensating winding provides the simplest relationship which is that of a
straight line through the origin. That is,

Nee K7® pl, ia Ksrla

where

Ka = K7®ysp,

When the shunt motor operates under the influence of a demagnetizing effect, the
decreasing pole flux with increasing armature current causes the characteristic to
fall below that of the linear relationship. This distinction is illustrated in Fig. 8-4.
For the cumulatively compounded motor the pertinent relationship is

f= K7(®s, a ®,)1q = KP yl te K7®,1, os K7® yi cts KE (8-21)

where

®, = K,I, and K, = K7K,

Clearly, Eq. (8-21) emphasizes the increasing contributions to the developed
torque as the armature current gets larger. Observe the increasing slope of the
characteristic as /, increases.
 Sec. 8-2 Equivalent Circuits and the Power Flow Diagram 345

Series
Cumulative compound

Shunt with comp. wdg

Shunt without comp. wdg

7Developed
torque,

Figure 8-4 Torque versus armature

current characteristic for various dc
Armature current, /, motor types.

For the series motor the expression for the developed torque is more simply
described by Eq. (8-14), which is parabolic as indicated in Fig. 8-4. It is instruc-
tive to note that as the series motor reacts to develop greater torques in response
to load requirements, the speed drops correspondingly. This feature is responsi-
ble for making the series motor very well suited to applications involving traction-
type loads.

8-2 EQUIVALENT CIRCUITS AND THE POWER FLOW DIAGRAM

can
The performance of the dc motor operating in any one of its three modes
circuit, a set of performa nce
conveniently be described in terms of an equivalent
The equivalent
equations, a power-flow diagram, and the magnetization curve.
note that now the armature
circuit is depicted in Fig. 8-5. It is worthwhile to
Observe that the defined
induced voltage is treated as a reaction or counter emf.
. The generated voltage
direction of I, is reversed from the case for a de generator
to the magnetic field, and
is caused by the relative motion of the rotor with respect
current. By imposing
its polarity is independent of the direction of armature
obtain the correct
constraints similar to those applied to the dc generator, we

Figure 8-5 Equivalent circuit of the de

motor.
346 Direct-Current Motors Chap. 8

equivalent circuit for the desired mode of operation. For example, for a series
motor the appropriate equivalent circuit results upon removing A; from the cir-
cuitry of Fig. 8-5.
The set of equations needed to compute the performance of dc motors
consists of the induced armature emf equation, the developed torque equation,
and Kirchhoff’s voltage and current laws as they apply to the equivalent circuit
displayed in Fig. 8-5. These are listed here as as matter of convenience.

E, = Ke®n (8-22)

T = K,@l, (8-23)
Vea £, + 1, Ra Ro (8-24)
Tp = Ip + Iq (8-25)

The first two equations are identical with those used in generator analysis. Note,
however, that the next two equations are modified to account for the fact that for
the motor V, is the applied or source voltage and as such must be equal to the sum
of the voltage drops. Similarly the line current is equal to the sum rather than the
difference of the armature current and field currents.
The power-flow diagram, depicting the reversed flow from that which occurs
in the generator, is illustrated in Fig. 8-6. The electrical power input V,L, origi-
nating from the line supplies the field power needed to establish the flux field as
well as the armature circuit copper loss needed to maintain the flow of I,. This
current flowing through the armature conductors imbedded in the flux field causes
torque to be developed. The law of conservation of energy then demands that the
electromagnetic power E,1, be equal to Tw,,, where w,, is the steady-state operat-

P,=V, I_= electrical power input

o-

ee a —— Useful power
tigi =-==—=< Power losses
aot
1

12(R,+R,)
2 Y fm =a Egla= Tw= T (7)
2mn
Vly -
ss
>

hd
ot,
=
‘z

I
Y
Prot R = Tow2= mechanical
|__| ' power output
2 losses
Figure 8-6 Power-flow diagram of the dc motor.
Sec. 8-2 Equivalent Circuits and the Power Flow Diagram 347

ing speed. Removal of the rotational losses from the developed mechanical power
yields the mechanical output power.
The strategy associated with the particular form of the power-flow diagram
illustrated in Fig. 8-6 is to start with the input power and then follow this power
through the machine to the place where the useful output appears. In the process
appropriate steps are included to underscore any transformations or conversions
that occur. Moreover, the flow of the useful power is represented by solid lines,
whereas the losses accompanied with each conversion or transformation are indi-
cated by broken lines. A particular advantage of this representation is the ease
with which the pertinent relationships that are used to provide a complete descrip-
tion of the behavior of the motor are displayed. In essence the power-flow dia-
gram so formulated furnishes an excellent model of the mathematical depiction of
the machine. The following examples illustrate these advantages.
Example 8-6
A 20-hp 230-V 1150-rpm shunt motor has four poles, four parallel armature paths,
and 882 armature conductors. The armature-circuit resistance is 0.188 ©. At rated
speed and rated output the armature current is 73 A and the field current is 1.6 A.
Calculate: (a) the electromagnetic torque, (b) the flux per pole, (c) the rotational
losses, (d) the efficiency, (e) the shaft load.

Solution (a) The developed electromagnetic torque can be found by employing the
relationship that appears on the second level of the power-flow diagram which is
simply a statement of the law of conservation of energy. It is at this point that the
power associated with armature winding, namely E,J,, is converted to mechanical
power, Tw,,. Accordingly, we proceed as follows:

E, = V, <1, Re = 2300—.73(0.188). = 230,.— 13,7 = 216.3,V

any 27 7
27(1150)
Wm MEE 60 = 120 rad/s

Therefore
EWE lye L1GIS(73) Mt
i ihe d aia ORGROE 132 N-m

(b) E, "2
= Kg®n IY= 60 a On NE
= 60 4 (1150)

* 216.3(60) -—
= 882(1150) = (0.0128 Wb

(c) From the power-flow diagram

Prot =P — Po = Egla — 20(746) = 15,790 — 14,920 = 870 W

(d) First find the sum of the losses. Thus

> losses = Prot + Ra +-Vily = 870 + (73)°(0.188) + 230(1.6)

= 870 + 1002 + 368 = 2240 W
 348 Direct-Current Motors Chap. 8

Hence

>= losses 9240

adie ce pe 17,160
= 0.869
‘ Sale ara bad upc Welhe
The efficiency is 86.9%.

(€) Ty = wae+ Pp 2) 120=< (20 x 746) = 124 N-m

The difference between this torque and the electromagnetic torque is the amount
needed to overcome the rotational losses.

Example 8-7
The shaft load on the motor of Example 8-6 remains fixed, but the field flux is
reduced to 80% of its value by means of the field rheostat. Determine the new
operating speed.

Solution Information about the speed is available from Eq. (8-22). However, in
turn, knowledge of flux and armature induced emf E,, is needed. By the statement of
the problem the new flux ®’ is related to the original flux by

®' = 0.80
To obtain information about E,, we must determine the change in /,, if any. By the
constant-torque condition we have

K;®l, = K,@'l;,
Hence

i 73 =91.3A
0.8
Consequently

E, = V, — [,Rq = 230 — 91.3(0.188) = 212.8 V

Returning to Eq. (8-22) we can now formulate the ratio

Ey _ Kp®'n'
E, Ke®n

from which the expression for the new operating speed becomes

ni = Ea ® _ 212.8 1
E,® 216308
1150 = 1414 rpm

8-3 SPEED CONTROL

One of the attractive features the dc motor offers over all other types is the
relative ease with which speed control over a substantial range can be achieved.
The various schemes available for,speed control can be deduced from the speed
Sec. 8-3 Speed Control 349

equation which is readily obtained from the formula for the induced armature
voltage and Kirchhoff’s voltage law for the armature circuit. Accordingly,
E, = Ke®n

and
E,= V;— 1,Ra + Re)

Hence
am Ey 1a V, ey T(Ra 5s R.)
n (8-26)
K_® K;-®

Observe that the armature circuit resistance is augmented by the inclusion of an

adjustable external resistor, R.. Inspection of Eq. (8-26) reveals that the speed of
the motor can be controlled by adjusting any one of three factors appearing on the
right side: ®, R,, and V,. Our goal next is to examine and illustrate these methods
of control for each motor type.

By Adjustment of Pole Flux (®)

Pole flux is the simplest to adjust and involves the least expense. We consider
of
first the shunt motor and for simplicity the analysis proceeds on the assumption
negligible demagneti zing effect from the armature winding mmf throughou t.

Shunt motor. In the case of flux control, R, is zero in Eq. (8-26) and the
flux field ® is assumed entirely due to current that flows to the field winding.
ofthe rheostat
Adjustment of the pole flux is conveniently accomplished by means
with the field winding in the manner depicted in Fig. 8-7.
which is placed in series
and the speed
If the field rheostat resistance is increased, the pole flux diminishes
shunt motors are designed to provide a 200% increase
increases. General-purpose
in base (rated) speed by this method of control.
e of R,,, influ-
We now undertake to derive the manner in which the presenc
of the shunt motor. The analysis begins by getting
ences the speed-torque curve
and torque when R,,, = 0. Moreove r, no saturation
the relationship between speed
flux can be directly related to the field current by
is assumed so that the pole

®,, = Ksnl =
Tee:V
K sh Ry
(8 oF )

Figure 8-7 Speed control of a shunt

motor equipped with a rheostat.
350 Direct-Current Motors Chap. 8

where K,, is expressed in units of Wb/A and the field current is replaced by its
equivalent expression using Ohm’s law. Then by inserting Eq. (8-27) into the
expression for the basic torque we get
V,
r= K;®l, a KrKsh — i; (8-28)
Ry
Also,

I, = 1 =!
V, =
Ne
E
(8-29)
and

E, a Ke®,,n a KP, (8-30)

The last equation indicates that the armature induced voltage may be represented
either with speed expressed in rpm or in rad/s.+ Because Eq. (8-28) already
involves the K; parameter, we choose the second alternative for use in Eq.
(8-29). Accordingly,

L= V; yx K7®,,@
a Ke

and so Eq. (8-28) becomes

T=
_ KrKsn
RAR; Vi
,,, — (KrKsiVi)?
RR} (8-31)

Rearranging yields the desired result,

03) Sper
Ry
K;Ky
R,R;
ag) fake eb V
hy lpn @)a oes— aT z
K2K°, (8-32)

where

wo = KK,
Ry '
= speed intercept (8-33)

and

a= iting. |
K2K2,V2~ slope of f speed-torque
d curve (8-34)

Clearly, Eq. (8-32) is that ofa straight line with no-load intercept speed of wp)
anda
negative slope described by Eq. (8-34). The plot of this curve is shown
in Fig. 8-8
for R,, = 0. (It is also plotted as the solid line in Fig. 8-1.)
How does Eq. (8-32) change when the field resistance is increased
by the
contribution of the rheostat? In such a case the total field resistanc
e changes from

:
+A comparison 2
of Eqs. (3-36) and (3-44) shows that Ky, = =A K; so that
replacing K, by its
equivalent yields the angular velocity w = ma n.
‘60
Sec. 8-3 Speed Control 351

205

Wo

Rsh = 0

Figure 8-8 The effect of rheostat field

Riposel iy resistance on the speed-torque curve of
. .

a dc shunt motor neglecting saturation

s

0 Torque, N-m and armature reaction.

curve can be
Ry to (Rp + Rs). The modified expression for the speed-torque
ce on both the
found by determining the effects of the increased field resistan
replacement of Rp by
intercept and the slope. The strategy to employ is the simple
(REAR). Accordingly, the expression for the new intercept becomes

y
paddah ya 2sR, (se Rat ) Ra= ) (8-35)
CR R, = ow \1 +
OP

sed in terms of the origi-

By appropriate manipulation the new intercept is expres
Ry, has the effect of doubling
nal value. Clearly, for the linear case, making R,, =
(8-26) also makes this self-evident because at
the no-load speed. Of course, Eq.
no-load (J, = 0) halving @ must cause n to double.
get
By proceeding in a similar fashion for the slope we

Ra(Ry
+Rus)? _ RoR (y 5Re oa (144) 836)
K7KS,shV; K}KS,V;
results. If Ry, = Rr, Eq. (8-36)
Again, note that by setting Ry, = 0, Eq. (8-34)
factor of 4. Figure 8-8 illustrates
discloses that the negative slope increases by a
this matter graphically.
speed to the developed torque
The modified expression of the relationship of
is presen t is
of a shunt motor when rheostat resistance

0 (1+ Fe)~ a (1 + Ba) 7 (8-37)

w=

(8-34).
where wy and a are given by Eqs. (8-33) and
Example 8-8
of 144 Q. Moreover, the values of wo
A shunt motor has a field winding resistance
rad/s/N-m respectively. Find the value of
and a in Eq. (8-37) are 129 rad/s and 0.07
N-m is supplied at a speed of 1650 rpm.
field rheostat resistance so that a torque of 70
352 Direct-Current Motors Chap. 8

Solution The new speed in rad/s is

21
o = 1650= 86.4 rad/s

etx — 1: Aah Inserting the foregoing data into Eq. (8-37) there results
f
86.4 = 129x — 0.07(70)x"
4.9x? — 129x + 86.4 = 0
X12 = 13.163 + 9.32 = 3.84, 22.48
Choosing the smaller value then yields

R sh
It ae = 3.84
Rr
Ry, = 2.84(R) = 409 0

Compound motor. An analysis similar to that just made for the shunt
motor can be employed to get an expression that relates w to T for the compound
motor. However, because the result depends on the relative strength of the series
field contribution vis-a-vis the shunt field and owing to the complexity of the
resulting expression, let it suffice to say that the plots for the compound motor
with and without a field rheostat resistance will generally be similar to those
displayed for the shunt motor in Fig. 8-8. In visualizing these characteristics it is
helpful to keep in mind that as the shunt field winding contribution to the pole flux
is diminished with the insertion of R,,, the importance of the series field contribu-
tion increases.

Series motor. Control of the pole flux for the series motor is achieved
through the use of a diverter resistance R, in the manner shown in Fig. 8-9. The
developed torque can be expressed as follows:
Ra
T = K7O,Iq = KrKglslq = K7Ks Ret
s>——R, re
= kp (8-38)

Figure 8-9 Field control of a series

motor by means of a diverter resistor.
Sec. 8-3 Speed Control 353

where

Rg (8-39)
k= Krk, and, p= R.+R,
Ss d

The expression for the armature current comes from Kirchhoff’s voltage law for
the armature circuit. Thus ;
E RsRa ad
aad KrK,I,@ ts plRs + TRa
V, oe Ea he Me R, ri Ru aa IR

= kpwl, cs I,pRs + TRa

= [,(kpo ar pR;s a Ra) (8-40)

Accordingly,
V,
‘* : (8-41)
es pR, + Ra + kp

which upon insertion into Eq. (8-38) yields

kpV? (8-42)
ER, + pR, + kpor
for the
Equation (8-42) is the general expression that relates speed and torque
configuration of Fig. 8-9. By Eq. (8-39)0 =p= 1. When the series motor is used
(8-42) becomes
without the diverter resistance, i.e., Ry = ~, p = | and so Eq.
more simply
%. kv? (8-43)
i
(R, + Ry + koy

where T is in N-m and a is in rad/s.

diverter resis-
To examine how the speed changes with the presence of a
resistan ce (p < 1) and find the new speed at which the
tance, introduce a diverter
Eq. (8-42) equal to Eq.
same torque of Eq. (8-43) is supplied. This requires setting
(8-43). The result is

Bie AC elys a eae (8-44)

is the predominant one, it
Because the first term on the right side of Eq. (8-44)
e of the diverter resistance causes
follows that at constant load torque the presenc
an increase in speed.
Example 8-9
A 250-V, 25-hp series motor produces a torque of 150 N-m at a speed of 868 rpm.
The armature winding and the field winding resistances are 0.12 © and 0.09 Q,
respectively. For a constant torque load find the new speed when a diverter resis-
tance of 0.09 Q is placed in parallel with the series field winding.
354 Direct-Current Motors Chap. 8

Solution Use Eq. (8-43) to find the value of k. Thus

k(250)? (250);
a= i i (0.21
+ 90.88k)2
(0.21 + a, 868k
which leads to

k2 — 0.04583k + 5.34 x 10°° =0

k — 0.04583k = 0

k = 0.0483

Also,
Ra 0.09 ~__
Pris
RR, Ome
Inserting these values into Eq. (8-44) yields

I 0.12 ( 1 I ) 0.09 1 )
= —— ——'— —_]} + ———___ |-—- ]
V0.5 ee 0.04583 \\/o.5 0.5 0.04583 ea
= 128.52 — 1.534 + 0.8134

mI 7eomraays

n= 2 (127.8) = 1220.4 rpm

Speed Control by External Armature Resistance

A second method of speed adjustment involves the use of an external resistor R,

connected in the armature circuit. The size and cost of this resistor are consider-
ably greater than those of the field rheostat because R. must be capable of han-
dling the full armature current. Equation (8-26) indicates that the larger R, is
made, the greater will be the speed change. Frequently the external resistor is
selected to furnish as much as a 50% drop in speed from the rated value. The chief
disadvantage of this method of control is the poor efficiency of operation. For
example, a 50% drop in speed is achieved by having approximately half of the
terminal voltage V, appear across R,. Accordingly, almost 50% of the line input
power Is dissipated in the form of heat in the resistor R.. Nonetheless, armature
circuit resistance control is often used—especially for series motors.

Shunt motor. The circuit arrangement for a shunt motor whose speed can
be adjusted by an external armature-circuit resistance R, as well as by shunt field
current control is illustrated in Fig. 8-10. The general expression for the shunt
motor speed as a function of torque without the presence of R, was previously
Sec. 8-3 Speed Control 355

Figure 8-10 Schematic arrangement

for speed control of a shunt motor both
by pole flux adjustment and armature-
voltage control.

found and displayed in Eq. (8-37). It is repeated here with the expressions for the
no-load speed (intercept) and the slope of the speed-torque curve included. Thus

= Ry (1 AN,
Rh R; ( Ha
@
KrKsn a in’Kj.KynVi
‘ ;t iB Ry A ie

The inclusion of the external resistance R, means that effectively the armature
circuit resistance is raised from a value of R, to one of value (R, + R.). Hence the
with an
expression for the speed-torque characteristic of a shunt motor equipped
by simply replacing R, by
external armature resistance follows from Eq. (8-45)
(R, + R.). Accordingly,

Ry ( =o) Rj ( eH).
@ OE 1 ert
Ka NY Te Reeser
R, PPE (Ra Ne y) 2K? V2 1 lee; meh -
(8-46)
sh

speed; however,
A study of Eq. (8-46) reveals that R, does not affect the no-load
straight line
its presence can have a noticeable effect on the slope of the drooping
This situation
represented by Eq. (8-46) especially if R, is much greater than Ry.
power to a constant
is graphically presented in Fig. 8-11. Observe that to deliver
decrease in speed.
torque load the insertion of R, can cause a considerable
Fig. 8-10 that
It is helpful to keep in mind for the arrangement depicted in
flux adjustm ent via
speeds above the rated or base value can be achieved by pole
control through R,
Ry,, whereas for speeds below this base value armature voltage

Constant torque
load

rad/s
Speed,

Figure 8-11 Speed-torque characteris-

R. >> R,

tic of shunt motor showing the effect of

Torque, N-m an external armature resistance.
356 Direct-Current Motors Chap. 8

is viable. Overall, by the combination of both schemes, speed control over a

range of 4:1, from the lowest to the highest, is easily managed.

Compound motor. As was the case for speed control by the adjustment
of pole flux, speed control of the compound motor by armature-circuit resistance
is similar to that which prevails for the shunt motor. Although the presence of the
series field winding modifies the results somewhat, the general trends remain the
same. This can be best appreciated by examining the general expression for the
speed of the compound motor which is

V, fa T(Ra ia RG Ae ee)

oe K(®y, a ®,) oe

Note that in addition to R, the numerator also includes the resistance of the series
field winding. Moreover, the denominator shows the contribution of the series
field mmf on the pole flux. At a given torque an increase in R, causes the numera-
tor to diminish which in turn drops the speed. Large changes in R, are followed
by correspondingly large changes in speed.

Series motor. Figure 8-12 illustrates the arrangement used to control the
speed of a series motor by means of an external resistance. The torque in this
case is described by Eq. (8-43) which is repeated here with a modification to
account for R,. Keep in mind that R = R, + R,. Thus

kV?
(8-48)
eg R, + kw)?]
where k is defined by Eq. (8-39). Examination of the denominator expression of
Eq. (8-48) makes it clear that for a constant torque load, increases in R, must be
accompanied by corresponding decreases in w in order to keep the expression
invariant. Similarly, to preserve constant speed, increases in R. must produce
reduced torque. The net effect of these responses is to cause the speed-torque
characteristic with R, present in the circuit to shift to the left in the manner
depicted in Fig. 8-13 relative to its position when no external resistor is used.
The influence of the external resistance in changing the speed for a constant
torque load can be more dramatically exposed by setting Eq. (8-48) equal to Eq.
(8-43). Thus

Figure 8-12 Circuit arrangement for

series motor speed control by external
resistor, R,.
Sec. 8-3 Speed Control 357

rad/s
Speed,

Figure 8-13 The effect of external

resistance on the speed-torque charac-
0 Torque, N-m teristic of the series motor.

kv? kV?
(R+R, + ko’ (R+ ko) ee?
is employed. The solu-
where w’ denotes the speed when the external resistance
tion of w’ in the last expression leads directly to
R.
jh wisi (8-50)
k
albeit scaled, decrease in
Consequently, the insertion of R. brings about a direct,
load.
speed when power is delivered to a constant torque
Example 8-10
initial operating point. Find
Refer to the motor of Example 8-9 and assume the same
torque of 150 N-m at a speed of
the external resistance needed to produce the same
400 rpm.

Solution The new speed in rad/s is

27 Qa
@ ny (600) = 41.89

The speed without R, is

wal iL
wW 30 868 = 90.9

Then from Eq. (8-50)

Ri Me a) = (.04583(90.9 — 41.89) = 2.246 0.

Voltage
Speed Control by Adjustable Armature

ves adjustment of the applied terminal

A third method of speed control invol
from the standpoint of flexibility and
voltage. This scheme is the most desirable
the most expensive because it requires its
high operating efficiency. But it is also
a motor-generator set or a rectified
own dc power supply. It means purchasing
358 Direct-Current Motors Chap. 8

adjustable voltage source with a capacity at least equal to that of the motor to be
controlled. Such expense is not generally justified except in situations where the
superior performance achievable with this scheme is indispensable, as is the case
in steel mill applications. When the adjustable de voltage source is obtained from
an independent motor-generator set it is referred to as the Ward-Leonard sys-
tem. Consult Chapter 13 for further details.

Shunt and compound motors. Figure 8-14 shows the schematic diagram
of shunt and compound motors that are operated with adjustable dc voltage
sources. The adjustable dc source can be an ac motor driving a de generator, a dc
motor driving a de generator or a single-phase or a three-phase source that is
rectified using thyristors. The latter is the subject matter of Sec. 8-8. A rectified
fixed dc source may also be employed for field excitation as indicated.
The manner by which speed control is achieved by adjustable armature
voltage is immediately apparent from Eq. (8-3) which at no-load reduces to

V, V,
Dt ee or O= FO (8-51)

Clearly, then, for constant field excitation the operating speed is directly propor-
tional to the applied armature voltage. Even under load conditions, which call for
accounting for the armature-circuit resistance drop, this proportionality holds
within 5% or so. Adjustable armature voltage is invoked to furnish speed control
below the base speed (i.e., the speed corresponding to rated armature voltage for
full field current), whereas field control is used for speeds above the base value.
Adjustable armature voltage achieves the same result for speed control below
base speed that is obtained by external armature circuit resistance. However, the
former can do it at much higher efficiencies and superior speed regulations albeit
at higher initial costs.

Adjustable
de
voltage
source

Figure 8-14 Speed control of a dc

Rectified dc compound motor by adjustable arma-
source
ture voltage.
 Sec. 8-3 Speed Control 359

Constant torque
load

rad/s
Speed,
Vey

Vi2 < Vin

Vi3 < Vie Figure 8-15 The effects of reduced

armature voltage on the speed of a
Torque, N-m series motor at constant load torque.

Series motor. In the case of the series motor the adjustable dc voltage
source is applied to the series connection of the series field and the armature
applied to the series
winding. The effect that reductions in the armature voltage
defines the approximate
motor have on speed is easily seen from Eq. (8-18), which
ion that any point
speed-torque curve of the motor. It is obvious from this express
adjustment of V,. Thus,
on the speed-torque curve can be directly scaled through
causes a corresponding
for a constant torque load, halving the armature voltage
. The results are display ed in Fig. 8-15.
decrease in speed by one-half
Example 8-11
In an industrial application the need arises to provide a constant torque of 105 lb-ft
to be considered negligible for
over a speed range of 250 to 1000 rpm. All losses are
simplicity.
a shunt motor to accomplish
(a) Determine the hp rating and speed rating of
to be obtain ed through adjustment of the
this task when the control of speed is
armature voltage.
is to be obtained from
(b) Repeat part (a) for the condition where speed control
adjustment of the field current.
are designed to operate at
(c) Assume that the motors in parts (a) and (b)
re current of both motors when they operate at
220 V. Find the value of the armatu
1000 rpm delivering the specified torque.
e voltage of both motors
(d) Plot the variation of armature current and armatur
over the specified speed range.
s in Eq. (8-23) [or K-® that
(e) Compute the value of the factor K7® that appear
of 1000 rpm. Also, plot the variation
appears in Eq. (8-22)] for both motors at a speed
range.
of this parameter for each motor over the speed
the delivered hp varies over the
(f) Show a plot of the manner in which
specified speed range.
360 Direct-Current Motors Chap. 8

(g) What is the maximum hp that can be delivered by each motor at an operat-
ing speed of 250 rpm?

Solution (a) When armature voltage control is used, the base speed is taken to be
the maximum value of the specified speed range. In this case, it is 1000 rpm. Since
the motor must deliver a torque of 105 Ib-ft = 105(1.356) N-m [see Fig. 8-16(a)], it
follows that the required hp rating is determined to be

T
: (105)(1.356)( 1000 a aa) _ (105)(1000)
hp = _ 59
746 5250
Thus the rating for motor A (for armature control) is

A: 20 hp, 1000 rpm

(b) When speed is varied by field current control, the maximum speed of the
specified speed range is attained corresponding to minimum field current (i.e., the
minimum flux condition). The minimum speed of 250 rpm is then obtained as the
field current is advanced to its maximum setting. Because the base speed of a motor
is defined as the speed that corresponds to maximum flux at rated armature voltage,
it follows that the base speed in this case must be 250 rpm. The required hp rating of
this motor can be calculated at the maximum speed value of the speed range.
Because the torque is to be constant over the speed range, which includes 1000 rpm,
the required hp is the same as for machine A of part (a), namely

— 105(1000) _
bps
gygy a0
Hence the complete specification for motor F (for field control) is

F: 20 hp, 250 rpm

A comment is in order at this point concerning motors A and F. First, how-

ever, it is useful to cite a general expression for the rating P of electrical machines
that puts the emphasis on physical dimensions. This is

P = kRD?Ln (8-52)
where D is the rotor diameter, L the effective axial core, n the speed in rpm, and R is
an appropriate design parameter. In essence the last expression states that the
power rating of a machine is a function of the product of its volume and speed.
Accordingly, the greater the speed, the smaller will be the volume. Therefore, motor
F in this application will be larger, heavier and so more expensive than motor A.
This greater volume of motor F is entirely consistent with the fact that at the lower
speeds it must be capable of accommodating values of flux that are much larger than
at higher speeds.
(c) Because losses are being neglected, we can set input power equal to output
power. Thus
I,(220) = 20(746)
or

I, = 67.8 ~ 688A
 1000 Speed, rpm
(a)

Armature
current,A

68A iS
F

\7A

1000 Speed, rpm

Armature
voltage, V

220 z

5S

1000 Speed, rpm

Flux factor,
K@,N-m/A

8.4 F

aM A

O 250 500 750 1000 Speed, rpm

(d)
Figure 8-16 Performance features of
Horsepower two motors used to deliver constant
torque over a 4:1 speed range: (a)
constant torque characteristic needed
20
by load; (b) armature current variation
A,F at constant torque; (c) armature voltage
variation at constant torque; (d) flux
3) factor, K®, variation; (e) horsepower
variation. A, motor with armature
O 250 500 750 1000 Speed, rpm voltage control; F, motor with field
current control.

361
362 Direct-Current Motors Chap. 8

This armature current is the same for both motors at 1000 rpm, the high end of the
speed range.
(d) The variation of the armature current for the motor (A) operated with
adjustable armature voltage is depicted in Fig. 8-16(b) as curve A. This motor oper-
ates with constant field current at all speeds. Hence, its air-gap flux is also con-
stant. The basic torque expression of Eq. (8-23) thus can be written as T= K;7®I, =
KI,. Consequently, a constant torque demands a constant armature current.
The situation is different for the motor (F) that employs field control. In this
case a drop in speed is the result of an increase in flux. At a constant torque condi-
tion this flux increase means a corresponding decrease in armature current. The
variation of this current over the speed range is illustrated as curve F in Fig. 8-16(b).
Appearing in Fig. 8-16(c) are the plots of the armature voltage with speed.
Motor F operates at a constant armature voltage. The armature voltage variation of
motor A, however, must change with the speed. In fact, the speeds below 1000 rpm
are achieved precisely because of the reduction in armature voltage consistent with
the equation for the counter emf of the armature which, by the assumption of zero
losses, becomes

V = K,@n = Ken = E,
Hence, if rated voltage corresponds to operation at 1000 rpm, then operation at 250
rpm is obtained when the armature voltage is reduced to 55 V.
(e) Motor A operates at a constant flux and a constant armature current. The
quantity K7® readily follows from Eq. (8-23) as

(K;®)I, = T

K7®(68) = 105(1.356) = 142.4 N-m

K;® = 2.1 N-m/A

Of course, this quantity stays fixed throughout the speed range as indicated by curve
A in Fig. 8-16(d).
In the case of motor F part (c) shows that the armature current is also 68 A at
1000 rpm. But because the lower speeds in this motor are the result of increasing flux
values, it follows that the quantity K;® for motor F increases with decreasing speed
as illustrated by curve F of Fig. 8-16(d). It is instructive to note that for constant
torque the variations of (K;®) and armature current for motor F are complemen-
tary. The increase in K;®@ is accompanied by a corresponding decrease in armature
current.
In a motor with negligible losses by the law of conservation of energy we can
write

Tim = Eqla
K;®lp@m = Kp®onl,
or

K,® = K,®
Therefore, the variation of K;® which bears units of N-m/A is identical to the
parameter K;®, which bears units of V/rad/s. As an illustration of the use of this
 Sec. 8-4 Speed Reversal 363

result, we can find the armature voltage of motor A at a speed of 250 rpm (= 26.18
rad/s) by simply multiplying the parameter 2.1 by the speed expressed in radians per
second. Thus,
2.1(16.18) = 55 V

(f) The plot for both motors is shown in Fig. 6-23(e). Of course, the hp
changes in direct proportion to the speed.
(g) The maximum hp for motor A is 5 because the motor is working as hard
electrically and magnetically as it has been designed to work. It is at full capacity
electrically since it is drawing rated armature current. It is also at full capacity
magnetically since it is operating at its fixed excitation value.
The situation, however, is different with motor F. Although this motor has
rated armature voltage applied, it is drawing only 17 A at 250 rpm. Accordingly, this
motor could be called on to deliver at 250 rpm an increased torque to the level of
4(105) = 425 lb-ft, corresponding to a fourfold increase in armature current thus
raising it to its rated value of 68 A. The corresponding delivered hp is then 20, which
is four times larger than that of motor A.

8-4 SPEED REVERSAL

How is the direction of rotation of a dc motor reversed? An examination of the

situation that prevails inside the machine reveals that the torque developed on
each conductor of the armature winding beneath the field poles can be reversed by
reversing either the armature current or the direction of the field flux (by reversal
A
of the direction of flow of the field current). The issue is illustrated in Fig. 8-17.
two-pole motor is shown with a single coil representi ng the armature winding.
current
The direction of field flux is from left to right. In Fig. 8-17(a) the armature
out of the plane of the paper on the right side (conductor b) and
is assumed to flow
occurs
into the paper on the left side (conductor a). Because a flux reinforcement
a as well as the bottom side of b, the net torque produces
on the top side of
winding are
counter clockwise rotation. When the connections to the armature
Fig. 8-17(b), a flux rein-
arranged so that the current is reversed as illustrated in

ampere-
Figure 8-17 Speed reversal of a dc motor. (a) Field flux and armature
; (b) reversed armature
conductor acting to produce counterclockwise rotations
e rotation.
current reverses the developed torque to cause clockwis
364 Direct-Current Motors Chap. 8

forcement now takes place on the bottom side of a and the top side of b so that the
net torque produces a clockwise rotation. If the field direction is reversed in each
part of Fig. 8-17 for the specified armature current directions, then the direction of
rotation in the configuration of Fig. 8-17(a) becomes clockwise while that of Fig.
8-17(b) becomes counterclockwise.
Is one method of reversal preferred over the other? Let us investigate the
advantages and shortcomings of each method. In industrial applications reversal
is often called for while the motor is operating under load (e.g., in hoists, punch
presses, and so on). Because of the large number of turns that are placed on the
field poles of dc motors, the inductance of the field winding is very high as Eq.
(1-40) readily indicates. Hence the reversal of the field current through a revers-
ing switch demands a design that will suppress the attendant sparking that would
otherwise occur. However, despite such precautions, deterioration of the contact
surfaces of the switch will occur over time. It is generally a poor idea to interrupt
current in a highly inductive circuit, if it can be avoided. Furthermore, if the
motor is equipped with a series field, then reversal of the shunt field must be
accompanied by reversal of the series field too otherwise the compounding type
will not be the same. Beyond these factors, however, there exists the very serious
one of reducing the safety of operation of the motor when a switch is placed in the
field circuit. Clearly, a faulty switch here can create an extremely dangerous
situation for the motor. This fact alone is sufficient to eliminate speed reversal by
field reversal as a viable alternative.
Are there similar disadvantages with armature current reversal? Let us
see. When the motor is operating, say, in a clockwise direction under load, it is
clearly developing an armature induced voltage, E,. To reverse the armature
current to change the direction of rotation, it is necessary to reverse the applied
armature voltage in the manner depicted in Fig. 8-18 through a double-pole dou-
ble-throw switch. During the instant that the switch is reversed, the armature
induced emf still exists. Hence, when the reversed armature voltage appears at
the armature terminals, the polarity of V, is opposed to the original E,. The effect
is to change the speed rather abruptly (called dynamic breaking) with little induc-
tance effect. It is only the armature leakage inductance that is active in this state
of affairs and it is quite a small quantity compared to the field inductance.
Because this feature is such an attractive one and in view of the serious shortcom-
ings associated with field reversal, speed reversal in dc motors in industrial and
commercial settings is accomplished exclusively by armature current reversal.

Figure 8-18 Preferred circuit arrange-

ment for reversal of the direction of
rotation of a shunt motor via armature
current reversal.
 Sec. 8-5 Permanent Magnet (PM) dc Motors 365

8-5 PERMANENT MAGNET (PM) DC MOTORS

For a dc motor to develop torque and thereby attain an operating speed, it is

necessary that there exist a flux density distribution along the air gap of the
machine. We already know that in conventional de motors this field distribution iS
established by the action of a de current that flows through the field winding which
wraps around the salient poles of the stator structure. However, circumstances
do arise in the application of dc motors where machines so equipped offer no
particular advantage and, in fact, may turn out to be impractical, inconvenient,
and/or more expensive than alternative ways of furnishing the machine with flux
density. A particularly striking example of this situation is the application of dc
motors in countless motor vehicles throughout the world, doing such tasks as
driving windshield wipers, fans, windows, air conditioners, radio antennas, and
the like. The selection of the dc motor is clearly a natural choice because of the
availability of a de source in the form of a rechargeable battery. The motor size
needed to accomplish most of these jobs is very small and in most cases is a small
fraction of a horsepower. Accordingly, the use of permanent magnets (PM) in
place of field windings offers the advantages of lower cost and simpler construc-
tion. It also means higher efficiency because of the absence of the field winding
copper losses but in these small sizes this is generally not a significant factor.
Permanent magnet de motors have also found an application niche in situa-
tions requiring integral horsepower sizes exceeding even 100 hp. These motors
only is
can be found in systems where motor speed control below base speed
required and is achieved by armature voltage control through electroni c methods
of the
like those described in Sec. 8-6. Of course, in applications where control
extend beyond base speed, adjustmen t of the field flux is neces-
motor speed must
motor does
sary and a wound-field machine must be specified because the PM dc
not offer this flexibility .
types:
The material used to construct a PM de motor may be one of three
(ferrite) magnets, or rare earth magnets. These mate-
Alnico magnets, ceramic
shing feature of very high residual flux density and high
rials have the distingui
8-19. Alnico has a
coercive force. A typical hysteresis loop is shown in Fig.
to the flux densities
residual flux density of the order of 1 T, which is comparable
machines , but it falls mostly on the low end of the range.
found in wound-field
dc motor has a slightly
Consequently, for a given hp rating the Alnico-type PM
magnet has a resid-
larger armature than its wound-field counterpart. The ceramic
and it is also the least
ual flux density that is usually less than half that of Alnico
are used for the
expensive of the PM materials. As a result ceramic magnets
Alnico magnets are
construction of PM dc motors in the fractional hp range while
used invariably for motors in the integral hp range.
strong permanent
There are certain disadvantages associated with the use of
lly when the ratings are in the integral hp range.
magnets in dc motors especia
at all times even when the motor is not being
Because the magnetic field is active
be more carefull y designe d to guard against at-
used, the motor enclosure must
366 Direct-Current Motors Chap. 8

Flux density,T

Figure 8-19 General shapes of the

hysteresis loops of two permanent
magnet materials: (a) Alnico; (b) ce-
ramic (ferrite).

tracting foreign matter that could be harmful to the motor. A more serious draw-
back concerns demagnetization of the permanent magnets. This can occur espe-
cially if the armature current should reach excessive values as might happen under
fault conditions such as an armature short-circuit. Equipping the motor with a
compensating winding is one effective way to avoid this problem but of course this
increases the cost of the motor.
Demagnetization is a problem that is inherently characteristic of permanent
magnets. To better understand this statement, let us examine the situation illus-
trated in Fig. 8-20. Depicted in Fig. 8-20(a) is a permanent magnet attached to a
very high permeability magnetic material to form a closed magnetic circuit. If the
magnetic material attached to the PM is assumed to offer zero reluctance, the flux

Flux density

Shear line

(a)
© (b)
L (c)
Figure 8-20 Application of the PM to magnetic circuits: (a) PM ina magnetic
circuit of zero
reluctance; (b) PM in a magnetic circuit with two air gaps; (c) locating the
operating flux
density value in the magnetic circuit of (b).
Sec. 8-5 Permanent Magnet (PM) dc Motors 367

density that then penetrates the magnetic circuit is B,, the residual flux density of
the PM. Consider next the situation that is shown in Fig. 8-20(b), which is more in
keeping with what is encountered in the actual magnetic circuit of the machine
because of the presence of the two air gaps. The insertion of the air gaps into the
magnetic circuit of Fig. 8-20(a) necessarily leads to a level of flux density in the
magnetic circuit of Fig. 8-20(b) which is less than that found in Fig. 8-20(a). The
presence of the air gaps increases the reluctance of the magnetic circuit and the
effect may be interpreted in terms of an equivalent demagnetizing field intensity
(H,) associated with the insertion of the air gaps. Expressed mathematically we
write

—H ln = 2F, (8-53)
one air
where %, is the mmf required to force the flux density of the PM to cross
H, is the equivalen t demagnet izing
gap, I, is the effective length of the PM and
quantity Hl, is that magnetom o-
field intensity associated with the air gaps. The
the PM ina direction opposite [hence
tive force which would have to be applied on
same effect as
the negative sign in Eq. (8-53)] to the residual flux to produce the
the insertion of the air gaps. The gap mmf is

Fy = heRe (8-54)

flux density
where ¢, is now the gap flux associated with a partially demagnetized
ctional area A of the magnetic circuit. Thus
B. (< B,) and the cross-se
de — BJA

Hence

peor
g bg g A (cE _ Bag
d
&
pA Lo

where jy denotes the permeability of air.

terms leads to
Introducing the last expression into Eq. (8-53) and rearranging

By = - (Hea) Hy (8-55)
the origin and
This is clearly the equation of a straight line that passes through
negative slope plots
exhibits a negative slope given by the coefficient of H,. The
nt magnet. The
the line on the demagnetizing side of the B-H curve of the permane
yields the new
intersection ofthis line with the hysteresis loop ofthe PM material
PM magnet is placed
operating flux density By when the configuration in which the
8-20(b). Figure 8-20(c)
changes from that depicted in Fig. 8-20(a) to that of Fig.
the shear line of the
illustrates the result. The straight line of Eq. (8-55) is called
to note that when the gap
magnetic circuit in which the PM acts. It is instructive
line becomes infinite.
length is allowed to approach zero, the slope of the shear
to the residual value.
This means that the operating flux density returns
368 Direct-Current Motors Chap. 8

In any practical PM dc motor the operating point is represented by P in Fig.

8-20(c) which means that the actual flux density is less than the residual flux
density that is associated with the B-H curve of the PM material. If an Alnico PM
motor is subjected to a severe armature current, a further demagnetization takes
place from the operating point P. Upon subsequent removal of this demagnetizing
force, the effect of a minor hysteresis loop prevents the operating point from
returning to point P. Instead the operating flux density assumes a somewhat
smaller level. It is useful to note, however, that the foregoing situation does not
duplicate itself when the PM material is ceramic. For permanent magnets con-
structed of ceramic, the operating point P is virtually re-established upon removal
of the demagnetization associated with an excessive armature current.

TABLE 8-1 CHARACTERISTICS AND APPLICATIONS OF DC MOTORS

Starting torque Max. running torque,

Type (%) momentary (%)

Shunt, constant speed Medium—usually limited to less Usually limited to about 200
than 250 by a starting resistor by commutation
but may be increased

Shunt, adjustable speed Same as preceding Same as preceding

Compound High—up to 450, depending Higher than shunt—up to 350

upon degree of compounding

a eee eee eee

Series Very high—up to 500 Up to 400

Source: By permission from M. Liwschitz-Garik and C. C. Whipple, Electrical Machinery,

vol. 1
(Princeton, N.J.: D. Van Nostrand Co., Inc., 1946).
 Sec. 8-6 Operating Features and Applications 369

8-6 OPERATING FEATURES AND APPLICATIONS

The dc motor is often called upon to do the really tough jobs in industry because of
its high degree of flexibility and ease of control. These features cannot easily be
matched by other electromechanical energy-conversion devices. The de motor
offers a wide range of control of speed and torque as well as excellent acceleration
and deceleration. For example, by the insertion of an appropriate armature cir-
cuit resistance, rated torque can be obtained at starting with no more than rated
current flowing. Also, by special design of the shunt-field winding speed adjust-
ments over a range of 4: | above rated speed are readily obtainable. If this is then
combined with armature-voltage control, the range of speed adjustment spreads to

a
Speed-regulation
or characteristic Speed control Typical application and
(%) (%) general remarks

5-10 Increase up to 200 by Essentially for constant-speed applications

field control; de- requiring medium starting torque. May
crease by armature- be used for adjustable speed not greater
voltage control than 2:1 range. For centrifugal pumps,
fans, blowers, conveyors, woodworking
machines, machine tools, printing presses
ti tlt Ni a

10-15 6:1 range by field con- Same as above, for applications requiring
trol, lowered below adjustable speed control, either constant
base speed by arma- torque or constant output
ture-voltage control
eal, ee. nae i SE ais le SeiE Te Ss
call
Not usually used but For drives requiring high starting torque
Varying, depending
of com- may be up to 125 by and only fairly constant speed; pulsating
on degree
pounding—up to field control loads with flywheel action. For plunger
25-30 pumps, shears, conveyors, crushers,
bending rolls, punch presses, hoists
EE
aan
I ———————E—
Widely variable, By series rheostat For drives requiring very high starting
high at no-load torque and where adjustable, varying
speed is satisfactory. This motor is
sometimes called the traction motor.
Loads must be positively connected, not
belted. For hoists, cranes, bridges, car
dumpers. To prevent overspeed, lightest
load should not be much less than 15 to
20% of full-load torque
ee
ee Se
 370 Direct-Current Motors Chap. 8

8:1. In some electronic control devices that are used to provide the dc energy to
the field and armature circuits, a speed range of 40: 1 is possible. The size of the
motor being controlled, however, is limited.
Table 8-1 lists some of the salient characteristics and typical applications of
the three types of dc motors. It is interesting to note that the maximum torque in
the case of the dc motor is limited by commutation and not, as with all other motor
types, by heating. Commutation refers to the passage of current from the brushes
to the commutator and thence to the armature winding itself. The passage from
the brushes to the commutator is an arc discharge. Moreover, as a coil leaves a
brush the current is interrupted, which causes sparking. If the armature current is
allowed to become excessive, the sparking can become so severe as to cause
flashover between brushes. This renders the motor useless.
Another point of interest in the table is the considerably higher starting
torque of the compound motor in comparison with the shunt motor. This feature
is attributable to the contribution of the series field winding. The same comment
is valid as regards the maximum running torque. In each case, of course, the limit
for the armature current is the same.

8-7 STARTERS AND CONTROLLERS FOR DC MOTORS

The limitations imposed by commutation as well as voltage-dip restrictions on the

source as set forth by the electric utility company make it necessary to use a
starter or controller on all dc machines whose ratings exceed 2 hp. A glance at
Eq. (8-2) discloses that at starting (n = 0) the armature current is limited solely by
the armature circuit resistance. Hence, if full terminal voltage is applied, exces-
sive armature currents will flow. This is especially so where large machines are
involved because the armature resistance gets smaller as the rating increases. In
addition to limiting the armature current, controllers fulfill other useful functions
as described in Sec. 4-12.
Depicted in Fig. 8-21 is a simple line starter which is used for small dc
motors. The operation is straightforward. Pushing the start button energizes the

0000000

Shunt field winding

_Istart
st ©

Figure 8-21 Line starter for a low-

Q horsepower dc motor.
Sec. 8-5 Permanent Magnet (PM) dc Motors 371

main coil which-then closes the main contactors M and the interlock M,. Note
that, when the main contactors close, voltage is applied to the armature winding
through the starting resistor R and simultaneously to the field winding. This ar-
rangement prevents ‘‘shock’’ starting because some time is needed before the
field flux is fully established. Note, too, that this starter has no provisions for
removing the starting resistor once the motor has attained its operating speed. In
small motors this is of little consequence and it makes for an inexpensive starter.
However, the starter resistor in this case does serve another purpose. When the
motor is stopped, the field winding is disconnected from the line. The energy
stored in the magnetic field then discharges through the starting resistor, prevent-
ing possible damage to the field winding.
Three types of magnetic controllers are in use today for controlling the
starting current, the starting torque, and the acceleration characteristics of de
motors. One type is the current-limit controller, which works on the principle of
keeping the current during the starting period between specified minimum and
maximum limits. It is not too commonly used because its success depends upon a
tricky electrical interlock arrangement which must be kept in excellent operating
condition at all times. No further consideration is given to this type. The second
type is the counter-emf type which is illustrated in Fig. 8-22. Two kinds of relays
are used in this controller. One is a light, fast-acting unit called an accelerating
relay (AR). The other is the strong, heavy-duty type previously discussed. The
accelerating relays appearing across the armature circuit in Fig. 8-22 are voltage-
sensitive devices. They are designed to close when the voltage across the coil
exceeds a preset value. For the controller under discussion the accelerating relay
Dynamic breaking
G M: resistor Lo
Y 1A 2A io

M OL.
f
Starting
(aR) resistors

Shunt field winding

560000

_|Stort Relay coil

Stop

(2a) AR Figure 8-22 Counter-emf magnetic

controller for a dc motor.
372 Direct-Current Motors Chap. 8

1AR is usually adjusted to ‘‘pick up’’ at 50% of line voltage and 2AR is adjusted to
close at 80% of line voltage.
Pressing the start button energizes coil M, which then closes interlock M,
and the main contactors M. Since the armature is initially stationary, the acceler-
ating relays are de-energized so that both steps of the starting resistor are in the
circuit. As the armature gains speed and develops an induced emf exceeding 50%
of line voltage, coil 1AR snaps closed, closing contactors 1AR. In turn, coil 1A is
energized, closing contactors 1A, which short out the first section of the starting
resistor. This then applies increased voltage to the armature, which furnishes
further acceleration. When the armature induced emf exceeds 80% of line volt-
age, accelerating relay 2AR closes. This excites coil 2A, which shorts out the
second section of the starting resistor. The motor then accelerates to its full-
voltage operating speed.
The controller of Fig. 8-22 is also equipped with a resistor that shunts the
armature terminals. It is called a dynamic breaking resistor because it serves to
bring the motor to a quick halt by dissipating energy obtained from the rotating
armature inertia. The diagram of Fig. 8-22 shows that a path is established for the
dynamic breaking resistor when the stop button is closed and coil M becomes de-
energized. This action closes the normally closed contactor M which is placed in
series with the dynamic breaking resistor. Since the motor is still turning at a high
speed with stored kinetic energy in its inertia and because the shunt field remains
energized, the armature behaves as a generator which in turn delivers current to
the dynamic breaking resistor. The subsequent expenditure of energy as heat in
the resistor acts quickly to drain the stored kinetic energy of the armature, thus
reducing the speed to zero rapidly.
The counter-emf controller has the advantage of providing a contactor clos-
ing sequence which adjusts itself automatically to varying load conditions. Fur-
thermore, this is accomplished in a manner that maintains uniform accelerating
current and torque peaks. There can be no question about the desirability of such
starting performance; however, there is one disadvantage. The contactor closing
sequence is based on the assumption that the motor will start on the first step. If it
fails to do so, all subsequent operations cannot take place. Furthermore, the
starting resistor is in danger of burning up. To avoid such occurrences general-
purpose dc motors are most often equipped with definite time-limit controllers.
Figure 8-23 shows the schematic diagram of such a controller. After a relay coil is
energized, the corresponding contactors do not close until the elapse of a preset
time delay. The time delay is achieved either by means of magnetic flux decay, a
pneumatic device, or a mechanical escapement.
Pressing the start button energizes coil M and closes interlock M, and the
main contactors M, thereby applying voltage to the armature winding through the
starting resistor. A definite time after the armature of relay coil M closes, contac-
tors Ty in the control circuit close, regardless of whether the rotor is turning or
not. This energizes coil 1A, which closes contactors 1A and thus shorts out the
first section of the starting resistor. At a preset time delay after coil 1A is ener-
 Sec. 8-8 Speed Control by Electronic Means 373

Figure 8-23 Definite time-limit control-

ler for a dc motor.

gized, contactors T;,4 close. Coil 2A becomes energized, removing the entire
starting resistor. The motor then assumes its normal operating speed.
In the definite time-limit controller the preset time intervals between the
closing of contactors are adjusted to obtain smooth acceleration and uniform
current peaks for average load conditions. If a heavy starting condition occurs
and the motor fails to start on the first step, the first accelerating (time-delay)
contactors close anyway. This allows an increased starting torque to be devel-
oped. Thus the motor is made to ‘“‘work harder”’ if it does not start on the first
step. Accordingly, whenever a controller must be selected for a general-purpose
motor, it is wiser to prescribe the definite time-limit type. The reason is that as a
rule for general-purpose applications the starting conditions are not well known.

8-8 SPEED CONTROL BY ELECTRONIC MEANS

and voltage rat-
The development of reliable thyristors (SCRs) with high current
adjustment
ings has made the speed control of dc motors by armature voltage
ent of the armature voltage offers the
more viable. Keep in mind that adjustm
speed as well as operatio n at high effi-
advantage of speed control below base
Moreove r, the control of speed can
ciency coupled with good speed regulation.
under load conditio ns or no-load.
be achieved whether the motor is operating
Single-Phase, Half-Wave Thyristor Drive
phase source is
Figure 8-24 shows the arrangement that is used when a single-
for motors which are equippe d with permanent
available. It is frequently used
or less. Because a single thyristo r is used in this
magnet fields and rated for 1 hp
374 Direct-Current Motors Chap. 8

Ir

Field source

Figure 8-24 Armature voltage control of a de motor through use of a single-

phase, half-wave controlled rectifier. The motor may have either a permanent
magnetic field or a separate source of excitation.

circuit, control of the firing point, 0;, by the gate signal can occur anywhere
between 0 and a radians. Illustrated in Fig. 8-25 is a typical situation that can
prevail for a firing angle of 6; = 90° during steady-state conditions. Time in this
diagram is assumed to be measured after steady state is reached. Moreover, the
load torque to be supplied by the motor is assumed to be at such a level as to
require an average armature current of J, projected over a full cycle. To produce
this average armature current, however, it is necessary for the actual variation of
the armature current to have such an instantaneous variation that integration over
the conduction period yields an average value of J,. The conduction period is the
time during which the thyristor is conducting and generally this is represented by

6. ie 6. * Of (8-56)

where 6 is the angle in the positive excursion of the supply voltage where the gate
signal fires the thyristor and 6, is the extinction angle of the thyristor.
The extinction angle is determined as that point in the cycle where the actual
armature current i, returns to zero, which in turn depends upon the relationship of
the instantaneous value of the supply voltage vu; and the variation of the armature
counter emf e,. This e, must always assume such a position relative to u; that it
produces a time variation in i, which yields a value of the average armature
current that allows a torque to be developed to meet the needs of the load. It is
instructive to note here that the instantaneous armature current i, undergoes an
increase in value from 6; to 0,,, where 6,, corresponds to the time instant where
Sec. 8-8 Speed Control by Electronic Means 375

U; = @,. In this interval the voltage across the armature winding inductance in Fig.
8-24 is positive. Consequently, an increase in current from zero takes place which
can be described by a form of Faraday’s law as

®m
wipes i i. (Wey) elt (8-57)
where L, denotes the armature winding inductance. Beyond 6,, the quantity (vu; —
€,) becomes negative. By Eq. (8-57) the effect of these negative volt-seconds,
appearing across the armature inductance, is to cause the armature current to
diminish. In fact, the extinction angle is that point in the cycle where the sum of
the negative volt-seconds (i.e., the double-crosshatched area in Fig. 8-25) is equal
to the sum of the positive volt-seconds (i.e., the single-cross-hatched area).
When the load calls for an increased developed torque by the motor, the
motor responds by reducing its speed (and therefore e,) by an amount that allows
a sufficient increase in positive volt-seconds to yield a greater instantaneous cur-
rent flow, i,. Correspondingly, a larger average value of J, is made to exist over a
full cycle thereby meeting the increased torque demand of the load. The variation
with time of e, and the motor speed, w,,, are represented in Fig. 8-25 by the same
curve. It is understood, of course, that the curve is drawn to a different scale.
The same applies to the curves marked J, and T,. In the interval from 6 to 6,
observe that the motor is receiving a pulse of energy from the source which
manifests itself as a slight increase in speed. The energy stored in the rotating
mass, consisting of the rotor of the motor and its attached mechanical load (1.e.,
the total inertia J), is thereby increased. However, because there is no armature
current flow in the interval from 6, to (27 + 6;), the motor finds itself coasting
along. During this period, energy is delivered to the load by extracting it from the
kinetic energy stored in the rotating mass. The result is a dip in speed. When the

Eg,w (to different

scales)

Figure 8-25 Waveshapes of the various electrical quantities and motor speed and
load torque associated with the system of Fig. 8-24. The thyristor firing angle is
illustrated for 4, = 90°.
376 Direct-Current Motors Chap. 8

motor is called on to deliver full load torque at low speeds, the speed dip will be
greater than when it operates at speeds close to its base value.
Although small dc motors do operate successfully in the circuitry of Fig.
8-24, it is not without notable shortcomings. For example, the need to meet load
torque requirements in terms of an armature current averaged over a full cycle of
the supply frequency means that large pulses of armature current must flow during
the conduction period. This leads to much higher heating losses. Unless special
arrangements are made for cooling, it may be necessary to operate the motor at a
reduced horsepower rating. Another disadvantage is related to the fact that an
average dc current is made to flow through the ac source. If this source should be
a single-phase transformer, then it is possible for the transformer to operate in a
state of partial saturation. These problems are essentially avoided by operating
from a three-phase source.

Three-Phase, Half-Wave Thyristor Motor Drive Circuit

The circuit diagram for this mode of operation is depicted in Fig. 8-26. A three-
phase voltage source is assumed to be available with an accessible neutral. It is
the line-to-neutral voltages which appear across the series combination of the
individual thyristors and the armature winding circuit. The motor field is pre-
sumed to be originating from a permanent magnet or from a separate dc source
which itself may involve rectification of an ac supply. The application of the
three-phase voltage set of Fig. 8-27 in the drive circuit of Fig. 8-26 causes the
thyristors S,, S,, and S, to fire in sequence over a potential conduction period of
120°. The actual conduction period depends on when the firing gate signal is

I¢
Three-phase source
Field source

Figure 8-26 “Schematic diagram of the three-phase, half-wave controlled rectifier

drive circuit for the dc motor.
Sec. 8-8 Speed Control by Electronic Means 377

Voltage ,V

Figure 8-27 Waveforms of the balanced three-phase source. The heavy lines
indicate the 120° intervals associated with each phase voltage during which thyris-
tor firing can be initiated.

applied during this 120° interval. A glance at Fig. 8-27 should make it apparent
that between 6 = 30° and 6 = 150° a positive potential appears across thyristor
S.. Accordingly, it can act as a switch (hence the notation S for the thyristor) that
serves to put a portion or all of v,, across the armature terminals of the motor.
Simultaneously, during this period, the voltage appearing across the other two
thyristors is negative and so S, and S, are kept ina cutoff state. Similarly, in the
interval from @ = 150° to 6 = 270°, thyristor S, is prepared for firing because here it
now possesses a positive forward voltage while thyristors S, and S, have negative
voltages on their anode-cathode terminals. In the interval from 270° to 390° thyris-
tor S. is ready for conduction. This sequenced firing of the thyristors results in a
rectification of the three-phase line-to-neutral voltages which has a controllable
average value that is determined by the point at which the thyristors are allowed
to conduct through action of the gate signals. It is this nonzero average voltage
that drives the dc motor.
Let us now obtain an expression that relates this controllable average volt-
age to the firing angles of the thyristors. Call 6, the thyristor firing angle and for
convenience let 6; = 0° correspond to 6 = wt = 30° in Fig. 8-27. Because 6,can be
assigned any value from 30° to 150° (or 7/6 to a/6 + $27), it follows that the
expression for the average voltage over the first conduction period corresponding
to firing at angle 6; is
1 @,+(S/6)r
Vin = Vac = V, sin (wt) d(wt) (8-58)
% O¢+7/6

where V, denotes the peak value of the line to neutral voltage and V,,, refers to the
upper
motor armature voltage. It is important to note in this equation that the
378 Direct-Current Motors Chap. 8

limit of the integral is written in recognition of the fact that the thyristor stays in
the ON state for a period of 120° (i.e., 377) once it is allowed to initiate conduc-
tion. Performing the integration and inserting the limits, the desired result is
found to be
Vic = Vn = 0.827V, cos 6 (8-59)
This equation clearly reveals the controllable aspect of the rectified ac source
voltage. A setting of the thyristor gate circuit to yield 6; = 0° means that a
maximum dc voltage appears across the armature terminals. At the other extreme
a setting of 67 = 90° reduces this voltage to zero. Accordingly, there is at our
disposal a conveniently adjustable armature voltage source to permit speed con-
trol below the base speed of the motor.
An illustration of the speed control capability of the drive circuit of Fig. 8-26
for three values of the thyristor firing angle is shown in Fig. 8-28. Part (a) of this
diagram depicts the situation in the drive circuit when the motor is at no load and
6, = 0°. Thyristor S, initiates conduction at 6 = 30° and so applies the value Of Van
to the armature winding of the motor over the period from 6 = 30° to 6 = 150°. A
similar situation prevails when phases b and c become most positive. The average
value of the rectified source is at its maximum. If the armature winding resistance
drop is assumed to be zero, then E, becomes equal to this value and w,, takes on a
corresponding value consistent with the state of its field excitation. Based on the
assumption that the maximum rectified voltage is the same as the rated dc motor
voltage and the flux field is at its maximum setting, the resulting speed is the base
speed. In Fig. 8-28(b) the firing angle is delayed 60°. Consequently, the initial
value of v,, that appears on the armature circuit is its peak value along with the
variation of u,,, that occurs over the next 120°. In other words, that section of v,,
is placed across the armature circuit that corresponds to curve def. The average
value of this section is 0.413V, and is consistent with a firing angle of 60°. Of
course, this process is repeated for each third of the cycle involving in succession
thyristors S, and S,. Now the speed is at a level one-half the base speed again
based on the assumption of negligible armature winding resistance drop and thy-
ristor drop. Figure 8-28(c) depicts the situation as it exists for 67 = 90°. Thyristor
S, initiates conduction at 6 = 120° (i.e., 8 = 30° + 4, = 30° + 90°) and continues in
the ON state until 6 = 240°. Accordingly, the portion of v,, that appears across
the motor armature winding is the section represented by d'e'f’, which already
has an average value of zero. Hence the motor speed becomes zero.
When the de motor is operating at a subbase speed under no-load conditions,
what is the effect of applying load? The demand for increased torque to be deliv-
ered to the attached mechanical load is met by a decrease in speed just as it is in
the conventional dc motor. Although the average armature voltage and current
are derived from rectification, Eqs. (8-22) to (8-24) continue to be valid provided
that the terminal average values are used. As a result in those instances where the
armature winding voltage drop is moderate or small, relatively small changes in
 NO%S—+1«—

"06 = 6 (©) ‘.09 = “6 (9) £.0 = “P IOJ S10}SLIAY} 94} Jo BuLIY-[[NJ (e) :“m poads 10j0uN

379
ay} pue ’y JWo poonpul sinjeULIe oY) JO sonjeA deVJOAe oY} oe UMOYS OSTY ‘o[sue SULIT
JO}SLIAY) OY} JO SaNTeA JAY} JOJ 9oINOS aseyd-se1Y} PoyNjsel oy} JO sodeysoaem 7-g ANSI
(9)

—+}«—
NO°s
Nos
—>
|.
3 *n998'0-
(9) (9)

°S—»}e—NO
%S—+}e—NO
NO°S—>}«—NO4S—+}«—-NO °S —+|

Ik--NO
—>+
o0=78
°S
oOL 2 OSI
380 Direct-Current Motors Chap. 8

the speed accompany the changes in load. This matter is illustrated in the follow-
ing example.
Example 8-12
A dc motor, which is rated at 550 V and 60 hp, draws 30 A when it delivers rated
torque at speed of 800 rpm. The field is set at a fixed value that gives the motor a
torque and speed parameter of 5.94 N-m/A (or V/rad/s). The motor is supplied from
a three-phase, 60-Hz, 1100 line-to-line source through a half-wave thyristor drive
circuit. The armature circuit resistance is 1 ©. Assume that the thyristors operate
with negligible voltage drop.
(a) Determine the required firing angle when the motor operates at rated load
conditions.
(b) Compute the change in speed that occurs when the load torque is reduced
to one-half the rated value. What is the percent speed change?
(c) Find the firing angle that is needed when rated torque is to be delivered at a
speed of 500 rpm.

Solution (a) Kirchhoff’s voltage law employing average values appearing at the
motor terminals allows us to write

Vac = Vin = Ea ar Ths

0.827V, cos 6 = (5.94)@m + 30(1)

Here

Wm m = 800 (Z)
30 == 83.78
5 Tad/srad/

1100
Vp= V2 (Hoe) = 898.1 V
4 V3
Thus

742.73 cos 0; = 497.65 + 30 = 527.65

6, = 44.73°
(b) With the load torque reduced to one-half, the required average armature
current is also halved. Since the firing angle is maintained, we have

527.65 = 5.94, + 15
or
®m = 86.3 rad/s
Expressed in revolutions per minute, the speed is

Nm = Wm (2) = 86.3 (=) = 824.1 rpm

The change in speed from the rated value expressed in percent is

824.1 — 800
800 xX 100 = 3.02%
Sec. 8-8 Speed Control by Electronic Means 381

(c) We proceed as in part (a). However, in this part the speed is

Om = ;(83.78) = 52.36 rad/s

Hence

742.73 cos O¢ = (5.94)(52.36) + 30 = 341 V

O¢ = 62.7°

Chopper Drives

There is an area of application of dc motors where dc energy is available, often in

the form of batteries, and where operation below base speed is extremely impor-
tant. This is the case for example with forklifts, golf carts, electrically powered
trucks, and even rapid-transit cars. Of course, as already pointed out in the
preceding section, one workable scheme is to use a series resistor. But we al-
ready know that this solution carries with it such serious shortcomings as poor
efficiency, poor speed regulation, and high cost. A popular method of providing
armature voltage control today in such applications is to use the thyristor in a
pulse width modulation mode or a pulse frequency modulation mode. In the
former the dc source voltage is switched on and off for a variable part of a cycle in
the interest of adjusting the level of average voltage applied to the motor armature
winding. In the latter a pulse of voltage of fixed width, which is derived from the
de source, is repeated more or fewer times over a given time period. A typical
schematic arrangement of the circuitry is shown in Fig. 8-29. Because the princi-
ples are the same, attention is directed here to just one of the schemes, namely,
pulse width modulation of the dc source voltage.
The control circuitry that is connected to the thyristor has a twofold pur-

Figure 8-29 Chopper drive circuit for

armature control of a dc series motor.
 382 Direct-Current Motors Chap. 8

Vp
Vde = Vin e as

O To Tp Time

loiles
I: Io
Ia
Ipf | Ip
| |
| |
| |
| |
oe us t O To Tp :
(a) | (b)

Figure 8-30 Voltage and current waveshapes appearing at the armature terminals of the
motor in Fig. 8-25: (a) thyristor fired for low-speed operation; (b) thyristor fired for high-
speed operation.

pose. First, it serves to turn the thyristor on over a variable part of a period Tp as
illustrated in Figs. 8-30(a) and (b). In the time interval from 0 to Jp the gate signal
puts the thyristor into a conductive state thereby passing the battery voltage to the
armature winding. Second, at time 7), a reverse voltage is instantaneously (mea-
sured in the tens of microseconds) applied to the thyristor thereby cutting it off.
At the end of the period represented by Tp, the process is repeated. In essence
then the control circuitry acts to ‘“‘chop”’ the de supply voltage before it is applied
to the motor armature winding. These drive circuits are called chopper drives for
obvious reasons. The average value of the dc voltage that is transferred to the
motor is quite simply given by

T,
Vika saan (8-60)
Pp

where the meaning of the quantities are self-evident from their identities in Fig.
8-30.
The frequency at which the thyristor is pulsed is critical to the design of the
control circuitry if large peaks in the armature current are to be avoided. Ina
properly designed chopper drive circuit, the variation of the armature current
looks very much like that illustrated in Fig. 8-30 during steady-state operation.
For equilibrium to exist the increase in current during time Ty is equal to the
decrease in current during the time (7p - Ty). In terms of the quantity of volt-
seconds applied to the armature inductance, we can write as the increase in
Sec. 8-8 Speed Control by Electronic Means 383

current in period Tp

AT = ck o
et or B a UY ni eR aRa)] ae
i
[Vp = Oe, Site T,Ra)\To
Te (8-61)

since the integrand is essentially constant. The motor voltage V,, is given by
Eq. (8-60). Of course, the effect of this increased current is to augment the mag-
netic energy stored in the armature field. Then, following cutoff of the thyristor at
time Tp, the energy is released to the motor through the circuit established by the
free-wheeling diode D,,. The rate of decay of the armature current during the
time (Tp — Ty) is determined by the time constant of the armature circuit which is
often very large compared to the cycle time of the pulses. For this reason the
decay is illustrated as a straight line. Because of the important role of the arma-
ture inductance in storing sufficient magnetic energy, chopper drives are fre-
quently used with series motors. The series motor has larger armature inductance
than the shunt motor, although perfectly acceptable performance can be had when
the latter machine is used with an external series inductor.

Example 8-13
Refer to the de motor described in Example 8-12 and assume that it is to be operated
with a chopper drive from a 500-V dc source.
(a) Find the motor speed when the thyristor is made to operate with a ratio of
ON time Ty to cycle time Tp of 0.4. Assume that the motor delivers rated torque.
(b) Determine the required pulse frequency of the thyristors that serves to limit
the total change of armature current during the ON period of 5 A. Assume an
armature winding inductance of 0.1 H.
(c) Repeat parts (a) and (b) for T/Tp = 0.7.

Solution (a) From Eq. (8-60) the average value of the armature terminal voltage is

Vin = Jo Vp = 0.5(550) = 220 V

Tp
The corresponding value of the induced armature voltage then becomes
B= V,, ~ Lh, = 220 — BOC) = 190 V

Therefore
190
594
Om == — =
32 rad/s or 305.4 cA. rpm

(b) The time Jy is found directly from Eq. (8-61). Thus,

T = E(AT,)
= = ema
0.1(5) ZT WL x
';
Oy pe gor 0550 = 190 ARO D ae
Accordingly,
To 1.4 3 x =
P 0. 0.4 x 10 3 5 S
 Direct-Current Motors Chap. 8
384

and this leads to the following required pulses per second:

f= 71 = 55
10 |
= 286 Hz
(c) Now Eq. (8-60) yields

V,, = 0.7(550) = 385 V

E, = 385 — 30 = 355 V

Om = = = 64.81 rad/s or 618.9 rpm

Also, from Eq. (8-61),

aS) kee =
Ty = =e 2.56 X 10

=
Tp = eee = 3.66 x 1073

f= pe 273 Hz
T

PROBLEMS

8-1. A 5-hp, 113-V, 1150-rpm, shunt motor has an armature resistance of 0.2 0. When
delivering rated output the motor draws a line current of 40 A. The field winding has
a resistance of 65 1. Neglect the demagnetizing effect of the armature mmf.
(a) Find the motor speed when the motor develops a torque of 15 N-m.
(b) Determine the speed regulation of this motor. Assume the armature current at
no-load is negligible.
8-2. A 15-hp, 230-V, 1750-rpm shunt motor draws 56 A from the line at rated output. The
resistance of the armature circuit is 0.16 (0; the field winding resistance is 153 ©. The
field current is reduced to 80% of its full excitation value by rheostat control.
(a) If the developed torque is correspondingly reduced in order not to exceed the
rated armature current, find the new value of speed. Neglect saturation.
(b) Determine the new speed when the load demands the same torque.
8-3. The magnetization curve of a 10-hp, 115-V, 1000-rpm shunt motor is shown in Fig.
7-17. Each vertical diversion represents 10 V and each horizontal unit represents
0.4 A. The armature circuit has a resistance of 0.07 © and the field winding resis-
tance is 48 ().
At a given load condition the motor armature current is 60 A. The demagnetiz-
ing effect of the associated armature winding mmf is found to be 0.4 A expressed in
equivalent field amperes. Find the motor speed.
8-4. A 115-V de machine is equipped with both a shunt and series field winding the
resistances of which are respectively 60 0 and 0.15 ©. The armature winding resis-
tance is 0.2 ©. When operating as a shunt motor, it draws a line current of 40 A and
Chap. 8 Problems 385

develops a-torque of 31 N-m to a shaft load. When operating as a compound motor

supplying the same torque, the series field contribution to the pole flux is known to be
25%.
(a) Find the speed as a compound motor.
(b) Find the speed as a shunt motor.
8-5. A 40-hp, 550-V, 500-rpm dc machine is equipped with a shunt and series field winding
the resistances of which are 225 © and 0.25 Q, respectively. The resistance of the
armature circuit is 0.36 ©. When the motor is operated as a series motor with an
armature current of 59 A, the motor speed is found to be 1120 rpm.
The machine is next operated as a cumulatively compounded motor drawing
the same armature current. At this load condition the shunt field contributes 1.5
times as much pole flux as does the series field.
(a) Find the speed of the compound motor.
(b) Compute the torque developed by the compound motor.
(c) What is the developed horsepower?
8-6. A 200-hp, 550-V, 450-rpm dc series motor draws a line current of 295 A at rated
output. The resistances of the armature and series field windings are 0.07 © and
0.04 Q, respectively. The torque is reduced to one-fourth the rated value. Find the
new operating speed.
8-7. Refer to the motor of Prob. 8-6.
(a) Evaluate the parameters that appear in Eq. (8-17) and show that the equation
that relates the speed w to torque T for this motor is
2843.6
= ———— — 2:95
ns i
(b) Next, evaluate K in Eq. (8-20) and compare the result with the result found in
part (a).
(c) Use the result of part (a) to find the speed corresponding to a torque of 3166 N-m.
(d) Repeat part (c) using the result of part (b).
(e) Compute the percent error caused by neglecting the winding resistances for the
series motor.
rated
. A 200-hp, 550-V, 450-rpm dc series motor draws a line current of 295 A at
output. The resistance of the series field winding is 0.11 ©. Determine both the
speed and torque when the load changes so that it draws a current of 180 A.
a field
A 10-hp, 230-V shunt motor has an armature circuit resistance of 0.5 0 and
and rated voltage the speed is 1200 rpm and the
resistance of 115 ©. At no-load
is 2 A. If load is applied, the speed drops to 1100 rpm. Determine :
armature current
(a) The armature current and the line current.
(b) The developed torque.
W.
(c) The horsepower output assuming the rotational losses are 500
Neglect armature reaction.
armature current
8-10. A 230-V, 50-hp, dc shunt motor delivers power to a load drawing an
is given by
of 200 A and running at a speed of 1100 rpm. The magnetization curve
l unit repre-
Fig. 7-17, where each vertical unit represents 20 V and each horizonta
sents 2 A. Neglect armature reaction. Also R, = 0.02 ©.
.
(a) Find the value of the armature induced emf at this load condition
(b) Compute the motor field current.
386 Direct-Current Motors Chap. 8

(c) Compute the value of the load torque. The rotational losses are 600 W.
(d) At what efficiency is the motor operating?
(e) At what percentage of rated power is it operating?
8-11. Refer to Prob. 8-10, and assume the load is reduced so that an armature current of
75 A flows.
(a) Find the new value of speed.
(b) What is the new horsepower being delivered to the load?
. The magnetization curve of Fig. 7-17 applies to a 230-V, 15-hp motor, where each
ordinate unit represents 20 V and each abscissa unit represents 0.2 A. At no-load,
the field circuit draws a current of 1.9 A from the line, while the armature current is
negligibly small. The speed at no-load is 1000 rpm. A mechanical load is then ap-
plied to the motor shaft. The measured armature current is found to be 40 A, causing
an armature circuit resistance drop of 16 V. Moreover, the demagnetizing effect of
the armature mmf is found to be 0.3 equivalent field ampere.
(a) Compute the load speed in rpm, neglecting the demagnetizing effect of the arma-
ture reaction.
(b) Find the speed for the specified load when the demagnetizing effect is accounted
for. (Hint: The procedure is similar to that outlined for generator behavior,
except that now the voltage equation for the motor is used in place of that for the
generator. Also, keep in mind that the magnetization curve is valid only for a
single speed, so that adjustments are needed.)
(c) Compute the percentage of change in speed from no-load to full-load for parts (a)
and (b); compare and explain the difference.
8-13. A 230-V, 15-hp, dc shunt motor has the magnetization curve shown in Fig. 7-17,
where each ordinate unit represents 20 V and each abscissa unit represents 0.2 A.
This curve is valid for operation at 1000 rpm.
(a) Rated voltage is applied to this motor, and at no-load the speed is found to be 900
rpm. Negligible armature current flows at this condition. Find the total field
circuit resistance for this point of operation.
(b) Load is applied to this motor, corresponding to an armature current of 40 A. If
the armature circuit resistance drop is 16 V and the demagnetizing effect of the
armature mmf is 0.3 equivalent field ampere, determine the new speed in rpm.
8-14. A 115-V, 15-hp, de cumulative-compound motor has the magnetization curve shown
in Fig. 7-17, where each ordinate unit represents 10 V and each abscissa unit repre-
sents 1 A. The motor speed at no-load and rated voltage applied is 1000 rpm.
(a) Assuming negligible armature current at no-load, find the field circuit resistance
for this operating point.
(b) The motor delivers power consistent with an armature current of 100 A, which is
also made to flow through the series field winding. The magnetization effect of
the series field is represented by 3 equivalent field amperes. The demagnetizing
effect of the armature winding is denoted by 0.7 equivalent field ampere. The
armature circuit resistance drop (including the series field) is 6 V. Find the speed
under load.
8-15. A 20-hp, 230-V, 1150-rpm, four-pole, dc shunt motor has a total of 620 conductors
arranged in two parallel paths and yielding an armature circuit resistance of 0.2 QQ.
When it delivers rated power at rated speed, the motor draws a line current of 74.8 A
and a field current of 3 A. Compute:
Chap. 8 Problems 387

(a) Theflux per pole.

(b) Thedeveloped torque.
(c) Therotational losses.
(d) Thetotal losses expressed as a percentage of the rated power.
Assume negligible armature reaction.
8-16. A 230-V, 10-hp, de series motor draws a line current of 36 A when delivering rated
power at its rated speed of 1200 rpm. The armature circuit resistance is 0.2 ©, and
the series field winding resistance is 0.1 ©. The magnetization curve may be consid-
ered linear. Effects of armature reaction may be neglected.
(a) Find the speed of this motor when it draws a line current of 20 A.
(b) What is the developed torque at the new conditions?
(c) How does this torque compare with the original value? Why?
8-17. A 250-V, 50-hp, 1000-rpm, dc shunt motor drives a load that requires a constant
torque regardless of the speed of operation. The armature circuit resistance is
0.04 Q. When this motor delivers rated power, the armature current is 160 A.
(a) If the flux is reduced to 70% of its original value, find the new value of armature
current.
(b) What is the new speed?
8-18. When a 250-V, 50-hp, 1000-rpm, de shunt motor is used to supply rated output power
to a constant-torque load, it draws an armature current of 160 A. The armature
circuit has a resistance of 0.04 , and the rotational losses are equal to 2 kW. An
external resistance of 0.5 9 is inserted in series with the armature winding. For this
condition compute:
(a) The speed.
(b) The developed power.
(c) The efficiency assuming the field loss is 1.6 kW.
Armature reaction is negligible.
8-19. A dc shunt motor has the magnetization curve shown in Fig. 7-17, where one vertical
unit represents 20 V and one horizontal unit represents 2 field amperes. The arma-
ture circuit resistance is 0.05. Ata specified load and a speed of 1000 rpm, the field
current is found to be 12 A when a terminal voltage of 245 V is applied to the motor.
The rotational losses are 2.5 kW, and armature reaction is negligible.
(a) Find the armature current.
(b) Compute the developed torque.
(c) What is the efficiency?
is
8-20. Answer whether the following statements are true or false. When the statement
false, provide the correct answer.
is necessarily
(a) The emf induced in a single coil of the armature of a dc machine
unidirectional.
most
(b) Putting a suitable limit on the amount of heat dissipated is the single
important design factor for all electric machines.
output shaft by
(c) A de motor responds to the need for increased power at its
its ampere-c onductor distribut ion relative to its field distributi on.
changing
e of 0.5 . At full-
8-21. A 230-V, dc shunt-wound motor has an armature circuit resistanc
armature winding draws 40 A and the speed is measure d at 1100 rpm
load the
corresponding to a field circuit resistance of 115 Q.
(a) Find the developed motor torque in newton-meters.
388 Direct-Current Motors Chap. 8

(b) The field circuit resistance is increased to 144 0. Find the new operating speed
assuming the developed torque remains constant in demand to load require-
ments. Neglect saturation.
(c) Calculate the efficiency in part (b). Assume that rotational losses amount to
600 W.
8-22. A 230-V, dc shunt motor has an armature circuit resistance of 0.15 © and a field
winding resistance of 68 ]. The pole flux at full field current is 0.1 Wb. The rated
torque is 100 N-m. This motor undergoes a change in speed of 6% from no-load to
rated load.
(a) Find the value of the torque constant Ky.
(b) Determine the no-load speed.
(c) What is the speed at rated load?
8-23. A 230-V, dc shunt motor has an armature winding resistance of 0.2 © and a field
winding resistance of 125 9. At no-load it operates at a speed of 1500 rpm.
Find the speed when it delivers a full-load rated torque of 75 N-m.
. Refer to the motor of Prob. 8-23. While operating at no-load, a field rheostat resis-
tance of 75 © is inserted in the field winding circuit.
(a) Find the new no-load speed.
(b) What is the value of torque that the motor can develop without exceeding the
rated armature current?
(c) Determine the speed at the load condition of part (b).
(d) Determine the percent change encountered in part (c) relative to the no-load
value.
(e) Find the speed regulation when the motor operates with the specified rheostat
resistance and delivers rated torque.
8-25. A 230-V, dc shunt motor has a field winding resistance of 115 0, an armature
resistance of 0.18 ©, and a no-load speed of 1250 rpm. Find the value of field
rheostat resistance that permits the motor to deliver a torque of 40 N-m at a speed of
2000 rpm.
8-26. A 460-V, 50-hp series motor delivers rated torque at a speed of 600 rpm. The arma-
ture circuit resistance is 0.06 © and the series field winding resistance is 0.04 0.
Determine the value of diverter resistance that is needed to permit the motor to
develop rated torque at 900 rpm.
8-27. Refer to the motor of Prob. 8-23 and answer the following questions.
(a) Determine the value of external armature resistance that is needed to allow the
motor to develop rated torque at 750 rpm. The field rheostat resistance is set at
zero.
(b) At this load condition how much larger is the loss in the external armature
resistance than in the armature winding resistance?
(c) If the rated armature winding copper loss is 5% of the rated output and is also
equal to the fixed losses, find the efficiency of the motor in part (a).
- A 230-V, 75-hp, 450-rpm shunt motor has an armature resistance of 0.043 Q and a
field winding resistance of 42 0. When operating at no-load with full excitation, the
speed is measured to be 1512 rpm.
(a) Determine the motor speed when it delivers rated torque and an external resis-
tance of 0.2 ( is placed in the armature circuit.
Chap. 8 Problems 389

(b) Find the speed regulation.

(c) Repeat parts (a) and (b) for an external resistance equal to 1 OF
8-29. A 550-V, 100-hp, 400-rpm shunt motor has an armature resistance of 0.19 Q and a
field winding resistance of 250 9. The no-load speed at full excitation is 420 rpm.
(a) An external armature resistance of 1.6 © is inserted into the armature circuit.
Determine the motor speed at rated load torque.
(b) What is the speed regulation for the condition of part (a)?
8-30. A 150-hp, 550-V, 690-rpm series motor has a series field and armature winding
resistances of 0.065 © and 0.1 , respectively. While the motor is delivering rated
output to a constant torque load, an external armature resistance is inserted to
reduce the speed to 450 rpm. Find the ohmic value of the external resistance.
8-31. Refer to the motor of Prob. 8-30. An external armature resistance of 1.2 0 is used
when the motor delivers one-half of its rated torque to a coupled load. Find the
motor speed.
8-32. A 15-hp, 230-V, 1750-rpm shunt motor draws 56 A from the line at rated output. The
resistance of the armature circuit is 0.16 Q and the field winding resistance is 153 0.
It is desired that 80% of the rated torque be delivered at a speed of 1000 rpm.
Determine the value of the external armature resistance that permits the motor to
operate at this condition.
to
8-33. A 230-V, dc shunt motor is used as an adjustable speed drive over a range of zero
zero to 1600 rpm are obtained by adjusting the armature
2400 rpm. Speeds from
full field
terminal voltage from zero to 230 V, with the field current kept fixed at
increasing the field cur-
value. Speeds from 1600 rpm to 2400 rpm are obtained by
maintained at 230 V. Ignore machine losses
rent, with the armature terminal voltage
and armature reaction effects.
speed range.
(a) The torque required by the load remains constant over the entire
Show the general form of the curve for armature current versus speed, over the
full speed range.
current is
(b) Instead of keeping the load torque constant, suppose that the armature
exceed a specified value. Show the general form of the curve for allow-
not to
able load torque versus speed, over the entire speed range.
a load requiring a
8-34. A customer is interested in buying a de shunt motor to supply
range of 500 to 2000 rpm.
constant torque of 525 lb-ft continuously, over a speed
(a) List three ways that this job can be done.
wer rating of the
(b) For each method of part (a), specify the base speed and horsepo
motor.
of control.
(c) List the outstanding advantage associated with each method
speeds of 2000 rpm and
8-35. Two adjustable-speed, dc shunt motors have maximum
g constant horsepower
minimum speeds of 500 rpm. Motor A drives a load requirin
requirin g constant torque. Neglect all
over the speed range. Motor B drives one
losses and armature reaction.
horsepower outputs are
(a) If speed adjustment is obtained by field control and if the
A, compute the arma-
equal at 500 rpm and the armature currents are each 100
ture currents at 2000 rpm.
of the individual motors to
(b) Specify the nameplate horsepower and speed ratings
jobs. Assume that at 500 rpm the torque for both loads
be purchased to do these
is the same, and equal to 1050 lb-ft.
390 Direct-Current Motors Chap. 8

8-36. Figure P8-36 shows a dc series motor, with a rheostat in parallel with the field
winding. The resistance of this rheostat is equal to two-thirds of the field winding
resistance. The motor is operating in the steady state, with the rheostat switch
open. The terminal voltage is 250 V, the armature current 40 A, the speed 700 rpm.
When the load torque is increased by 50% and the rheostat switch is closed, find the
new steady-state values of speed and armature current. Given values:
Armature resistance 0.2 , series field resistance 0.15 (.

Neglect saturation, armature reaction, and rotational power losses.

Figure P8-36

8-37. A dc series motor runs at 300 rpm, and draws a current of 75 A from the 500-V line.
The total resistance of the armature and field circuits is 0.4 2. Find the new steady-
state values of armature current and speed if the line voltage is increased from 500 to
600 V. Assume that the developed torque is to remain constant.
8-38. Read the following sentence carefully, and correct any wrong statement that you
may find in it:

A dc shunt motor operates under load in the steady state. At the time t = 0,
the field rheostat resistance is reduced; this increases the flux and thereby the
torque. So the motor accelerates and reaches a new steady-state speed which
is higher than it was before.
8-39. A dc shunt motor draws 10.0 A from its 220-V supply line. The resistance of its
armature circuit (including brushes, brush contacts, and interpoles) is R, = 19. The
field winding resistance is Rr = 300 (2. (There is no field rheostat.) The rotational
losses (hysteresis, eddy currents, friction, and windage) are 10% of the output
power. Find the efficiency, as accurately as the given information and the use of a
calculator permit.
8-40. A de shunt motor is rated at 3 hp, 115 V, 1000 rpm. At rated operating conditions,
the efficiency is 82.5%. The armature resistance (including brush contact) is 0.44 0,
the field circuit resistance is 145 (.
(a) Calculate the value of the induced voltage E, at rated operating conditions.
(b) If this motor is tested at no-load, with the supply voltage adjusted to the value
calculated under (a) and the speed kept at 1000 rpm (so that the rotational losses
remain the same as at full load), how much is the armature current under this
condition? (Note: The I,r, drop at no-load can be neglected.)
Chap. 8 Problems 391

8-41. Two identical dc machines are tested in ‘‘opposition’’ as shown in Fig. P8-41 (arma-
tures in parallel, shafts coupled, losses supplied from an electrical source). The
letters G and M stand for generator and motor.
(a) Which machine has the higher armature copper losses?
(b) Which machine has the higher field copper losses?
(c) Which machine has the higher core losses?
To each answer, add a brief explanation.

Figure P8-41

8-42. The magnetization curve of Fig. 7-17 shows the no-load characteristic for a dc shunt
ordinate
generator at 900 rpm, where each abscissa unit represents 0.5 A and each
field circuit
unit represents 20 V. The armature circuit resistance is 0.2 , the
equivalent
resistance (without field rheostat) 30. The effect of armature reaction is
rns per pole for every ampere of armature current.
to four demagnetizing ampere-tu
The field winding has 1250 turns per pole.
for the
(a) Find the resistance of the field rheostat (in series with the field winding)
900 rpm and carrying an armature
machine operating as a generator driven at
current of 125 A at a terminal voltage of 220 V.
ted. The
(b) The field rheostat remains as for part (a). The load is now disconnec
terminal voltage?
speed remains unchanged. What is the new value of the
of 420 V at a speed of
(c) Find the field rheostat resistance for a no-load voltage
1800 rpm.
a de shunt motor.
8-43. Depicted in Fig. P8-43 is the reversing controller circuitry for
g operation. Ex-
Identify the unmarked armature contacts and explain the reversin
interlock contacts F, and R, are placed in series with the
plain, too, why the auxiliary
Rev and Fwd switches.
shown in Fig. P8-44. The
8-44. A current-limit type controller for a dc shunt motor is
These relays have a light-
accelerating relays (AR) respond to armature current.
their contacts whenever the
weight armature so that they easily pick up and open
By means of a step-by- step procedure,
armature current exceeds the rated value.
controlle r. Be careful to identify the proper closing
describe the operation of this
wheneve r two or more contacto rs appear in series.
and opening sequence
of Fig. 8-26, sketch the wave-
8-45. In the three-phase half-wave controlled rectifier circuit
armature circuit when the firing
form of the source voltage that appears across the
value of this voltage?
angle of the SCRs is set at 45°. What is the average
of the SCRs is set at 75°.
8-46. Repeat Prob. 8-45 for the case where the firing angle
392 Direct-Current Motors Chap. 8

Le
+ Shunt field 1-

R, Ro Rs

1A 2A 3A

leat)
Forward

Stop

Figure P8-43

+9Ly Le
M Fuse S (iar) Ry Ro R3

1A 2A a 3A
}+—J

Shunt field

OLR Start

Figure P8-44
Chap. 8 Problems 393

8-47. A 25-hp, 250-V, dc motor is driven from a three-phase, 550-V source through a half-
wave thyristor drive circuit as displayed in Fig. 8-26. The armature circuit resistance
is 0.1 © and the developed torque parameter of this motor is known from design data
to be 6.25 N-m/A. The thyristor voltage drops and losses are negligible.
(a) When the firing angle of the SCRs is set at 30° and the motor is operating at no-
load, find the speed in rpm.
(b) Rated load torque of 575 N-m is applied to the motor shaft. Determine the new
operating speed in rpm.
(c) Calculate the speed regulation of this electronically controlled dc motor.
. It is desirable that the motor of Prob. 8-47 be made to deliver its rated torque of
575 N-m at a speed of 250 rpm.
(a) Find the required firing angle of the SCRs.
(b) Compute the speed regulation.
8-49. The motor of Prob. 8-47 is driven by the chopper circuitry of Fig. 8-29 from a 250-V
dc source. The thyristor control circuitry is arranged to provide a ratio of ON time to
cycle time of 0.5.
(a) Find the operating speed when rated torque is delivered by the motor.
(b) Determine the no-load speed and the speed regulation for the specified switching
ratio.
(c) Repeat parts (a) and (b) for To/T, = 1.0.
8-50. The motor of Prob. 8-47 is driven by a chopper drive of a 250-V source where the
ratio of ON-time to cycle time is 0.5. Determine the required pulse frequency of the
thyristors in order that the change in armature current during the ON period be
limited to 5 A when operating at rated load. The armature inductance is 0.05 H.
8-51. Repeat Prob. 8-50 for the case where the ratio of ON time to cycle time is 0.8.
 3
Single-Phase
Induction Motors

By far the vast majority of single-phase induction motors are built in the frac-
tional-horsepower range. Single-phase motors are found in countless applications
doing all sorts of jobs in homes, shops, offices, and on the farm. An inventory of
the appliances in the average home in which single-phase motors are used would
probably number beyond a dozen. An indication of the volume of such motors
can be had from the fact that the sum total of all fractional-horsepower motors in
use today far exceeds the total of integral-horsepower motors of all types.

9-1 HOW THE ROTATING FIELD IS OBTAINED

In its pure and simple form the single-phase motor usually consists of a distributed
stator winding (not unlike one phase of a three-phase motor) and a squirrel-cage
rotor. The ac supply voltage is applied to the stator winding, which in turn creates
a field distribution. Since there is a single coil carrying an alternating current, a
little thought reveals that the air-gap flux is characterized by being fixed in space
and alternating in magnitude. If hysteresis is neglected, the flux is a maximum
when the current is instantaneously a maximum and it is zero when the current is
zero. Such an arrangement gives the single-phase motor no starting torque. To
understand this in terms of the concepts discussed in Sec. 3-1 refer to Fig. 9-1. As

394
Sec. 9-1 How the Rotating Field is Obtained 395

Stator winding

Figure 9-1 Simplified diagram of the

single-phase motor.

a matter of convenience the distributed single-phase winding is represented by a

coil wrapped around protruding pole pieces; some motors do in fact use this
configuration. Assume that instantaneously the flux-density wave is increasing in
the upward direction as shown. Then by transformer action a voltage is induced
in the rotor having that distribution which enables the corresponding rotor mmf to
oppose the changing flux. To accomplish this, current flows out of the right-side
conductors and into the left-side conductors as illustrated in Fig. 9-1. Note that
this resulting ampere-conductor distribution corresponds to a space phase angle of
Ww = 90°. By Eq. (3-27) the net torque is therefore zero. Of course what this means
is that beneath each pole piece there are as many conductors producing clockwise
torque as there are producing counterclockwise torque. This condition, however,
prevails only at standstill. If by some means the rotor is started in either direc-
tion, it will develop a nonzero net torque in that direction and thereby cause the
motor to achieve normal speed. The problem therefore is to modify the configura-
tion of Fig. 9-1 in such a way that it imparts to the rotor a nonzero starting torque.
The answer to this problem lies in so modifying the motor that it closely
approaches the conditions prevailing in the two-phase induction motor. In accor-
dance with the general results developed in Sec. 4-1 we know that to obtain a
revolving field of constant amplitude and constant linear velocity in a two-phase
induction motor two conditions must be satisfied. One, there must exist two coils
(or windings) whose axes are space-displaced by 90 electrical degrees. Two, the
currents flowing through these coils must be time-displaced by 90 electrical de-
If the
grees and they must have such magnitudes that the mmf’s are equal.
in time but greater than 0°, a rotating field will still
currents are less than 90° apart
the resultant flux vector will be an ellipse rather than
be developed but the locus of
point
a circle. Hence in sucha case the linear velocity of the field varies from one
currents are 90° apart but the mmf’s of the two
in time to another. Also, if the
locus for the rotating field again results. Finally, if
coils are unequal, an elliptical
ced nor of a magnitude to furnish equal
the currents are neither 90° time-displa
but now the locus
mmf’s, a rotating magnetic field will continue to be developed
396 Single-Phase Induction Motors Chap. 9

will be more elliptical than in the previous cases. However, the important aspect
of all this is that a revolving field can be so obtained even if its amplitude is not
constant during its time history, and satisfactory performance can be achieved
with such a revolving field. Of course such performance items as power factor
and efficiency will be poorer than for the ideal case, but this is not too serious
because the motors are of relatively small power.
Appearing in Fig. 9-2 is the schematic diagram which shows the modifica-
tions needed to give the single-phase motor a starting torque. A second winding

Auxiliary winding

Squirrel —coge Figure 9-2 Schematic diagram of the

rotor resistance split-phase motor.

called the auxiliary winding is placed in the stator with its axis in quadrature with
that of the main winding. Usually the main winding is made to occupy two-thirds
of the stator slots and the auxiliary winding is placed in the remaining one-third.
In this way the space-displacement condition is met exactly. The time displace-
ment of the currents through the two windings is obtained at least partially by
designing the auxiliary winding for high resistance and low leakage reactance.
This is in contrast to the main winding, which has low resistance and higher
leakage reactance. Figure 9-3 depicts the time displacement existing between the

Figure 9-3 Phasor diagram showing

the time-phase displacement between
the auxiliary and main winding currents
at standstill.
auxiliary winding current J, and the main winding current Jy at standstill. Fre-
quently in motors of this design the J, and Jy phasors are displaced by about 45° in
time. Thus with the arrangement of Fig. 9-2 a revolving field results and so the
motor achieves normal speed. Because of the high-resistance character of the
auxiliary winding, this motor is called the resistance-start split-phase induction
motor. Also, the auxiliary winding used in these motors has a short time power
rating and therefore must be removed from the line once the operating speed is
 Sec. 9-2 Equivalent Circuit by the Double-Revolving Field Theory 397

reached. To do this a cut-out switch is placed in the auxiliary winding circuit

which, by centrifugal action, removes the auxiliary winding from the line when the
motor speed exceeds 75% of synchronous speed.

9-2 EQUIVALENT CIRCUIT BY THE DOUBLE-REVOLVING

FIELD THEORY

In this analysis the focus is on the operation of the single-phase induction motor
with only the main winding energized by a single-phase source. When this voltage
is applied to the coil at standstill, there is no starting torque because the pulsating
field distribution produced by this distributed winding merely produces trans-
former action as already described. It is useful at this point to introduce an
alternative explanation of this condition by replacing the alternating flux field by
oppositely revolving fields having half the amplitude of the alternating field and
each traveling at the same frequency at which the pulsating field alternates. We
resort to mathematics to illustrate this equivalence in precise terms.
As described in Sec. 4-1 the equation for an alternating magnetic field whose
axis is fixed in space is given by
b(a) = By», sin wt cos a (9-1)

where B,, denotes the amplitude of the sinusoidally distributed air-gap flux density
produced by a properly distributed stator winding carrying an alternating current
of frequency w, and a denotes a space-displacement angle measured from the axis
of the stator (or field) winding. If the trigonometric identity
sin x cos y = 3 sin (x — y) + 3 sin (x + y) (9-2)

is introduced into Eq. (9-1), the expression for the magnetic field becomes alterna-
tively
Ba
b(a) = 2 sin (wt — a) + > sin (wt + a) (9-3)

g field that
The first term on the right side of Eq. (9-3) is the equation of a revolvin
(see Sec. 4-1) and has an amplitud e equal to one-
moves in the positive a direction
field. The second term is the equation of a
half that of the original alternating
also one-half that of the alternati ng field and
revolving field whose amplitude is
a direction . The field that moves in
whose direction of rotation is in the negative
referred to as the forwardl y rotating field or
the positive a direction is sometimes
nds to the
the positive sequence. Similarly, the backwardly rotating field correspo
the forward direction is that direction in
negative-sequence field. By definition
which the single-phase motor is initially started.
not equipped
The absence of a starting torque in a single-phase motor that is
be underst ood in terms of effects
with an auxiliary starting device may now
of the two revolv-
produced by the forward and backward fields. The amplitudes
 Single-Phase Induction Motors Chap. 9
398

no net revolving field to

ing fields in Eq. (9-3) are the same and thereby yield
ng fields in the motor produce effects
produce a starting torque. The two revolvi
that are equal and opposite.
at standstill is
The equivalent circuit of the single-phase induction motor
as a transfo rmer with the secondary
easy to identify because it is behaving simply
our experie nce with transformers
short-circuited. Accordingly, drawing from
circuit as shown in Fig. 9-4. Appear-
(see Fig. 2-12), we can draw the equivalent
the exact equival ent circuit of a
ing in Fig. 9-4(a) is the conventional form of
the presenc e of the air gap in
transformer. The exact version is cited here because
and consequ ently its effect is
the motor calls for appreciable magnetizing current
note that the core loss resistor ,
no longer small and cannot be ignored. Also,
omitted for
which is found in parallel with the magnetizing reactance, is here
simplicity. It can always be added later, if it is needed. Moreover, r5 denotes the
winding
rotor winding resistance per phase referred to the stator and x} is the rotor
9-4(b) depicts the equival ent
leakage reactance per phase, stator referred. Figure
circuit of Fig. 9-4(a) modified to bring into evidenc e the interpr etation based on
the double-revolving field representation of the alterna ting field produc ed by the
energized main winding, which is designated by subscript m. In recognition of the
fact that the forward and backward revolving fields have amplitudes that are one-
half the value of the alternating field, it is necessary to apply a factor of 0.5 to the
parameters appearing between the terminals where the emf induced by these fields
appears. This is necessary to preserve the equivalence of E,+ E, in Fig. 9-4(b) to
E,, in Fig. 9-4(a). Of course, E; denotes the emf produced by the forward field in
the rotor winding referred to the stator, and, similarly, E, is the emf generated by
the oppositely rotating field also referred to the stator winding.

ry %

(a) (b)

Figure 9-4 Equivalent circuit of the single-phase induction motor at standstill

with only the main winding energized. (a) Conventional configuration; (b) modi-
fied configuration to stress the double-field equivalence of the alternating field
caused by the main winding.
Sec. 9-2 Equivalent Circuit by the Double-Revolving Field Theory 399

When the single-phase motor is given a starting torque and allowed to reach
a speed n in the-forward direction, the rotor is operating at a slip s which is defined
in the usual way as

ce (9-4)
In the part of the equivalent circuit associated with the forwardly revolving field
the effect of s is to put into evidence the resistance r3/2s and the leakage reactance
x3/2. The rationale that yields this result is the same as that for the balanced
three-phase induction motor since the situation between the rotor and the forward
field is the same. However, the situation is different between the forwardly turn-
ing rotor and the backwardly revolving field. Now the slip of the rotor relative to
this field is
oe Oy n
Sh:=z vas Lt +i: (9-5)

which also reveals that this slip is greater than unity. By employing Eq. (9-4) in
Eg. (9-5) it is possible to express the backward slip s, in terms of the forward slip
s. Thus
SS piles. = 2. o28 (9-6)

Accordingly, the parameters associated with the section of the equivalent circuit
where the effects of the backward field are indicated must be expressed in terms of
resistance (r3)/2(2 — s) to represent the power transferred across the air gap along
with a leakage reactance of x3/2. Thus, the equivalent circuit of the single-phase
motor operating at a slip s assumes the final form shown in Fig. 9-5. The current
I, is the forward rotor current per phase referred to the stator and J, is the
backward rotor current per phase referred to the stator.
In normal operation of the single-phase motor the slip can be expected to lie
in the range 0 < s = 0.15. This means that the impedance of the parallel combina-

ry mG

Figure 9-5 Equivalent circuit of a

single-phase induction motor operating
at a slip s base on the double-revolving
field theory.
 Single-Phase Induction Motors Chap. 9
400

tion associated with the forward field, namely

Lites leis (0-7)

Z Z. Is, 5 I
a lke +okD)

will be much larger than the impedance associated with the backward field
. ( habia x5)
Zrdappioitia aa: (9-8)
ee
2 2 ro 5 ’
> ade he Bea

Consequently, since each of these impedances carries the same current, the mag-
nitude of E;> E,. In turn the magnitude of the forward field, ¢;, which generates
E;is much greater than the magnitude of the backward field, @, , which is responsi-
ble for E,. Figure 9-6 shows a simplified representation of Fig. 9-5 expressed in
terms of the forward and backward impedance symbols introduced in Eqs. (9-7)
and (9-8).
Figure 9-7 depicts the forward and backward fields moving in opposite direc-
tions but frozen in time at a specific instant. This representation shows how dy
and ¢, cooperate to produce a resultant pole flux ¢,. It is especially instructive to
note that operation of the single-phase motor at s # 1 does in fact result in a pole
flux that rotates with time. The magnitude of this resultant is not constant. For
example, as time elapses in Fig. 9-7, an instant is reached when ¢, and @, will lie
along the horizontal opposing one another to yield a resultant that is the differ-
ence, (¢¢ — ¢,). Ninety degrees later in time the two fluxes will again be collinear
but now the resultant is their sum. Accordingly, the resultant flux is a revolving
one with an elliptical locus.

Reference line
& Pa , (axis of stator
winding)

by = 4 |

bp = 2

Figure 9-6 Simplified rep- Figure 9-7 Illustrating how a re-

resentation of Figure 9-5 sultant revolving field is obtained
through use of Eqs. (9-7) _ when the motor operates at a
and (9-8). slip # 1.
 Sec. 9-3 Equivalent Circuit Via Symmetrical Components 401

It is useful to-note too that at standstill when s = 1 and dr = @,, the resultant
flux , always. lies along the vertical in Fig. 9-7 and so yields an alternating field
rather than a revolving one.

9-3 EQUIVALENT CIRCUIT VIA SYMMETRICAL COMPONENTS

The single-phase motor can be analyzed as a special case of an unbalanced two-

phase motor where in the extreme case the second phase is entirely missing. In
turn, unbalanced two-phase motors can be treated in terms of balanced two-phase
systems through the use of symmetrical components. The underlying theory of
symmetrical components for this purpose is fully exposed in the next chapter and
the interested reader is encouraged to read Secs. 10-1 to 10-3 prior to continuing
with this section if the intention is not to proceed further without first seeing the
derivation of the results cited here.
Chapter 10 demonstrates that any two-phase unbalanced set of voltages, V,
and V,, can be replaced by two sets of balanced two-phase voltages of opposite
sequence. Here subscript a refers to phase a and subscript b to phase b. The first
set is called the positive sequence set and the reversed set is identified as the
negative sequence. (Respectively, these are analogous to the forward and back-
ward fields described in Sec. 9-2.) The balanced positive sequence set is repre-
sented by V; and —jV, where subscript | is used to denote the positive sequence
and the —j attached to the second component of the balanced set stresses the 90°
lag in time between the two balanced voltages. Similarly, the second balanced set
is represented by V2 and jV2 where subscript 2 denotes the negative sequence
which, of course, is assured by attaching j to the second component. By this
scheme the unbalanced voltages of phases a and b can now be expressed in terms
of the two balanced sets as
v, = Vi ai V7 (9-9)

V, = -jVi + jV2 (9-10)

The theory of symmetrical components also provides the instructions on
how to calculate the positive and negative sequence components from the original
unbalanced set. The results are

Vi; = (Va + iVb) (9-11)

V2 = 3(Va — jVo) (9-12)
The results contained in Eqs. (9-9) through (9-12) lead to a very useful and
interesting physical interpretation of the analysis of the unbalanced two-phase
motor. Each phase voltage consists of a component of a balanced two-phase set,
one positive and the other negative. The positive-sequence components in each
unbalanced phase voltage, V; and —j V,, can be considered to cooperate in creat-
ing a revolving magnetic field that interacts with the rotor winding to produce a
 Single-Phase Induction Motors Chap. 9
402

e compo-
positive-sequence torque 7,. At the same time the negative-sequenc
nents in each phase, V2 and —j V>, can be considered to combine in creating a
revolving magnetic field of reversed direction that interacts with the rotor winding
to produce a negative-sequence torque 7>. Of course the rotor responds to the
resultant torque, which is the difference between 7; and T>. A schematic repre-
sentation of this situation is depicted in Fig. 9-8. Note that for convenience the
rotor is treated as consisting of two rotors that are mechanically coupled. The
positive-sequence voltage set (V, and —jV,) produces 7;, and the negative-se-
quence set produces 7). All the known techniques of balanced operation can now
be used to determine 7, and also 7), and the results can then be superposed to
obtain the resultant. Unbalanced operation is thereby treated entirely in terms of
the techniques of balanced operation. Accordingly, no new theory is needed to
treat the motor analysis beyond this point.
The equivalent circuit of the unbalanced two-phase motor actually consists
of two parts: one representing the motor response to the positive-sequence volt-
age set and the other representing the response to the negative-sequence set.
Because each voltage sequence is a balanced set, the equivalent circuit expressed
ona per-phase basis is similar to that used for the balanced three-phase motor [see
Fig. 4-7(f)], which is repeated in Fig. 9-9(a). The applied stator voltage is the
positive-sequence voltage V,. The quantity r, is the stator winding resistance.
The quantity x, is the total magnetizing reactance associated with the magnetic
circuit of the motor. The stator referred rotor leakage reactance per phase is
denoted as x, and r} is the rotor resistance per phase referred to the stator. Use
of r3/s implies that the rotor travels in the direction of the positive-sequence
rotating field.
The equivalent circuit of the unbalanced two-phase motor in response to the
balanced negative-sequence set is shown in Fig. 9-9(b). However, since the rotor
travels opposite in direction to the negative sequence field, the backward slip is
given by Eq. (9-6) and so the quantity r3/2 — s appears in Fig. 9-9(b). By superpo-
sition the same negative-sequence voltage set is assumed applied to the same
stator windings and producing its own reactions in the same rotor winding. This is
the meaning that is intended in the schematic diagram of Fig. 9-8. There is really
only one pair of stator coils and one rotor, but the effect is the same as if there
were two identical pairs of stator coils and two identical rotors mechanically
coupled.

Figure 9-8 Representation of unbal-

anced two-phase motor by a positive-
sequence motor coupled to an identical
negative-sequence motor.
Sec. 9-3 Equivalent Circuit Via Symmetrical Components 403

r% x

(a) (b)

Figure 9-9 Equivalent circuit of unbalanced two-phase motor. (a) Circuit for
computing motor performance of the positive-sequence voltage set; (b) circuit for
the negative sequence.

In accordance with the theory of symmetrical components the phase cur-

rents of the unbalanced two-phase motor can now be found by using the equiva-
lent circuits of Fig. 9-9 to find the positive-sequence current I, and the negative-
sequence current /,. Then
len!iS ll +h (9-13)
I, = -jl, + jb (9-14)
It is at this juncture that we can now extend the analysis of the unbalanced
two-phase motor to the single-phase induction motor. The latter can be treated as
that special case of the two-phase motor where only one phase is used. In other
words, we have a situation where the current I, of Eq. (9-14) is necessarily zero
because of the absence of phase b. By Eq. (9-14) this forces the positive- and
negative-sequence currents in the single-phase motor to be identical. It then fol-
lows from Eq. (9-13) that
(9-15)

Figure 9-10(a) depicts the equivalent circuit of the single-phase motor in

terms of the positive- and negative-sequence voltage sets. But, because the posi-
tive- and negative-sequence currents are the same, a series connection of these
Fig.
individual equivalent circuits is now permissible. The result is shown in
the current and voltage equations for the
9-10(b) and it is notable that now both
by a single circuit employing appropria te posi-
stator winding can be represented
at the termi-
tive- and negative-sequence components. The impedance looking in
sequence impedance Z,, which specifically
nals at which V, appears is the positive
is given by
Z\ Sa) gi + jx, te Zoi (9-16)

where ie ei (9-16a)
 Single-Phase Induction Motors Chap. 9
404

(a) (b)

(c)
Figure 9-10 Equivalent circuit of the single-phase induction motor, based on symmetrical
components: (a) the equivalent circuits associated with the positive- and negative-sequence
fields; (b) series connection representation that is permitted because the positive- and nega-
tive-sequence currents are always the same; (c) simplified representation in terms of the
positive-sequence input impedance Z, and the negative-sequence input impedance Z,. Core
loss is neglected throughout.

Similarly, the negative-sequence impedance is described as

Zs 9) | TX] 4 Zp (9-17)

é r i
‘ IX¢ (;cys ixs)
where Li eee (9-17a)
FQ : ’
Ptah + jp + x2)

Examination of Fig. 9-10(b). discloses that:the values of V, and V> are influ-
enced by the slip at which the motor is operating. Varying slip values yield
Sec. 9-3 Equivalent Circuit Via Symmetrical Components 405

varying values for Z; and Z). Because Z; and Z; can be series connected (see Fig.
9-10(c)) for the single-phase motor, it follows that the positive- and negative-
sequence voltages are directly related to Z, and Z, through voltage division.
Accordingly, we can write
: ids v
Vv;1 = =——
7, Ia VJ, r
(9-18)

Sears Zr
Z+Z (9-19)

It should be noted from these expressions that the values of the positive- and
negative-sequence components change with each value of slip.
Several simplifications of the equivalent circuit of Fig. 9-10(b) can be intro-
duced. The first involves combining the stator impedance drop into a single total
value which is permissible for the single-phase induction motor because the posi-
tive- and negative-sequence currents are the same. Thus,

total impedance drop = h(n, + jx) + L(n + jx)

°

(9-20)
sz

— (rj + J) + oy(r, + jxy) = La(ri + jx)

a . Iq .

= w|!

The second simplification comes about upon noticing that the voltage drop across
Z,1 [the parallel combination of jxg and (r3/s + jx3)] and Z,2 can be written as
Ts a EZ
5) Zi = tq a (9-21)

and

eal
Le 5
Hose
T y, 2
(9-22)

Whenever the stator impedance drop is expressed by the result of Eq. (9-20), the
preferred way of expressing the voltage drops appearing across the parallel ele-
ments of the positive- and negative-sequence circuits is by the right side of the last
two equations. With the formulations represented by Eqs. (9-20), (9-21), and
(9-22), the equivalent circuit of Fig. 9-10(b) can be replaced by the simpler version
in the
depicted in Fig. 9-11. Note that it is now the total stator current /, that flows
series arrangement of the circuit elements, r; + jx1, Zp:/2, and Z,/2 in response to
and
the driving force represented by V,. The manipulation involved in Eqs. (9-21)
(9-22) calls for the use of one-half the values of the actual rotor winding resistance
magne-
and leakage reactance per phase as well as one-half the actual value of the
tizing reactance.
Once the single-phase induction motor is started and achieves normal speed,
onal
the slip frequently lies in the range 0 < s = 0.15. Moreover, for conventi
x4 > |r2 + jx}|. Accordin gly, very little error is made if the
single-phase motors
 Single-Phase Induction Motors Chap. 9
406

Figure 9-11 Simplified approximate Figure 9-12 A modified simplified equiva-

equivalent circuit of the single-phase in- lent circuit applies for 0 < s = 0.15.
duction motor based on symmetrical
components.

ce Z,./2. Also,
magnetizing reactance is entirely eliminated from the impedan
carries the same current as r;, the
r5/[2(2 — s)] ~ 13/4 and, since this resistor
Similarl y, jx2/2 may be combine d with
former may be combined with the latter.
circuit of Fig. 9-11 leads to a still simpler
jx;. Imposing these assumptions on the
form as depicted in Fig. 9-12.
e
This form of the configuration of Fig. 9-12 is like that of the per-phas
inductio n motor. A consider able and
equivalent circuit of a balanced polyphase
at first
useful simplification has clearly been made possible in a situation that
appeared quite complex.
A comparison of the configurations appearing in Figs. 9-5 and 9-11 reveals
these circuits to be identical except for the notation. The comparison makes it
obvious that phase a is the same as the main winding m, and subscript 1 for
positive sequence plays the role of f for forward field and subscript 2 for the
negative sequence corresponds to b for backward field. Finally, E; and V, are
corresponding quantities as well as E, and V>.

9-4 TORQUE-SPEED CHARACTERISTIC

We have seen that the curve relating torque as a function of speed is a very
important characteristic of electromechanical energy-conversion devices operat-
ing as motors. The single-phase motor is no exception. It is to be expected,
however, that the resultant torque-speed curve that applies over the entire operat-
Sec. 9-4 Torque-Speed Characteristic 407

ing range is not as straight-forward to obtain as it was for the three-phase induc-
tion motor. The reason is that the positive- and negative-sequence voltage compo-
nents change in value with slip [refer to Eqs. (9-18) and (9-19)]. Accordingly,
when the resultant torque-speed curve is found, care must be taken to compute 7;
and T) at a given slip by first calculating those values of V, and V> that are valid for
the slip concerned.
To emphasize this point further, refer to the plot appearing in Fig. 9-13(a). It
shows the plot of the positive-sequence torque-speed curve 7; corresponding to
V, = V,/2, and the negative-sequence torque-speed curve 7) corresponding to
V, = V,/2. When this plot is made to apply to the single-phase induction motor,
the only point on the resultant curve shown that has any validity is the one for s =
1. It indicates that the resultant torque is zero—a familiar result by now. When
the motor is assumed to be operating at a normal slip, say s,,, it is incorrect to
identify the corresponding torque as T;, in Fig. 9-13(a), because V, and therefore 7,
are much larger than the values prevailing at standstill. Rather, the correct value
of the resultant torque at slip s, is determined by first finding the 7; and 7,
quantities using the appropriate values of V, and V> and then subtracting to obtain

O,slip

(a) (b)

s, O, slip

(c)

motor: (a) valid

Figure 9-13 Torque-speed curves of the single-phase induction
tion of re-
representation of resultant torque at s = 1 only; (b) valid representa
(c) resultant curve that is valid over the entire speed
sultant torque at s = s, only;
range (solid line only).
 408 Single-Phase Induction Motors Chap. 9

T, as illustrated in Fig. 9-13(b). The true resultant torque-speed curve of the

single-phase induction motor assumes a form like that depicted in Fig. 9-13(c).
The broken-line curve of Fig. 9-13(c) denotes the torque-speed curve of the
two-phase motor operated under balanced conditions. A comparison with single-
phase operation reveals two interesting facts: (1) operation as a single-phase
motor at small slips is quite acceptable in spite of increased rotor losses; (2) the
single-phase motor will run in the direction in which it is started.

Effect of Rotor Resistance on the Torque-Speed Curve

In the case of the polyphase induction motor it was shown that neither the shape
nor the maximum value of the torque-speed curve was influenced by the value of
rotor resistance. Refer to Eq. (4-33). However, this is not so for the single-phase
induction motor and the reason is obvious from a glance at the equivalent circuit
of Fig. 9-12. Recall that the slip at which the maximum torque is developed
corresponds to the conditions for which maximum power is delivered to the
resistor r3/2s. The slip at which this occurs is found by setting r3/2s equal to the
magnitude of the impedance looking back into the equivalent circuit from the
terminals of this resistance. Clearly, the ensuing value for s,, will be found to
depend upon r3. Accordingly, the resultant torque too is so influenced by r3, and
the manner in which this occurs is depicted in Fig. 9-14. Note that it not only
causes a reduction in the peak developed torque but it brings about a flattening of
the curve as well. This behavior of the torque-speed curve restricts the use of
external rotor resistance for controlling the speed of the single-phase induction
motor.

Torque 12
3r2
6r2

Figure 9-14 The effect of rotor resis-

tance on the torque-speed curve of a
1 O, slip single-phase induction motor.

9-5 PERFORMANCE ANALYSIS

The calculation of the performance of the single-phase induction motor can be

readily determined through the use of the equivalent circuit and the appropriate
power-flow diagram. The approximate equivalent circuit of F ig. 9-12 may be used
Sec. 9-5 Performance Analysis 409

whenever it is desired to compute the behavior of the motor in the slip range of
0 < s < 0.15.’ Otherwise use is made of the circuit depicted in Fig. 9-11.
The flow of power from the line to the shaft in the single-phase motor is a bit
more complicated than it is in the balanced polyphase motor because of the need
to work with the positive- and negative-sequence fields. The complete power-
flow diagram appears in Fig. 9-15. The electrical input power is applied to the
motor at the stator winding. After removing a small part of this power to supply
the stator winding copper loss and the core loss produced by the resultant field,
the remainder is the power transferred across the air gap. However, note that the
total gap power is now expressed in two parts, one associated with the positive-
sequence field, P,;, and the other associated with the negative-sequence field,
P,.. Because the rotor travels in the same direction as the positive-sequence
field, the negative-sequence gap power is necessarily consumed entirely as rotor
copper loss. It is worthwhile to note too that this gap power P,2 only partially
supplies the total copper loss associated with the rotor resistance of the negative-
sequence impedance. From the equivalent circuit of Fig. 9-11 it is seen that the
negative-sequence gap power is the power associated with r3/2(2 — s) and is
expressed by
(9-23)

Upon rearrangement of this equation, it follows that the expression for the total
copper loss in the rotor resistance of the negative sequence becomes

(In)? > = 2 — s)Pe2 = Per + A — 8)Po2 (9-24)

r4

P, =\V,I,cosé
—— Useful power
-
=-== Power losses
oo" ‘i

Py' 0 peemeeee Woe

:
jh a Ne Fy (Io4) F
R i
R= (22) 2(2-s) E.
H it
i touig
(Ip)?
'
Pn = Pan (15)
H ?
!1 cg ¢
a
: 4
eet tr Td.
i
i 1
i]
i] 1
q t Po
(1-s) Po2 or
Pg2 +
Veena poems

Total rotor copper loss

tive sequence
__ area Figure 9-15 Power-flow diagram of the
> losses single-phase induction motor.
 Single-Phase Induction Motors Chap. 9
410

to emphasize that a portion of

The right side of Eq. (9-24) is written in two parts
h the medium of the
this copper loss, P,2, is supplied directly from the line throug
— s)Pg2, is supplied me-
negative-sequence field, while the remaining portion, (1
uence field. At normal
chanically by the rotor in response to the positive-seq
times greater than that of
operating slips the positive-sequence gap power is many
in the rotor winding
the negative-sequence. Upon subtracting the copper loss
associated with the positive-sequence [i.e., 5,)°(r3/2) from P,1], there results the
useful mechanical developed power P,,;. Specifically,
Pm = Pill — 5) (9-25)
examination of
and this quantity represents the force that drives the load. Further
can be deliv-
the power-flow diagram indicates, however, that before useful power
e-sequence field
ered to the load the mechanical power developed by the positiv
to drive the
must first supply the rotational losses as well as the power needed
This latter quan-
rotor against the opposing action of the negative-sequence field.
tity is already denoted in Eq. (9-24) as ChixisPa.

Example 9-1
loss of
A }-hp, 110-V, 60-Hz, four-pole single-phase induction motor has a rotational
s are known to be as fol-
15 W at normal speeds. The equivalent circuit parameter
lows:
Pus 16320; rg =3.20

x, = 2.50, x4 = 22 0

Xp = 48.0

Determine the performance (i.e., line current, line power factor, horsepower output,
and efficiency) of this motor when it operates at a slip of 4%.

Solution The approximate equivalent circuit of Fig. 9-12 is used. It is repeated in

Fig. 9-16 with all parameter values shown. The impedance of the parallel circuit is
given by
ji a ee
Zane, es 2/ _ j24(40 + j1.1) (9-26)
faye ( ) 2 Ee 10.8 + 17.2 = 20.4/57.9
A aE a
The total input impedance is therefore

2r= (n+) ilur9) 4

4 5 2Ateg sve 1058 + f17.2

12.9 + j20.8 = 24.4 /58.4° (9-27)

Accordingly, the line current is

I, eerie
= Z, 7 34/58 4.51 /-58.4° ! A (9-28)
Sec. 9-5 Performance Analysis 411

Figure 9-16 Equivalent circuit for

Example 9-1.

and the line power factor is

power factor = cos 58.4° = 0.525 lagging (9-29)

Such a low value for the line power factor is typical of single-phase motor operation.
The input power is found to be
P; = VaIq cos 6 = 110(4.51)0.525 = 261 W (9-30)
The positive-sequence gap power is found from

Poi ale :Ry; = (4.51)710.8 = 220 W (9-31)

Also, the negative-sequence gap power is determined to be

Py = (In) = ~ [2 2 = (4.51)°0.8 = 16.3 W (9-32)

From the power-flow diagram the expression for the output power is

Po = Puy — Pro — C1 = $) P29

= Py — SPg1 — Pro (1 = 5) P02 (9-33)

Inserting values
Py = 220 — (0.04)220 — 15 — (0.96)16.3
PEN Menge ct Sra tar Al
Pia)
182.4 W (9-34)

The horsepower delivered to the load is therefore

horsepower = sit
746RETO,
746 0.245 (9-35)

This result indicates that the motor is delivering almost rated output power.
The efficiency is found to be

1 _=B=
Py _ 6
182.4 = (0-70 oF 70% (9-36)
 Single-Phase Induction Motors — Chap. 9
412

Speed, rpm

Torque,
n—m Current,
amps

Rated output

(0) 80 160 240

Output, watts

Figure 9-17 Typical performance characteristics for a single-phase induction mo-

tor, rated at 4 hp, 60 Hz, four poles.

By repeating the calculations described in the foregoing example for various

values of slip, the performance characteristics of the motor can be determined
over the full operating range. Curves for a typical 4-hp, 1725-rpm motor are
depicted in Fig. 9-17.

9-6 APPROXIMATE DETERMINATION OF THE EQUIVALENT

CIRCUIT PARAMETERS

The data obtained in a blocked-rotor test and a no-load test can be used to furnish
information about the parameters appearing in the equivalent circuit. The
blocked-rotor test data are used to reveal information about the leakage reac-
tances and the stator-referred rotor winding resistance. The determination of
these parameters is approximate because the associated procedure involves sim-
plifying assumptions. One assumption is to consider that for s = 1 the impedance
of x,/2 in the equivalent circuit of Fig. 9-11 is so large compared with (13/2) +
j(x3/2) that for all practical purposes it may be eliminated entirely from the equiva-
lent circuit. Accordingly, the equivalent circuit of Fig. 9-11 reduces to the form
depicted in Fig. 9-18.
If we let the measured values of voltage, current, and power obtained in the
rotor-blocked test be denoted respectively by V,,, J;., and P;,, it then follows that
the equivalent impedance referred to the stator coil is
Vise
Le yaa (9-37)
Sec. 9-6 Approximate Determination of the Equivalent Circuit Parameters 413

" ix

Figure 9-18 Equivalent circuit of the

single-phase induction motor as it
applies at standstill.

Moreover, the equivalent resistance component of Eq. (9-37) is readily deter-

mined from the wattmeter and ammeter readings. Thus

R=n+ri= Pe (9-38)
Ti.
To determine r} from the result of Eq. (9-38), it is necessary to introduce a second
important assumption, 1.e.,
ioe a (9-39)

This assumption is valid for all well-designed machines, i.e., those machines for
which a minimum amount of iron and copper is used for a specified output rating.
Unless otherwise stated, the assumption stated in Eq. (9-39) is always made in
analyzing the single-phase motor. Therefore, the rotor winding resistance term
appearing in the equivalent circuit can be identified as
Roe Re
Lies (9-40)

where R, is determined by Eq. (9-38).

+An alternative procedure is to find r; in the manner described for the three-phase induction
motor. See Sec. 4-7.
 Single-Phase Induction Motors Chap. 9
414

from
The value of the equivalent leakage reactance is obtained
Wes ViZe ay (9-41)
The stator-referred rotor
where Z, and R, are found from Eqs. (9-37) and (9-38).
winding leakage reactance is then found by assumin g that

x, = x} (9-42)

and this leads directly to

(9-43)

ed is
The sole parameter of the equivalent circuit that remains to be determin
the magnetizing reactance x4. Its value can be established from the no-load test
The
data and knowledge of x,, x3, and r; as derived from the rotor-blocked test.
referring to the equivale nt circuit of Fig. 9-19 which is
procedure is illustrated by
derived from Fig. 9-11 upon imposing simplifyi ng assumpti ons. Recall that at no-
load the slip is very close to zero; it then follows that the quantity r/2s represents
a very high resistance and so may be omitted without much error. Therefore, at
no-load Z,;/2 is assumed to be equal simply to jx4/2. Similarly, the impedance
Zp2/2 associated with the negative-sequence field can be simplified to the low
impedance represented by (r3/4) + j(x3/2)—which is many times smaller than
jxq/2. Certainly this assumption is even more valid than the corresponding as-
sumption used in the rotor-blocked test.
Determination of x4 essentially requires finding the voltage across a and b in
Fig. 9-19. By letting the instrument readings for voltage, current, and power in
the no-load test be denoted respectively by V,, J,, and P,, the power factor angle
of the current can be found from
cos 0, = (9-44)

Figure 9-19 Equivalent circuit of the

single-phase induction motor at no-load
(s ~ 0).
Sec. 9-6 Approximate Determination of the Equivalent Circuit Parameters 415

Then,

Vi». =a = In[=65|(rs + 2) +a (u+t 2) (9-45)

Since all the quantities on the right side of this equation are known, the magnetiz-
ing reactance is readily computed as

X¢5 _ Vab
ip :
(9-46)

The foregoing procedure can be somewhat simplified by using an approxi-

mate expression for the voltage across a and b that uses magnitudes instead of
phasors. Thus

V,$PaW, = il, (x,+ =) (9-47)

The validity of this assumption is based on the fact that the single-phase induction
motor operates at very poor power factor at no-load. A glance at Fig. 9-20 indi-
cates that the three quantities involved in Eq. (9-47) are practically collinear.
When Eq. (9-47) is used to determine V,,, there is no need to use Eq. (9-44); x4 is
found directly through the application of the last two equations only.
Example 9-2
Laboratory tests on a single-phase induction motor yield the following data:
Blocked-rotor test: V,. = 110 V, I,, = 14.8 A, P,. = 1130 W
No-load test: V, = 110 V, n= 2.8 A, P,, = 60 W
Rotational losses: 17 W

Determine the five parameters of the equivalent circuit and the value of the core loss
in watts.
Solution From Eq. (9-38) the equivalent resistance is

faye foraldsnbeon A120 = S150

Sh ee Biles 014.8
Hence,
r= 75 = 2.65O

My

Figure 9-20 Phasor diagram for the

1 circuit of Fig. 9-19.
 Single-Phase Induction Motors Chap.
416

and

ra
5 1.33"

Also,
s
fais = 7.440,
Ze = Viega

X, = V7.44 — 5.15? = 5.36 0

Therefore,

xy = xp = 23° = 2.68.0

and

call)
5 1.340

Moreover, by Eq. (9-47) the voltage across the magnetizing reactance at no-
load is

Vab = Vn — In (x ap 2) = 110 — 2.8(2.68 + 1.34) = 110 — 11.2 = 98.8 V

Hence,

=>
= 33.30

=2.65 jx,= j2.68

Figure 9-21 Equivalent circuit showing

parameters computed in Example 9-2.
 Sec. 9-7 Single-Phase Motor Types 417

The core loss is found from

P.= P, —I2 (, : 2) yah,

= 60 — 2.8°(3.17) — 17 = 18.2 W (9-48)
The equivalent circuit showing the computed parameters appears in Fig. 9-21.

9-7 SINGLE-PHASE MOTOR TYPES

Many different types of single-phase motors have been developed primarily for
two reasons. One, the torque requirements of the appliances and applications in
which they are used vary widely. Two, it is desirable to use the lowest-priced
motor that will drive a given load satisfactorily. For example, a high-torque ver-
sion of the split-phase induction motor discussed in Sec. 9-1 is designed almost
exclusively for washing-machine applications. This motor is available in a single
speed rating and in just two horsepower ratings. This, coupled with the large
volume of sales, enables it to be the least expensive motor in its category. See line
two of Table 9-1 for more details.

Capacitor-Start Induction-Run Motor

The chief difference between the various types of single-phase motors lies in the
method used to start them. In the case of this motor a capacitor is placed in the
auxiliary winding circuit so selected that it brings about a 90-degree time displace-
ment between J, and Jy. See Fig. 9-22. The result is a much larger starting torque
than is achievable with resistance split-phase starting, because the starting torque
is essentially proportional to the magnitudes of Ty and I, as well as the sine of the
phase angle between these quantities i.e., Ts a Iyi, sin < ta,
A typical torque-speed curve for this motor is deciaindain Fig. 9-22(c). The
cut-out switch serves to remove the auxiliary winding and the starting capacitor
from the line once the motor has reached operating speed. Both items have
intermittent ratings and could easily be damaged if allowed to remain energized.
The starting capacitor is of the electrolytic type and generally has a value in the
range of 100 wF for motor ratings of § to } hp. The capacitor can be easily
damaged by a faulty cut-out switch; so proper maintenance of this switch is
important.

Permanent-Split Capacitor Motor

At normal speeds the capacitor-start induction-run motor behaves as a conven-

tional single-phase motor which means it has a resultant revolving field that has an
elliptical locus. The result is a motor that operates less smoothly and in a noisier
 418 Single-Phase Induction Motors Chap. 9

Auxiliary

400

Both windings
300

200 Main wdg alone

rated)
(%
Torque
100

Figure 9-22 Capacitor-start induction-

he run motor: (a) schematic diagram; (b)
Activating
Fa |ay speed
ars = ‘
phasor diagram 1%
at starting;§ (c) typical
:
torque-speed characteristic of capaci-
(c) tor-start induction-run motor.

fashion than one where the resultant field has a constant amplitude with time.
This latter condition can be closely duplicated by designing the single-phase motor
with an auxiliary winding that is identical to the main winding and then placing in
series with the auxiliary winding a full-duty, oil-filled capacitor whose value is
selected to provide the constant amplitude rotating field in the vicinity of rated
output. The schematic arrangement appears in Fig. 9-23 and for obvious reasons
this motor is called the permanent-split capacitor motor. Basically, this motor is
designed to operate as a balanced two-phase motor near rated load.
A typical torque-speed curve is illustrated in Fig. 9-24. There is no need for
a cut-out switch because of the continuous-duty rating of the auxiliary winding.
Part of the price paid for the improved running performance is the much lower
torque at starting and breakdown. The starting torques are usually 60% to 75% of
rated value. This is attributable to the fact that the capacitor value that gives
balanced performance at normal slips is far different from the value needed for
large starting torques. This is entirely understandable in view of the large differ-
ence between the input impedance at standstill compared to running conditions.
Besides the quiet, smooth running performance and the increased operating
reliability (owing to the absence of a cut-out switch), the permanent-split capaci-
tor motor also makes reversible operation a routine matter since basically it is a
two-phase motor. Reversible operation of the capacitor-start induction-run motor
is trickier to achieve. (Why?) ° —
Sec. 9-7 Single-Phase Motor Types 419

Auxiliary
winding

Main
winding
Figure 9-23 Permanent-split capacitor
motor.

300

200

rated)
(%
Torque

Figure 9-24 Typical torque-speed

0 02 ia 06 08 102 characteristic of permanent-split capaci-
Ns tor motor. No cut-out switch.

Capacitor-Start Capacitor-Run Motor

The high starting torque of the capacitor-start motor can be combined with the
good running performance of the permanent-slit motor by employing two capaci-
tors in the manner shown in Fig. 9-25. Of course, the auxiliary winding must carry
a continuous rating as in the case of the permanent-split motor and C; must also
have a continuous-duty rating. Capacitor C), which is used only during starting,
can be a less expensive, intermittently rated element. A typical torque-speed
curve for this motor is illustrated in Fig. 9-26.

Shaded-Pole Motor

The shaded-pole motor is used extensively in applications that require zo hp or

less. The construction is very rugged as a study of Fig. 9-27 reveals. It consists
essentially of copper wound on iron, a squirrel-cage rotor, and not much else. It
contains no cut-out switch, which can be a potential source of trouble.

Auxiliary
winding

Cutout ; ,
switch Figure 9-25 Capacitor-start capacitor-
run motor.
 420 Single-Phase Induction Motors Chap. 9

Copper ri
3 With C, and C,
(shaded coil
8
=
® Al Squirrel cage
c Vt rotor
°
kr

Salient -pole
piece

Stator coil

Figure 9-26 Typical torque-speed characteristic of ca- Figure 9-27 Construction features of
pacitor-start capacitor-run motor. See Fig. 9-25 for the shaded-pole motor.
identity of C; and C).

The manner of obtaining a moving flux field in this motor is different than it is
in those considered previously. Here use is made of a copper ring that is embed-
ded in each salient-pole piece as depicted in Fig. 9-27. The action of this ring is to
oppose flux changes; and, in the process, over the course of the cyclic variation of
the coil’s exciting current, it causes the flux to move in a sweeping fashion across
the face of the poles from the unshaded to the shaded sections. Reference to Fig.
9-28 illustrates the procedure.
In the discussion that follows hysteresis is neglected; hence flux and current
are in time phase. Appearing in Fig. 9-28(a) is the situation that prevails at the N-
pole flux at time wf, in the cyclic variation of the magnetizing coil current i [see
Fig. 9-28(d)]. At this time instant flux is increasing, but the reaction in the copper
ring is to produce a current which acts, by Lenz’s law, to prevent the increase in

(a) (b) (c) (d)

Figure 9-28 The action of the copper ring in producing a moving flux across the pole face.
(a) Field distribution at time wt, ;(b) time wt); (c) time wf;; (d) coil current.
Sec. 9-7 Single-Phase Motor Types 421

flux. This calls fer a current direction in the copper ring that is into the plane of
the paper on the right side and out on the left side. The result is to crowd the flux
to the unshaded portion. Later, at wf,, the induced emf in the copper ring is zero
because dd/dt is zero. Hence there occurs a uniform distribution of flux across
the pole face as illustrated in Fig. 9-28(b). Next, at wf;, the pole flux is decreasing
in response to a decreasing magnetizing current. Since the reaction in the ring is
always to prevent change, which at this instant means to prevent a decrease in
flux, the induced ring current now takes the direction shown in Fig. 9-28(c). The
result is a crowding of flux in the shaded portion of the pole. The net effect,
therefore, is to produce a sweeping flux motion in the air gap beneath the pole face
which in turn induces currents in the squirrel-cage rotor that react with the pole
flux to produce torque.
A typical torque-speed characteristic is shown in Fig. 9-29. Note the modest
value of starting torque of these motors. The breakdown torque is also of moder-
ate value.

Universal Motor

The notion of a universal motor is tied to the capability of the motor to operate
with dc or ac voltages applied, i.e., to operate over the range of frequencies from
zero to 60 Hz. By way of introducing this topic we ask the question: Can a shunt
motor, designed for dc operation, be made to deliver power with a 60-Hz voltage
connected to its input line terminals? Recourse to the basic torque equation re-
minds us that we need to identify a flux field, an ampere-conductor distribution
and a proper field orientation of the two that is maintained over time. The am-
pere-conductor distribution is ensured by the flow of armature current through the
armature winding. There will be a pole flux but its magnitude will be very small
for two reasons. One, since the field winding is equipped with many turns and the
inductance is proportional to the number squared [see Eq. (1-40)], the inductive
reactance of the field winding is extremely high thus producing very little field
current. Two, the air gap of dc machines, unlike ac machines, is quite large.
Hence the net effect is to produce a very small pole flux. Furthermore, although
the brushes are correctly placed along the interpolar axis, still during every cyclic

Figure 9-29 Typical torque-speed

characteristic of a shaded-pole motor.
0 0.2 0.4 0.6 0.8 10 Note low starting and breakdown
: Ns torques.
 Single-Phase Induction Motors Chap. 9
422

variation of the field and armature currents, the flux field and the armature am-
pere-conductor distribution will not produce a unidirectional torque on each con-
ductor. This is because the armature current is not in time phase with the field
current; and the reason is that the inductance of the armature circuit is not as large
as that of the field circuit. Therefore, despite the fact that the ac operated shunt
motor produces some starting and running torque, its performance would be far
from acceptable.
What happens when the single-phase ac voltage is applied to a de series
motor? The situation now is very much improved. First, because the series field
and armature winding carry the same current in the same circuit, the field orienta-
tion for producing torque is optimum. Second, the pole flux is now greater be-
cause the series field inductance is low owing to the use of a small number of
series turns compared to the number of field turns in a shunt motor. The series
field mmf is chiefly dependent on the flow of a large armature current. Accord-
ingly, acceptable performance can be expected for this mode of operation. Core
losses, however, can be exorbitant if the stator iron is unlaminated. Also, com-
mutation is likely to be poor because of the induced currents that occur through
transformer action when a coil undergoes commutation.
The term universal motor is applied to a de series motor the design of which
has been modified to mitigate the shortcomings associated with ac operation.
These modifications include the following:

1. A completely laminated stator structure including the poles.

2. High resistance connections from the coils to the commutator segments to
reduce the transformer induced voltages.
3. A smaller number of series field turns to reduce the reactance drop in the
armature circuit.
4. Anincreased number of armature conductors to compensate for the reduced
pole flux associated with the smaller number of series field turns.

It should come as no surprise that the universal motor is an excellent dc series

motor.
The torque developed by the universal motor is proportional to the square of
the armature current just as it is for the conventional dc series motor. But, be-
cause the armature current in the universal motor is sinusoidally varying, the
instantaneous torque as a function of time varies in accordance with double the
line frequency as illustrated in Fig. 9-30. Although the average torque is responsi-
ble for the useful work at the motor shaft, this double frequency variation of the
instantaneous torque makes the universal motor more prone to noise and vibra-
tion.
Typical torque-speed curves for the universal motor and the conventional dc
series motor are depicted in Fig. 9-31. Observe that at any given torque the speed
of the universal motor is less than that of the dc series motor. The reason lies in
Sec. 9-7 Single-Phase Motor Types 423

wh

rary
rem
ie
Instantaneous torque
5
oO
Ko 2
o 3 Average torque
20
Fe=
ad
Time, wt

ig
Armature
current

Figure 9-30 Instantaneous torque of the universal motor.

the reactance drops that are associated with the series field and armature wind-
ings.
It is useful to note that the universal motor develops more torque for a given
size than any other single-phase motor because the torque is proportional to the
armature current squared. Hence, it finds considerable popularity in those appli-
cations where space is at a premium and high torque is needed. The whole area of
home appliances is typical of such applications. Furthermore, speed control of
such motors is easily achieved either through the use of a series resistor (for the
smaller ratings) or through the use of an adjustable rectified source employing
silicon controlled rectifiers.

Hysteresis Motors

This topic is included here primarily because the hysteresis motor is most often
used on single-phase power supplies in competition with the other motors de-
scribed in this section. However, it is sometimes designed for three-phase opera-
tion and it always runs at synchronous speed in steady state.

dc series motor

Average
torque,
Universal ~ —

PURSE Figure 9-31 Typical torque-speed

0 Speed, rpm curves.
 Single-Phase Induction Motors Chap. 9
424

The single-phase hysteresis motor depends for its success on a rotor that has
a smooth,
no teeth or windings but is constructed of high-retentivity steel in
essential ly
cylindrical form. The high-retentivity feature of the rotor material
furnishes the motor with the equivalent of permanent magnet poles that permits
ob-
operation at synchronous speed. The rotating field of the stator is usually
tained by employing two well-distributed windings with a fixed capacito r perma-
nently connected to one of the windings to provide the time displacement needed
to generate a rotating field of constant amplitude. Care is taken in the stator design
to ensure a nearly sinusoidal field distribution in order to minimize losses.

Starting torque. When power is applied to the stator windings at stand-

still, the resulting rotating stator mmf generates hysteresis loops in the rotor steel
with a cyclic variation of line frequency. The magnetized rotor exhibits induced
magnetic poles that lag behind the magnetizing stator mmf by the hysteretic angle
y.t Materials of high retentivity have large values of y. Accordingly, at standstill,
these induced poles follow the rotating stator mmf but do so with a lag of y
electrical degrees. The torque produced in this fashion by hysteresis can be de-
scribed by
T, = Kndsoe sin y
where K;,, is an appropriate design parameter, ds denotes the flux produced by the
stator mmf, ¢z is the induced flux per pole in the rotor caused by the hysteresis
effect and y is the space displacement between these two fields expressed in
electrical degrees. Observe that 7), is a fixed quantity for a given machine.

Accelerating torque. If the starting torque exceeds the load torque, the
excess torque accelerates the load toward synchronous speed. During this time
the frequency of generation of the hysteresis loop in the rotor steel is diminishing
but neither ¢z nor y are affected because these quantities depend on the retentiv-
ity property of the hysteresis loop of the rotor material. Therefore, T;, remains
constant from zero to synchronous speed. This is indicated in Fig. 9-32 by the
horizontal line. The resultant torque-speed curve over this range is actually slop-
ing negatively because of the contribution to the net rotor developed torque that is
associated with eddy-currents that are produced in the rotor steel during the cyclic
variation of the rotor flux. Of course, once synchronous speed is attained, the
cyclic variation of flux in the rotor ceases and so too does the eddy-current torque
contribution. However, the hysteresis torque remains because despite the ab-
sence of cyclic variation of flux in the rotor, the high-retentivity feature of the

+See p 55.
 Sec. 9-7 Single-Phase Motor Types 425

Eddy-current torque contribution

2
T,
, HSS

N-m
Torque,

Va : ;

0 n, n,rpm__ tic of the hysteresis motor.

rotor steel provides the rotor with what can now be considered to be permanent
magnets. Moreover, if at this synchronous speed the hysteretic torque exceeds
that required by the load, the rotor will readjust the physical lag between the axis
of the stator poles and the axis of the rotor poles to ensure equilibrium. This
operating point is indicated by a in Fig. 9-32.
Two outstanding features of the hysteresis motor are quiet operation and
good accelerating torque. Also, the capability of this motor to operate at synchro-
nous speed is advantageous especially where the electric power system furnishes
power at virtually constant frequency. In such circumstances the hysteresis mo-
tor is ideal for driving clocks and other speed sensitive devices such as record,
tape, and compact disc players.

9-8 OPERATING FEATURES AND TYPICAL APPLICATIONS

The principal performance features as well as typical applications of single-phase

ac motors appear in Table 9-1. Note that for applications involving less than 5 hp
the shaded-pole motor is invariably used. On the other hand, for applications
involving # hp to ? hp the choice of the motor depends upon such factors as the
starting and breakdown torque and even quietness of operation. Where a mini-
mum of noise is desirable and low starting torque is adequate, such as for driving
fans and blowers, then the motor to choose is the permanent-split type. If quiet-
a
ness is to be combined with high starting torque, as might be the case where
tart
compressor drive is needed, then the suitable choice would be the capacitor-s
is lo-
capacitor-run motor. Of course, in circumstances where the compressor
tart
cated in a noisy environment, clearly the choice should be the capacitor-s
induction-run motor because it is less expensive.
 Single-Phase Induction Motors Chap. 9
426

-PHASE A-C MOTORS

TABLE 9-1 CHARACTERISTICS AND APPLICATIONS OF SINGLE

Starting Approx. Breakdown

comparative torque Starting
torque
(% of price (% of current
Type
(%) normal) at 115 V
designation normal)

90-200 85 185-250 23
General-purpose split-phase 1/4 hp
Medium Medium
motor

200-275 65 Up to 350 32
High-torque split-phase motor High
High
1/4 hp

60-75 155 Up to 225 Medium

Permanent-split capacitor motor
Low

Up to 200 155 260

Permanent-split capacitor motor
Normal

ll y
bi EA TES
O
Js ita
Up to 435 100 Up to 400
Capacitor-start general-purpose
motor Very high
a i ee ee Pee ToT
ee
380 190 Up to 260
Capacitor-start capacitor-run
motor High

a Oe
te
Shaded-pole motor 50 — 150

ee
ee
vol. I
Source: By permission from M. Liwschitz-Garik and C. C. Whipple, Electric Machinery,
(Princeton, N.J.: D. Van Nostrand Co., Inc., 1946).

PROBLEMS

9-1. Two field coils are space displaced by 45 electrical degrees. Moreover, sinusoidal
currents that are time displaced by 45 electrical degrees flow through these wind-
ings. Will a revolving flux result? Explain.
9-2 Two coils are placed in a low reluctance magnetic circuit with their axes in quadra-
ture—the first along the vertical and the second along the horizontal. Sinusoidal
currents that are time displaced by 90 electrical degrees and of the same frequency w
are made to flow through the coils. Moreover, the mmf of the first coil is twice that
of the second.
(a) Assuming that at time wf) the mmf of the first coil is at its positive peak value,
determine the relative magnitude and direction of the resultant mmf.
(b) Repeat part (a) for a time wf, that is 45° later than wf.
Chap. 9 Problems 427

TABLE 9-1 (Continued)

Power
Factor Efficiency Horsepower Application and
(%) (%) range general remarks

56-65 62-67 1/20 to 3/4 Fans, blowers, office appliances, food-preparation

machines. Low- or medium-starting torque,
low-inertia loads. Continuous-operation loads.
May be reversed.

50-62 46-61 1/6 to 1/3 Washing machines, sump pumps, home work-
shops, oil burners. Medium- to high-starting
torque loads. May be reversed.

80-95 55-65 1/20 to 3/4 Direct-connected fans, blowers, centrifugal

pumps. Low-starting torque loads. Not for
belt drives. May be reversed.

80-95 55-65 1/6 to 3/4 Belt-driven or direct-drive fans, blowers, cen-

trifugal pumps, oil burners. Moderate-starting
torque loads. May be reversed.

80-95 55-65 1/8 to 3/4 Dual voltage. Compressors, stokers, conveyors,

pumps. Belt-driven loads with high static fric-
tion. May be reversed.
30. 5. + ROA SA 2A eA sly 7) Se eI) a Dep TIE SUS SEE
80-95 55-65 1/8 to 3/4 Compressors, stokers, conveyors, pumps. High-
torque loads. High power factor. Speed may
be regulated.
ee eee
30-40 30-40 1/300 to 1/20 Fans, toys, hair dryers, unit heaters. Desk fans.
Low-starting torque loads.
ee
eee

(c) Repeat part (a) for a time wt, that is 90° later than wt.
(d) What conclusion can you draw from parts (a), (b), and (c)?
9-3. The two coils of Prob. 9-2 are now energized with sinusoidal currents that are 45
electrical degrees out of phase but yield equal mmfs.
(a) Assuming that at time wf) the mmf of the first coil is at its positive peak value,
find the relative magnitude and direction of the resultant mmf.
(b) Repeat part (a) for a time wt, that is 45° later than wf.
(c) Repeat part (a) for a time wf, that is 90° later than wf.
(d) What conclusion can you draw from parts (a), (b), and (c)?
9-4 In Fig. P9-4, I, II, and III denote the three coils of a balanced three-phase winding
is
which are spaced displaced by 120° from one another. If a single-phase voltage
applied to the three-phase, squirrel-cage motor in the manner shown, will the motor
operate? Justify your answer.
quadra-
9-5. The stator windings of a two-phase, two-pole induction motor are in space
the effective values of the phase currents are J, = 10 /0° and J, = 10 [90°
ture. When
is 2000 A-t per
A, the resulting stator mmf wave rotates clockwise, and its peak value
 Single-Phase Induction Motors Chap. 9
428

Single - phase
voltage

Figure P9-4

pole. If the phase currents are changed to Fo [0° and J, = 10 [93. 1°, find the peak
values of the clockwise and counterclockwise rotating mmf waves.
9-6. The phase currents in the stator windings of a two-phase induction motor are I, =
8 [0° and I, = 8 el A. Together they produce an mmf wave rotating counterclock-
wise and having a peak value of 1000 A-t per pole. These currents are then changed
to the following: 7, = 10/0° and J, = 6/+36.9°. Find the peak values of the resulting
clockwise and counterclockwise rotating mmf waves.
9-7. At standstill, the currents in the main and auxiliary windings of a single-phase induc-
tion motor are given by Iy = 10 [0° and I, = 5 [60°. There are five turns per pole on
the auxiliary winding for every four turns on the main winding. Moreover, these
windings are in space quadrature. What should be the magnitude and phase of the
current in the auxiliary winding to produce an mmf wave of constant amplitude and
velocity?
. The armature mmf of a machine is described by the equation
F, = K, cos wt cos a + K2 cos wf sina

Is this a single-phase or a polyphase machine? Explain your answer.

9-9. An unbalanced two-phase voltage system is described by
V, = 120/0°
V, = 80 /—53°
(a) Find the positive-sequence voltage set of this system.
(b) What is the negative-sequence set?
(c) Draw the complete phasor diagram of this two-phase system, showing each
positive- and negative-sequence component.
Repeat Prob. 9-9 for the case where V, = 120/0° and V, = 80/-—90°.
Find the unbalanced two-phase system that is represented by a positive-sequence
component of 95.2 /14.6° and a negative-sequence component of 31 /40.6° expressed
in volts.
. The results of a load test on a 4-hp, 115-V, 60-Hz, four-pole, single-phase induction
motor are as follows:
applied voltage = 115 V
main winding current = 3.7 A
power input = 270 W
stator core loss = 10 W
slip = 90 rpm
Chap. 9 Problems 429

The stator winding resistance is 2 1 and the stator referred rotor resistance is 1.6.
The starting winding was disconnected during this test. Calculate the power output
during the test. Assume that the rotational losses are equal to the stator core losses.
9-13. A 110-V, 60-Hz, single-phase induction motor has the following parameters:
r, = 1.860, ry = 1.780
x, = 2.560, x3 = 1.280
Xo = 25 0

At a slip of 5% find:
(a) Input line current and power factor.
(b) Horsepower output.
(c) The efficiency.
9-14. A 115-V, 4-hp, 60-hz, single-phase induction motor has the following design parame-
ters:
r, = 1.48 0, re = 1.68 0
Xi ood A), x3
= 160
Xp = 25 0

Also, stator core loss is negligible.

(a) Determine the operating performance of this motor when it runs at a slip of 10%.
(b) Would you permit continuous operation of this motor under the conditions of
part (a)?
9-15. The rotor of the machine described in the preceding problem is blocked and rated
voltage is applied. Compute the readings of a line wattmeter and ammeter. Make
approximations where proper.
9-16. A 3-hp, 115-V, four-pole, single-phase induction motor presents the following test
data:
At no-load: 115 V, 4.81 A, 105 W

With rotor blocked: 115 V, 27.2 A, 1308 W

The stator winding resistance is 0.83 . The stator core losses are equal to the
rotational losses. Determine the parameters of the equivalent circuit.
9-17. The following no-load and blocked rotor test data are available for a single-phase
induction motor:
Blocked rotor: J, = 6.6 A, Vsce = 53 V, Pse = 210 W
No-load: I, = 3.85 A, V, = 208 V, P, = 160 W
The stator resistance is 1.48 © and the rotational losses are 60 W.
(a) Determine the parameters of the equivalent circuit.
(b) What is the value of the core loss?
(c) Compute the value of resistance that can be used to represent this core loss in
the equivalent circuit. Show its location.
9-18. A }-hp, 110-V, 60-Hz, four-pole, single-phase induction motor has the following
parameters:
mn = 1.50, r5 140
x) ll 2.40, x4 lI isp
Xb a 40 10)
 Single-Phase Induction Motors Chap. 9
430

The stator core loss is 20 W.

(a) Calculate the developed torque at a slip of 5%.
torque?
(b) If the rotational losses are equal to 20 W, what is the shaft
e, single-p hase inductio n motor has the following
9-19. A 3-hp, 220-V, 60-Hz, four-pol
equivalent circuit parameters:
rn = 2.80, re = 2.40

x = 3.5 0, x5 = 2.80,

Xb = 90 QO,

s = 0.05:
The core loss is 30 W and the rotational losses are 46 W. Determine for
(a) Line current and line power factor.
(b) Horsepower output.
(c) Efficiency.
9-20. The following data apply to a }-hp, 110-V, 60-Hz, four-pole, single-phase induction
motor.
No-load test: V, = 110 V, f= 2.73 A, Py = 56 W

Blocked rotor test: V,. = 110 V, Is. = 16.6 A, Pse = 1260 W

The resistance of the main winding is 1.5 9. Assume the rotational losses are equal
of
to the stator core loss. On a carefully drawn equivalent circuit, show the value
each parameter for a slip of 5%.
9-21. A two-phase, 60-Hz, squirrel-cage induction motor has a blocked resistance per
phase of 6 0 and a reactance of 8. When connected as a normal two-phase motor,
its starting torque is 200%. It is desired to use this motor on a single-phase circuit.
(a) A resistance of 10 © is inserted in series with one phase and the machine is
operated as a split-phase motor. Determine the developed torque at standstill.
Consider that the torque is proportional to the product of the phase currents as
well as the sine of the phase angle between them.
(b) A capacitor is to be used in place of the resistor. Calculate the microfarad rating
of this capacitor so that the phase angle between the currents is 90°.
(c) Determine the starting torque for the condition of part (b).
9-22. Discuss in some detail the capability offered by the following motors in permitting
reversible operation. Be sure to cite the ease or difficulty that is likely to be encoun-
tered in each case as well as any precautions to be observed. In those instances
where reversible operation is deemed practical show a complete schematic diagram
of the motor including the circuit that allows the reversible mode.
(a) Permanent-split capacitor motor.
(b) Capacitor-start capacitor-run motor.
(c) General-purpose split-phase motor.
(d) Universal motor.
(e) Hysteresis motor.
9-23. The following questions relate to the single-phase shaded-pole induction motor:
(a) Can the direction of rotation be reversed by reversing the line leads? Explain.
(b) Devise a scheme that permits reversible operation of the single-phase shaded-
pole motor. ;
 PART 2
Topics for Further Study

10
Unbalanced
Two-Phase Motors:
Servomotors

The two-phase servomotor has gained widespread use as an output actuator in

feedback control systems. A high rotor resistance is one of its distinguishing
design characteristics. This feature ensures a negative slope for the torque-speed
characteristic over its entire operating range and thereby furnishes the motor with
positive damping for good stability. The ac servomotor often operates with unbal-
anced two-phase voltages. A fixed reference voltage is applied to its main winding
and a varying voltage is applied to its control winding. The control winding is
frequently fed from the output terminals of a servoamplifier with a signal that is
usually displaced 90° from that on the main winding.
The analytical treatment of the two-phase servomotor supplied with unbal-
anced voltages can be conveniently accomplished by the use of symmetrical com-
ponents. This technique is described in the first section of this chapter for the
more general case of an unbalanced four-phase system and is subsequently ap-
plied to the two-phase case. The symmetrical components approach allows the
unbalanced voltages to be replaced by two or more sets of balanced voltages. In
this way the unbalanced machine is treated in terms of theoretical considerations
that have already been fully exposed. (The computation of the performance of
polyphase motors is described in Chapter 4.)

431
 s Chap. 10
432 Unbalanced Two-Phase Motors: Servomotor

10-1 SYMMETRICAL COMPONENTS

system with balanced
The method of replacing an unbalanced four-phase voltage
te formulation of the
voltage sets is described first chiefly because a more comple
voltage system were treated
problem can be made. If the unbalanced two-phase
e voltage s or balanced
directly, there would be no need to deal with zero-sequenc
thereby be lost. In Fig.
single-phase voltage sets so that some generality would
In a balanced system the
10-1(a) is shown an unbalanced four-phase voltage set.
90° displaced from one
phase voltages would be equal in amplitude and exactly
d in Fig. 10-1(a). Figure
another. Neither condition is satisfied by the set depicte
10-1(b) shows that the phasor sum is not zero but adds, rather, to a resultant
e
quantity which is identified as 4V). The quantity V> is called the zero-sequenc
d to have been
voltage’ and the factor 4 states that this unbalance can be assume
each phase voltage. It is
brought about by the existence of an equal unbalance in
is in-phas e with the
called zero-sequence because the V, in any one phase voltage
V» quantity of the other phase voltages.
be read-
For any given unbalanced system the zero-sequence component can
ily found from

1V,+V,+ Ve + Va) (10-1)

The removal of Vy from each of the phase voltages results in a four-phase voltage
system that adds up to zero. Thus

(V, — Vo) + (Vs — Vo) + (V. — Vo) + (Va — Vo) = 9 (10-2)

For convenience let

Va ae Ve ¥ Vo

Ve = V, ie Vo
bs - ef (10-3)
Ve=V.— Va

Vp = Va iF Vo

Va

V, %

Ye
Vb yy Vp

Ve
(a) (b)

Figure 10-1 Unbalanced four-phase voltage system: (a) star notation; (b) identi-
fying the zero-sequence component Vo.
Sec. 10-1 Symmetrical Components 433

Equation (10-2) then becomes

Va aE Vp aia Ve ae Vp = 0 (10-4)

The foregoing is graphically represented in Fig. 10-2. It is important to keep in

mind that, although the phasor sum of V4, Vg, Vc, and Vp is zero, these quantities
still represent an unbalanced four-phase system in the sense that the voltages are
unequal in magnitude and have unequal phase-displacement angles.
Attention is next directed to the conditions that must be imposed in order
generally to represent the unbalanced voltage system of Eq. (10-4). Simplicity is
served if we can succeed in representing this unbalanced system by groups of
balanced voltages. Specifically, because we are dealing with a four-phase system,
it is desirable to make this representation in terms of balanced four-phase systems
to the extent possible. In this connection then a good starting point is to assume
that each of the phase voltages of the system of Eq. (10-4) contains an appropriate
component of a balanced positive sequence set as illustrated in Fig. 10-3. The
quantity V, is assumed to be acomponent of V4. The subscript 1 is used to denote
the positive sequence. Note that in a positive sequence set of four-phase balanced
voltages, the phase voltages are equal in magnitude and 90 degrees displaced from
one another and the rotation is counterclockwise (i.e., positive). The quantity aV;
is the corresponding component of Vz where the symbol a denotes the rotational
operator defined by

Gea bee DY) (10-5)

Similarly, a?V, is the positive sequence component of Vc and a3V, is the positive
sequence component of Vp. Observe that a? = —1 and a3 = 1 [90°.
What can we now add to the positive-sequence four-phase voltage set of Fig.
10-3 to ensure at least some unbalance in the resulting phase voltages? The inclu-
sion of a second set of balanced positive-sequence voltages is fruitless, as a glance

Ve Ene otafe Vo

x
-Vo Vp

Figure 10-2 Removing the zero-se- Figure 10-3 Balanced positive-se-

quence voltage V) from each phase quence four-phase voltage set.
voltage.
 omotors Chap. 10
434 Unbalanced Two-Phase Motors: Serv

two positive-sequence voltage sets still

at Fig. 10-4 reveals. The result of adding
fore seems worthwhile at this point
yields a balanced four-phase system. It there
ad a balanced negative-sequence four-
to investigate the effect of adding inste
Now the result is that illustrated in
phase voltage set to the system of Fig. 10-3.
ite sequence has the effect of
Fig. 10-5. Note that the selection of the oppos
preserving a balanced system, 1.€.,
unbalancing two pairs of voltages while still
the negative-sequence component of
yielding a zero resultant. The quantity V> is
ive sequence. Note too that for
V,. The subscript 2 is used to denote the negat
is a component of Vg, and aV> is a
the negative-sequence set it is a3V> that
achieving our goal of representing
component of Vp. This marks a further step in
le balanced sets.
an unbalanced voltage system in terms of suitab
to each phase of the system
It is now desirable to add a third set of voltages
each of the phase voltages (and
depicted in Fig. 10-5 but in such a way as to cause
in magnitude. In this connection
not just pairs) to be different in phase as well as
-phase set, keeping in mind that
let us explore the effect of adding a balanced single
zero in order to preserve balance.
the resultant of the four components is to yield
to V, and V> to yield
Accordingly, if we assume that a third component V; is added
added to aV, and av»
V,, then an equal and opposite quantity, a2V3, is assumed
be V3, while the third
to yield Vg. Similarly, the third component of Vcis taken to
of these additions are depicted
component of Vp is taken to be a’V;. The results
set differs from the zero-
in Fig. 10-6. Note that this single-phase reversed-phase
to zero. An inspection of
sequence set in that the former set taken together adds
balanced voltage sets, it
Fig. 10-6 discloses that, with the use of the indicated
m of unbalanced voltages.
becomes possible to represent any four-phase syste
sufficient flexibility to
The reason is that with such a composition there exists
ude from the others.
make each phase voltage different in angle and magnit

ov;

Figure 10-4 The addition of two positive-se- Figure 10-5 Addition of positive- and nega-
quence voltage sets yields a balanced resul- tive-sequence balanced four-phase voltage
tant set. sets.
Sec. 10-1 Symmetrical Components 435

Figure 10-6 Illustrating how an unbal-

anced set of four-phase voltages can be
represented generally by balanced sets.

However, our interest really focuses on the reverse situation, i.e., finding the
balanced voltage components in each phase when an unbalanced system is speci-
fied. The foregoing treatment was described as presented in order to emphasize
better how each component affects the total voltage.
The unbalanced phase voltages of the four-phase system can now be ex-
pressed mathematically as follows:

=V, + V2 + V3 (10-6)
= aV, + a3V> + a?V; (10-7)
= a’V; + a2V,+4+ V3 (10-8)
a3V, + aV2 + a2V3 (10-9)

The first term on the right side of each equation represents a component of the
balanced positive-sequence set, the second term represents a component of the
balanced negative-sequence set, and the third term represents a component of
the balanced single-phase set. Since each component set is a balanced set, to-
gether they yield a resultant that is zero. This is not surprising in view of the fact
that the unbalanced component was previously removed. [Refer to Eq. (10-3).]
If Eq. (10-3) is inserted into Eqs. (10-6) to (10-9), expressions for the original
unbalanced voltages in terms of the appropriate components are obtained. Thus

=V, + V.+ V3 + Vo (10-10)

V, = aV,; + a3V2 + a2V3 + Vo (10-11)
V. = a?V, + a2V2 + V3 + Vo (10-12)
= a@V, + aV2 + a2V3 + Vo (10-13)
 s Chap. 10
436 Unbalanced Two-Phase Motors: Servomotor

V, in terms of the given

The manner of finding the zero-sequence component
(10-1). An appropriate matter
unbalanced four-phase system is indicated by Eq.
the remain ing three balanced
of concern at this point is consideration of how
unbala nced system. Note
components can be determined in terms of the given
a is determ ined, then the
that, if the positive-sequence component V, of phase
phases are also known be-
corresponding components in each of the remaining
situati on obtain s for the
cause they differ by a rotation factor only. A similar
ents.
negative-sequence and the balanced single-phase compon
through (10-13) must
To find the positive-sequence component, Eqs. (10-10)
of these equations reveals
be manipulated in such a way as to isolate V;. A study
by a3, Eq. (10-12) by a’,
that this can be accomplished by multiplying Eq. (10-11)
ons with Eq. (10-10).
and Eq. (10-13) by a and then summing each of these equati
Thus
V.= Vit Vo+ Vst+ Vo (10-14)
av, = V;+ a2?V2 + aV3 + aVo (10-15)

a’V.. Vit V2+ a2V3 + a*Vo (10-16)

aVa= V; + aV2 + a3V; + aVo (10-17)

V,+aV,+ @2V.+aVy=4Vi+.0 + 0 + O (10-18)

Keep in mind that the quantities (1 + a + a2 + a3) and (1 + a?) are identically
the positive-
equal to zero. From the last equation it follows that the value of
sequence component of voltage is

(10-19)

since a3 = +j, a” = —1, anda = —j.

Applying a similar procedure to the negative-sequence and the balanced
single-phase set leads to the following results:

(10-20)

and
(10-21)

values of
Therefore by means of Eqs. (10-1), (10-19), (10-20), and (10-21) the
se system of
balanced voltage sets can be found to replace an unbalanced four-pha
voltages.
Example 10-1
The phase voltages of an unbalanced four-phase system are given by
V, = 120/0° = 120 + j0 (10-22)
Sec. 10-1 Symmetrical Components 437

V, = 75 /—60° = 37.5 — j65 (10-23)

Ve
V ll100 /—150° = —86.6 — j50 (10-24)
Va = 50/80° = 8.7 + j49 (10-25)
(a) Find the zero-sequence, positive-sequence, negative-sequence voltages
and the balanced single-phase set.
(b) Draw the phasor diagram showing the manner by which the individual
phase components add to yield the given unbalanced voltages.

Solution (a) By Eq. (10-1) the zero-sequence component for each phase is found
to be

Vo = 4(Vq + Vy + Ve + Va)
= 1(79.6 — j66) = 19.9 — j16.5 = 25.8/-39.8° (10-26)
The positive-sequence component as obtained from Eq. (10-19) is
VY, = 1(V, + iV, ia V.. — jV4a)

4(120 + jO0 + j37.5 + 65 + 86.6 + j50 — j8.7 + 49)

80.2 + j19.7 = 82.6 /13.8° (10-27)

The negative-sequence component is determined from Eq. (10-20) to be

Figure 10-7 Unbalanced four-phase voltage system of Example 10-1, and the
balanced voltage components.
 omotors Chap. 10
438 Unbalanced Two-Phase Motors: Serv

Vy = (Va — JV — Ve + iVa)
— 49)
= 4(120 + jO + j37.5 — 65 + 86.6 + j50 + j8.7
(10-28)
= 23.2 + j5.3 = 23.7/12.9°
Finally, the V3 component is computed to be
V; = 1(V. — Vp + V- — Va)
8.7 — j49)
= 4(120 + j0 — 37.5 + j65 — 86.6 — j50 —
(10-29)
32s = 8.9 /—107.8°
10-7.
(b) The required phasor diagram appears in Fig.

LANCED
10-2 SYMMETRICAL COMPONENTS APPLIED TO UNBA
TWO-PHASE MOTORS
treated as a special case of the
The unbalanced two-phase voltage system can be
nced symmetrical four-
unbalanced symmetrical four-phase system. The unbala
tively equal to — Va
phase system is one in which the voltages V.and V, are respec
of V, and Vp.
and —V,. Thus in the phasor diagram V, and V, are mirror images
on of an unbalanced two-
Appearing in Fig. 10-8(a) is the schematic representati
\Val = 1Val;
phase system. Keep in mind that for a balanced two-phase system
s. If either condi-
also these two quantities are exactly out of phase by 90 degree
in Fig. 10-8(b) is the
tion is not satisfied, the system is unbalanced. Depicted
system. The use
schematic diagram of the corresponding symmetrical four-phase
of any two successive phases yields a two-phase system.
represent an
What are the balanced components that can now be used to
compon ents as
unbalanced two-phase system of voltages? Do we need as many

(a) (b)
Figure 10-8 Two-phase system’ as a special case of the four-phase system:
(a) unbalanced two-phase system; (b) unbalanced symmetrical four-phase system.
Sec. 10-2 Components Applied to Unbalanced Two-Phase Motors 439

were required to represent the unbalanced four-phase system? The answers to

these questions follow immediately by treating the two-phase system in terms of
the symmetrical four-phase system and then applying Eqs. (10-1) and (10-19) to
(10-21). Introducing V. = —V, and V, = —V, into Eq. (10-1) leads to the conclu-
sion that the zero-sequence voltage is zero. This is entirely consistent with the
fact that, as a symmetrical system, the resultant of the four-phase voltages must
be identically equal to zero. Also, the insertion of these same quantities into Eq.
(10-19) discloses that the value of the positive-sequence component of phase a is
given by

or

Vi = (V. + jVs) (10-30)

Similarly, by Eq. (10-20) the negative-sequence component is found to be

V2 = 4(Va — jVp + Va — iV»)

or

V2 = (Va — iV) (10-31)

The value of V3 is determined from Eq. (10-21). Thus

V; = 4(V, — Vs — Va — Vs) = 0 (10-32)

On the basis of the foregoing analysis we reach the conclusion that any
unbalanced two-phase system can be completely represented by a two-phase
balanced positive-sequence component and a two-phase balanced negative-se-
quence component. The unbalanced phase voltages can then be expressed simply
as

(10-33)

(10-34)

where V, and V; are given by Eqs. (10-30) and (10-31).

Example 10-2
An unbalanced two-phase voltage system is described by

Vi= 120 /0° = 120+ j0 (10-35)

V, = 75 /—60° = 37.5 — j65 (10-36)

Find the positive- and negative-sequence components that can be used to represent
the unbalanced voltage system.
 s Chap. 10
440 Unbalanced Two-Phase Motors: Servomotor

Eq. (10-30) to
Solution The positive-sequence component of phase a is found from
be

V, = (Va + jVp) = 3(120 + jO + f37.5 + J65)

9.25 + j18.8 = 94.4/11.5° (10-37)

It first requires
Figure 10-9(a) depicts a simple graphical procedure for finding V,.
by 90° in the positive direction to yield jV,, then finding the phasor
that V,, be rotated
of V, and jV,, and finally taking one-half of this resultant. The desired quantity
sum
is indicated by OA.
follows
The analytical determination of the negative-sequence component
from Eq. (10-31). Thus

V2 = (Vq — jVs) = 4120 — 0 — f37.5 — 65)

27.5 — j18.8 = 33:3:/--34.4° (10-38)

The graphical determination of V, is illustrated in Fig. 10-9(b). It first requires that

V, be rotated in the negative direction by —90° to yield —jV,, then adding V, and
—jV,, and finally taking one-half of this sum. The desired quantity is identified as OB
and is equivalent to the result expressed in Eq. (10-38).
The complete expressions for the phase voltages in terms of the positive- and
negative-sequence components are then
V,=Vi + V2 = 9434 (115° -F BSS 4 eae Uns (10-39)

V, = aV, + a3V_ = —jV; + jV2 = 94.4 (518.50 5 33.3 /55.6° (10-40)

The graphical representation of the last two equations appears in Fig. 10-9(c).

The results embodied in Eqs. (10-33) and (10-34) lead to a very useful and
interesting physical interpretation of the analysis of the unbalanced two-phase
motor. Each phase voltage consists of a component of a balanced two-phase set,
one positive and the other negative. The positive-sequence components in each
phase, V, and aV,, can be considered to cooperate in creating a revolving mag-
netic field that interacts with the rotor winding to produce a positive-sequence

(a) (b) (c)

Figure 10-9 Phasor diagrams for Example 10-2: (a) graphical determination of the
positive-sequence component;'(b) graphical determination of the negative-
sequence component; (c) unbalanced voltages with balanced components.
 Sec. 10-3 Circuit and Performance Based on Symmetrical Components 441

Figure 10-10 Representation of the unbalanced two-phase servomotor in terms

of positive- and negative-sequence voltages.

torque T,. At the same time the negative-sequence components in each phase, V,
and a?V>, can be considered to combine in creating a revolving magnetic field of
reversed direction that interacts with the rotor winding to produce a negative-
sequence torque 7). Of course the rotor responds to the resultant torque, which is
the difference between 7; and T,. A schematic representation of this situation is
depicted in Fig. 10-10. Note that for convenience the rotor is treated as consisting
of two rotors that are mechanically coupled. The positive-sequence voltage set—
V; and aV,—produces 7;, and the negative-sequence set produces 7>. All the
known techniques of balanced operation can now be used to determine 7; and also
T>, and the results can then be superposed to obtain the resultant. Unbalanced
operation is thereby treated entirely in terms of the techniques of balanced opera-
tion. Accordingly, no new theory is needed to handle the two-phase motor be-
yond this point.

10-3 EQUIVALENT CIRCUIT AND PERFORMANCE BASED ON

SYMMETRICAL COMPONENTS

The equivalent circuit of the unbalanced two-phase motor actually consists of two
parts, one representing the motor response to the positive-sequence voltage set
and the other representing the response to the negative-sequence set. Because
each voltage sequence is a balanced set, the equivalent circuit configuration ex-
pressed on a per-phase basis is similar to that shown in Fig. 4-7(f). The circuit is
repeated in Fig. 10-11(a). The applied stator voltage for this case is the positive-
sequence voltage V,. If the main winding (phase a) of the motor is assumed
identical with the control winding (phase b), then the equivalent circuit for each
winding is identical. In this chapter we restrict attention to motors that have
identical stator phase windings. However, in those instances where this condition
 Chap. 10
442 Unbalanced Two-Phase Motors: Servomotors

(b)
(a)
un-
Figure 10-11 Two-part equivalent circuit of the two-phase servomotor with
balanced applied voltages: (a) equivalent circuit per phase of the balanced posi-
tive-sequence set; (b) the per-phase equivalent circuit of the negative-sequence
balanced set.

is not met, the use of a suitable transformer can be made to bring about the
balanced condition.
The parameters appearing in the equivalent circuit of Fig. 10-11(a) have the
same meanings as described in Chapter 4. The quantity r; denotes the stator
winding resistance per phase. For the motor with balanced windings this repre-
sents the winding resistance of either phase. The total magnetizing reactance per
phase is denoted by x4, and the total leakage reactances per phase of the stator
and rotor windings are respectively denoted by x; and x3. The prime notation
indicates that the last quantity is referred to the stator winding. The total rotor
resistance per phase is identified as r;. Use of the quantity r5/s in the equivalent
circuit implies that the rotor travels in the direction of the positive-sequence
rotating field. Finally, note that a core loss resistor is not included. Hence the
assumption throughout is that core loss is negligible.
The equivalent circuit of the motor in response to the balanced negative-
sequence set is depicted in Fig. 10-11(b). Each circuit parameter has the same
meaning as it has for the equivalent circuit of the positive-sequence set. By
superposition the negative-sequence voltage set is assumed to be applied to the
same stator windings and producing its own reactions in the same rotor winding.
This is the meaning that is intended in the schematic diagram of Fig. 10-10. There
is really only one pair of stator coils and one rotor, but the effect is the same as if
there were two identical pairs of stator coils and two identical rotors mechanically
coupled. A comparison of Fig. 10-11(b) with 10-11(a) shows differences occurring
in two places. One, the stator applied voltage now has the value V> instead of
V,. Two, the slip of the actual rotor rotation with respect to the revolving field of
the negative-sequence voltage set is 2 — s instead of s. Keep in mind that even
with unbalanced voltages the rotor can revolve at only one speed. The slip of the
rotor with respect to the positive-sequence field is then clearly
Sec. 10-3 Circuit and Performance Based on Symmetrical Components 443

i 2OR
s= (10-41)
Ns

Because the negative-sequence voltage set establishes a field that revolves in the
opposite direction to that of the positive sequence, it follows that the slip of the
rotor with respect to the negative-sequence field, sy is given by
n, — (—n) n
> Ss
rs 1 + y
— (10-42)
rs

In the interest of dealing solely in terms of s—the slip with respect to the positive-
sequence field—use is made of Eq. (10-41) in Eq. (10-42) to yield

u=tsZa2-s (10-43)

Therefore in all future work that involves the use of positive- and negative-se-
quence fields, the rotor slip with respect to the negative-sequence field will always
be denoted by 2 — s.
The computation of performance can be readily determined through the use
of our knowledge of the operation of a polyphase induction motor in response to
balanced applied voltages. Let us find the expression for the resultant torque and
the individual phase currents when the motor operates at a slip s with unbalanced
stator voltages. First we find the behavior of the motor in response to the posi-
tive-sequence voltages. This is easily accomplished by means of the equivalent
circuit of Fig. 10-11(a). In a typical problem V,, s, and the circuit parameters are
specified. As the initial step the input impedance occurring with the positive-
sequence voltage is found. Calling this impedance Z, we have

A= t+Z,=n+in ixgl(ro/s) + jx2]

+ SS
So vain thpelt de a) aN OETIERED (10-44)

where Z,; obviously denotes the impedance of the parallel arrangement of the
circuit elements. The second subscript, 1, identifies the positive sequence. With
the positive-sequence input impedance known, the positive-sequence component
of the input current in phase a is given by
opie
i= > (10-45)
Z|
The rotor current produced by the action of the positive-sequence rotating field
follows from
Py L,Z
Bya or
(F515) + JX3 ead
where the second subscript denotes the positive sequence. A little thought dis-
closes that an alternative expression for /5, can be had by applying the current
 Unbalanced Two-Phase Motors: Servomotors Chap. 10
444

usually preferred
divider rule for the parallel circuit. However, Eq. (10-46) is
in a convenient
because the quantities on the right side are frequently available
form.
Pa ,18
The gap power associated with the positive-sequence rotating feld
then found from
ia oe:
Pe = qn? 2 = Un S (10-47)
The corresponding developed torque is then

T; = Pa N-m (10-48)
@

where w is the synchronous angular frequency expressed in radians per second.

is
The performance of the motor in response to the negative-sequence set
e input impedanc e can be
found in a similar fashion. Thus the negative-sequenc
expressed by
prgl(ro/2 — 5s) + jx2] (10-49)
[eee eat Zain Sa koi ee A eo
Z)= 2 +2p — 71 + J (rh /2 sus) ati ected)

where Z, is the impedance of the parallel arrangement appearing in the circuit of

Fig. 10-11(b). The negative-sequence component of the current in stator phase a
becomes
a ary
= (10-50)
Ly
The associated rotor current takes on the value

LZ,»
In = (rs/2 — 8) + Jxh (10-51)

so that the expression for the power transferred across the air gap by the negative-
sequence flux field becomes
,

Pin ag)
NC eG
ea et |
(10-52)

and

I, =
_ Pe
a (10-53)

The resultant torque produced by the simultaneous effects of the positive-

and negative-sequence voltages expressed in newton-meters is

1 |
T= T, a To = o (Pai = P,2) (10-54)
Sec. 10-3 Circuit and Performance Based on Symmetrical Components 445

For servomotors; operating as unbalanced two-phase motors used in instrument

servomechanisms the value of torque in newton-meters is a very small quantity.
It is customary in such instances to refer rather to the torque in terms of its
proportional quantity, power. Hence it is not uncommon for control engineers to
refer to a servomotor torque as ‘‘so many units of synchronous watts.’’ We too
adopt this practice.
The resultant line current in phase a is

,=i,+h (10-55)

and that in phase b is similarly obtained as

Information about the power factor of each stator phase current is implied in the
polar expression of the phase currents. In normal applications of the servomotor,
efficiency is not very important and so this matter receives no further attention
here.
Example 10-3
A 5-W, 60-Hz, 120-V, two-pole, two-phase servomotor has the following
parameters:

ry, = 2850, ry = 8500

x, = 600, x4, = 600

xg = 995 O

When it is operating at a slip of 0.6, determine (a) the resultant torque in synchronous
watts and (b) the stator phase currents. Assume the servomotor operates with the
following unbalanced two-phase voltages: V, = 120 [0° and V; = 75 /—60°.

Solution (a) The positive- and negative-sequence voltages for the specified un-
balanced voltages are computed in Example 10-2 to be as follows:
V7, = 94.4 /11.5° and Vy, = ier ee

We first compute the motor response to the positive-sequence set. Thus

= — {995[(850/0.6) + j60] _ 79551415)
1760/36.8° 800 /53.2° (10-57)
0.
P| ~"“(850/+ 55
j106)
and

Z, = Z + Zp, = 285 + j60 + 481 + 7644 = 1038 /42.6° (10-58)

Hence
Bs piled 944 [11:99 Te (10-59)
h= 3 = 7038/40.6 ~ 0.091 /=31.1
 s Chap. 10
446 Unbalanced Two-Phase Motors: Servomotor

and

TZ 0.091 /—31.1° (800 /53.2°) ne £0514 /22.1° (10-60)

ees 1415 + j60 1415

The positive-sequence gap power then becomes

P= 2(0.0514)21415 = 2(3.7) = 7.4 W (10-61)

to the negative-sequence voltage set is computed to be as

The response
follows:
_ j995{(850/1.4) + j60] _ J995(606) _ r 10-62
=
1218 /60.1° py aes C068)
p2 ~~ (850/1.4) + j(995 + 60)
Zp = 1 + Zyy = 285 + f60 + 431 + j248 = 779 /23.3° (10-63)

oe Vie 779oe
I= Z, = 3 = 0.042 spun
/-57.7 (10-64)

ay Mh 0.042 /—57.7° (496 /29.9°) hi : :

mL (10-65)
+ j00 ~ Ti, mENeaT Yes Je 0.034 /—33.5
In = 606

Pyo = 2(0.0344)? 5 re HE 2(0.0344)7(606) = 2(0.716) = 1.43 W (10-66)

Hence the resultant gap power is

P, = Pg — Per = 7.4 — 1.43 = 5.97 synchronous watts (10-67)

(b) By Eq. (10-55) the expression for the current in stator phase a is
I, = 1, + b = 0.09/—31.1° + 0.042 /-57.7°
= 0.0778 — j0.047 + 0.0224 — j0.0355 (10-68)

= 0.1002 — j0.0825 = 0134739,55

Similarly by Eq. (10-56) the phase current in the control winding b is

I, = —jl, + jl, = —j0.0778 — 0.047 + j0.0224 + 0.0355
= —0.0115 — j0.0554 = 0.0566 /—101.7° (10-69)

Attention is next turned to the derivation of an oftentimes useful equation

for computing the resulting torque. The insertion of Eqs. (10-45) and (10-46) into
Eq. (10-47) leads to the following expression for the gap power of the positive-
sequence field

aeca ta
Vi\
a= 217) lt
nsral
Gals) + dual ies,
Zp1 2/1)
\s es
Let us now determine the expression for the gap power when the motor is
operated with balanced two-phase voltages, i.e., V, = V, and each displaced from
the other by 90 degrees. In this instance the phase current is given by
Sec. 10-3 Circuit and Performance Based on Symmetrical Components 447

Lae
a Vo
7 (10-71)

where Z, is described by Eq. (10-44). Moreover the rotor current expression is

5 bes
| (10-72)
where Z,; is specified as before. The expression for the total gap power for
balanced operation, P,, becomes therefore

AD rar ezeai()
Ss ; 74) ONT 31S) IKOiy\S Te an
By formulating the ratio of Eq. (10-70) to Eq. (10-73) there results

Vi)? P,
P= (>) (10-74)
By proceeding in a similar fashion it can be shown that
V. 2
P= (2) P} (10-75)
where P; denotes the gap power computed for balanced two-phase operation at a
slip of 2 — s. The resultant gap power is accordingly expressed as
2. V 2
Pins Pi goPi = (7) P, - ) Pi (10-76)
Correspondingly, the resultant torque can then be written as

= SIG a= (B)
T= Bye P, V, te (10-77)

or

72 (an (0) a
¥,\* V2\?
40.7%
where 7z; = P,/w and denotes the developed torque computed at slip s for
balanced two-phase operation at rated voltage, and Tg, = P,/w and denotes the
developed torque computed at slip 2 — s for balanced two-phase operation at rated
voltage.
The usefulness of Eq. (10-78) lies in the fact that the resultant torque at any
slip and for any condition of unbalanced two-phase voltages may be computed in
terms of the torque for full-voltage balanced operation. This information is usu-
ally supplied by the motor manufacturer. Hence the torque-speed characteristic
 Unbalanced Two-Phase Motors: Servomotors Chap. 10
448

for various values of the control winding voltage is readily determined from
Eq. (10-78) without resorting to the equivalent-circuit computational procedure.
This matter is further elaborated in the next section.

Example 10-4
Compute the resultant torque of the servomotor of Example 10-3 by employing
Eq. (10-78).

Solution Under conditions of balanced two-phase operation the phase current be-
comes

ae Va 120
I, = he= 7038 /—42.6° = 0.116/—42.6 (10-79)

The corresponding rotor current is

Tr
ipvaS
pl
_ 0. 116 /—42.6° (800 /53.2°) A tess °
te (iis) + xh 1415 ay
P, = 2(14) a = 2(0.0655)7(1415) = 12.1 W

Therefore the gap power of the positive-sequence set becomes by Eq. (10-61)

Y i) = (3
it(12.1) = 7.45 (10-80)
Lake ke g 120
A comparison with Eq. (10-61) indicates favorable agreement.
To find P;, the motor is assumed to be operating at a slip of 2 — s = 1.4 with
balanced two-phase voltages applied. Then

eee ate meio) .

I, = Z, = 779
779 [23.3°
3 = 0. 152/23. 3 (10-81)

and

i TZ 0.152 /—23.3° (496 /29.9°)

ees
/
ea <b
2
a I a ee °

= 2(0. 1245) = = 2(0.0155)606= 18.8 W (10-82)

Hence by Eq. (10-75) the gap power of the negative-sequence field is

V>\? 33.3
Po= ee Ph = (33) (18,8).= 1.45 (10-83)
which compares favorably with Eq. (10-66).
The resultant torque is therefore
1 ,
oe (7.45 — 1.45) = 375 (6) = 0.0159 N-m (10-84)
 Sec. 10-4 Servomotor Torque-Speed Curves 449

10-4 SERVOMOTOR TORQUE-SPEED CURVES

The torque-speed curve of a conventional two-phase induction motor supplied by

a balanced two-phase voltage has a shape similar to that of the three-phase induc-
tion motor. This characteristic is depicted as curve a in Fig. 10-12. The use of
such a motor in control systems is intolerable because of the positive slope that
prevails over most of the operating speed range. The positive slope represents
negative damping in the control system, which in turn can lead to a condition of
instability. Therefore for control systems applications the motor must be modi-
fied in a way that ensures positive damping over the full speed range. A conve-
nient way to achieve this result is to design the rotor with very high rotor resis-
tance. The torque-speed characteristic then assumes the shape shown by curve b
of Fig. 10-12. It is sketched over a range of slip from zero to 2 because of the
pertinence of that portion of the curve lying between s = 1 and s = 2 when
negative-sequence voltages exist in unbalanced operation. Whenever the two-
phase induction motor includes the high rotor resistance design feature, it is
referred to as a two-phase servomotor.
In many applications of the servomotor in feedback control systems, phase a
is energized with fixed rated voltage (often called the reference voltage), while
phase 5 is energized with a varying control voltage that is usually obtained from a
preceding amplifier stage. Moreover, the arrangement in these configurations is
such that the control voltage is frequently adjusted to be exactly 90° out of phase
with the voltage applied to phase a. In these instances unbalanced operation is
really an unbalance in voltage magnitude only. If we proceed under the assump-
tion that the reference voltage V, and the control voltage V, are always 90° apart
in phase and if we define the quantity p such that

V,
(es
Pie -85
(10-85)
then the phasor expression for the control voltage becomes

Vp = —JjpVi (10-86)

Figure 10-12 Torque-speed curves:

(a) for the balanced two-phase induc-
tion motor with low rotor resistance;
(b) for the same motor with high rotor
<— Slip resistance.
 Servomotors Chap. 10
450 Unbalanced Two-Phase Motors:

een the two stator phase voltages, and

The —j factor accounts for the 90° lag betw
nit ratio of the reference voltage. Upon
p expresses the magnitude of V; as a per-u e
the expression for the positive-sequenc
inserting Eq. (10-86) into Eq. (10-30),
voltage becomes

(10-87)

Similarly

(10-88)

e-speed curves for a two-phase

Appearing in Fig. 10-13 is a set of actual torqu
s of the control voltage expressed as a
servomotor corresponding to various value
marked 115 V (or p = 1) repre-
fraction (p) of the reference voltage. The curve
tion. However, note that, as the con-
sents the characteristic for balanced opera
e-speed characteristic becomes se-
trol voltage is reduced, the associated torqu
torque-speed characteristic of the
verely restricted. If information about the
balanced rated operation over the
servomotor is supplied by the manufacturer for
ed characteristic for any value
full range of slip from zero to 2, then the torque-spe
(10-87), (10-88), and (10-78). The
of p can be easily derived by means of Eqs.
procedure is illustrated in Fig. 10-14.

Figure 10-13 Torque-speed curves of

6000 ~-—sctthe: two-phase servomotor for various
rpm values of p.
Sec. 10-4 Servomotor Torque-Speed Curves 451

Torque

— —_—_——
— — —

n,, speed
mm
ee ee
-_—_ _ O, slip

p=,
neg. sequence

Figure 10-14 Graphical determination of the servomotor torque-speed curves for

p= 0.5.

Consider that the servomotor operates with the control voltage fixed at one-
half the value of the reference voltage, i.e., p = 2. Then by Eq. (10-87) the
positive-sequence voltage has a magnitude of
a V.
Via,
Ul tele (l+2)=4V, (10-89)
and the negative-sequence voltage is found by Eq. (10-88) to be
rey Cl eat!
maz Va '
— 4) =4/, (10-90)

Inserting these results into Eq. (10-78) leads to the following expression for the
resultant torque:

V,\? V2)? ‘ :
Ty = \=>) Taye \>) Te = felon — eT (10-91)
Vi, Vi,
The quantity *% 7p, is readily obtained over the entire speed range by taking nine-
sixteenths of the ordinate value of the characteristic for full-voltage balanced
operation at all values of slip. This result is identified as curve a in Fig. 10-14. In
like manner the quantity 7s 7g is easily determined over the entire speed range by
 omotors Chap. 10
452 Unbalanced Two-Phase Motors: Serv

teristic supplied by the manufac-

taking one-sixteenth of the value of the charac
is represented by curve b in Fig.
turer for full-voltage balanced operation. This
se the negative-sequence torque
10-14 and is drawn below the abscissa axis becau
resultant torque-speed charac-
is opposed to the positive-sequence torque. The
curves a and b, which leads
teristic over the useful speed range is found by adding
ted with a voltage V, = —j0.5Va,
to curve c. So long as the servomotor is opera
ing the foregoing procedure
curve c describes its operating behavior. By repeat
of p can be determined.
the resultant torque-speed characteristic for other values
obtainable without the neces-
It is interesting to note that these characteristics are
with the equivalent circuit as
sity of making repeated calculations in conjunction
procedure described here
outlined in Example 10-3. However, the alternative
balanced operation at full
does require having the torque-speed characteristic for
voltage over the slip range from 0 to vi
tion of the devel-
Another useful characteristic of the servomotor is a descrip
control voltage with speed as a parame ter. These
oped torque as a function of the
the use of Eq. (10-78) and the availab ility of the
curves, too, can be derived with

Torque

KS
we
Eig) Rasa
ma

Figure 10-15 Graphical construction for deriving the curves of Fig. 10-16. The
abscissa axis is slip or speed.
Chap. 10 Problems 453

Figure 10-16 Variation of resultant torque versus control voltage with slip as a
parameter.

torque-speed characteristic for balanced operation at rated voltage. Appearing in

Fig. 10-15 are the torque-speed characteristics corresponding to various values of
p for both the positive- and negative-sequence voltages. The curves for p = 0 are
identical but merely reversed in position consistent with their association with the
positive- and negative-sequence voltage sets. As p is increased toward unity the
positive-sequence contribution becomes increasingly larger than it is for p = 0,
while the corresponding negative-sequence contribution becomes increasingly
smaller than it is for p = 0. To obtain a plot of resultant torque as a function of
control voltage expressed as a ratio of the rated reference voltage at zero speed, it
is merely necessary to subtract the negative-sequence contribution from the posi-
tive-sequence torque. Thus at zero speed (or s = 1) the resultant torque is a; —
a, = 0 for p = 0. For p = 0.2 the resultant torque is b; — b2, and this quantity is
greater than zero. The net torque for p = 0.4 is c; — cz, and it should be apparent
that this quantity is greater than b; — b,. The complete plot is depicted in Fig.
10-16 as curve (a). Repeating the procedure at a second value of slip leads to
curve (b). Note the region of negative torque.

PROBLEMS

10-1. An unbalanced four-phase system is described by the following voltage set:

110/15°
1107-50"
= 110 j-13)

Sean
aay |=Il 110/90°
 Unbalanced Two-Phase Motors: Servomotors Chap. 10
454

(a) What is the value of the zero-sequence voltage?

voltage set, and draw the
(b) Find the phasor expression of the positive-sequence
corresponding phasor diagram.
(c) Repeat part (b) for the negative-sequence set.
(d) Repeat part (b) for the balanced single-phase set.
individual components
(e) Draw the complete phasor diagram, showing how the
combine to yield the given unbalanced voltage set.
phasors:
10-2. A four-phase voltage system is described by the following
¥, = 110/15"
V, = 60/—50°
V. = 140 /-135°
Va = 40/90°
the unbalanced
(a) Determine that set of balanced components that can replace
system.
showing all com-
(b) Draw the complete phasor diagram of this four-phase system,
ponents.
10-3. Show the derivation that leads to Eq. (10-20).
10-4. Show the derivation that leads to Eq. (10-21).
e voltage system is
10-5. The positive-sequence component of phase a of a four-phas
known to be equal to 60 [=30" Vv
of phase c?
(a) What is the phasor expression of the positive-sequence component
(b) What is the positive-sequence component of phase b?
(c) Draw the complete phasor diagram for the positive-sequence set.
system to
10-6. Under what conditions is it possible for an unbalanced four-phase voltage
yield a resultant voltage that is zero?
s:
10-7. One phase of a four-phase voltage system has the following component
V, = 60/-60°
V2 = 30 = 90)
Vz = "20 "at in
Vo = 0
(a) Find the phasor expression of phase b voltage.
(b) Is this a balanced four-phase system? Explain.
10-8. A four-phase system has the positive-, negative-, and zero-sequence voltages speci-
fied in Prob. 10-7. Draw to scale the complete phasor diagrams of the resulting four-
phase system. A graphical solution is permissible.
10-9. A four-phase voltage system is described as follows:
V. = 100/0°
Vp= 30/-90°
V..= 100 /—180°
Va ae 50 /90°
Chap. 10 Problems 455

(a) Is there a need for a balanced single-phase set when replacing this voltage
system with balanced sets? Explain.
(b) Find the positive-sequence voltage set.
(c) Find the negative-sequence voltage set.
10-10. The positive-sequence component of phase a of a four-phase system is 75 /Uns
and the negative-sequence component is 25 /—30°. The zero-sequence and the bal-
anced single-phase sets are both zero. Determine the expressions for the phase
voltages of the original four-phase system.
10-11. A four-phase voltage system is given by:
V, = 100 [0°
V, = 60/—90°
V.= 60 /— 180°
Vi= 60 /90°
(a) Determine the phasor expressions of that four-phase system which has a zero
resultant.
(b) Find the positive-sequence voltage set for the given system.
(c) Is the positive-sequence set the same for the system of part (a)? Explain.
(d) Find the negative-sequence voltage set.
(e) Compute the balanced single-phase set.
(f) Draw the phasor diagram for phase b, clearly identifying each component.
10-12. Compute the zero-, positive-, and negative-sequence sets as well as the balanced
single-phase set that can be used to replace the following four-phase system:

V. = 100/0°
a {
V, = 60/—65°
Va= 100 /— 180°
Va = 60/115°
What generalization can you draw concerning the component sets that make up the
original voltage system?
10-13. Find the unbalanced two-phase system that is represented by a positive-sequence
component of 95 2 [14.6° and a negative-sequence component of 31 /40.6°.
10-14. An unbalanced two-phase voltage system is described by:
Vs 120 /0°
V, = 80/-53°
(a) Find the positive-sequence voltage set of this system.
(b) What is the negative-sequence set?
(c) Draw the complete phasor diagram of this two-phase system, showing each
positive- and negative-sequence component.
10-15. Repeat Prob. 10-14 for the case where V, = 120/0° and V, = 80/-90°.
10-16. A two-phase, four-pole servomotor is energized by an auxiliary winding voltage
that is 90° out of phase with the main winding voltage. The main winding voltage
 Unbalanced Two-Phase Motors: Servomotors Chap. 10
456

rating at 60 Hz is 75 V. The motor parameters are as follows:

r, = 50Q, r5 = 1000
x, = 1200, x» = 1000
a slip of
The rotor leakage reactance is negligible. Compute the developed torque at
0.4 when the auxiliary voltage is 37.5 V.
two
10-17. Appropriate no-load tests on a two-phase servomotor rated at 115 V, 60 Hz,
poles, reveal the following parameters:
r, = 302 0, r5 = 1380 0
x4 = 695.0
x, = 385 0, x4 = 385 0,

This servomotor is operated with V, = 115/0° and V, = 80/—90°, and a rotor slip of
0.25.
(a) Draw the appropriate equivalent circuit for the positive- and negative-sequence
voltage sets. Show the proper applied voltage in each case.
(b) Compute the developed torque, expressed in synchronous watts and newton-
meters.
(c) What is the rms value of the current that flows through the control winding
(phase b)?
10-18. Refer to the servomotor of Prob. 10-17 and assume that it continues to operate at a
slip of 25%.
(a) Compute the torque developed by this motor when it operates as a balanced
two-phase motor, i.e., V, = 115/0° and V, = 115/-90°.
(b) Compute the developed torque when the motor operates as a balanced two-
phase motor at a slip of 1.75 per unit.
(c) By employing the results of parts (a) and (b), compute the torque developed by
this motor when V, = 115/0° and V, = 80/—90°.
10-19. Refer to the servomotor of Prob. 10-17. When this motor operates at a slip of 0.5,
the positive-sequence current is found to be 80.5/—51.1° ma and the negative-
sequence current is 38.5 [—98.4 ma.
(a) Determine the voltages applied to the main and control windings.
(b) Find the control winding current.
10-20. A two-phase servomotor has 115/0° V applied to the main winding. At a particular
point of operation corresponding to a control winding voltage of 75/—80°, the
positive- and negative-sequence impedances are, respectively, 220/67° and
RA SWlee
(a) Determine the positive- and negative-sequence components of the main and
control winding currents.
(b) Find the total values of the main and control winding currents.
10-21. The servomotor of Prob. 10-17 is operated with the main and control winding
voltages always in quadrature.
(a) Compute the stall torque when the control winding voltage equals the main
winding voltage, i.e., p = 1.
(b) Determine the stall torque for p = 0.5.
Chap. 10 Problems 457

(c) Assuming the no-load speed is 3400 rpm, calculate the ratio of stall torque to
speed, in radians per second, for p = 1.
(d) Of what usefulness is the quantity computed in part (c)?
10-22. A two-phase, two-pole, ac servomotor equipped with a drag-cup rotor has the
following parameters at 60 Hz:

rn = 360 0, r3 = 260 0
x, = 500, x,
= 500

X4 = 890 0
The main winding is identical to the control winding, and both are rated at 115 V.
The source of the control winding voltage is arranged to provide a quadrature
voltage at all times. Moreover, the internal impedance of this source is negligibly
small.
(a) Determine the developed torque in synchronous watts for p = 0.8 and s = 0.3.
(b) How much power is supplied to the main winding in part (a)?
(c) What is the power in the control winding in part (a)?
10-23. (a) For the servomotor of Prob. 10-22 compute the value of the developed torque
corresponding to operation with p = 0.1 and s = 0.05.
(b) Repeat this calculation for the servomotor of Prob. 7-17. Compare and com-
ment.
10-24. Refer to the servomotor of Prob. 10-22.
(a) Determine the developed torque in synchronous watts for p = 1 and s = 0.3.
(b) Repeat part (a) for p = 1 and s = 1.7.
(c) Find the developed torque for p = 0.7 and a rotor slip of 30%.
10-25. For the operating condition specified in Prob. 10-17, compute the ratio of the posi-
tive-sequence flux wave to the negative-sequence flux wave. Neglect saturation.
10-26. For the operating condition specified in Prob. 10-22(a) determine the ratio of the
positive-sequence flux wave to the negative-sequence flux wave. What significance
can you draw from this result?
 11

Stepper Motors

an
The explosive growth of the computer industry in recent years has meant
enormous growth too for stepper motors because these motors provide the driving
force in many computer peripheral devices. They can be found, for example,
driving the paper feed mechanism in line printers and printing terminals. Stepper
motors are also used exclusively in floppy disk drives where they provide precise
positioning of the magnetic head on the disks. The x- and y-coordinate pens in
expensive plotters are driven by stepper motors. Here they offer good dynamic
performance while eliminating the need for a heavy maintenance schedule associ-
ated with alternative drive schemes that employ gears, slide wires, and brushes.
The stepper motor is especially suited for such applications because essentially it
is a device which serves to convert input information in digital form to an output
that is mechanical. It thereby provides a natural interface with the digital com-
puter.
The stepper motor, however, can be found performing countless tasks out-
side the computer industry as well. In many commercial, military, and medical
applications, the stepper motors perform such functions as mixing, cutting, stir-
ring, metering, blending, and purging. They provide many supporting roles in the
manufacture of packaged foodstuffs, commercial end products, and even the
production of science fiction movies. The advantages offered by the stepper mo-
tor in these applications can easily be exploited with the use of two pieces of

458
 Sec. 11-2 Method of Operation 459

auxiliary equipment: the microprocessor and controlled switches in the form of

transistors.
Although the stepper motor is driven by a properly sequenced set of digital
signals, the motor itself exhibits the characteristics of a synchronous motor. In
one of its most popular forms, it acts as a doubly excited machine, which in turn
means it produces a steady-state torque at one speed.

11-1 CONSTRUCTION FEATURES

The construction of the stepper motor is quite simple. It consists of a slotted

stator equipped with two or more individual coils and a rotor structure that carries
no winding. The classification of the stepper motor is determined by how the
rotor is designed. If it is provided with a permanent magnet attached to its shaft, it
is called a permanent magnet (PM) stepper motor. If the permanent magnet is not
included, it is simply classified as a reluctance-type stepper motor. Attention here
is directed first to the PM stepper motor. Of course, the presence of the perma-
nent magnet furnishes the motor with the equivalent of a constant dc excitation.
Thus, when one or more of the stator coils are energized, the machine behaves as
a synchronous motor.
The basic construction details of the PM stepper motor are illustrated in Fig.
11-1(a) for a four-pole stator and a five-pole rotor structure. Observe that the
action of the permanent magnet on the particular orientation of the rotor structure
is to magnetize each of the poles (or slots) at one end of the rotor with an N-pole
polarity and each of the poles at the other end with an S-pole polarity. Moreover,
the N-pole set of rotor poles is arranged to be displaced from the S-pole set of
rotor poles by one-half of a pole-pitch and is readily evident by a comparison of
Figs. 11-1(b) and (c). The symmetry that exists between the stator poles and each
set of rotor poles makes it apparent that each rotor pole set behaves in an identical
fashion. For example, if coil A-A’ is assumed to be energized to yield an N-pole
polarity at A and an S-pole polarity at A’, then the relationship between N of the
stator and S1, S2, and SS of the rotor in Fig. 11-1(b) corresponds exactly to that of
S of the stator and N1, N2, and NS of the rotor in Fig. 11-1(c). A similar statement
can be made for the remaining sets of poles.

11-2 METHOD OF OPERATION

The manner in which the PM stepper motor can be used to perform precise
positioning is explained by examining the sequence of diagrams that appear in Fig.
11-2. The starting point is represented in Fig. 11-2(a) with coil A-A’ energized to
carry positive dc current so that the upper stator pole is given an N-pole polarity
and the corresponding lower stator pole is made to behave with an S-pole polar-
ity. Coil B-B’ is assumed to be de-energized. The rotor maintains this orientation
 *Aytrejod ajfod-N Jo sajod
BUIMOYS 4-q Je MIIA [CUOT}DAS-SSOID (9) ‘AjIeJOd aJod-g Jo sajod 1001 SUIMOYS D-D Je MIA
JeUOT}deS-ssoId (q) {WYeYS 10j01 oy} 0} poyoeye JouseW JUOURULIOd OY) SUIMOYS META TeIXE (2)
‘10}0ul Joddajs Wad ey) JO SoINjeay UONONNsuUOS oY) BUNeNSNT! wesiselq T-IT omsiy
(9) (0) (9)
joubow
34
JUDUDW

460
Sec. 11-2 Method of Operation 461

as long as the stator coil currents remain unchanged. At this position, the net
torque developed by the stepper motor is zero. This conclusion is readily appar-
ent in this case from symmetry considerations. For example, at the S-pole end of
the rotor it is seen that S1 is in alignment with the N-pole of the stator. The torque
produced by the action of the field distributions of the stator pole and the rotor
pole SI is dependent on the sine of the angular displacement of these two flux
fields. With exact alignment the angle is clearly zero and so too is the torque
between these two fields. Proceeding on the basis of superposition, we note that a
torque of attraction exists between the stator N-pole field and the field associated
with S5, which acts to cause counterclockwise (ccw) rotation. The magnitude of
this torque is proportional to the sin 72°, which is the rotor pole-pitch angle.
However, the action between the N-pole stator field and S2 is also one of attrac-
tion but it is exactly equal and opposite to that produced between the stator
N-pole and S5. A similar result is obtained upon examining the situation that
exists under the S-pole stator field. It is useful to note here that a difference in
detail does exist in that the torques involved are of a repulsive nature rather than
attractive as occurs beneath the N-pole stator field for this orientation. Because
the repulsive force has its maximum effect when a rotor flux field is in exact
alignment with a stator flux field, the cosine of the displacement angle between the
axes of two flux fields of opposite angle must be used. Accordingly, there is a
torque existing between the stator S-pole and the rotor S3 pole which is propor-
tional to the strength of these fields as well as the cos 36°, which acts to produce a
clockwise rotation. But the action between S4 and the stator S-pole is exactly
equal and opposite, thus yielding a resultant of zero torque produced by conditions
beneath the stator S-pole. The net torque is zero and so the rotor remains in a
position of equilibrium.
Let us now examine the effect of alternately energizing the two quadrature
stator coils in sequence through positive and negative values of dc excitation. The
first step in the sequence is represented in Fig. 11-2(b) by placing coil A-A’ in a de-
energized state and coil B-B’ in a fully energized state. The rotor is shown in the
final steady-state position that results as long as coil B-B’ is energized with posi-
tive current (i.e., so that the left pole becomes an N-pole and the right pole
assumes an S-pole polarity). The net rotor displacement for this switch in coil
excitation in this case is 18°. Once the rotor assumes the position depicted in Fig.
11-2(b), the orientation of the rotor poles relative to the newly located stator poles
is identical to that appearing in Fig. 11-2(a) and so the rotor is again in equilib-
rium. It is instructive, however, to examine the torque conditions that develop
when coil A-A’ is de-energized and coil B-B’ is energized while the rotor is still in
the position corresponding to Fig. 11-2(a). We proceed on the assumption that the
currents in these coils change instantaneously, i.e., the time constants of the
windings are very small compared to the mechanical time constants involved.
This is not an unrealistic assumption. Because the magnitude of the flux field
produced by a given coil excitation in the fixed geometry of a given machine is
determinable and since the permanent magnet flux field of each rotor pole is also
 (b)

Figure 11-2 Illustrating the indexing of the rotor in a ccw direction for a full cycle
of coil A-A’/coil B-B’ excitation in sequence through positive and negative val-
ues. Observe that the orientation between rotor and stator in (e) exactly corre-
sponds to the starting point in (a) except for rotation. (a) Coil A-A’ energized for
an N-S orientation; coil B-B'’ deenergized. Comment: starting point. (b) Coil
A-A' deenergized; coil B-B’ energized for an N-S orientation. Result: Rotor ad-
vances ccw 18° from its position in (a). (c) Coil A-A’ energized with reversed
current for an S-N orientation; coil B-B’ de-energized. Result: Rotor advances
ccew an additional 18° from its position in (b). (d) Coil A-A’ de-energized; coil B-B’
energized with reverse current for an S-N orientation. Result: Rotor advances
ccw an additional 18° from its position in (c). (e) Coil A-A' energized with positive
current for an N-S orientation; coil B-B’ de-energized. Result: Rotor advances an
additional 18° for a total of 72° or one rotor pole pitch corresponding to one cycle
of + excitation of stator coils.

462
Sec. 11-2 Method of Operation 463

N tc) N

(d)

(e)
Figure 11-2 (Continued)
 Stepper Motors Chap. 11
464

calculable, we can represent the maximum torque produced by these fields as

T,,. When the rotor orientation is such that the torque developed is other than
maximum, the appropriate sine or cosine function is introduced depending upon
whether the torque is one of attraction or one of repulsion. Thus with coil B-B’
n
energized and the rotor still in the position depicted in Fig. 11-2(a), the expressio
for the torque employing superposition is

S2 Sl S3 OE) S4
T=T; sing + sin ee — sin ee + cos Kae — COs es | (11-1)

where the plus sign is used if the torque acts to produce ccw movement and the
negative sign is used if the movement would be cw. Inserting the proper angles
corresponding to the stator pole orientation of Fig. 11-2(b) and the rotor position
of Fig. 11-2(a), Eq. (11-1) becomes

T = T,,(sin 18° + sin 90° — sin 54° + cos 18° — cos 54°)
= T,,(0.309 + 1.0 — 809 + 0.951 — 0.588)
= 0.8637,, (11-2)

Because this is a positive torque, the rotor in Fig. 11-2(a) moves to the new
equilibrium position shown in Fig. 11-2(b).
Consider next that coil B-B' is switched off and coil A-A' is again energized
but this time with current flowing in the opposite direction. The upper stator pole
assumes an S-pole polarity and the lower stator pole takes on an N-pole polarity.
Accordingly, the new stator flux configuration [in Fig. 11-2(c)] exerts a torque on
the rotor orientation of Fig. 11-2(b) which is again described by Eq. (11-2). The
ensuing torque displaces the rotor by an additional 18° counterclockwise in order
to reach the new equilibrium position shown in Fig. 11-2(c). In the subsequent
step in the sequence, the current in coil A-A’ is switched off while the current in
coil B-B’ is switched on but in the reversed direction. The stator poles then
assume the polarity depicted in Fig. 11-2(d), which in turn produces a torque on
the rotor to yield a further 18° displacement in the ccw direction. In other words,
the rotor moves to the new equilibrium position that finds S4 in alignment with the
N-pole associated with coil B-B’.
It is instructive to observe here that once a sequence is initiated and main-
tained through the proper positive and negative alternate excitation of the stator
coils, the action of the torque produced by the subsequent coil excitation is to
displace the rotor in the established direction. This is done in a manner to ensure
that the rotor pole having the least angular alignment with a stator pole of opposite
polarity is brought into exact alignment with that pole. Accordingly, when coil
A-A’' is next energized, the result is to move SS into alignment with the upper
stator N-pole. The new equilibrium position is shown in Fig. 11-2(e). Observe
that this last orientation is identical to the first one except that the entire rotor
structure has been indexed through a total displacement equal to one rotor pole
 Sec. 11-3 Drive Amplifiers and Translator Logic 465

pitch, or 72° in this case. Note too that for a two-coil (or two-phase) stator with
two states (positive and negative), it requires four switching operations to com-
plete a switching cycle. Thus, this motor is made to step four times during each
cycle of excitation which in turn produces a net displacement of one rotor pole
pitch. Or, stated differently, each switch of the stator excitation produces a rotor
displacement of one-fourth a rotor pole pitch. The stepping action exhibited by
this motor in response to alternate sequenced excitation of stator coils is the
reason it is called a stepper motor.
The direction of rotation of the stepper motor can be conveniently reversed
by merely reversing the switching sequence. Thus, if in Fig. 11-2(b) the excitation
of coil B-B' were reversed, then the action of the developed torque would be to
displace S5 in Fig. 11-2(a) into a position of exact alignment with the N-pole flux
which would now be associated with coil B’. The result is a clockwise displace-
ment of 18° in contrast to the counterclockwise displacement that occurs when
coil B is made to produce an N-pole polarity.

11-3 DRIVE AMPLIFIERS AND TRANSLATOR LOGIC

The foregoing description of the method of operation of the stepper motor makes
it plain that the stepping action of the motor is dependent on a specific switching
sequence that serves to energize and de-energize the stator coils. A practical
scheme that achieves this objective is the one of a drive amplifier consisting of
transistors that are switched on and off sequentially in accordance with the signals
that originate from an appropriate translator logic control circuit. A block dia-
gram of the arrangement is shown in Fig. 11-3. The drive amplifier provides the dc
excitation to coils A-A’ and B-B’ through switched circuitry that is controlled by
the translator logic device. Appearing in Fig. 11-4 is a detailed description of the
composition of one type of drive amplifier, namely, the bipolar type. A total of
four switching transistors is needed for each phase coil whenever a single power

Drive a Stepper Mechanical

amplifier motor load

Translator
logic

Figure 11-3 Block diagram of the stepper motor and its switched supply.
 Stepper Motors Chap. 11
466
Drive amplifier

.
Figure 11-4 Diagram illustrating the details of a bipolar drive amplifier

supply is used and current reversal in the phase coils is desired. The term bipolar
coil
is used for such a circuit because the circuitry permits either end of the phase
to be connected to the positive side of the power supply. There is a drive circuit
available that employs two power supplies in which case it becomes necessary to
use only two transistors per phase coil. In addition to the eight transistors, the
drive amplifier of Fig. 11-4 also includes eight diodes. Of course, the diodes are
required to provide paths for the inductive coil currents to flow whenever the
transistors are switched off by logic control circuitry. For example, if transistors
Q1 and Q2 are ON during the first quarter of a switching cycle thus causing
current to flow in the phase coil from A to A’, the transition into the next quarter
of the cycle calls for Q1 and Q2 to be switched OFF. Because of the magnetic
storage associated with the current in coil A-A’, the current cannot be reduced to
zero instantaneously. But, as it attempts to do so, a large emf is induced in the
coil with terminal A’ positive and A negative. This induced emf puts a positive
voltage across diodes D3 and D4 thus furnishing a path for the switched-off
current in coil A-A’. The current decays rapidly because of the small time con-
stant of the discharging circuit.
How is the logic schedule of the translator control circuit determined? This
is best explained by developing thé logic required to perform the operations of the
Sec. 11-3 Drive Amplifiers and Translator Logic 467

stepper motor described in connection with Fig. 11-2. In this case, a complete
switching cycle required that the stator coils be energized in accordance with the
schedule in the table below for a counterclockwise rotation.

Coil A-A' Coil B-B'

ON+ OFF
OFF ON+
ON- OFF
OFF ON-

Here ON+ is distinguished from ON— by a current reversal in the coil. The logic
schedule that achieves this switching sequence for the drive amplifier of Fig. 11-4
expressed in binary language where | denotes ON and 0 denotes OFF is as
follows:

Coil A-A’ Coil B-B’

Qtl! 222h Qt! 04 Q Q6 Q7~ 8
1 1 0 0 0 0 0 0
0 0 0 0 1 I 0 0
0 0 I 1 0 0 0 0
0 0 0 0 0 0 1 I

Each repetition of this switching cycle provides an advancement of the stepper

motor equal to one rotor pole pitch. Hence in this case five cycles of this switch-
ing sequence are needed to complete one full revolution of the rotor.
A very popular circuit for the drive amplifier of stepper motors is the one
illustrated in Fig. 11-5. This circuit clearly has the advantage of fewer electronic
components and simpler bias circuitry for the switching transistors. Each phase
coil of the stepper motor that is used with these drives is equipped with a center
tap and the switching logic for the associated transistors arranged so that current
flows only in one section at a time. Reversal of the magnetic pole polarity is
achieved by directing the current either to one section or to the other. The disad-
vantage of such a scheme is that the copper is not made to work continuously in
producing useful output. This arrangement is called the bifilar unipolar drive. It
is referred to as unipolar because the ends of both coil sections stay connected to
the negative side of the power supply during operation unlike the bipolar case
where the connection alternates between the high and low side of the source.
The logic schedule to be furnished by the translator control circuitry for each
full switching cycle of the operation of a stepper motor such as the one illustrated
in Fig. 11-2 is the following:
 Stepper Motors Chap. 11
468

1 0
0 0
0 0
0 1

the point of
It is assumed here that coil A is equipped with a center tap by using
11-1. A similar arrangem ent is assumed to
connection of coil A and coil A’ in Fig.
be used for coil B.
The diodes D1 through D4 appearing in the circuitry of the drive amplifier
the
shown in Fig. 11-5 are for the purpose of maintaining a current path following
cutoff of a particular transistor switch. For example, when current is flowing
through the left-half section of coil A in Fig. 11-5 and Q1 is subsequently cut off, a
circuit is established by transformer action in the second section of coil A which
allows the current in the first section to drop to zero as the current in the second
section instantaneously rises to the level of the current in the first section. The
latter current then flows through a path that involves the power supply and diode
D2. Diode D2 is made conductive since a positive voltage appears across it as a
Drive amplifier

Wel
Ioihes

Figure 11-5 Drive amplifier of the bifilar, unipolar type. Each phase coil of the
stepper motor comes equipped with a center tap.
 Sec. 11-4 Half-Stepping and the Required Switching Sequence
469

result of the transformer voltage induced in the right-half section of coil

A. The
transfer of current from the left section to the right section of coil A is allowed
to
occur instantaneously because there is no change in magnetic energy
in this
process. The mmf of both coils acts on the same magnetic circuit.

11-4 HALF-STEPPING AND THE REQUIRED

SWITCHING SEQUENCE

The description of the operation of the stepper motor so far has been limited to
situations where either one stator coil or the other was allowed to be energized at
any one time. What would be the consequence if a switching schedule is devised
which permits both stator coils to be excited simultaneously at certain periods ina
cycle? To examine this matter, let us return to the configuration depicted in Fig.
11-2(a). Recall that a full 18° ccw displacement resulted from switching coil A-A’
off and coil B-B’ on. However, suppose that coil B-B’ is switched on while main-
taining the excitation on coil A-A’. In other words, coil A-A’ is not de-energized
when excitation is applied to coil B-B’. Intuition, based on the symmetry preva-

Figure 11-6 Half-step equilibrium position that occurs when the rotor is initially
at the position illustrated in Fig. 11-2(a) and both coils A-A’ and B-B' are ener-
gized to produce the field orientation shown here. The result is a 9° displacement
of the rotor as indicated here.
 Stepper Motors Chap. 11
470

the motor takes a half-step, i.e., it

lent in this situation, leads us to suspect that
a rotor pole-pitch. An examination
advances in the ccw direction by one-eighth of
at this displacement angle reveals the
of the resultant torque on the rotor structure
ed by writing the expression for
motor to be in equilibrium. This is readily verifi
this condition. The situation is
the net torque on the rotor corresponding to
Note that all four stator poles are
illustrated in Fig. 11-6 for a half-step angle of 9°.
superposition, the expression
energized with the polarities indicated. Applying
for the net torque can be written as
N N Sa’ Sa’
N co s + cos ay
|-sinKale — sin Su + sin se —
Tees

Np Ng : Np Sp’ Sp’
+ si — sin + cos = "cos
+ si 4
a See3 sin 5 a <i)

by the stator poles created by

where the first line represents the torque produced
rotor and the terms on the
coil A-A’ interacting with the S-pole section of the
ated with the stator poles of
second line represent the torque contributions associ
inserting the angular displace-
coil B-B’ and the S-pole section of rotor. Upon
we get
ments existing between the axes of the indicated poles,
Te = Te(-sin 9 — sin 81" sin 63° — cos 27° + cos 45°

+ sin 9° + sin 81° — sin 63° + cos 27° — cos 45°)

=0
an equilibrium position.
Hence, this half-step displacement of the rotor represents
B-B’ is main-
If coil A-A’ is now de-energized while the excitation of coil
ly expected 18°
tained, the motor steps an additional 9°, thus yielding the original
d on. An exten-
that results when one coil is switched off and the other is switche
execute d by half-step
sion of this analysis through a complete switching cycle
counter clockwi se rota-
increments leads to the following switching sequence for
tion:
Coil A-A’ Coil B-B'

ON+ OFF
ON+ ON+
OFF ON+
ON- ON+ One complete
ON- OFF switching cycle
ON- ON-
OFF ON-
ON+ ON-

ON+ OFF
 Sec. 11-5 Reluctance-Type Stepper Motor 471

A full switching cycle now includes a total of eight switching states, which yield
eight half-steps, to give a total displacement of one rotor pole pitch. The switch-
ing schedule required to half-step the motor of Fig. 11-1 arranged to operate in the
bifilar mode and driven by the unipolar drive amplifier of Fig. 11-5 is detailed in
the following table:

Coil A Coil B

Qi Q2 Q3 Q4
i 0 0 0
1 0 1 0
0 0 1 0
0 1 1 0
0 1 0 0
0 1 0 1
0 0 0 1
1 0 0 1

The advantage of half-stepping of course is that it improves the precision of the

stepper motor in applications where more precise positioning is demanded.

11-5 RELUCTANCE-TYPE STEPPER MOTOR

The reluctance-type stepper motor is not equipped with a permanent magnet on

the rotor. It develops an indexing torque in response to the sequenced dc excita-
tion of stator coils by virtue of the large difference in magnetic reluctance that
exists between a direct-axis path and a quadrature-axis path. The resulting
stationary flux fields produced by the dc current in selected stator coils always
exert a force which acts to minimize the reluctance of the flux path in accordance
with the characterization of Eq. (1-65). A typical configuration of the reluctance-
type stepper motor is illustrated in Fig. 11-7. The stator in this case is designed for
eight poles and is equipped with four phase coils each of two sections. The rotor
is slotted to produce the effect of six poles corresponding to which is a pole pitch
of 60°. The rotor is at a position of equilibrium which occurs when coil A-A’ (i.e.,
phase A) is energized and the remaining phases are left unexcited. Observe that
the constant flux field produced by phase A encounters a path of minimum reluc-
tance by passing through rotor poles | and 4 which are aligned with the stator pole
structures associated with phase A.
For the combination of stator and rotor poles that is designed into this
motor, it is instructive to note that the axes of both rotor poles 2 and 5 are 15°
displaced from the axes of the two stator poles associated with phase B. Hence, if
phase A is de-energized and phase B is energized to produce an N-pole polarity at
coil B and an S-pole polarity at coil B’, a reluctance torque is developed that acts
 Stepper Motors Chap. 11
472

Figure 11-7 Construction details of a reluctance-type stepper motor equipped

with four stator phases (i.e., eight poles) and six rotor poles.

to align the axes of rotor poles 2 and 5 with the axes of the stator poles at B-B’.
The rotor advance is 15° ccw which in this machine too is one-quarter of a rotor
pole pitch. Successive excitations of phases C, D, and A yield a total advance in
the ccw direction equal to a rotor pole pitch. A sixfold repetition of this switching
sequence for the excitation of the stator phases results in one revolution of the
motor shaft.
The equilibrium position of the rotor in Fig. 11-7 also prepares the motor to
move in a clockwise direction. However, it now requires that coil D’ be energized
for an N-pole polarity and coil D for an S-pole polarity following the removal of
excitation on phase A. A reluctance torque now develops that pulls rotor teeth 6
and 3 into alignment with the poles produced by coil D’-D. Clearly this results ina
clockwise rotation.
The design of the drive amplifiers and the translator logic for the reluctance-
type stepper motor differ in some details from those reported for the PM stepper
motor. This is necessary in order to account for differences in the design aspects
of the two types of motors. The principles, however, remain the same and are
easily transferable from one motor type to the other.
 Sec. 11-6 Ratings and Other Characteristics 473

11-6 RATINGS AND OTHER CHARACTERISTICS

On the basis of the material that is related in the preceding pages concerning the
stepper motor, it should come as no surprise that these motors are used in circum-
stances where the need for large horsepower is not a factor. A motor’s output
power is dependent upon the extent to which the copper is worked electrically and
the iron is worked magnetically. The stepper motor ranks low on both counts. To
achieve the characteristics which make the stepper motor unique, it is necessary
that during each complete switching cycle each of the phase coils spends time in a
de-energized state and that some parts of the magnetic flux paths be essentially
free of field flux. Consequently, stepper motors are found to range in output
power from about 1 W to a maximum size of 3 hp. In terms of physical dimen-
sions, the largest machines are built with diameters up to seven inches while the
smallest ones are constructed in a pancake form with diameters as small as | in.
Many step sizes are also available with stepper motors. They can be pur-
chased with step sizes as small as 0.72° or as large as 90°. The most common step
sizes are 1.8°, 7.5°, and 15°. To procure a step size of 1.8° in a PM type, it is
necessary to design the stator for 40 poles and the rotor for 50 poles (or slots).
The shape of the torque-speed curve of the stepper motor takes on the
general form depicted in Fig. 11-8. The speeds indicated in this particular motor
are achieved as a result of programming the switching sequence for appropriate
repetition. For example, to realize one revolution per second, it is necessary for
the translators’ logic to switch a 1.8° step motor at the rate of 200 steps per
second. Any attempt to operate this motor at a rate of five revolutions per second
fails because the torque drops to zero. This decrease in torque is attributable to
the effects of the speed voltages induced in the phase coils at the higher speeds as
well as to the inductance of these coils. Compensation circuits are available

300

200

oz—in.
Torque,
100

fe) 200 400 600 800 1000 Figure 11-8 Torque-speed curve of a
Speed, |.8°- steps/s PM stepper motor operated at full step.
 Stepper Motors Chap. 11
474

is at the cost of increased circuit

which can double the useful speed range but this
complexity for the drive amplifiers.

PROBLEMS

position depicted in Fig. 11-2(b).

11-1. Consider the situation corresponding to the rotor coil A’ is
magnetic material on which
Assume that coil A-A’ is energized so that the
of coil A is of an S-pole polarity.
wrapped is of an N-pole polarity and that
produc ed on the rotor by the S-pole section
(a) Write the expression for the torque
dal flux distrib utions throughout.
of the rotor. Assume sinusoi
degrees of rotatio n occur before a positio n of equilibrium is reached?
(b) How many
N-pole section of the rotor.
11-2. Repeat Prob. 11-1 for the conditions that exist at the
cooperation of a single PM rotor
11-3. Let T,,, denote the maximum torque produced by the
in the configu ration of Fig. 11-2. Calculate the net torque
pole with the stator pole
by Fig. 11-2(c).
produced on the rotor shaft for the condition represented
motor is shown in Fig.
11-4. The schematic diagram of a four-phase, two-pole PM stepper
of a translat or logic circuit.
P11-4. The phase coils are excited in sequence by means
(a) Write the logic schedule for full-stepping of this motor.
(b) What is the displacement angle of the full step?
half-step mode.
(c) Write the logic schedule for operating this: motor in the
a sketch of the stator poles and the relative rotor position when coils B and
(d) Make
C are simultaneously energized.
d from the position
(e) What is the total rotor displacement in part (d) measure
depicted in Fig. P11-4?

Figure P11-4
Chap. 11 Problems 475

11-5. The excitation signals used to drive the coils in the motor of Fig. P11-4 originate from
digital signals that drive the supply amplifier. A continuous sequencing of these
signals will cause the stepper motor to run at a constant speed (thus behaving as a
synchronous motor). Determine the frequency rate of these digital signals if the
speed of rotation is to be 360 rpm. Assume a half-stepping pattern.
11-6. Determine the frequency rate at which the digital signals that drive the amplifier of
the motor of Fig. 11-2 must be set in the half-step mode if the motor is to run at a
speed of 360 rpm.
11-7. The drive amplifier of Fig. 11-4 is used to drive the stepper motor of Fig. 11-2 in the
full-step mode. Write the logic schedule for switching the transistors in a manner
such that it yields a clockwise direction of rotation.
11-8. A PM stepper motor is designed to provide a half-step size of 0.9°.
(a) Determine the rotor pole pitch in degrees.
(b) Find the number of stator poles.
11-9. Draw the diagram that illustrates the orientation of the stator poles with the N-pole
section of the rotor for the half-step case illustrated in Fig. 11-6.
 12

Synchros

servomech-
Synchros are ac electromagnetic devices that find wide applications in
instrume nts for registering
anisms.+ However, they can also be used as indicating
n is first directed toa
mechanical displacements remotely. In this chapter attentio
as well as an explanat ion
description of the construction features of the synchros
of
of their theory of operation. This is then followed by a discussion of the types
errors,
errors that are found in such devices which include static errors, dynamic
and residual voltages.

12-1 CONSTRUCTION FEATURES

There are four basic types of synchros: the control transmitter (denoted by cx),
the control transformer (cr), the control differential (cD), and the control receiver
(cr). The essential construction characteristics of the control transmitter are rep-
resented in Fig. 12-1(a). It consists of a stator and a rotor. The inner surface of
the stator is slotted to accommodate a balanced three-phase winding, which is
usually of the concentric-coil type. The rotor is of the dumbbell construction with

vari-
+A servomechanism is an error-actuated feedback control system in which the controlled
able is mechanical.

476
Sec. 12-1 Construction Features 477

Balanced
three-phase
distributed as
winding

Rotor
winding

(a) (b)

Figure 12-1 Synchro control transmitter: (a) construction features; (b) schematic
diagram.

many turns of wire wrapped around the stem. An ac voltage is applied to the rotor
winding through a pair of slip rings, which causes an excitation current to flow and
produce an alternating flux field with respect to the rotor structure. By definition
the electrical zero position of the rotor corresponds to maximum coupling with the
turns of phase winding 2 of the stator. Moreover, since there are two rotor posi-
tions for maximum coupling, a unique electrical zero is identified by matching dots
as shown in Fig. 12-1(b). Figure 12-1(b) depicts the rotor axis displaced from the
electrical zero by the angle a. Note the similarity in construction to the three-
phase synchronous generator; the synchro can be aptly described as a miniature
generator. It is important, however, to understand the distinction in the way the
two machines are used. The synchronous generator is excited with a de voltage
and driven at constant speed, and produces a three-phase voltage at the stator
terminals. In contrast, the synchro is excited with an ac voltage and is often
merely displaced by finite amounts from the electrical zero, and thus induces
single-phase voltages through transformer action in the stator winding. Some-
times, in control system applications, the synchros are driven at low tracking
speeds.
The construction details of the synchro control transformer differ somewhat
air gap
from those of the control transmitter. One important difference is that the
construct ion. Refer to
is uniform owing to the rotor’s cylindrical or umbrella-like
magnetizi ng current
Fig. 12-2(a) and (b). This feature is included to keep the
is the way in
drawn by the control transformer to a minimum. Another difference
er it is defined as
which the electrical zero is defined. For the control transform
2 of the stator
that position of the rotor which makes the coupling with winding
The stator winding
zero. This particular configuration is depicted in Fig. 12-2(c).
the transmitt er in
is also a balanced three-phase winding; however, it differs from
 Synchros Chap. 12
478

Balanced
three-phase
distributed
winding

Cylindrical
rotor

Distributed
single—ph Umbrella
Stator ing ti Concentrated
single-phase rotor
winding

(a) (b)

(c)
rotor
Figure 12-2 Control transformer: (a) construction features of cylindrical
diagram with
type; (b) construction features of umbrella rotor type; (c) schematic
rotor winding at electrical zero.

that the transformer has a higher impedance per phase. This latter feature permits
several control transformers to be fed from a single transmitter.
The differential synchro has a balanced three-phase distributed winding in
both the stator and the rotor. Moreover, the rotor is cylindrically shaped as
shown in Fig. 12-3(a). Although three-phase windings are involved, it is important
to keep in mind that these units deal solely with single-phase voltages.
The basic structure of the control receiver is the same as that of the control
transmitter. It has a balanced three-phase stator winding and a salient-pole ro-
tor. The one detail in which the receiver differs from the transmitter is the inclu-
sion of a mechanical viscous damper on its shaft. In normal use both the rotor and
 Sec. 12-2 Voltage Relations 479

Balanced
three-phase
distributed
winding

Stator Rotor

(a) (b)

Figure 12-3 Differential synchro: (a) construction features; (b) schematic representation.

stator windings are excited with single-phase currents. Accordingly, the field
distribution of the rotor interacts with the ampere-conductor distribution of the
stator to produce torque and hence rotation. The purpose of the damper is to
permit the receiver rotor to respond to the changing stator ampere-conductor
distribution without causing the rotor to overshoot its mark. If the overshoot is
large, average torque may be produced such that the receiver runs as a single-
phase motor.

12-2 VOLTAGE RELATIONS

is
A description of how the stator phase voltages vary with rotor displacement
synchros. In this connectio n
useful in understanding the various applications of
Fig. 12-4.
refer to the schematic diagram of the control transmitter shown in
ase ac voltage and that
Assume that the rotor winding is excited with a single-ph
by the angle a and held fast.
the rotor is displaced from its electrical zero position
the same frequency as the
By transformer action single-phase voltages having
phase. The value of the
rotor applied voltage will be induced in each stator
with the rotor winding.
induced voltage depends upon the coupling of each phase
To describe these relationships mathematically let
_ effective stator turns (12-1)
ti
roto
¢ = ‘effec turns
r ve
(12-2)
E, = rms value of the induced emf in the rotor winding

and
(12-3)
E = aE,
voltage that can be in-
The last expression denotes the maximum value of rms
duced in any balanced stator phase.
 Synchros Chap. 12
480

Figure 12-4 Circuit for deriving the

stator phase and line voltage variations
3 as a function of rotor displacement a.

The rms value of the emf induced in winding 2 of the stator when the rotor is
at the angle a is described by
EsqaE cose (12-4)
Note that, when a is zero, the rotor is completely linked with winding 2 and so
yields the maximum rms voltage E. This situation is entirely consistent with the
definition of the electrical zero for the control transmitter. The expression for the
instantaneous value of this voltage can be written as
€>,(t) = V2 E sin wt cos a (12-5)
where w is the frequency of the applied rotor voltage. A study of the last expres-
sion should make it apparent that the sin wt term accounts for the transformer
action that takes place in the synchro and that cos a describes the coupling
between the rotor winding and winding 2 of the stator. The use of a properly
distributed and pitched stator winding that yields a quasisinusoidally distributed
field is the reason for the cosine function in Eq. (12-4). If the winding were
distributed differently, a different functional relationship would be needed.t
However, throughout this chapter, we assume the use of sinusoidally distributed
windings only.

The value of the rms voltage induced in winding 1 is given by

E,, = E cos (a — 120°) (12-6)
The argument is modified by —120° to account for the fact that the axis of winding
1 is located 120° ahead of the axis of winding 2. The corresponding expression for
the instantaneous voltage then becomes
ein(t) = V2 E sin ot cos (a — 120°) (12-7)

+Some synchros are designed to provide a linear output with rotor displacement.
Sec. 12-2 Voltage Relations 481

Note that the sin wt factor remains the same as in Eq. (12-5), which emphasizes
that this induced emf is of the same frequency and of the same phase as that
occurring in winding 2. Only the amplitudes differ because of the difference in
coupling.
Finally the emf induced in winding 3 can be expressed as

E;, = E cos (a + 120°) (12-8)

The presence of 120° in the argument accounts for the fact that winding 3 is
located behind the axis of winding 2 by this amount. The expression for the
instantaneous voltage in winding 3 is accordingly

€3,(t) = V2 E sin wt cos (a + 120°) (12-9)

Equations (12-4) to (12-9) describe the stator phase voltages. Information
about the associated line voltages is readily obtained from the phase voltages.
Thus the induced voltage appearing between terminals 1 and 2 in Fig. 12-4 can be
written as

Ep Eo Ein i Eo = Ein = En (12-10)

Inserting Eqs. (12-4) and (12-6) into Eq. (12-10) yields

E, = Ecos a cos 120° + E sin a sin 120° — E cos a

= (- 3cos a + Me sin a)E (12-11)

Or combining the sine and cosine functions into an equivalent cosine function
leads to
Ey» = V3 E cos (a — 150°) (12-12)

The expression for the line voltage between terminals 2 and 3 can be simi-
larly determined. Thus,

Ey; = Eo, + Ens = Ena = Es (12-13)

Substituting Eqs. (12-4) and (12-8) leads to

EE = (5cos a + a sin a)E (12-14)

or

E,; = V3 E cos (a — 30°) (12-15)

The line voltage between terminals 1 and 3 is

Ey = Es, + Ea = Ean — Esc (12-16)

 Synchros Chap. 12
482

Upon inserting Eqs. (12-6) and (12-8) there results

E;, = Ecos (a + 120°) — E cos (a — 120°)
— V3 E sina (12-17)
or

E;, = V3 E cos (a + 90°) (12-18)

Equation (12-17) is put in the cosine form to facilitate comparison with Eqs.
(12-12) and (12-15) to show that these line voltages are such that the arithmetic
sum of two of them must be equal to the third.
As a result of the variations in the line voltages that can be effected by
displacement of the rotor, it is possible to feed these voltages to the stator of a
control transformer, a receiver, or even a differential synchro and thereby pro-
duce useful effects. Such applications are discussed in the next section.

12-3 APPLICATIONS

The use of synchros in servomechanisms represents one of the biggest areas of

applications. Shown in Fig. 12-5 is the arrangement most often found. This con-
figuration constitutes that part of the servomechanisms called the error detector.
Its chief function is to convert a command in the form of a mechanical displace-
ment of the cx rotor to an electrical signal appearing as a voltage at the rotor
winding terminals of the control transformer. It offers the distinct advantage of
furnishing information about shaft displacement by means of wires, which allows
the control transformer to be located remotely from the control transmitter.

e=E, sina

Figure 12-5 Synchro arrangement for converting a CX displacement to a corre-

sponding voltage signal at the rotor winding of the CT.
Sec. 12-3 Applications 483

What is the mechanism by which this is brought about? Keep in mind that
the rotor of the control transmitter is excited with an ac voltage so that an alternat-
ing field is created whose axis coincides with the cx rotor axis. The time-varying
nature of the air-gap flux and its linkage with the stator coils induce in them
corresponding voltages. Thus, if a equals 30° measured in the counterclockwise
direction, it follows that the voltages induced in the stator phases are as specified
in Fig. 12-5. When the cx stator terminals are connected to the cT stator terminals
in the fashion shown in Fig. 12-5, the transmitter is made to supply magnetizing
currents in the cr stator windings, which in turn create an alternating flux field in
its own air gap. The values of the cr stator phase currents must be such that the
resulting air-gap flux induces voltages that are equal and opposite to those prevail-
ing in the stator of the transmitter. This denotes the equilibrium condition satisfy-
ing Kirchhoff’s voltage law in the circuit between the stator windings of the two
synchros. Accordingly, the direction of the resultant flux produced by the cx
stator phase currents is forced to take a position of exact correspondence to that
of the rotor axis of the control transmitter.
If the ct rotor is assumed to be held fast at its electrical zero position as
depicted in Fig. 12-5, then the rms voltage induced in the rotor can be expressed
as
e=E,sina (12-19)
when
where E, denotes the maximum rms voltage induced by the ct air-gap flux
a is the displace ment angle of
the coupling with the rotor winding is greatest and
l function appears because the
the rotor of the control transmitter. The sinusoida
along
distributed stator winding produces a sinusoidally distributed flux density
indicate those sides of the stator
the air gap. The polarity markings are used to
established
coils that have the same instantaneous polarity. For equilibrium to be
that both phase and magnitude
in the stator circuits it is therefore necessary
conditions be satisfied. :
an
It is important to note that the value of the cr rotor voltage is in essence
rotor position s of the transmit -
indication of a lack of correspondence between the
Thus, if the cr rotor were
ter and the transformer from their electrical zeros.
= 30°, then clearly
manually turned counterclockwise in Fig. 12-5 by an amount a
corresp ondence is pro-
the rotor emf would become zero. In this manner exact
ve electrica l zero positions
vided with the transmitter measured from the respecti
of each synchro.
(e) induced
A general expression for the instantaneous value of the voltage
that of the rotor of the CT
in the ct rotor when the displacement of the cx is ay and
iS aT is

e = E, sin wt sin (ay — ar) (12-19a)

o pair of Fig.
Figure 12-6 is the schematic diagram showing how the synchr
for the load to be moved a
12-5 is used in a servomechanism. If it is desirable here
form of a displacement
specified amount in response to a command applied in the
484 Synchros Chap. 12

Ne

SI! 9S3
Gearing

ste] fone Nm
N

“aJu» fe

Figure 12-6 Servomechanism illustrating the use of synchros as the error detec-
tor. Dashed lines denote mechanical connections. ;

of the cx rotor, an error voltage e should initially exist because of the inertia on
the output shaft associated with the servomotor and the load. As time elapses, the
servomotor will begin to respond to this error voltage and thus displace the load.
As the output displacement increases, the mechanical connection between the
servomotor shaft and that of the cT rotor will cause the cT rotor to advance in the
direction to reduce the error voltage e. The servomotor will cease to turn when a
position of exact correspondence occurs—at which time e is zero. It is worth-
while to note that with such a control system a large load (e.g., a 100-ft radar dish
antenna) can be controlled with very little effort applied at the rotor of the cx
synchro.
A second application of synchros is for torque transmission over an appre-
ciable distance without the use of a rigid mechanical connection. The connection
diagram is illustrated in Fig. 12-7. The arrangement requires the use of a transmit-
ter and a receiver unit. The manner in which the system works is quite straight-
forward. When the rotor of the transmitter is displaced, say, by a = 30°, the
stator behaves like the secondary of a transformer and delivers current to the
three-phase stator winding of the receiver; this produces an ampere-conductor
distribution whose axis is fixed by a. By exciting the rotor of the receiver from
the same ac source as the rotor of the transmitter, a flux field is created which
interacts with the stator ampere-conductor distribution to produce an electromag-
netic torque. The action of this torque brings the rotor axis of the receiver, which
Sec. 12-3 Applications 485

Figure 12-7 Use of synchros for remote position indication.

is free to turn, to a position of correspondence with the rotor of the transmitter.

Hence the cr rotor revolves by a degrees in the counterclockwise direction. Note
that at this position the induced voltages in the stator windings of the receiver
ideally bear the same phase and magnitude as those prevailing in the stator coils of
the transmitter.
The use of a differential synchro to add or subtract an angle in the torque
transmission system is illustrated in Fig. 12-8. In part (a) of this figure the voltage
distribution in the stator phase windings of each synchro is shown corresponding
to a cx rotor displacement of ay = 30° in the positive (or counterclockwise)
direction. The differential synchro is assumed to be kept at the electrical zero
position in this instance and so does not alter the transmission characteristic from
that represented by Fig. 12-7. The rotor of the receiver, of course, responds by
turning counterclockwise by an amount equal to an ax of 30°. It is helpful to note
that a counterclockwise displacement of the cx rotor yields a counterclockwise
displacement of the cr rotor.
Depicted in Fig. 12-8(b) is the effect of introducing a displacement of ap
in the differential synchro superimposed upon the situation represented in Fig.
12-8(a). For the convenience of illustration ap is taken equal to —60°. The minus
sign means that the rotor of the differential synchro is turned clockwise. Since the
axis of the flux field in the differential synchro is determined by ay, the displace-
ment of the cp rotor by —60° means that the induced emf in rotor winding 2 must
be zero, while in the remaining windings the rms value of the induced emf is (V3/
2)E having the polarities indicated. Immediately after the displacement of the cb
rotor, a voltage unbalance exists between the rotor of the differential and the
stator of the receiver. This unbalance causes currents to flow so that the resulting
ampere-conductor distribution in the cR stator produces an electromagnetic
torque that brings the cr rotor to the position that ensures voltage balance. An
examination of Fig. 12-8(b) discloses that the resulting equilibrium position of the
cr rotor is a position that is 90° displaced counterclockwise from the electrical
zero. Note that a clockwise displacement of the cp rotor yields a counterclock-
 *wiasAS UOIS
FIP [0-0 uod 8 jo Is 8-zT amsiy
-sTusuet} anbio} & 0} [sue Ue SuIppe ul OyOUAS [eTUoJaJ
(9)
|OIyUE1a4J!P |O4jUOD

40}0}S

486
 Sec. 12-4 The Advantage of Sinusoidally Distributed Windings 487

wise displacement of the cr rotor. A mathematical description of the transmitter-

differential-receiver configuration of Fig. 12-8 is given by
Orn — Gy > ap (12-20)

where ap is the net displacement of the cr synchro corresponding to an ay dis-

placement of the transmitter and an ap displacement of the differential synchros.
Positive values of the angles expressed in degrees are used for counterclockwise
displacements and negative values for clockwise displacements.

12-4 THE ADVANTAGE OF SINUSOIDALLY DISTRIBUTED

WINDINGS

Rather than consider separately the various schemes that are used for synchro
windings, attention is directed in this section to show that no harmonic induced
emf’s can occur in a sinusoidally distributed winding even though such harmonics
do exist in the field distribution created by the field winding. Synchro residual
voltages originating from this source can accordingly be kept to a minimum by
employing winding arrangements that approximate the sinusoidal distribution.
The concentric type of winding is such a scheme and so is frequently employed in
synchro design.

Ne= Np sin o¢

Flux
density

Figure 12-9 (a) Sinusoidally distrib-

uted armature winding; (b) air-gap field
distribution, showing the fundamental
(b) cand the third harmonic components.
 Synchros Chap. 12
488

? of a
Depicted in Fig. 12-9 is the sinusoidally distributed armature winding
synchro. Let
N = turns/pole (12-21)

Na = N, sin a = sinusoidal distribution of turns (12-22)

These quantities are related by

N= |"Nada (12-23)
Performing the indicated integration then yields
N = N,[-cos a]g = 2N, (12-24)

Hence the distribution of turns may be expressed in turns per pole as

Nz = . sin a (12-25)

In general the field distribution produced by the cT synchro field winding

contains a large fundamental component plus odd harmonics. No even harmonics
exist because the field patterns beneath the north and south poles of the synchros
are mirror images of one another. The expression for the nth harmonic flux
density is given by
b,(x) = B, sin (na — y) (12-26)
where B,, is the amplitude of the nth harmonic field component. In the case of the
transmitter synchro, B, is a sinusoidal function of the line frequency w applied to
the rotor winding. Also, a denotes a space angle measured in the manner illus-
trated in Fig. 12-9. Moreover, the quantity y is proportional to the cx rotor speed
w,. A comparison of Eq. (12-26) with Eq. (4-8) discloses that both are expressions
for traveling waves.
Because the induced emf in the synchro armature winding is determined by
the time rate of change of the flux linkage, it is worthwhile to find the expression
for the flux linkage for both the fundamental field component and the harmonic
field components. But first the equation for the flux is needed. For the nth har-
monic this is

bn = |” by(a)Ir da (12-27)
where r denotes the rotor radius and / is the axial length. For simplicity a unit
axial length is considered. Equation (12-27) may then be written as

2B,r
ee ie B, sin (na — y)r da = cos (na — y) (12-28)

+Armature winding is the stator for the cx synchro and the rotor for the cr synchro.
 Sec. 12-5 Stator MMF of the Control Transformer in the CX-CT Mode 489

The associated expression for the flux linkage then becomes

dn = |"Nu dar by(a) = NB, r

[°sin a cos (na - y)da (12-29)
n

Inserting the trigonometric identity

sin a cos (na — y) = ¢sin[(n + a — y] — sina — la — y] (12-30)

leads to

‘4 ie NB, r
[7{sin a + Da = y] ~ sin [m - Da See (12331)
2n
ing
The flux linkage of the fundamental field component is found upon substitut
ng the indicated integratio n. This leads to
= | into Eq. (12-31) and then performi
ra: B,Nar
Ay 5 sin y (12-32)

rotor and that

Keep in mind that y varies with the angular position of the cx
variatio n of B,, the ampli-
superimposed upon this is a sinusoidal line-frequency
worthwh ile to note that the
tude of the fundamental field component. It is also
ly equal to zero.
fundamental flux linkage by Eq. (12-32) is not identical
is obtained
A more useful expression for the flux linkage of the nth harmonic
d limits. Thus
by integrating Eq. (12-31) and then inserting the specifie

yt2n ({n+ee e a — yl 3 - cos [(n — Da — yl})

1
(12-33)
or
=
NBon Z|Frcery 1 \- ( )
ed COR Pi a \COSy cos y)
Die
for a sinusoidally distrib-
This last result is a significant one. It states that,
of all the field harmonics are zero
uted armature winding, the induced emf’s
despite their presence in the field distribution.t

SFORMER
12-5 STATOR MMF OF THE CONTROL TRAN
IN THE CX-CT MODE
g currents that flow in the trans-
In Sec. 12-3 it is pointed out that the magnetizin
that the axis of the resultant air-gap
former of the cx-cT configuration are such
This conclusion was reached on the
flux follows the position of the Cx rotor axis.
of the
state that all the pitch and distribution factors
+Another way of describing this result is to
harmonics are zero.
 Synchros Chap. 12
490

basis of satisfying Kirchhoff’s voltage law in the circuit involving the three stator
phases of each synchro. Now attention is directed toa verification of this result in
appropriate mathematical formulations. For the sake of generality the analysis is
carried out for the nth harmonic. This procedure offers the advantage of making
the general result available in the next section to determine the static errors
caused by the cT space harmonics.
To focus attention solely on the stator mmf harmonics, several simplifying
assumptions are made concerning the control transformer. These are: no satura-
tion of the iron, uniform air gap between rotor and stator, and balanced phase
windings. In addition, the control transmitter is assumed to be perfect, which
means that it is equipped with a sinusoidally distributed winding as well as bal-
anced stator windings and a uniform air gap. In Fig. 12-10 is a typical stator mmf
wave for the actual winding distribution employed in synchros. Specifically only
the mmf curve of winding 2 is illustrated here. A Fourier-series analysis applied
to this stepwise function makes it clear that mmf space harmonics do exist and
must be of odd order because of half-wave symmetry. Also depicted in Fig. 12-10
is the location of the rotor axis of the transmitter measured from a zero reference
line. Note that displacements measured in the cT air gap are denoted by I’, while
the cx displacements are represented by a.t

mmf distribution in Location

CT stator winding 2 of CX rotor
for one pole

- ra}pe O° line

(a)

Figure 12-10 (a) Magnetomotive force distribution in phase 2 of the control transformer,
showing one pole; (b) transmitter rotor position that produces the magnetizing current
leading to the mmf distribution of (a).

The expression for the mmf per pole for winding 2 in the stator of the control
transformer for the nth harmonic is
fra = F, sin wt cos nT (12-35)
where F,, is the amplitude of the mmf harmonic contained within the stepped wave
shape. The sin wt factor represents the alternating nature of the mmf wave; the

+This @ is not the same as that used in Sec. 12-4.

Sec. 12-5 Stator MMF of the Control Transformer in the CX-CT Mode 491

angle [ permits finding the value of the mmf at any fixed point in space for a
specified time instant. The harmonic amplitude F,, can be expressed in terms of
the winding constants and the effective value of the stator phase current. In
balanced polyphase machines used in conjunction with three-phase voltages this
éffective current is a fixed quantity. However, in the case of the control trans-
former this current is a variable quantity dependent upon the value of a. If J
denotes the maximum rms current in winding 2 corresponding to the maximum
applied voltage and if N, denotes an equivalent number of effective turns per pole,
then Eq. (12-35) can be more specifically written as

fnr = (N-1)cos @ sin wt cos nT (12-36)

Keep in mind that I and @ are independent quantities. For given values of a and
wt we may choose different values of I’.
By introducing the fact that windings | and 3 are located, respectively,
ahead and behind winding 2 by 120°, the expressions for the mmf of windings 1
and 3 are correspondingly

fay = (Nel) sin wt cos (a + 120°) cos n(T + 120°) (12-37)

and
fa3 = (N21) sin wt cos (a — 120°) cos n(T — 120°) (12-38)

To find the resultant mmf in the stator of the control transformer, it is merely
necessary to add the last three equations. Thus
far = (Ne1) sin wt [cos a cos nT’ + cos (a — 120°) cos (nT — 120°n)
+ cos (a + 120°) cos (nT + 120°n)] (12-39)

Inserting the trigonometric identity

cos x cos y = $.cos (x — y) + cos (x + y) (12-40)

into Eq. (12-39) yields

far = (Nel) sin wt cos (a — nT’) + 4 cos (a+ nl)

+ 4cos [(a — nT) — (n — 1)120°] + 3 cos [(a + nl) + (n + 1)120°]
+ 4cos [(a — nT) + (n + 1)120°] + 3 cos [(@ + nI) — (n + 1)120°]}
(12-41)
After some manipulation the last equation reduces to

Snr = “ sin wt[A cos (a — nT) + Bcos (a + nl)] (12-42)

where
A= + 2 cos (a 1120" (12-43)

B = 1 +2 cos (n + 1)120° (12-44)

492 Synchros Chap. 12

Equation (12-42) can be further simplified by determining the values of A and

B for various values of n. In this connection recall that n is restricted to odd
values. Accordingly,
for n=l, A=
B=

n = 3, A=0
B=0
n=5, A=0

B=3
or more succinctly
A=3 forn =3k+1 (12-45)
where
k = 0, 2, 4,6,...,
and
B=3 for n = 3k —- 1 (12-46)
where
k = 2, 4,6,....

Therefore, the expression for the resultant mmf may be written as

Snr = 3(N-L) sin wt cos (a — nT) forn = 3k+1 (12-47a)
= 3(N_J) sin wt cos (a + nD) forn = 3k —1 (12-47b)
where the k’s are defined as in Eqs. (12-45) and (12-46). Examination of Eq.
(12-47) discloses that a single expression can be used provided that k is allowed to
take on negative as well as positive values. Accordingly, we may write

far = 3(Ne1) sin wt cos (a — nT) (12-48a)

where

n=3k+1 (12-48b)
and

= 10 ch Diese AE GY aM (12-48c)

When a negative number is used for k, it yields a negative value for n. This sign
must be carried along with the magnitude of n. When it is, Eq. (12-47b) immedi-
ately results.
The use of the negative sign offers still another advantage. If for the moment
the cx rotor is assumed to be driven at a constant velocity w,, it follows that a =
w,t. Insertion of this result into Eq.,(12-48a) yields the equation of a traveling
 Sec. 12-6 Displacement Errors Caused by CT Space Harmonics 493

wave having a fixed amplitude of $(N,J). The use of k = —2 in Eq. (12-48a) leads
to

fsr = 3(NeL) sin wt cos (w,t + ST) (12-49)

Examination of the last expression discloses that it is the equation of a wave
traveling in the negative I direction. Therefore the minus sign that results from
Eq. (12-48b) corresponding to k = —2 can be used to denote those harmonic
components of the mmf wave that travel in the negative I direction.
How can Eq. (12-48) be used to show that the fundamental component of the
resultant field in the air gap of the control transformer follows exactly the cx rotor
displacements? First let us write the expression for this fundamental term. Thus,
fir = #(N-D) sin wt cos (a — T) (12-50)

Now assume that an observer is placed at [ = 0 degrees in the air gap of the
control transformer and stays fixed there. Then with the cx rotor at its zero
position (i.e., a = 0°), Eq. (12-50) states that the observer at T = 0 degrees sees the
amplitude of the field mmf. Of course, this amplitude will be at its maximum value
when the cT stator phase current is a maximum. Next assume that the cx rotor iS
displaced by a = 30°. In accordance with Eq. (12-50) the observer at T = 0
degrees now sees a value of the mmf which is given by
fri = #(N-J) sin wt cos a = 3(N-J) sin wt cos 30° (12-51)

A comparison of Eq. (12-51) with Eq. (12-50) reveals a reduction by the factor cos
a. Hence one can conclude that the axis of the mmf in the air gap of the control
transformer (i.e., the location of its amplitude) has revolved in the positive direc-
tion by the amount a = 30°. Therefore the stator field of the control transformer
follows exactly the position of the transmitter rotor axis.
Perhaps a more convincing demonstration of this conclusion can be had by
placing the observer at a point T = 30°. Then when a is 0°, the largest mmf seen
by the observer is \/3/2 times the peak value. However, when a is increased to
the
30°, Eq. (12-48) indicates that the observer now sees the peak mmf. Hence
axis of the air-gap flux has moved in the positivet I direction by the amount a.

12-6 DISPLACEMENT ERRORS CAUSED BY CT

SPACE HARMONICS
ble for
The presence of harmonics in the stator mmf of the ct synchro is responsi
in the Cx-CT configur ation. The variatio n of
the existence of a displacement error
rmonic function of the cx rotor displace -
this error can be shown to be a sixth-ha
is first given to a physical demonst ration
ment angle. In this section consideration

would see only one-

+If the wave would have moved in the negative I direction, the observer
half the peak mmf.
 Synchros Chap. 12
494

A mathe-
of why the presence of harmonics leads to such a displacement error.
matical description then follows.
from
The physical explanation requires the use of a suitable zero reference
ments are measured . For convenie nce this
which the cx and ct rotor displace
defined electrica l zero positions . At this point
reference is taken as the previously
correspon-
it is assumed that the rotors of the cx and cT synchros are in exact
illustrat ed through a graphical interpret ation
dence. This situation can be readily
of Eq. (12-48). In the interests of simplifyi ng the represen tation, only the funda-
in the
mental component and the fifth and seventh harmonics are assumed present
ct stator mmf wave. At the electrica l zero position, a is zero. Similarly , to place
the cT rotor at electrical zero it is assumed that I is also zero. Equation (12-48)
then yields the following results:

n= 1,

3 F 3 :
fir = ) (NJ) sin wt cos (a — T) = 5 (N-,1) sin wt cos 0°

3 5
2% (N,1) sin wt = Fy (12-52)

n= —S,

N. ;
fo = ;(~=) sin wt cos (0° + 5 X 0°)

= 53 (AS)
(Ne
sin; ot = Fs (12-53)
Ti -Fil.,

3 (N. 7
fr = 7)(A+) sin wt cos (0° — 7 X 0°)

Ys rey : F; (12-54)
= > 7 sin wt

where F,, Fs;, and F; are defined as indicated. A glance at these expressions
shows that all three components are in space phase as well as in time phase. The
space-phase condition is graphically illustrated in Fig. 12-1 1(a). Note that the line
of action of the resultant mmf Fe coincides with that of the fundamental F;
because the harmonic terms are in phase at this point too. The coupling with the
CT rotor winding is therefore zero so that no emf occurs. The control transformer
is accordingly at the null? position.
Next consider the case where the cx rotor is displaced by 20° in the positive
counterclockwise direction. Then imagine that the rotor of the cT synchro is

+The null position is the position of minimum voltage in the cT rotor winding, which in this case
is Zero.
Sec. 12-6 Displacement Errors Caused by CT Space Harmonics 495

Fra AFR

Figure 12-11 Graphical representation

of the displacement error caused in a
CX-CT synchro configuration by the
CT mmf harmonics: (a) condition at
electrical zero; (b) location of the
resultant mmf when a = 20° and the
(a) CT rotor is displaced by T = a = 20°.

displaced by a like amount in the same direction, i.e., © = 20°. If the resultant
field mmf continues to be in space phase with the fundamental component, then a
null voltage occurs at [ = 20° and so no displacement error exists. On the other
hand, if the resultant mmf, which produces the net field flux, does not lie along the
20° line, then some coupling with the cT rotor winding exists and so causes a small
voltage to appear between its terminals. To reduce this voltage to zero it will then
be necessary to displace the cT rotor by an additional amount, «,. The Cx-CT
20°.
synchro configuration is then said to have a displacement error of ¢;, at a =
By resorting once again to Eq. (12-48) we can ascertain which condition prevails.
Accordingly, we get
n=l],
3 : 0 °
Tink 5 (Nel) sin wt cos (a — T) = F; cos (0° — 0°)

= F, cos 0° (12-55)

n =’5;

fy WS (Ae) sin wt cos (« + ST) = Fs cos [20° + 5(20°)]

3 N. - re} te)

(12-56)
= F; cos 120°

n= 7,

fr = 5(A+) sin wt cos (a — 71) = F7 cos [20° — 7(20°)]

= F, cos (—120°) (12-57)

 Synchros Chap. 12
496

. The fifth and

A graphical representation of these results is shown in Fig. 12-11(b)
component as
seventh harmonics are now no longer in phase with the fundamental
in the foregoing
can be seen by the difference in the argument of the cosine terms
the amount
equations. Keep in mind that the angle T° in these equations denotes
or not a
by which the rotor of the cr synchro is displaced. To determine whether
ent the argu-
null results, I is set equal to a. Hence for the fundamental compon
t of the cosine
ment of the cosine term becomes identically zero. The argumen
term can thus be conside red to be an angle that is measure d relative to the position
means that there
of the cx rotor axis. The zero-degree value for the fundamental
the cx
is exact correspondence between the axis of the fundamental cT field and
means
rotor axis. In the case of the fifth harmonic the argument is 120°, which
of the
that the axis of the fifth-harmonic mmf lies 120° displaced from the location
fundamental component. In the diagram, Fs is depicted behind F; by 120° because
the fifth harmonic is known to rotate in a direction opposite to that of the funda-
al
mental component. However, note that Fs is displaced 100° from the electric
For
zero reference line, which is consistent with the value nT’ = —5(20°) = —100°.
ntal, F7
the seventh harmonic, which travels in the same direction? as the fundame
is shown ahead of F, by 120°. Note again that this is consistent with the fact that
F; is displaced by n(20°) = 7(20°) = 140° in the positive direction from the zero
reference line. It is also worthwhile to note that the positions of F's and F) are
sixth-harmonic functions of the cx rotor displacement angle a.
In general the sum of the fundamental mmf component and the harmonics
yields a resultant mmf that is displaced from the fundamental by the angle e,.
Accordingly, if the cx-CT arrangement is used as an error detector in a servomech-
anism and a command is applied at the transmitter in the form of a = 20°, the
servomechanism achieves a position of equilibrium (i.e., a null position) when the
cT synchro is displaced through the mechanical feedback connection by an
amount equal to a + «,. Therefore, the output follows the command but with an
error of e,.

Mathematical Analysis

An equation for the displacement error ¢, can be found by first determining the
expression for the null voltage. This can be conveniently done by finding the
voltage induced in the transformer rotor winding by each harmonic component of
the revolving mmf. Keeping in mind that at the electrical zero position the axis of
the cr rotor winding is displaced by 90 electrical degrees from the assumed zero
position of the stator mmf axis (i.e., the ' = 0 degree point), the voltage induced in
the rotor winding by the fundamental field component is

E, = K, cos E 2 (r = )| (12-58)

+The minus sign in the argument of the cosine term of Eq. (12-48) represents positive rotation.
Sec. 12-6 Displacement Errors Caused by CT Space Harmonics 497

where

K, = 4.44f N,K,,® (12-59)

Note that K, actually denotes the rms voltage induced in the rotor winding by the
fundamental field component corresponding to maximum coupling. Accordingly,
Eq. (12-58) describes the rms value of the voltage induced in the rotor winding for
a given Cx rotor displacement a and a cT rotor displacement [. The cosine term
of this equation is simply a factor that indicates the degree of coupling that exists
between the fundamental field flux ©, produced by the fundamental field mmf and
the cT rotor winding. Equation (12-58) may be rewritten as
E, = —K, sin (a — T) (12-60)
This result can now be used to find the voltage induced in the rotor winding when
the cT rotor is placed at the null position. It is important to keep in mind that in
general the coupling of the fundamental flux component with the rotor winding
will not be zero at the null point. This is graphically illustrated in Fig. 12-11,
where the rotor winding is at null when its axis is in quadrature with the resultant
mmf. However, note that there is some small amount of coupling of the winding
with the fundamental field component even though the coupling with the resultant
field is zero. Since the null position is found by displacing the cT rotor by l = a +
e, rather than by [ = a, it follows that the rms value of the voltage induced by the
fundamental field at null is
E, wi = —K; sin (a — T) = —K, sin (a — a — &) = Kj, Sin &p (12-61)

Because such displacement errors are often very small, the last expression may be
written as
Ey mut = Kien (12-62)

A similar procedure can be followed to find the emf induced in the CT rotor
by the fifth harmonic of the field mmf. Thus the general expression for the rms
value of this voltage is

Es = Ks cos a +5 (r ~ ) | (12-63)

where
Ks = 4.44f N-Kwrs®s (12-64)

Here K,,,5 is the rotor winding factor for the fifth harmonic and ®, denotes the
fifth-harmonic flux per pole. In Eq. (12-63) the 7/2 quantity is treated in the same
way as I because it is a space angle associated with the cT rotor. By means of an
appropriate trigonometric identity Eq. (12-63) can be rewritten as
Es; = Ks sin (a + SI) (12-65)

When the ct rotor is assumed to be placed at its null position, I takes on the value
 Synchros Chap. 12
498

a + e,, which upon insertion into Eq. (12-65) yields

+ Se,) (12-66)
Es au) = Ks sin (a + 5D) = Ks sin (a + 5a + Se,) = Ks sin (6a
induced by the
An important observation to make at this point is that the voltage
ntal
fifth-harmonic field mmf is of the same frequency as that of the fundame
r, it is in time phase with the fundame ntal. Therefo re the
component. Moreove
to
fundamental and the harmonic voltage contributions can be added directly
obtain the resultant voltage.
The expression for the rms value of the induced voltage for the seventh
harmonic is found to be
E, = K7 sin (a — 71) (12-67)
equa-
where K; is defined in a manner similar to that for Ks. The corresponding
tion for the null voltage associate d with this harmonic is then

E, null — —Kj7 sin (6a ar Ten) (12-68)

a
It is interesting to note that the argument in Eqs. (12-66) and (12-68) involves
sixth-harmonic function of a. It is reasonabl e to expect therefore that extension
of the foregoing procedure to the eleventh and thirteenth mmf harmonics leads to
expressions that involve twelfth-harmonic functions of a. Specifically it can be
shown that
E\, null — Ky; sin (12a TU lle,) (12-69)

Ey null — Ky sin (12a ae 13e,) (12-70)

Of course, as higher order harmonics are considered, their importance diminishes

because the magnitudes of the K coefficients vary inversely with the order of the
harmonic. Hence the harmonics that bear the greatest influence in causing dis-
placement errors are the fifth and seventh.
The complete expression for the null voltage that occurs when the CT rotor is
placed at the position T = a + e, may be written as
Euw= Ej ant Eson + Fo mat (12-71)

An algebraic sum is permitted because these voltages are of the same frequency
and in time phase. Inserting Eqs. (12-62), (12-66), (12-68), (12-69), and (12-70)
yields
Eau = eK SUF Ks sin (6a as Sén) oa K, sin (6a + Ten)

+ Ky, sin (12a + lle,) — Ki; sin (12a + 13e,) ++ > - (12-72)

This expression can be simplified by noting that the quantity ne, appearing in the
arguments of the sine terms can be neglected with little or no error. Accordingly,
we have

Enun = €nK, + (Ks — Kz) sin 6a + (Ki — Ki3) sin 2a ++: - (12-73)
 Sec. 12-7 Static Errors and Residual Voltages 499

However, with the cT rotor placed at the null position the resultant emf induced is
zero. Hence
Eau =0= e,K, =f (Ks — K3) sin 6a + (Ky; oe K,3) sin 2a, °° > (12-74)

from which it follows that the expression for the displacement error is

Ky .
sin 6a + On
Ki
K,ee ta De e (12-75)

An examination of Eq. (12-75) discloses that the magnitude of the displace-

ment error caused by the space mmf harmonics in a cx-cT synchro configuration
can be controlled by properly manipulating the K coefficients. Thus, if Ks; and K7
are made equal, the sixth harmonic displacement error is entirely eliminated. This
result can be achieved in some synchros through the judicious selection of the
winding factors for these harmonics. When cancellations cannot be accom-
plished, the winding design is chosen so that the winding factors for the harmonics
are kept as small as possible.

12-7 STATIC ERRORS AND RESIDUAL VOLTAGES

The accuracy of a servomechanism can be no better than the synchro pair that is
used as the error detector. Hence the static displacement errors must be kept as
small as manufacturing tolerances will allow, consistent of course with economi-
cal production. A typical static error curve for a Cx-cT synchro error detector is
depicted in Fig. 12-12. This curve is obtained by displacing the rotor of the trans-
mitter by an amount a and then displacing the cT rotor to that position which

Error,
min. of arc

-20
Note
Figure 12-12 Typical displacement error curve for a CX-CT synchro pair.
the predominance of a six-cycle and a two-cycle error variation.
 Synchros Chap. 12
500

The difference be-

yields a minimum voltage across its rotor winding terminals.
I and the cx rotor displac ement a constitutes the
tween the cT rotor displacement
signed synchro s this error rarely exceeds
static displacement error. In well-de
the predom inance of a six-cyc le error. As
4 degree. A study of Fig. 12-12 reveals
ng section, this is due to the effect of the mmf
already described in the precedi
transfo rmer. Note that elimina tion of this
harmonics in the stator of the control
towards increas ing the accurac y of the Cx-CT
source of error goes a long way
synchro pair.
the static
Another dominant component that can readily be distinguished in
cy error of the cx rotor displac ement a. The
error curve is a double frequen
lies in the slight unbalan ce that necessar ily
source of this second harmonic term
of the synchros . It can be shown that, if the
occurs in the stator phase windings
of the phase impedan ces are not identical , then
resistive and reactive components
of a. The
a displacement error exists which varies as a second harmonic function
is very often attribut able to manufac turing
nonidentical nature of the windings
one or more of the followin g factors: slight devia-
difficulties. These often involve
tions in the turns used for each phase, varying tensions on the wire spools during
the winding process, slight deviations in the cross-sectional area of the wires
the
obtained from different spools or different manufacturers, slight differences in
leakage reactance of each phase winding, and finally the effect of tooth iron
saturation. The relationship between the static error and the unbalanced imped-
ances discloses that unbalances in the reactive parts of the impedances exert a
greater influence on the static error than unbalance in the resistive parts.
Another source of a double frequency error in the error curve is the absence
of a perfectly uniform air gap. This can happen for several reasons: ellipsing of
the rotor, ellipsing of the stator, an eccentrically placed rotor, or a skewed rotor
axis. Usually production methods are such that there is no difficulty in getting a
perfectly round rotor. However, the same is not true for the stator structure;
there is a greater tendency for the stator to come off the production line slightly
ellipsed. Actually the ellipsing need not be very large to cause a significant static
error. For example, a difference between the major and minor axes of the stator
of only 0.0001 in. can cause a static error spread of as much as 3 to 4 minutes of
arc.
Besides static displacement errors, the cx-cT synchro pair has the additional
shortcoming of residual voltages at null. Whenever residual voltages do in fact
exist, they cannot contain a component of fundamental frequency that is in phase
with the applied line voltage. After all, the null position is achieved whenever the
in-phase fundamental frequency component is zero. Hence, any voltage that ex-
ists at null must be due to other sources. Figure 12-13 depicts a typical curve
showing how the residual voltage varies with a (as the transformer is placed at the
null position for each value of a).
Analysis shows that the residual voltage contains two components: time
harmonics of the fundamental frequency and quadrature voltages. The time har-
monics in turn arise from two sources, i.e., the saturation of the iron used in the
Chap. 12 Problems 501

Residual
voltage,
mv rms

200

100

2 3a en oe, rad.

Figure 12-13 Typical variation of a residual voltage content in a CX-CT configu-

ration at null.

magnetic circuit of the synchros as well as the nonuniform field distribution in the
air gap of the control transmitter. The quadrature voltage is defined as the mini-
mum voltage of fundamental frequency existing at null. Its presence is caused
either by unbalanced stator phase windings or a nonuniform air gap or by a
combination of both. In each instance, however, it can be shown that the quadra-
ture voltage is a double-frequency function of a. A glance at Fig. 12-13 shows the
prevalence of the second harmonic term. Moreover, analysis also shows that
resistance unbalance in the synchro phase windings exerts a greater influence than
reactance unbalance does. It is for this reason that switches in the stator line
connections of the cx-cT synchro pair should be avoided if at all possible.
Good synchro design focuses on keeping the residual voltages as small as is
economically possible especially when these units are used as the error detectors
in servo systems. Unusually large residual voltages can easily cause saturation of
servoamplifiers which provide the control voltages that drive the output devices.
When saturation sets in, the system gain is sharply reduced, which adversely
affects the total system behavior. Moreover, although the residual voltages are
incapable of producing any useful output in such systems, they can nevertheless
cause currents to flow in the output devices that create unnecessary heat losses.

PROBLEMS

12-1. Explain why an umbrella-like rotor construction keeps to a minimum the magnetiz-
ing current drawn by the control transformer.
12-2. Prepare a rough sketch to show how the magnetizing current of a control trans-
former equipped with a salient-pole rotor varies with rotor displacement.
12-3. When the rotor of the synchro depicted in Fig. P12-3 is at the electrical zero
 Synchros Chap. 12
502

d in winding 2 is measured to be
position, the rms value of the phase voltage induce
g from neutral-to-line and from
52 V. Determine the voltage induced in each windin
60° from electri cal zero in a counterclock-
line-to-line when the rotor is at a position
e the polarit y markin gs for each voltage.
wise direction. Clearly indicat
2

n Stator

Figure P12-3

the following positions:

Repeat Prob. 12-3 for the instance where the rotor occupies
(a) 180° counterclockwise.
(b) 90° clockwise.
are connected in
A control transmitter, a control differential, and a control receiver
12-8(a). The control transmit ter is displace d 60° counter-
the manner shown in Fig.
and fastened. The control receiver is displace d 30° clockwis e and fas-
clockwise
different ial synchro move?
tened. In which direction and by what amount does the
for the new position of the synchro, and
In your solution, redraw the diagram
value of induced emf in each synchro winding. Assume that
clearly indicate the rms
the maximum rms voltage from line-to-neutral in each unit is E.
a differential
12-6 The output voltage of a control transmitter is fed to the stator of
synchro whose rotor output is then applied to the stator of a control transformer.
zero position. If the
Assume that, initially, each synchro is at its proper electrical
countercl ockwise and the rotor of the
rotor of the transmitter is displaced 90°
turned 60° countercl ockwise, determine the value of the
differential synchro is
voltage at the rotor terminals of the control transform er, expressed in terms
induced
of its maximum voltage E.
12-7. The output voltage of a control transmitter is fed to the stator of a differential
synchro whose output is then applied to the stator of a control transformer.
(a) Draw the schematic diagram of the system with each unit at the electrical zero
positions.
(b) If the rotor of the transmitter is displaced 30° counterclockwise and that of the
differential synchro is turned 60° clockwise, find the value of the induced volt-
age at the rotor terminals of the cr in terms of its maximum voltage E.
A synchro transmitter has its stator connected to the stator of a differential synchro
whose rotor, in turn, is connected to the stator of a synchro control transformer.
The rotor of the transmitter is properly energized and all three synchros are ad-
Chap. 12 Problems 503

justed to their zero positions. The transmitter rotor is then turned 15° clockwise,
and the differential rotor is turned 75° counterclockwise.
(a) Find the angle and direction of the flux axis in each synchro.
(b) Determine the voltage across the rotor terminals of the control transformer in
terms of the maximum possible value of this voltage E.
12-9. Repeat Prob. 12-8 for the case where the rotor of the cx is turned 45° clockwise and
that of the cp is displaced 90° counterclockwise.
12-10. With the aid of Eq. (12-48a) demonstrate that the resultant mmf associated with the
fifth harmonic travels in the negative I direction. Assume that an observer is
placed at [ = 30°.
12-11. Determine the maximum value of the sixth harmonic error in a Cx-CT synchro
transmission unit where the effect of the fifth space harmonic is 20% and that of the
seventh harmonic is 14% of the fundamental. Assume that all other errors are
negligible.
12-12. The mmf distribution in the air gap of the control transformer in a Cx-CT synchro
transmission configuration is known to have a fundamental component as well as
fifth and seventh harmonics. The fifth harmonic flux per pole is one-fifth that of the
fundamental, and the seventh harmonic flux per pole is one-seventh that of the
fundamental. Moreover, the rotor winding factor is one-fourth and one-fifth for the
fifth and seventh harmonics, respectively. Assume for convenience that the dis-
placement error due to the space harmonics is zero at the electrical zero position.
(a) For a cx rotor displacement of 30°, compute the value of displacement error
that occurs when the cT is turned to the null position.
(b) Repeat part (a) for a cx displacement of 240°.
(c) Prepare a rough diagram of the conditions prevailing in part (b).
. An ideal control transmitter is used with a control transformer in a typical Cx-CT
voltage indicating system. The air-gap flux of the cT contains a fifth and seventh
harmonic component in addition to the fundamental. The magnitude of the flux per
pole of the harmonics is inversely proportional to the order of the harmonic. A
rotor winding is used for the cr unit, which causes a 5% reduction in effective
induced voltage. This winding also yields identical winding factors for the harmon-
ics. Compute the harmonic winding factor that corresponds to a sixth harmonic
displacement error which never exceeds 1°.
12-14. The predominant space harmonics in the control transformer of a Cx-CT arrange-
ment are known to be the Sth, 7th, 11th, and 13th. The winding factors for the
fundamental and the 11th and 13th harmonics are each 0.96, while for the 5th and
7th they are 0.25 and 0.2, respectively. Assume that the harmonic displacement
error is zero at the electrical zero of the cT.
(a) Compute the value of the harmonic displacement error for the following values
of the cx rotor displacement: 5°, 10°, 15°, 20°, 25°, and 30°.
(b) Sketch the general nature of the curve that passes through these points.
12-15. Refer to the cx-cT synchros of Prob. 12-11, and assume that the harmonic displace-
ment function passes through zero at the electrical zero position of the cr unit.
Also, assume that the maximum rms voltage induced by the fundamental flux is
E, V.
504 Synchros Chap. 12

(a) What is the value of the voltage induced in the rotor of the ct by the fundamen-
tal flux, when the cT is at the null position corresponding to a cx rotor displace-
ment of 10°?
(b) What is the value of the fifth harmonic induced voltage in part (a)?
(c) Is the quantity computed in part (b) of the same frequency as the fundamental?
Explain.
12-16. Repeat Prob. 12-15 for the conditions specified in Prob. 12-12.
 13
Dynamics of
Electric Machines

In this chapter we study the dynamic behavior of electromechanical devices. In

all the preceding chapters the analysis was concerned with steady-state perfor-
mance only. However, nowadays, a knowledge of the behavior of electrome-
chanical devices in the transient state is at least equally important because this
equipment is often found as part of a total system. For example, generators and
motors are frequently found in the same interconnected electric power grid sys-
tem. As loads change and shift, it is the dynamic characteristic of the individual
machines taken together that will determine whether the system can react in a
stable manner to such disturbances. Similarly, in many feedback control systems
such as the servomechanism, it is not uncommon to find dc motors, two-phase
servomotors, pilot generators, amplidynes, and other electromechanical devices
as important components of such systems. The overall behavior and stability of
the entire control system is often greatly influenced by the dynamic characteristics
of these devices. An understanding of the steady-state characteristics alone is not
sufficient to provide a useful study of total performance in these situations. After
all, if the system of interest fails to achieve steady-state operation because of poor
dynamics, an understanding of the steady-state performance is really only aca-
demic.
The transfer function and the state variable method are the vehicles by
which the dynamic characteristics of electromechanical devices are described in

505
 506 Dynamics of Electric Machines Chap. 13

this treatment. However, the development proceeds on the assumption that the
reader already has a background in Laplace transforms.

13-1 THE TRANSFER FUNCTION

As the name implies, the transfer function is a mathematical formulation that

relates the output variable of a device to the input variable. For linear devices the
transfer function is dependent solely upon the parameters of the device plus any
operations of time, such as differentiation and integration, which it may possess.
It is independent of the input quantity—both its magnitude and variation.
The general procedure for deriving the transfer function is always the same
irrespective of the type component or engineering device under study. It involves
the following three steps: (1) determine the governing equation for the device
expressed in terms of the output and input variables; (2) Laplace transform the
governing equation, assuming all initial conditions are zero; and (3) rearrange the
equation to formulate the ratio of the output to input variable.
An example is useful at this point to illustrate the procedure and to show
how the dynamic behavior is thereby described. Refer to Fig. 13-1. Assume that

Rs

\ !
' Ip Sle Figure 13-1 Field winding circuit of an
electromechanical energy converter.
Example of a first-order system.

the input quantity is the voltage applied to the field winding of a machine, V;, and
that the desired output quantity is the field winding current, Jy. The equation that
relates Jy to Vy in terms of the appropriate parameters of the circuit is by Kirch-
hoff’s voltage law
dl
V> = IpRp + Ly (13-1)
Equation (13-1) is the governing equation and specifically is a first-order differen-
tial equation because it involves one energy-storing element, the inductance Ly.
Laplace transforming Eq. (13-1) for zero initial field winding current leads to the
following transformed differential equation:
V/s) ms I/(s) Rr + LrslI/(s) (13-2)

where V;(s) denotes the Laplace transform of V;, [;(s) denotes the Laplace trans-
form of /y, and the Laplace operator s denotes the first derivative of time. Upon
formulating the ratio of output to input variable in the last equation, the desired
transfer function results. Thus
Sec. 13-1 The Transfer Function 507

Is) _ 1 ie) 1
V/(s) f Ry a sL¢ ae Ry 1 oF STF (13-3)

where
L
= R, (13-4)

and is the time constant of the field winding.

Inspection of the right side of Eq. (13-3) discloses that the transfer function
is solely dependent on the circuit parameters R;and L;and the derivative operator
s. Moreover, the dynamic behavior of this circuit is completely determined by the
form of the denominator of the transfer function. This is because the denominator
of the transfer function is the characteristic equation of the governing differential
equation when set equal to zero. Reference to Eq. (13-1) makes this readily
apparent. To obtain the characteristic equation by the classical method of solving
differential equations, it is necessary first to set the forcing function V; equal to
zero and then to put the remaining terms in operator form. Thus
0= Rel “lp Lysl¢ (13-5)

or
Ry + sLy = 0 (13-6)

A comparison of Eq. (13-6) with the denominator of Eq. (13-3) reveals the two
expressions to be identical.
The complete time description of the output variable can always be found
from the transfer function. This includes steady-state as well as transient behav-
ior. In the case of the R-L circuit, e.g., the expression for the transformed solu-
tion of the field current can be written from Eq. (13-3) as
1
I(s) = V/(s) (13-7)
Ry 4. sLy

If Vis now identified for convenience as a step voltage of magnitude V; applied to

the field winding, the corresponding Laplace transform is

VAS) ee
V
(13-8)
Inserting this expression into Eq. (13-7) yields
Vy 1
9) =Tesla + RilEp
eae hd i eee
on
13-9

A partial fraction expansion allows Eq. (13-9) to be written as

SS eC Ve —— aSs + eS
sf (13-10)
Is) Lys(s_+ R,/Ly) Siest- Rl Ly
 Dynamics of Electric Machines Chap. 13
508

Evaluation of Ky and K;, then leads to

Is) = 2alt
(— — I 13-1
Ca
R; \s ra
By employing a table of Laplace-transform pairs the corresponding time solution
is found to be

id it) = A potas (13-12)

where t, = Ly/Ry.
A comparison of Eq. (13-12) with Eq. (13-11) shows that the transient term
—e-"t is directly related to that term of the transformed version of the solution
that results from the transfer function [the second term of Eq. (13-11)].
As a matter of convenience the transfer function is frequently represented
graphically by the block diagram notation depicted in Fig. 13-2(a). Sometimes the
signal flow diagram shown in Fig. 13-2(b) is used. The output quantity of the
block is always the Laplace transformed version of the output variable and the
input quantity is the Laplace transformed version of the input variable.
Systems often are represented by a block diagram showing the intercon-
nected transfer functions of the individual components of which they are com-
prised. Large complex systems contain many direct transfer functions of the type
depicted in Fig. 13-2, but they may also contain feedback transfer functions.
Whenever a feedback transfer function appears around a direct transfer function,t
the total transfer function of the combination is readily found by applying the
feedback formula which is now derived. Refer to the block diagram appearing in
Fig. 13-3. The quantity G(s) denotes the direct transmission function between the
reference input signal R(s) and the output signal C(s). The quantity H(s) repre-
sents the feedback function. Note that the output quantity C(s) is operated on by
this function and then subtracted from the reference input R(s) to generate the

Vals) Ir(s) Mee vee: | ae eins ) Figure 13-2 Graphical representation

Re 1+ ST; od.et ot of the transfer function of the circuit of
Ry 1487, Fig. 13-1: (a) block diagram form;
(a) (b) (b) signal flow diagram form.

Figure 13-3 Block diagram for a

feedback system.

+A direct transfer function of a system component is one that appears in the direct path linking
the input and output variables.
Sec. 13-1 The Transfer Function 509

actual input signal E(s) to the direct transfer function G(s). The quantity E(s) is
often called the actuating signal of the feedback system. The reference input
signal and the feedback signal must be of opposite signs. This is a desirable
condition for suitable operation of the feedback system and is commonly termed
negative feedback. Whenever a feedback path such as H(s) is wrapped around a
direct path such as G(s), the significant transfer function is the one that relates the
output C(s) to the input R(s). The ratio of these two quantities can be readily
expressed in terms of H(s) and G(s). A glance at Fig. 13-3 makes it apparent that
the output quantity can be expressed as

E(s)G(s) = C(s) (13-13)

But
E(s) = R(s) — H(s)C(s) (13-14)

so that upon its insertion into Eq. (13-13) there results

[R(s) — H(s)C(s)]G(s) = C(s) (13-15)

This last equation now involves just the input and output variables. Accordingly,
after terms are collected and the ratio of output to input is formulated, the desired
closed-loop transfer function T(s) results. Thus

G(s)
T(s) = (13-16)
1 + H(s)G(s)

The plus sign appears in the denominator whenever negative feedback is used.
The quantity H(s)G(s) is called the loop transfer function since it is obtained by
taking the product of the transfer functions encountered in traversing a closed
loop.
In dealing with systems that involve several energy-storing elements identi-
fied in the form of Eq. (13-3), it is sometimes useful to represent the known
transfer function in terms of an equivalent transfer function having a direct trans-
fer function that is pure integration (as denoted by the Laplace operator 1/s). The
advantage of such a block diagram representation is that the respective output
quantities of all the pure integrators in the direct transmission path are then in fact
the state variablest of the total system in terms of which the complete system
to use as the
+When dealing with systems containing energy-storing elements, it is convenient
be described at any
dependent variables those variables by which the system’s energy state can
can always be conve-
specified time. For example, in an R-L circuit the particular state of the circuit
This is because a
niently identified in terms of the current irrespective of the past history of the circuit.
yields the energy state by the relationship 2Li?. A similar
knowledge of the current immediately
the instantaneous value
situation prevails for the capacitor. Here, however, it is convenient to choose
state of the network
of the voltage across the capacitor vu, as the state variable. Once v, is known, the
C is known through the relationship 4Cv2. The state function depends solely upon the final
containing
state of the element.
to represent
The principal advantage of working with state variables is that it makes it possible
 Dynamics of Electric Machines Chap. 13
510

that leads
behavior can be described. In the simple case of the first-order system
through appropri ate manipula tion to rewrite the direct
to Eq. (13-3), it is possible
of Eq. (13-16). Keeping in mind that it is desirable to
transfer function in the form
nt formulat ion of Eq. (13-16), it is helpful to
identify G(s) as 1/s in the equivale
divide numerator and denominator of Eq. (13-3) by sv. This leads to
lila enh atl Ste. exe h., | I/s
(13-17)
Rel + stp Re (l/ste) + 1 Rete LL + (1/s)(A/tp)
A block diagram representation of the bracketed term of Eq. (13-17) appears in
Fig. 13-4(b). Note that it is identical to the representation depicted in Fig. 13-4(a),
but the former does identify a pure integration block in the direct transmission
path.
The transformed equation associated with the block diagram of Fig. 13-4(b)
by inspection can be written as
te. 1 (13-18)
VAS) RaW 5,1As) = sls)
The corresponding time equation is obtained by replacing s with d/dt and by
dropping the ‘‘of s’’ notation.+ Accordingly, we get
diy “Vy 1
dt = Rete ptt Ge)

Equation (13-19) is called the state equation for this system since it involves a
relationship between the state variable J; and its first derivative. A little reflection
of course discloses that this equation is identical with the original governing
differential equation [see Eq. (13-1)].

13-2 DYNAMIC BEHAVIOR OF THE DC MOTOR WITH CONNECTED

MECHANICAL LOAD

As already pointed out in the preceding section, the dynamic behavior of electro-
mechanical devices is determined by first finding the appropriate transfer function

the state of a system containing n independent energy-storing elements in terms of n state equations,
which are always first-order differential equations. Thus, if a system contains three independent
energy-storing elements, the total system behavior can be described by three state equations. In each
instance the state equation is a first-order differential equation involving the state variables and
possibly an external forcing function. In contrast, if the total system behavior were to be represented
by a single governing equation, it would be a third-order differential equation. An outstanding advan-
tage of the state approach to system analysis is that it lends itself naturally to easy simulation on the
analog and digital computers.
+ Although it is customary to use lower-case letters for time functions and capital letters for the
Laplace-transformed functions, that practice is not followed in this chapter because of the desire to
preserve the identity of the variables with the notation used in preceding chapters. As a substitute this
distinction is made by the absence or présence of the ‘‘of s’’ notation.
Sec. 13-2 Dynamic Behavior of the DC Motor with Mechanical Load 511

Figure 13-4 Alternative block diagram representations of transfer functions:

(a) direct representation; (b) equivalent representation, bringing into evidence
pure integration in the direct transmission path.

and then applying the techniques of the Laplace transform theory. In the process
suitable block diagram representations are developed and the proper state equa-
tions identified in order to enhance a computer study of the machine dynamics for
those readers who are so inclined. A study of the dynamics of the dc motor is
made for two cases—one where the armature winding leakage inductance? is
negligibly small and the other where it is large enough to be considered.
Figure 13-5 gives the schematic diagram of the dc motor with its connected
load. The combined inertia of the load and the rotor of the motor is denoted by
J.t The equivalent viscous friction of the motor and the load is denoted by F.
The opposing load torque is called T,. The quantity La is the armature leakage
inductance which is initially assumed equal to zero; R, is the armature winding
resistance; and w denotes the speed of the rotor expressed in radians per second.

Ro Lo

Ig

Vv ~W Mech.
; Eo JF, 1 | load

Figure 13-5 Schematic diagram of the

i dc motor with mechanically connected
Re f MOLE a
load. Field current is maintained con-
Vy stant.

to
Assume we wish to find the manner in which the motor speed responds
voltage for constant field current. In
changes in the applied armature winding
produced by the armature winding
+This refers to the part of the flux that can be considered
mmf and not linked with the field winding.
n for J, F, and K. For example,
+ A convenient set of units is assumed throughout this discussio
be expresse d in lb-ft-sec” , F in lb-ft-sec, and K in lb-ft/rad.
J would
512 Dynamics of Electric Machines Chap. 13

short, we want to find the transfer function that relates w to V,. To do this, it is
necessary to establish a governing equation that relates w to V,. At the motor
shaft the electromagnetic torque of the motor must be equal to the sum of the
opposing torques. Thus

T= 12 + Fo + Ty, (13-20)

or, expressed more conveniently,

Ti; Big La Hee (13-21)

dt

It is helpful to note at this point that Eq. (13-21) is a first-order differential equa-
tion; the sole energy-storing element is denoted by the total inertia J. Further-
more, by Eq. (8-23) the electromagnetic torque is expressed as
T = K,®I, (13-22)

where ® denotes the air-gap flux. Since flux is held fixed in the system of Fig.
13-5, Eq. (13-22) can be more succinctly written as
T= Kf. (13-23)

where K, is a torque constant expressed in units of torque per ampere. Also, the
armature current is related to the terminal voltage by
L~Ra=V; — Ea (13-24)

which follows from Eq. (8-24) for R, equal to zero. Moreover, the armature
induced emf is directly proportional to the speed as shown by Eq. (8-22). Thus
E, = Ke®n = K,o (13-25)
where K,, is a speed constant expressed in units of volts per radian per second.
Inserting Eq. (13-25) into Eq. (13-24) yields
I,ZRa = Vi — Keo (13-26)
It is interesting to note that the right-hand side of the last equation involves both
the input variable V, and the output variable w of the transfer function being
sought. Accordingly, Eq. (13-26) lends itself to the partial block diagram repre-
sentation of Fig. 13-6, expressed as a negative feedback arrangement (see solid
portion of the diagram). Note further that, if the quantity /,R, is multiplied by
1/R,, the armature current J, results. Subsequent multiplication by K, then yields
the developed electromagnetic torque. The last two statements are represented in
Fig. 13-6 by the dashed-line blocks.
In order to complete the block diagram, it is necessary to relate torque to
speed. Since this relationship is already available in Eq. (13-21), the desired result
is found by Laplace transformation. Thus
T(s) — Ty(s) = Jsw(s) + Fo(s) (13-27)
Sec. 13-2 Dynamic Behavior of the DC Motor with Mechanical Load 513

Figure 13-6 The block diagram repre-

sentation of Eq. (13-26) is drawn with
solid lines.

or

w(s) ee eh oe
FAL + stp (13-28)
Tis) — Tis) ST +E
where

Tm = mechanical time constant = F (13-29)

Upon adding the block diagram representation of Eq. (13-28) to the partial block
diagram of Fig. 13-6, the complete block diagram shown in Fig. 13-7 is obtained.
Finally, by applying the feedback formula of Eq. (13-6) to the configuration of Fig.
13-7, the transfer function relating speed to terminal voltage results. Accordingly,
w(s) _ (K,/RaF)U/A + stm))
(13-30)
Vs) 1+ (K,K./RaFU/C + stm)I
Simplifying and collecting terms yields

ol) Be sh teengra ! (13-31)

Vis) RF + K,K,1+ sri,

where

Gre = Kiko
Repay
(13-32)
It is possible of course to derive the desired transfer function without first
developing the block diagram. A single equation relating the input and output
T(s)

Figure 13-7 Complete block diagram of the system of Fig. 13-5.

 Dynamics of Electric Machines Chap. 13
514

sion for J, from Eq. (13-26)

variables can be obtained by substituting the expres
sion into Eq. (13-21). This
into Eq. (13-23) and then inserting the latter expres
leads to

: = ue o Knew + Ti, (13-33)

Laplace transform Eq. (13-33),

If we now let 7, be zero for convenience and then
we get
Kk (13-34)
é VAs) = Jsw(s) + Fo(s).+ = w(s)

Hence
K,/Ra K, 1
w(s)
Vis) sJ + F +(K:K./Ra) FRa + K:Ka1 + 5%, Uae
two to be identical.
A comparison of Eq. (13-35) with Eq. (13-31) shows the
change in the
What is the dynamic response of the motor speed to a step
the step change in
applied armature voltage V, for zero load torque? Assuming
transfer functio n of
applied voltage is V,, it follows that V.(s) = V,/s. From the
become s
Eq. (13-31), the Laplace transformed solution for the speed
os KV; 1 1
1 1

(bee
K,V;

OS) RF+K.K,slt+st,, JR, ss llth,

or
_ KV, 1 1 _ Ko K,
AD hii 9 OEMS AIT TRAN © et lerh ee
to
Evaluation of the coefficients of the partial fraction expansion leads

= + KK, \s_
FR, s + Ir, eel
The corresponding time solution is found to be

FR, + aKK,
w(t) Pte= cat sneato (= Feat 25s) (13-39)

Examination of the last expression reveals that the speed rises as depicted in Fig.
13-8 in accordance with a time constant 7/, to a steady-state speed given by the
coefficient on the right side of Eq. (13-39).
How does one obtain the dynamic response of the motor speed to a step
change in load torque for a constant applied terminal voltage? The answer is
readily obtained once the transfer function between speed and load torque is
known. A glance at Fig. 13-7 provides this information. Note that between 7;(s)
and w(s) there is a direct transmission function of (1/F)[{1/(1 + stm)]. The corre-
Sec. 13-2 Dynamic Behavior of the DC Motor with Mechanical Load 515

. wit)

KV
FR, +K,K
Figure 13-8 Response of motor speed
to a step change in V, for the system of
0 Time Fig. 13-5.

sponding feedback function in this instance is merely K,;K,/R,. Hence the use of
Eq. (13-16) yields as the appropriate transfer function

@(s) (1/F){A/( + stm] s Ra ( 1 )

Ts) 1 P(R RIRIU & Sta) RE KR 1 SET 57, ay

that
where 7’, is defined by Eq. (13-32). Then by following a procedure similar to
expression for w(t) is obtained. The
employed for a change in V, the complete
details are left as an exercise for the reader.
The block diagram of Fig. 13-7 can be redrawn by employing the manipula-
direct
tion illustrated in Eq. (13-17) to bring into evidence pure integration in the
This leads to the configura tion
transmission path of the system block diagram.
a pure integrato r is a state variable, it
shown in Fig. 13-9. Since the output of
state variable. The correspo nding state
follows that the motor speed w is the
of
equation is found by writing the equation that applies at the summing junction
the integrator. Thus for Fig. 13-9 we get (neglecti ng 7, for convenie nce)

{SL
Et)
OALOe yal re a (13-41)

Summing point

into evidence
Figure 13-9 Modification of the block diagram of Fig. 13-7, to bring
pure integrati on in the direct transmiss ion path.
 Dynamics of Electric Machines Chap. 13
516

The associated time-domain equation is

1

Inserting the appropriate expression for T leads to

Go ii Vers Kool) ont Inthe
dt R, eS Fr, TR 2M,

Ree Ae (i KK.)
Tiere OT parE RY Nes
The last expression is the state equation. Note that it involves the state variable
w(t) and its first derivative. Of course the solution of this equation has already
been shown to be Eq. (13-39).

Motor Dynamics with Armature Leakage Inductance

Not Negligible

Since the general solution procedure has already been just outlined, attention here
is focused only on those parts of the analysis where the presence of L, bears an
influence. Actually this occurs in just one place, Kirchhoff’s voltage law for the
armature circuit. Thus, instead of the difference between the terminal voltage and
the armature induced emf being equal to the J,R, drop alone, it is now equal to the
armature impedance drop. Expressed mathematically,

Teka tt Lie ine=V,- E, (13-44)

Laplace transformed, this becomes

I,(s)Ra(1 + sta) = Vis) — Kwo(s) (13-45)
where

T= R,
La (13-46)

or
I,(s) 1
Vi(s) — Kaw(s) Ra(l + sz) end
Equation (13-47) discloses that, to obtain the armature current from the difference
between the terminal voltage and the counter armature emf, this difference must
be operated upon, not by 1/R,, but rather by 1/[R,(1 + s7,)]. Hence the block
diagram of Fig. 13-7 becomes that illustrated in Fig. 13-10 when L, # 0.
Application of the feedback formula of Eq. (13-16) to Fig. 13-10 discloses
that the motor speed is now related to a change in terminal voltage by
Sec. 13-2 Dynamic Behavior of the DC Motor with Mechanical Load 517

-T(s)

Figure 13-10 Block diagram of the system of Fig. 13-5 when L, # 0.

K,
ols) RoF(1 + st.) + stm)
VG) ie KK. (les)
Rou + sr) 2 Sta)

which simplifies to

LS ENTSK
VAS) TaTmRaF : aicavan in. cope) Ke (13-49)
Ta Ting Tate, | Tatmltol

A glance at Eq. (13-49) indicates that now the characteristic equation is second
order. This is not surprising in view of the presence of an additional energy-
storing element in the form of the armature leakage inductance. The nature of the
dynamic response to a step change in applied voltage depends upon the character
of the roots of the denominator quadratic in Eq. (13-49). If they are complex
conjugate roots with negative real parts, the response is a damped oscillation. If
the roots are both negative real, the solution consists of a fixed term plus two
exponentially decaying terms.
Appearing in Fig. 13-11 is the block diagram of Fig. 13-10 redrawn to bring
into evidence the two pure integration terms associated with the second-order
system of Fig. 13-5 when L, # 0. The two state variables are the output quantities
of the two integrators, armature current J, and motor speed w. The two state
equations, which are linear first-order differential equations, are obtained by writ-
ing the expression for the input signal to each integrator, multiplying by 1/s, and
then equating to the specific state variable. Thus at the first integrator in Fig.
13-11 the transformed version of the state equation is

sIq(s) = Go
1
[Vi(s) — Kyets)] ~ =Lats)
R a Ta
(13-50)
 Dynamics of Electric Machines Chap. 13
518

-T(s)

Figure 13-11 Alternative form of the block diagram of Fig. 13-10, to bring into
evidence pure integrators in the direct transmission path. The output quantities of
the integrators are the state variables of the system.

At the second integrator the result is

1 1
sw(s) = ——
Fr,
[Kila(s) — T1(s)] — —Tm o(s) (13-51)

The corresponding time-domain form of the state equations is clearly

dipeseralae! (1K, Uy deol
dt T Ta £ Rala Site RaTa ee acta

d K. 1 1
dt
Nee] Pe, Fo
Pa TA
OyTy
Frm
(13-53)
In both equations, J, and w are functions of time. In matrix notation the time-
domain formulation of the state equations may be written as follows:
dIq 1 K, 1
dt Tele,Di Ma Rita : Ms
= + (13-54)
dials | Koga, tooth oi ofgeclaaltan -
dt Frm Tm e Fim

Equation (13-54) is called the normal form of the state equations describing the
complete behavior of the system depicted in Fig. 13-5.

13-3 DC GENERATOR DYNAMICS

How does the armature induced emf of a dc generator respond to a change in field
current? How does the armature winding current of the dc generator respond to
changes in field current? These are questions that are treated in this section. As
Sec. 13-3 DC Generator Dynamics 519

Re
Ty
Le

ye Constont
spee
Prime Figure 13-12 Direct-current generator
mover with connected load.

with motor dynamics, the answers lie in establishing the appropriate transfer
function relating the variables of interest. Two cases of generator dynamic behav-
ior are treated: one where no external load is applied to the generator (i.e., switch
S in Fig. 13-12 is open), and the other where external load is applied (S closed).

No External Generator Load

Consider the generator of Fig. 13-12 driven at constant speed by the prime
mover. Let it be desirable to find the transfer function relating the armature
induced voltage to the field winding voltage. The differential equation for the field
winding circuit is given by

Vi IpRp + 1,2 (13-55)

In general J; is a function of time whenever V; or Ry is changing. The Laplace

transformation of Eq. (13-55) for zero initial conditions yields
VAs) = Ip(s)Rp + Les p(s) (13-56)

where J,(s) denotes the Laplace transform of the time function J; and V;(s) is the
Laplace transform of V;. Upon formulating the ratio of I/(s) to V,(s), the transfer
function between these quantities results. Thus
(pC mae ge a | :
VAs) Re + sty) yak
where

TF = 7a= field winding time constant (13-58)

The relationship between the field current and the armature induced emf is
described by Eq. (7-10), which is repeated here for convenience. Thus
E, = Kp®n = Kyl; (13-59)
 Dynamics of Electric Machines Chap. 13
520

The right side of Eq. (13-59) is a valid equivalence provided that no saturation
in
occurs. For consistency the units of Ky are volts per field ampere. Expressed
Laplace notation, Eq. (13-59) becomes merely
E,(s) = Kylf(s) (13-60)

so that the transfer ‘‘function’”’ relating field current to the corresponding induced
emf is just a constant. Accordingly,
E,(s) _ (13-61)
I;(s) ’
The complete transfer function is then obtained by combining Eqs. (13-61) and
(13-57). Thus

E,(s) Ee 22] Ky
Vs) + s7/)
LVAsyJLI(s) 1~ Re (13-62)
The block diagram representation of the last equation appears in Fig. 13-13.

Vy (s) 1 I,(s) E(s) Figure 13-13 Block diagram of the

R. (1+ ee |Ke| “ system of Fig. 13-12, with load switch
i ! open.

Depicted in Fig. 13-14 is the form of the block diagram that brings into
evidence the pure integrator in the direct transmission path. The output of this
integrator is the state variable, namely, field current. The corresponding state
equation in transformed notation is found by taking the input to this integrator and
setting it equal to s/;(s). Thus
1 VAs) — :I/(s)
sis) = Re (13-63)
The corresponding time-domain form is then clearly

Figure 13-14 Block diagram of Fig. 13-13 modified to show the pure integrator. I;
is the state variable. j
Sec. 13-3 DC Generator Dynamics 521

: epee
gigi: (S hat (13-64)

Because state equations are only associated with energy-storing elements, E, does
not appear in Eq. (13-64). However, the transition from the state variable J; to the
desired output variable E,.is readily achieved through multiplication by a con-
stant, Kr.

Generator Dynamics Including the Effect of the Load

This situation is represented in Fig. 13-12 by closing switch S. The generator load
is assumed to contain inductance as well as resistance. Also, R, and L, are
respectively the armature winding resistance and leakage inductance.
The transfer function that relates E,(s) to V;(s) is identical with Eq. (13-62).
Always keep in mind that transfer functions are dependent upon the system
parameters—in this instance Ky, Ry, and Ly—and these have not been altered. If
it is now desirable to describe how the armature current is influenced by changes
in field voltage, then an additional expression is needed that relates E, to Iq.
Clearly this is readily available through the application of Kirchhoff’s voltage law
to the armature circuit. In differential equation form we have
dl, dl,
E,=1,R,+ L on eR +L Ae (13-65)

where R and L are the resistance and inductance of the load respectively. Laplace
transforming and collecting terms yield
E,(s) = Iq(s)(Ra + RYU + sta) (13-66)

where

a
Ta = armature circuit time constant = (13-67)
R,+R

The transfer function then becomes

I,(s)_. 1
(13-68)
E.(s) (Ra + R)U + sta)
It is worthwhile to note here that the existence of s in the last equation is occa-
sioned by the presence of a second energy-storing element in the system of Fig.
13-12 with the load connected.
The total transfer function relating the armature current to the field voltage is
determined by formulating the product of Eqs. (13-62) and (13-68). Accordingly,

ioe) Ky 1 1 (13-69)
VAs) RARa + R)1 + stp1 + sta
 Dynamics of Electric Machines Chap. 13
522

sik tae e 1 1(s)

Ry (1+ ST) (Ro+R) (1+8%)

Figure 13-15 Block diagram of the system of Fig. 13-12 when the load switch is
closed.

Again note that the denominator of the complete transfer function is now second
nt
order in s, which is consistent with the fact that there are now two independe
energy-storing elements in the system. The block diagram of the system is de-
picted in Fig. 13-15.
The modified form of the block diagram of Fig. 13-15 is illustrated in Fig.
13-16. Two pure integrators appear because the system is second order. Hence
there are two state variables, which a glance at Fig. 13-16 discloses to be the field
current and the armature current. The two state equations are found by employ-
ing the procedure previously described. In the s notation these are
1 1
si(s) = — — If(s) + =— VAs) (13-70)
i tga ARinve
sI,(s) eye
= ND
= I,(s) + i
(R,sal ides I/(s)
+ R)ta (13-71)

The corresponding time-domain versions are

Hig auras parame

dite, 4, digit: Rytr Ke (13-72)

dl, Ky 1
ir = ions Pret (13-73)

In matrix notation these state equations are written as

dl, nothGobiga| es
dt =
Tpanooed + Vr (13-74)
da Be eee ae As if 0
dt (Ra 2 R)t4 TA ‘

Figure 13-16 Modified version of the block diagram of Fig. 13-15 to bring into
evidence the state variables, J; and J,, as the outputs of pure integrators.
Sec. 13-3 DC Generator Dynamics 523

Example 13-1
The dc generator depicted in Fig. 13-12 has a field winding resistance of 40 0 and a
field inductance of 8 H. The generated emf per field ampere is 100, and the magneti-
zation curve is linear. Moreover, the armature winding resistance and leakage in-
ductance are respectively 0.1 © and 0.2 H. The load resistance is 5 0 and the
., associated load inductance is 2.35 H.
Determine the time solution for the armature current when a field voltage of
102 V is applied to the field winding. Assume that the prime mover is running at
rated speed and that the load switch is closed.

Solution The transfer function that relates the armature current to the field voltage
is given by Eq. (13-69). Thus
eyo" Ky 1 1
VAs) RAR, R) 1 Hise 1 + 74
100 1 1
1 + 2) (13-75)
~ 40(0.1 + 5) 1 + (s/5) (8/
where
Lists af
evbéiz, 28 has!
PE A010 bask :
(13-76)
and
£ SE EL, OLA BLL
TAsioRisel Ril laduics5)21 Qe (13-77)
The Laplace transform of the field voltage for a step change of 102 V is described by

V/(s) = u = (13-78)

Hence the transformed form of the solution for the armature current can be written
as
_ 102 (00 1 1 4 500 :
TAS) mea NFpAgee es Me DI ee SIGEND) (13-79)
Evaluation of the last equation is facilitated by a partial fraction expansion. Accord-
ingly,
500 Ko K, K,

dS) oe oe) Oe at 2
sh eed
These coefficients are found to be

500 13-81
= |———_——~ = ( )
Ko Fe+ 5)(s + 2)Js=0 30

500
ce |mee = +33, 13-82
Kis Fe + 2)Js==5 3343 ( )

=
500
——— =s— = 13-83

Ee + 5)Js=-2 83.3 ( )
K>
 524 Dynamics of Electric Machines Chap. 13

Insertion into Eq. (13-80) allows the transformed solution to be written as

50 =.33.3 83.3
(13-84)
OO) ea

The corresponding time solution as obtained from a suitable table of Laplace-trans-

form pairs becomes
LG) 50 93.32" 89.36. (13-85)

which is the desired solution. This last expression completely describes the process
of armature current buildup when a field voltage of 102 V is suddenly applied to the
field circuit. Note that the steady-state (or final-state) value of the armature current
is 50A—a result that can be verified through use of the procedure described in
Chapter 7 for finding steady-state performance.

13-4 DYNAMICS OF A GENERATOR-MOTOR SYSTEM

A method of speed adjustment that is employed in industry whenever control over

a wide speed range is required is shown in Fig. 13-17. Speed adjustment above
base speed? is obtained by field current (J;,) variation with rated voltage applied
to the armature terminals. Below base speed the speed is adjusted by changing
the armature voltage below its rated value. A convenient way to control the
armature voltage is to control the field current of a supplying generator as depicted
in Fig. 13-17. Large amounts of power at the load can be controlled by easy
adjustment of the generator field current. This method of speed control is referred
to as the Ward-Leonard system.

An constant

Prime
mover ||

Figure 13-17 System diagram for studying dynamic behavior of the Ward-
Leonard generator and motor system of speed control.

A little thought should make it apparent that adjustments of the field current
in the system of Fig. 13-17 do not translate immediately into appropriate speed

+Base speed corresponds to operation with rated armature voltage and rated field current at
rated motor output.
Sec. 13-4 Dynamics of a Generator-Motor System 525

changes. The change in speed must occur smoothly from one commanded level to
another. The change cannot occur abruptly because of the presence of energy-
storing elements. Examination of this system reveals the presence of three en-
ergy-storing elements, each of which influences system dynamics. These are gen-
erator field inductance, combined generator and motor armature winding leakage
inductances, and the inertia of the load and the motor.
A description of the system’s dynamic behavior lies in identifying the trans-
fer function that properly relates the motor speed w to the generator field voltage
V,. As a first step in deriving this transfer function, we write the differential
equation for the generator field winding circuit. Thus

Ve = Ip Ry st 1, (13-86)

Laplace transforming yields

V-(s) = Ip(s)ReU1 + stp) (13-87)
where 7; = L;/Ry. Formulating the ratio of field current to field voltage leads to

Jie
Vy(s)
Samet
Re + stp)
(13-88)
The generated armature voltage is related to the field current by
E,(s)
=Ky VIA (13-89)
I;(s)
In turn the generated emf is related to armature current by Kirchhoff’s voltage
law. Specifically,
dl,
= he
Eel ele —a + Whe oe
~ bpPF quae (13-90)

where R, and R,,, are the armature winding resistances of the generator and motor,
L, and L, denote the leakage inductances of the armature windings, and E,
denotes the counter emf of the motor. More explicitly E,, can be written as
En = Kyo (13-91)

where K,, is expressed in volts per radian per second and the motor speed @ is
expressed in radians per second. Inserting the last expression into Eq. (13-90) and
Laplace transforming gives
E,(s) — K,o(s)
I,(s) oe! MLR ISELT
Rileeves) eG Z
(13-92)

where
R=R, + Rn (13-93)

ee R,
ee
+ Rm (13-94)
 526 Dynamics of Electric Machines Chap. 13

It is helpful at this point to rewrite Eq. (13-92) as

I,(s) 1
s) RC + sta)
— K,o(s)
E,( (13-95)
This is in the form of a transfer function that lends itself to block diagram repre-
sentation. In particular, note that the input to the transfer function block consists
then of the difference between the generator source voltage E,(s) and a quantity
that is a function of the output variable K,,w(s). Hence the latter quantity can be
considered as coming from a feedback path originating from the output.
Continuing the development in a logical stepwise fashion, we next note that
the armature current combines with the motor field flux to produce an electromag-
netic torque that can be expressed as

f= KI, (13-96)
where K, bears units of torque per armature ampere. The transfer function nota-
tion is simply

TS)
K, (13-96a)
TAs) oy
Finally, to relate the developed torque to the motor speed, we note that the former
must be equal to the sum of all opposing torques. If we now let J denote the
total
inertia of the load and the motor and F denote the total viscous friction, we
can
write

dw
iy Grades
T=J i Br’ Fo -
(13-97)

The load torque is omitted in this analysis because it is assumed to be

constant.
Our interest here, remember, is to find the manner in which the
speed changes in
response to changes in the generator field current (or field voltage).
Upon Laplace
transforming Eq. (13-97) and formulating the ratio of speed to torque,
we have

os)
T(s) F(1 + stm) soe
where

. P J
Tm = mechanical time constant = F (13-99)

A successive block diagram representation of Eqs. (13-88),

(13-89), (13-95),
(13-96), and (13-98) taken in order leads to the complete system
block diagram
depicted in Fig. 13-18(a). By applying the feedback formula
of Eq. (13-16) to the
feedback loop, the complete expression of speed to generated
voltage is obtained.
Sec. 13-4 Dynamics of a Generator-Motor System 527

Vy (s)
O

(a)

Figure 13-18 (a) Block diagram of the generator motor system of Fig. 13-17. (b) Alternate
form of the block diagram of Fig. 13-17, to show explicitly the pure integrators of the
system. There are three; hence the number of state variables is also three.

Thus

SEcen es
w(s) RFQ + sta) + STm) ¥ K,
E(s)% K,K. RF(1 + sta) + stm) + KoK,
DF RR TR EE Gre)

= EP
IG,
Me e 1
aay ie
-100 )
RF + Kika TATmRE fe: (74 + TmJRE
* RE + K,K; RF + KK;

This result can also be found analytically without the use of the block diagram by
inserting Eq. (13-92) into Eq. (13-96) and then substituting for T(s) in Eq. (13-98).
Accordingly,
;
+5m)
F@Ril)— Kusc.a) ¢ FO
w(s) is 1
(3-100
 hines Chap. 13
Dynamics of Electric Mac
528

or
K,.K,o(s) (13-102)
RF(i + sta) + STm)@(S) = K,E,(s)—

from which
iee+csKKe,e
arm)il :
OG t e s s U
a st r
st a)
e+ l
St
(13-103)
E,(s) RFC +
discloses the two expres-
ediate step of Eq. (13-100)
Comparison with the interm
sions to be identical. by multiplying Eq.
on is determined of course
The complete transfer functi
and (13-89). Hence
(13-100) by Eqs. (13-88)

52 ee (ta + Tm)RF , y
V;(s)_ Re(RF + KaKs) 1 + Str) | RE + KK:
S°RF+K.K;
(13-104)

equation of this system is

of the governing differential
The characteristic equation to zero. This yields a
nator of Eq. (13-104) equal
obtained by setting the denomi lags in the system,
cates the presence of three time
third-order equation which indi three energy-storing
istent with the presence of
a result which is entirely cons
elements. k diagram of
the modified version of the bloc
Appearing in Fig. 13-18(b) is the outp uts of which
explicitly the pure integrators,
Fig. 13-18(a) in order to show this system are
rdingly, the state variables of
identify the state variables. Acco tions are found
and motor speed. The state equa
field current, armature current, g transformed
ions. This leads to the followin
as described in the preceding sect
forms:
1 1 (13-105)
sh(s) = ~ 7 W) + Rey V;(s)

sla(s) = PL Ys) ~ 7,19) ~ Beg 2

1 Ke (13-106)
(13-107)
sw(s) = = LAS). [ w(s)

the equations are

Expressed in the time domain
.
dij~
dt
ffeB fis 1
Ree Vr
(13-108
 Sec. 13-5 DC Motor Transfer Function 529

lia Kyo poitslet

dt > Rt, i TA Ia
punoiKs
Rt ¢ (13-109)

dw K, 1
aaere | Pee a w@ (13-110)

Finally, the matrix formulation becomes

dl;
at aa 0 0 Ir Ra

Se
dl,
Sil se
pal Ky
leap1 LS
Ol ee Pe (13-111)
da K, 1
dt : Ft, are ¢ v

There are three state equations associated with the three independent energy-
storing elements.

13-5 DC MOTOR TRANSFER FUNCTION

Two forms of the dc motor transfer function are of interest. One is the armature-
controlled mode with fixed field current; the other is the field-controlled mode
with constant armature current flowing.

Armature-Controlled DC Motor

Figure 13-19 depicts the armature-controlled de motor. The motor speed is made
to respond to variations in the applied motor armature voltage V,. For simplicity,
armature winding leakage inductance is assumed to be negligible. Our concern
here is to develop a transfer function that relates motor speed to applied armature
voltage. Let J denote the total inertia of the load and the rotor of the motor. Let

Figure 13-19 Armature-controlled de

motor, field current maintained con-
stant.
 Dynamics of Electric Machines Chap. 13
530

point in
F denote the total viscous friction of the motor and load. A good starting
at the motor shaft.
the development is to write the torque equation as it applies
Accordingly,
T= Kil,
dw (13-112 )
= J J—+
a Fo

The
where T is the motor electromagnetic torque and w is the motor speed.
expression for the armature current is given by

i V; = K,o
= R, (13-113)

where E, is the motor counter-emf, which is also expressed as a speed voltage

constant K, in volts per radian per second. In the interest of leading to the
appropriate block diagram representation, it is worthwhile to write Eq. (13-113) in
Laplace notation. Keeping in mind that both J, and w are the variables, we have

Ia(3) = (Vis) ~ Kyots)] 3 (13-114)

The right-hand side has been written as indicated to emphasize that the reference
input quantity V,(s) and a quantity derived from the output variable, K,,w(s), are
being compared and that the difference is then multiplied by 1/R, to yield the
armature current variable. Hence the symbol for a summing point followed by a
block marked 1/R, can be made to represent Eq. (13-114). It is interesting to note
that the counter-emf of a motor can always be treated as a negative feedback
quantity.
Once I,(s) is available, the transfer function relating it to the electromagnetic
torque is

———~. = K;} (13-115)

Finally, the transfer function that describes the variation of the motor speed with
torque is obtained from Eq. (13-112) after it is Laplace transformed. Thus

T(s) = Jsw(s) + Fo(s) (13-116)

or
w(s) _ 1 (13-117)
T(s) as F(iehakt,.)
Sec. 13-5 DC Motor Transfer Function 531

where F

Tm = mechanical time constant = > (13-118)

The successive block-diagram notation of Eqs. (13-114), (13-115), and (13-117)

leads to the complete block diagram depicted in Fig. 13-20.

ep
ph
F(l+STm)

Figure 13-20 Block diagram represen-

tation of Fig. 13-19. The output vari-
able is motor velocity.

ap-
The desired transfer function between the motor speed and the motor
of Fig. 13-20.
plied voltage is found by applying Eq. (13-16) to the block diagram
Accordingly,
w(s) _ (K,/RaF)U/A + stm]
VAs) 1+ (KoKi/RaF)U/G + Stm)]

or

w(s) _ K; 1
Vis) K.Ko + RaF 1 + sth (13-119)

where

Ral = nestil Resinitay (13-120)

tm 3h Pee 4K, Ka RpF +KyKo
around the direct
Equation (13-120) discloses that the feedback path that appears
is respons ible for reducin g the motor time constant
transmission path of Fig. 13-20
observation is that the
over the nominal value of J/F. The practical aspect of this
expect by looking at 7,,
motor responds dynamically more quickly than one would
alone.
(13-119) without resort-
If it is desirable to derive the transfer function of Eq.
entati on follow ed by the applica tion of the feed-
ing to the block diagram repres
inserting Eq. (13-113) into Eq.
back formula, this may be readily accomplished by
(13-112) and then rearranging terms. Thus

V, =
R,Kuo K; dw + Fo
= er (13-121)
532 Dynamics of Electric Machines Chap. 13

or

= RF [Jdo, Rok + RoR: |

testers ell RuF
eee RoF ea (13-122)
i K, ™ RF + KuK; dt
Laplace transforming and formulating the ratio of output to input yields a result
that is identical to Eq. (13-119).
Sometimes the dc motor Fig. 13-19 is used as the output actuator in feedback
control systems where the controlled variable may be position rather than veloc-
ity. Since velocity is the derivative of the position variable c, the transfer function
relating the two is obtained as follows:
dc
aa (13-123)

Laplace transforming for zero initial conditions gives

c(s) = -=«(8 (13-124)

The complete block diagram now becomes that shown in Fig. 13-21. Moreover,
the complete transfer function between the motor displacement variable and the
applied armature voltage becomes

CUS K, 1
VAs) KaKi + RaF s( + sri,) seth

where 7,, is specified by Eq. (13-120).

Appearing in Fig. 13-22 is the form of the block diagram of Fig. 13-21 that
allows convenient identification of the state variables, which in this case are

Figure 13-21 Block diagram of the

motor of Fig. 13-19 when the output
variable is motor displacement.

Figure 13-22 Alternative form of Fig.

13-21, drawn to facilitate writing the
appropriate state equations.
Sec. 13-5 DC Motor Transfer Function 533

clearly the motor velocity and the motor displacement. The first state equation in
transformed form is found by taking the input to the first integrator and setting it
equal to sw(s). Hence

bsK,V,(s) = KK; 1
SR Birla Rca ee ea
sSry
ight oet Ob i (aan
Kok ee
1) (13-126)
The second state equation is given by Eq. (13-124). The corresponding time-
domain expressions are
dw ai z eos as
5 Faded “317 dina’ 5 ck ae a)
dc
see @ (13-128)

The matrix notation in normal form is then

dw a (jek - = 0 K,
dt RF Rpaetii “ RiF a,
= - V, (13-129)
20 1 0 0
dt a

Field-Controlled DC Motor

The circuit configuration for this mode of operation is shown in Fig. 13-23. The
armature is supplied with a constant current from a suitable current source. In
light of this fact the variation of electromagnetic torque is dependent upon the

Re

Ve Le

Figure 13-23 Field-controlled de

motor.

variation in field current. This is apparent upon recalling that T = K;®J, and that
the air-gap flux ® is directly related to the field current for the assumed condition
of no saturation. Therefore, the torque in this instance may be written simply as
T = Kyls (13-130)
 Dynamics of Electric Machines Chap. 13
534

where Ky is a motor constant expressed in units of torque per field ampere. The
torque equation at the motor shaft then becomes

Kylp = J ‘e + Fo (13-131)

where J and F have the same meanings as in the preceding. Laplace transforming
and formulating the ratio of output to input yields the transfer function that relates
motor velocity to field current. Thus
w(s) <i Ky
(13-132)
Ip(s) F(1 + stm)

In turn the field current is related to the applied field voltage by the usual
transfer function. Accordingly,
I;(s) * 1
(13-133)
Vr (s) - Ry(1 + stp)

where 7; = Lr/Ry. By combining the last two expressions, the transfer function
between speed and field voltage results:

w(s) I-(s) w(s)_ Ky 1 1

(13-134)
Vi(s) Ve(s) Ip(s) ReF 1+
1+ st¢1 + stm

Finally, to obtain the relationship between motor displacement and applied field
voltage, it is merely necessary to multiply the last equation by Eq. (13-124) ex-
pressed in the form [c(s)/w(s)] = 1/s. This yields

c(s) a Ky 1
(13-135)
V;(s) ReF s(1 + stp) + stm)

A glance at the denominator of Eq. (13-135) discloses that now the characteristic
equation is third order. Figure 13-24 depicts the block diagram of Eq. (13-135).

V4(s) Tar] I¢(s) arss s) A cls) Figure 13-24 Block diagram of the
f az mer system of Fig. 13-23.

Consistent with the third-order nature of the characteristic equation, there

are three state equations which describe the complete dynamic behavior of this
system. These equations are more easily identified by referring to the alternative
block diagram illustrated in Fig. 13-25. Clearly, the three state variables are field
current, motor speed, and motor displacement, which are respectively the output
quantities of each of the integrators in Fig. 13-25. The transformed versions of the
three state equations as associated with the three integrators are then
1 1
sIp(s) = Fer; Ve(s) — poe d: (13-136)
 Sec. 13-6 Two-Phase Servomotor Transfer Functions 535

Figure 13-25 Alternative form of Fig. 13-24.

K 1
sw(s) = am Ip(s) — = w(s) (13-137)

sc(s) = @(s) (13-138)

The corresponding first-order differential equations then become

dip. 1
cau ee Vy sh (13-139)

def AOlK 71098] (13-140)

ap aE I; rT w

os 13-141
din we (1814)
Finally the matrix formulation is
dl; 1 1

a l=|po
dw a Ky
-z lel
1
+] o | (13-142)
dc 0
an 0 le O:|| ¢

13-6 TWO-PHASE SERVOMOTOR TRANSFER FUNCTIONS

discussed in
The steady-state characteristics of the two-phase servomotor are
to the dynamic behavior of these motors as
Chapter 10. Attention is directed here
described by their transfer functions.
by following
Although the servomotor transfer functions could be derived
in the precedi ng section, in the interest of
the procedure outlined for the dc motor
d using the external torque- speed
versatility an alternative approach is employe
ristics such as those depicte d in Fig.
characteristic. Typical torque-speed characte
more, as indicate d in Fig. 13-26, the
10-13 are assumed to be available. Further
a fixed referenc e voltage. The input
servomotor is assumed to be operated at
What quantity then
variable in this arrangement is the control winding voltage V..
536 Dynamics of Electric Machines Chap. 13

Squirrel—cage

+f
rotor
(eo)

Control oo/ J
winding FL
Reference
winding

|: Figure 13-26 Two-phase servomotor

Vret with fixed reference winding voltage.

shall we consider to be the output variable? Shall it be velocity w or displacement

c? Examination of Fig. 13-26 certainly leaves no doubt that the application of a
given control winding voltage manifests itself as an output velocity. However,
this alone is not necessarily the determining factor in establishing what the output
variable shall be. Keep in mind that servomotors are used in feedback control
systems so that a more important consideration is the function which the servo-
motor will fulfill. If the output of the control system is to be velocity, then it is
appropriate to derive a transfer function that relates velocity to the control wind-
ing voltage. On the other hand, if the motor output is to be used to drive a
position-sensitive transducer such as a potentiometer or a resolver or a synchro,
then clearly the motor displacement should be considered the output variable.
The primary concern at the moment is the general form that the servomotor
transfer function assumes. Consequently, the torque-speed curves shown in Fig.
10-13 are assumed to be linear for simplicity. A study of these curves reveals that
an increased motor torque comes about either by an increase in control winding
voltage or by a decrease in speed for a fixed V.. Expressed mathematically the
developed motor torque is

T = Ky V. — Fu (13-143)

where Ky = motor torque constant in units of torque/volt

Fy = motor equivalent viscous-friction constant in units of torque/rad/sec
It is helpful to note that Fy is nothing more than the slope of the torque-speed
curve at constant V,. Also, once the torque-speed curves are known, the parame-
ter Ky is merely the change in torque per unit of change in control voltage at
constant speed. Equating this torque to the sum of the opposing load torques
yields the governing differential equation that relates motor displacement to con-
trol voltage. Thus

de_ herrea?
Sl gcale @ Ngcpieen a? Std (13-144)
Sec. 13-6 Two-Phase Servomotor Transfer Functions 537

where J = motor inertia + load inertia referred to motor shaft

F,, = viscous friction of load referred to motor shaft

Rearranging and Laplace transforming lead to

Ku Vs) = (Js? + sF)c(s) (13-145)

where
Fe= Fu oF Fy, (13-146)

Hence the desired transfer function becomes

ROY mies in
Vis) sl + Stm) (13-147)

where
K (13-148)
Km = = (V-s)7!

im = 7 S (13-149)

Frequently the motor time constant 7,, plays an important role in determin-
is a
ing the character of the dynamic response of the control system of which it
establish ed in terms of the inertial
part. This is because the magnitude of 7,, is
and the
properties of physical elements such as the size of the servomotor rotor
that they cannot be reduced to
reflected load inertia. These elements are such
goes hand in hand with the required
negligence primarily because the size of J
is the deter-
level of output power. Accordingly, in many cases the motor inertia
which the system will be capable of
mining factor that establishes the extent to
following variations of the command signal with time.

Effect of the Control Winding Time Constant

time lag caused

The result of Eq. (13-147) is approximate because it neglects the
descript ion of the servo-
by the inductance of the control winding. A more exact
that the control winding
motor transfer function can be obtained by recognizing
in turn produce s the
voltage really produces a control winding current J,, which
equatio n as it applies at
flux that enables the motor to develop torque. The torque
the motor shaft then becomes
dc d’c dc (13-150)
Kelle — Fuh = de + Fr a
 Dynamics of Electric Machines Chap. 13
538

where K,, denotes a torque constant expressed in units of torque per field am-
pere. The corresponding transfer function is given by
c(s) _ Ke 1 (13-151)
Lisy° oF s+ stn)
Next compare Fig. 13-27 with Fig. 10-13. At a fixed speed the input imped-
ance at the control winding terminals remains constant. If operation is say at V.2,

Torque

I¢q > Ic3> Ica > ler

Figure 13-27 Linearized torque-speed

curves showing variation of control
_ winding current with control winding
voltage at constant speed. Each
straight line corresponds to a fixed
control voltage: the higher the line, the
0 Speed _ larger the control voltage.

then a sudden increase in control voltage to V,3 at the same speed increases the
torque along an ordinate line. If the field winding time constant is not negligible
then, as V,2 increases to V,.3, so too does the control current increase from J,2 (in
Fig. 13-27) to 7,3, but with a time delay.
The field current is therefore related to the control voltage by the usual
transfer function for an R-L circuit. Thus

LS) hentia (13-152)

VAs), «Rel tiste)
where 7, = L,/R,. Upon combining Eqs. (13-151) and (13-152), the desired trans-
fer function results:

EE NGR
{Soot COeen
V.(s) FR, s(1 + stm)(1 + ste)
or

LENO. Noti SIREN

Gend (13-153)
Vis) sC + stm) + ste)
It is interesting to note in this development that

K, = Ky
; R. (13-154)
Sec. 13-6 Two-Phase Servomotor Transfer Functions 539

A comparison of Eq. (13-153) with Eq. (13-147) indicates that the difference does
indeed lie in the time constant of the control field winding.
Figure 13-28 depicts the block diagram representation of Eq. (13-153). Note
again that, whenever the motor output variable is taken to be displacement and

1 eile @
F(1+ STR)
Figure 13-28 Block diagram of the system of Fig. 13-26.

the motor field winding time constant is included in the analysis, the result is a
third-order characteristic equation. Of course this means that there are three state
variables and three state equations. Reference to Fig. 13-29 reveals these state
equations in transformed form to be as follows:

sI.(s) = hes Vs) — * 1s) (13-155)

so(s) = = I(s) - “ w(s) (13-156)

sc(s) = w(s) (13-157)

c(s)

pure
Figure 13-29 Alternative form of Fig. 13-28, modified to show explicitly the
integrators, the outputs of which identify the state variables of the system.

as
The corresponding first-order differential equations can be written
dl, i 1 2
=aA ER: V. —-—z i, 13-15
(13-158)

do Ka 1
—=—*],.-— 13-159
at PT. Te Tm ¢ ( )

dc 2
os 13-160
n° (13-160)
 540 Dynamics of Electric Machines Chap. 13

The matrix formulation is then:

dl. 1 1
‘dt ae © ne RT
dw Ky 1
peel
AF PAPE Rell
recLA Ee mm 0 || @ fa 0 Vs (13-161 )

dc
Bi 0 f Ol c 0

13-7 AMPLIDYNE GENERATOR

The amplidyne generator is a power-amplifying device capable of generating high

output levels, usually in the range of 1 to about 50 kW. In control system applica-
tions it often serves as a power stage that drives a dc motor. The control field
winding of the amplidyne (of which there are several sections) is usually driven
from the output stage of an electronic amplifier.
The amplidyne is basically a dc generator. It is driven at constant speed by a
suitable motor which serves as a source of energy for the unit. The magnitude of
its output voltage is controlled by the amount of field current flowing through the
control winding. Since the armature winding of a dc motor is frequently placed
across the output terminals of the amplidyne, it is important to account for the
motor loading effect whenever the transfer function is derived.
The principle of operation of the amplidyne is straightforward. A small
current flowing in the control winding is made to create a flux ®, directed as
illustrated in Fig. 13-30. Because the armature winding is being driven at full
speed, a voltage E, is induced and it is collected at the brushes marked q-qg. By
short-circuiting these brushes and designing an armature winding of low resis-

Load switch

Quad. axis
series field

Control
field
winding
Motor field
winding
i|

Figure 13-30 Amplidyne generator, showing windings and connected dc motor

and load. ‘
Sec. 13-7 Amplidyne Generator 541

tance, very large armature currents can be made to flow. For example, if E, is 1 V
and the armature winding resistance is 0.01 ©, an armature current of 100 A
results. As a consequence of the huge armature reaction associated with such a
large current, a very strong flux field ®, is produced and directed in quadrature to
the original field ®,. By placing a second set of brushes, d-d, normal to the ®,
field, a high-level voltage source is made available that can be applied to the dc
motor, which in turn provides the desired output at the mechanical load. In
practice it is customary to short-circuit the brushes q-q through a series field
which enables the required values of ®, to be established with smaller quadrature-
axis armature current. Another modification involves the inclusion of a compen-
sating winding placed in the direct-axis circuit. Its purpose is to prevent the
armature reaction flux produced by the armature current J, from diminishing the
original field ®,. The compensating winding is designed to furnish 100% compen-
sation, a condition that is essential in view of the small values that ®; assumes
during normal operation.
The transfer function of the amplidyne as an isolated unit—without its con-
nected load—is found by considering the direct-axis induced voltage Eq as the
output quantity and the control voltage V. as the input quantity. The derivation
that follows is based on the assumptions that there is zero coupling between the
flux in the direct axis and the flux in the quadrature axis and that the magnetization
curve is linear. The differential equation for the field winding is as usual

dl,
ites ud bb este al ai (13-162)

where R, = control winding resistance, 0

L. = control winding inductance, H

Expressed in transformed form, this equation becomes

IS yey 1
+ ste)
Re (13-163)
VAs)
.
where 7, = L,/R, and denotes the control winding time constant
flux ®y
The presence of the control winding current produces the direct-axis
across the q-q brushes and which specifi-
which in turn causes an emf to appear
cally is expressed as

E, = K\®1 (13-164)
of the arma-
where K;, includes winding factors as well as the speed of rotation
of Eq. (13-164) results by recalling that ®, is
ture. A more convenient form
proportional to J, so that we can write

E, = Kal- (13-165)
542 Dynamics of Electric Machines Chap. 13

where K, is a proportionality factor for the number of volts generated per control-
winding ampere. The Laplace transformed version of this expression is simply

E,(s) _ (13-166)
I.(s) :
The governing differential equation for the quadrature-axis circuit is readily
seen from Fig. 13-30 to be
dl,
BE, = Rp eles (13-167)
dt
where R, = total resistance in the quadrature axis circuit, 0
L, = total inductance in the quadrature axis circuit, H
|
> = current flowing in the quadrature axis, A
The transformed version of the last equation expressed as a ratio then be-
comes
tape 1
E,(s) Rg(1 + stq) Lette
where

Tq = zs (13-169)

Finally, with the amplidyne output circuit open it follows that the output voltage
appearing at the direct-axis brushes is simply
Ey= Kd, (13-170)
where K, denotes the number of volts induced in the direct axis per quadrature-
axis ampere. The Laplace transformed expression is

Eq(s)
= Ka (13-171)
I,(s)
The transfer function that applies for the unloaded amplidyne generator
readily follows by multiplying Eqs. (13-163), (13-166), (13-168), and (13-171).
Thus

E,(s)__ KoKa (13-172)

Viisha Ra, (1 + st F 7)

The characteristic equation is second order in this mode of operation. Clearly the
usefulness of this result is limited to those applications where the load current of
the amplidyne, J, is negligible. This is certainly not the case in the situation
depicted in Fig. 13-30 when the load switch is closed.
Sec. 13-7 Amplidyne Generator 543

Amplidyne Transfer Function with Connected Load

The transfer function for this arrangement must be found by considering the dc
motor as part of the direct-axis circuitry of the amplidyne and by this means
accounting for the loading effect. The resulting transfer function then provides
the correct description of the dynamics of the amplidyne-motor combination. For
this derivation the additional assumption is made that the output circuit is com-
the
pensated perfectly, which ensures that the load current J, has no effect on
control flux ®,.
is
A few more equations are needed to complete the analysis. The first
obtained by applying Kirchhoff’s voltage law to the direct-axis circuit. Thus
Eq = Kale = lgRa + Kuo (13-173)

speed,
where R, is the total resistance found in the direct-axis path, w is the motor
d by the dc
and K,, denotes the number of volts per radian per second generate
negligibl y
motor. For simplicity the inductance in the direct axis is assumed
becomes
small. Expressed in the Laplace transform notation this equation

Eq(s) —RiKoos) _ Iq(s) (13-174)

Iq(s) is the Laplace

Keep in mind that E,(s) is the Laplace transform of Ez,
Also, Eq. (13-174) is
transform of J;, and w(s) is the Laplace transform of w.
ze the need for a
written as indicated on the left-hand side in order to emphasi
that w(s) is treated asa
summing point in the block diagram representation. Recall
feedback quantity in the formulation of Eq. (13-174).
torque to the di-
A second required equation relates the motor-developed
transform notation
rect-axis current through the torque constant K,. In Laplace
we have
T(s) _ (13-175)
Iq(s) S
follows from writing the
Moreover, the relationship between torque and speed
torque equation as it applies to the motor shaft. Thus
dw 13-1
(13-176)
J — tak
T = Joi

to the motor shaft and F denotes

Here again J denotes the total inertia reflected
shaft. The transfer function for
the total viscous friction reflected to the motor
Eq. (13-176) thus becomes
w(s) _ 1
Ts) FU + stm) ge?)
544 Dynamics of Electric Machines Chap. 13

where 7,, = J/F. Finally, in those applications where displacement rather than
velocity is of primary interest, use is made of
eisy te
AG ak (13-178)

The total system block diagram can now be drawn by employing a succes-
sive block representation of Eqs. (13-163), (13-166), (13-168), (13-170), (13-174),
(13-175), (13-177), and (13-178). The result is depicted in Fig. 13-31. The one
feedback path exists because of the presence of the dc motor and its associated
counter-emf during running conditions. By employing Eq. (13-16) the single

Figure 13-31 Block diagram representation of the amplidyne motor system of

Fig. 13-30.

transfer function that relates motor speed to the amplidyne induced emf in the
direct axis is obtained. Accordingly,
w(s) be K,/[RaFC + stm)]
E,(s) . 1+ LK,K/RoFQ + st)l
SER SSOP,
ON) taerniyISVS
FR,{1 + (K.K,/FRa)l 1 + sz’, Balt)
where

, 1
oe) te 1 + (K.K/FR) (13-180)

The complete amplidyne-motor-system transfer function can now be written

by formulating the product of Eqs. (13-163), (13-166), (13-168), (13-170), (13-179),
and (13-178). The result is

c(s) a K,Ka K, 1
Vis) |RR, Ri (KORUER ps st ea ee ee 318)
A comparison of this expression with Eq. (13-172) reveals the precise man-
ner in which the presence of the motor modifies the transfer function. It intro-
Sec. 13-7 Amplidyne Generator 545

duces a change‘in direct transmission gain along with pure integration and a simple
time lag as represented by 7/,. If the loading effect that the motor has on the
amplidyne generator had been neglected, the result would have been a larger
mechanical time constant and a different direct gain.
The magnitudes of the amplidyne time constants 7, and 7, vary with the size
of the unit, but often they lie in a range of 0.05 s for smaller units to 0.5 s for the
larger ones. The value of 7’, of course depends upon the size of the motor and the
nature of the connected mechanical load, but it easily can be of the order of
several seconds. The use of such a combination in control systems leads to stabil-
ity problems because of the time lags associated with the composite transfer
function.
A glance at Eq. (13-181) discloses the characteristic equation of the ampli-
dyne motor system to be of fourth degree. Accordingly, there are also four state
variables and four state equations. Figure 13-32 reveals the state variables to be
the control winding current /., quadrature axis current J,, motor speed o, and
motor displacement c. The state equations in Laplace notation are found by

Figure 13-32 Alternate form of the block diagram of the amplidyne motor combi-
nation. The output of the pure integrators discloses the state variables.

of
setting the input signal to each of the four integrators equal to s times the output
the respective integrator . This leads to

1 1
sI.(s) = ee V.(s) — z I.(s) (13-182)

K 1
sleisr= Rat, ES es 5 I,(s) (13-183)

= 4
KK,
7 (5)- (2
(K,.K;,
+— 1 )
13-184
I,{s) FRaTm Tm oe) ( )
Shs) FRatm

sc(s) = w(s) (13-185)

 Dynamics of Electric Machines Chap. 13
546

The corresponding time-domain equations expressed in the matrix formulation are

1
dl
Ae hoy1 1.
dt Te : : : Rot

ai 2 0 0) |1, 0
dt Rate si
a + Vv. (13-186)
dw Kakipae (Kok: hy
dt Uc Fassel i Te stingchevtes a aA Ta :
dc
a 0 0 1 0 Cc 0

13-8 A FEEDBACK VOLTAGE CONTROL SYSTEM

To illustrate how knowledge of the steady-state and dynamic performance of a de

generator can be applied in a general way, attention is directed to the elementary
voltage control system depicted in Fig. 13-33. The purpose of this control system
is to provide a terminal voltage across the load that will remain essentially invari-
ant in spite of changes in load current. Here the load current and the armature

Io

Reference
voltage
source

Figure 13-33 Elementary feedback voltage system.

current are assumed to be the same because the current drawn by the 10-kQ
potentiometer is negligible for all practical purposes. Also, assume that the gener-
ator external characteristics are properly represented by Fig. 13-34. The curves
are linearized for simplicity. These characteristics state, e.g., that, if the genera-
tor were operating isolated and at a constant field current of J; , a 20-V drop would
occur in terminal voltage for a load demand corresponding to 20 A. This is consis-
tent with an armature circuit resistance R, of 1 for the generator. Of course the
purpose of the arrangement of Fig. 13-33 is to prevent this drop from occurring
and to maintain stability in doing so.
 Sec. 13-8 A Feedback Voltage Control System 547

Y =

re)
>

D> Itq > I¢3 > Teo > Iq

$2 200
2
8 180

Figure 13-34 Generator output charac-

Lood current, amp teristics.

Before proceeding with a description of how this automatic regulating action

is achieved, let us first calculate the reference voltage needed to make available a
specified no-load voltage. Assume the remaining system parameters are as fol-
lows:

Ka II amplifier gain constant = 2000 V/V

II generator field winding resistance = 200 0
Ry
Ky = generated emf constant = 300 V/field ampere
h = feedback factor = 0.1 numeric

The generated emf constant is valid for a specified constant prime mover speed.
The feedback factor refers to the slider-arm setting of the feedback potentiome-
ter. Assuming the desired output at no-load to be 200 V, an amplifier input signal
e, is needed which satisfies the equation
€oKa (13-187)
—— Ky = 200 V
Ry
or
€,2000 a.
700 300 = 200

Hence
€y = 15 = 0.065 V

It should be apparent that Eq. (13-187) merely identifies how the amplifier input
in
signal at no-load is operated upon to yield the desired no-load terminal voltage
amplifier output voltage that is applied to the
steady state. Thus e,K, denotes the
field circuit to produce a field current. The steady-state value of this field current
is found by dividing the field voltage by Ry, which yields eoK,/Ry. Finally, by
548 Dynamics of Electric Machines Chap. 13

multiplying this quantity by Ky the desired no-load voltage is obtained. The re-
quired reference voltage is found from the equation that indicates how the ampli-
fier input voltage is obtained. A glance at Fig. 13-33 discloses that it is the differ-
ence between the reference voltage and the portion of the output voltage that is
fed back. Expressing this mathematically we have generally
Erp —-hV=e (13-188)

Therefore for the case at hand

Er = e, + hV = 0.065 + (0.1)200 = 20.065 V (13-189)

This expression states that, if the slider arm of the potentiometer at the reference
source is set at 20.065 V, the generated output voltage at no-load will be 200 V.
Obviously, as a matter of convenience, the reference potentiometer can be cali-
brated to convey this information.
Let us now describe how the closed-loop system of Fig. 13-33 automatically
furnishes a self-correcting action as load current to the extent of 20 A is permitted
to flow. From Eq. (7-12) we know that the terminal voltage is related to the
generated emf by

Ves Epa Tole, (13-190)

Hence, for J, = 20 and R, = 1, the terminal voltage attempts to fall to a value of

180 V. However, as it moves in this direction the feedback quantity hV becomes
correspondingly smaller, which increases the amplifier input signal as indicated by
Eq. (13-188). The increased amplifier input means a greater field current, which in
turn generates an increased E, to offset the drop in terminal voltage. Complete
neutralization does not occur in this system because some small drop in terminal
voltage must be allowed in order for the corrective action to exist.
The block diagram of this system may be developed by following the proce-
dure employed in the preceding sections. The transfer function of each part of the
system is found and these are then combined to yield the total system transfer
function and block diagram. The Laplace transformed version of the equation
that provides the actuating signal for the amplifier is obtained from Eq. (13-188).
Thus

Er(s) — hV(s) = e(s) (13-191)

Since V(s) is the output variable, a summing point is used to identify this equation
in the block diagram with hV(s) coming from a feedback path. The transfer func-
tion for the amplifier is clearly just the associated gain factor. Hence

VAS) S
e(s) a (13-192)

where V;(s) denotes the amplifier output voltage which is applied to the field
circuit. Next, because of the R-L nature of the field circuit, our previous experi-
Sec. 13-8 A Feedback Voltage Control System 549

ence} tells us that field current and field voltage are related by
I/(s) ib 1
(13-193)
V/(s) Re + sz)
where ty = Ly/Ryr. Proceeding on through the system, we next note that the field
current of the last equation yields a generated emf described by the transfer
function

Ex(s) 28 (13-194)
DG)
The final step in the development is the Laplace transformed version of Eq. (13-
190), which provides the relationship between the E,(s) of the last equation and
the output quantity V(s). Thus

E,(s) a I,(s)Ra = V(s) (13-195)

A successive block interpretation of Eqs. (13-191) to (13-195) leads to the com-

plete block diagram depicted in Fig. 13-35. Note that the increase in armature
current is treated as a negative input disturbance that acts to reduce the output
voltage.
—I,(s)Rg

Figure 13-35 Block diagram for the system of Fig. 13-34.

It is important to understand that from the system block diagram it is now

possible to write conveniently expressions which describe the manner in which
the output voltage responds to changes in either the reference signal or load
reveal
disturbances or in both. Moreover, these expressions can be made to
te performan ce. To
information about dynamic behavior as well as steady-sta
illustrate, let us first write the transfer function relating output voltage to reference
input. Application of Eq. (13-16) to the block diagram yields
Vis) ee +
[KaK/RA1e stp] _ _K 1 13-196
DO
Po MURR AR I LAK ste
+See derivation of Eq. (13-57).
550 Dynamics of Electric Machines Chap. 13

where

K= MAL = direct transmission gain (13-197)

A

t= ftRR
‘DAP (13-198 )
In a similar fashion the transfer function between the output voltage and the
armature voltage drop can be shown to be given by

Vs) ems 1 poidrhiurdastiny (13-199)

—R,I,(s) 1+ [AKaK;/RpA + stp] 1+ hK 1+ sry
The complete expression for the output voltage corresponding to Ep and J, is
found by superposing the results obtained from Eqs. (13-196) and (13-199). Ac-
cordingly, we have

K 1 Rala(s) 1 + sty
V(s) =
1+ hK1 + sz} ER(s) —
1+ hK 1 + sz} Spee!

To illustrate the use of the last equation, let it be desirable to find the value of
the output voltage for a reference input of 20.065 V and zero armature current.
The Laplace transform of the reference input is given by

ER(s) = ee (13-201)

Hence

1 1 20.065
Vs). = : (13-202)
eh te aS

Keep in mind that Eq. (13-202) furnishes total information about the response,
dynamic as well as steady-state. Since for the moment our interest is focused on
the steady-state value, applying the final value theorem to Eq. (13-202) yields the
desired result. Thus

V eat
= lim sV(s) Gates
= ppg
7 20.065 =_ 357
3000 20.065 js= 200 V (13-203)
This, of course, is exactly the value used previously in connection with Er =
20.065 V. Refer to Eq. (13-189).
If, in addition to Ep, it is further assumed that a current of 20 A is delivered
to the load, then application of the final value theorem to Eq. (13-200) discloses
Chap. 13 Problems 551

that the new steady-state value is

o ys Ce 20) | ay 3
V > ars 20.065 giv ae 200 301 > 200 — 0.0655 = 199.9345 V
(13-204)

Clearly, only a very slight drop in output voltage occurs in the closed-loop sys-
tem. Note that the placement of the generator in a closed-loop system caused the
normal 20-V drop in terminal voltage to be reduced to 0.0655, a reduction by the
factor 1/(1 + hK) where hK is the loop gain factor. What is more, all of this has
been achieved automatically.
If it should be desirable to describe the dynamic behavior of the system in
response to Ep(s) and J,(s), the usual procedure must be applied. First, the appro-
priate expressions for Er(s) and /,(s) are introduced into Eq. (13-200). This is then
followed by a partial fraction expansion and subsequent evaluation of the associ-
ated coefficients. Evaluation of the inverse Laplace transforms by means of a
table of Laplace transform pairs furnishes the complete time solution for the
output voltage. In this instance the output voltage will be seen to rise in accor-
dance with the usual buildup that takes place in any system containing a single
energy-storing element. See Eq. (13-12) for example.

PROBLEMS

13-1. Obtain the transfer function relating output voltage to input voltage for each circuit
shown in Fig. P13-1. Put the result in the form that involves factors such as
(i + sz).

Ee R,

E; Re Eo

(a) (b)

Figure P13-1

as follows: L = 0.5 H,
13-2. The values of the parameters in the circuit of Fig. P13-1 are
R, = 2, and R, = 3 ©. Find the complete expressi on for the output voltage for a
unit step input voltage.
input voltage for each
13-3. Find the transfer function relating the output voltage to the
circuit depicted in Fig. P13-3. Manipula te the result into the standard form.
552 Dynamics of Electric Machines Chap. 13

(c) (d)
Figure P13-3

13-4. Refer to Fig. P13-4.

(a) Derive the transfer function of this circuit relating E, to E;.
(b) Obtain the transfer function that relates the voltage across capacitor C, to the
input voltage E;.
(c) Represent the transfer function between E, and E; by a block diagram that
brings into evidence a pure integration function in the direct transmission path.
(d) Identify the state variable for this network and indicate its location in the block
diagram.

E, e (G3

cnet sera Figure P13-4

13-5. The block diagram relating output to input of an electric circuit is shown in Fig.
P13-5.
(a) Find the transfer function between input E; and output E,.
(b) Can you suggest a circuit for which this transfer function applies? The C’s
denote capacitance and R resistance.
13-6. Refer to Fig. P13-3(d).
(a) Derive the transfer function E,(s)/E;(s).
(b) Find the transfer function relating the voltage across the capacitor to the input
‘voltage.
Chap. 13 Problems 553

E,(s)

1
C(R, + Ro)
Figure P13-5

(c) Represent the transfer function between input and output in terms of a pure
integration in the direct transmission path.
(d) Identify the state variable in the block diagram of part (c).
£; in
13-7. Determine the transfer function that relates the voltage E, to the input voltage
the system depicted in Fig. P13-7. The system parameters are as follows:

Generator: Ly = 50H, Ry = 50.0, Ra = 10,L,=1H

emf constant: K, = 100 V/field ampere

Low-pass filter: L = 1H
R=10

a aa speed | | |

ow ass filter
|
1 OE org et Lowe
Figure P13-7

field current. The motor parameters

13-8. The motor of Fig. 13-5 is operated at constant
are: R, = 0.50, L, = 0, K, = 2 V/rad/s, K, = 5 lb-ft/A, and F = 1 Ib-ft/rad/s. The
equivalent inertia is 42 lb-ft-s?.
554 Dynamics of Electric Machines Chap. 13

(a) What is the numerical expression for the transfer function that relates motor
speed in radians per second to the voltage applied to the armature circuit?
(b) If the applied armature voltage is a step of 200 V magnitude, find the steady-
state speed of the motor.
(c) Obtain the complete expression for the motor speed from the instant the voltage
of part (b) is applied to the time that virtual steady state is reached.
(d) Show the block diagram of this system, using the numerical values.
(e) Write the state equation for the system.
13-9. The application of 100 V to the armature of the machine of Prob. 13-8 at no-load
generates an output velocity of 47.6 rad/s.
(a) If a load torque of 40 lb-ft is applied to the motor shaft, find the new operating
speed.
(b) How long does it take for the new speed to be reached? Use three time con-
stants as a satisfactory indication of settling time.
13-10. Repeat Prob. 13-8, taking into account the motor armature leakage inductance.
Assume that L, = 0.05 H.
13-11. The parameters of the system of Fig. P13-11 are as follows:

K, = 0.0234 Ib-ft/A = torque constant

K,, = 0.2 V/rev/s = speed constant

K, = 0.15 V/V = amplifier gain

R, = 0:2 0

L, = 0.0

Figure P13-11
The viscous friction coefficient of the load is negligible.
(a) Develop the block diagram for this system, relating output velocity to amplifier
input.
(b) Find the expression for the closed-loop transfer function, using the numerical
values.
Chap. 13 Problems 555

(c) Write the differential equation for this system.

sys-
(d) Is there an equivalent viscous friction (i.e., damping coefficient) in this
tem? Explain.
P13-11
13-12. Find the complete solution for the motor velocity in the system of Fig.
corresponding to a unit step input to the amplifier.
of
13-13. Determine the differential equation of the system shown in Fig. P13-13 in terms
the letter symbols denoting the system parameters . Consider that the armature
take the
leakage inductance of the generator and motor are both negligible. Also,
constant as K,, in volts per field ampere, the motor speed con-
generator voltage
as K; in lb-
stant as K, in volts per radian per second, and the motor torque constant
ft per armature ampere.

Rm
Kw
(7 +
Dinpitllat Wh. aan, Shelia wan: mane feat \iskooe<e |Load|
gain, Kg (m) {toad
[/ J, F

Figure P13-13

P13-13 in terms of the state

13-14. Express the dynamical behavior of the system of Fig.
equations expressed in normal matrix form.
The generator field winding has
13-15. A speed control system is depicted in Fig. P13-15.
resistance R, and inductan ce Ly. Let R denote the armature circuit resistance of

c= Bw; O Amplifier

Bw

Figure P13-15
556 Dynamics of Electric Machines Chap. 13

both the motor and generator. The armature winding leakage inductances are negli-
gible and zero-load torque is applied. Also, the output voltage of the generator is
described by

K,ly = 1,R + Kao where K, = V/field amp

and the output speed and developed motor torque are related by

ropemT, + Fw where K, = lb-ft/arm. amp

Moreover, 8 is a tachometer constant expressed in volts per radian per second.

(a) Develop the block diagram showing the transfer function of each block.
(b) What is the characteristic equation of this system?
(c) Write the state equations in the standard matrix form.
13-16. A system is represented by the block diagram shown in Fig. P13-16.
(a) Find the expression for the closed loop transfer function.
(b) What is the characteristic equation?
(c) What influence does the parameter K, exert on the system dynamic response?

Potentiometer Amplifier Servomotor

Figure P13-16

13-17. A system is represented by the block diagram depicted in Fig. P13-17.

Potentiometer Amplifier Servomotor

Figure P13-17
Chap. 13 Problems 557

(a) Determine the closed loop transfer function.

(b) What is the expression for the characteristic equation?
(c) Write the defining differential equation for this system.
(d) What role does Ky play in determining the dynamic response?
13-18. A system is represented by the block diagram depicted in Fig. P13-18. The amplifier
section is designed to provide an output signal that contains a proportional as well
as an integral component of the input signal.
(a) Find the closed loop transfer function.
(b) What is the characteristic equation?
(c) How does the result of part (b) compare with part (b) of Probs. 13-16 and 13-17?

Potentiometer Amplifier Servomotor

Figure P13-18

13-19. In the system of Fig. P13-19 assume that the reference and feedback tachometer
generators are identical. In addition, the system has the following parameters:
K, Il= amplifier gain, V/V

Ry = generator field winding resistance, 0

Ly = generator field winding inductance, H

Ky = generator emf factor, V/field amp

R = armature circuit resistance of motor and generator

K, = motor torque constant, lb-ft/arm. amp

Kw =motor speed constant, V/rad/s

B = dc tachometer speed constant, V/rad/s

J = system inertia in consistent units

F = system viscous friction coefficient in consistent units
ng clearly the transfer
(a) Draw the complete block diagram of this system, indicati
function of each block.
(b) Identify the state variables.
(c) Write the state equations in matrix form.
system.
(d) Determine the characteristic equation of this
in the dynami c state? Explain.
(e) Are oscillations possible
558 Dynamics of Electric Machines Chap. 13

Figure P13-19

13-20. A 10-hp armature-controlled de motor drives a load whose viscous friction coeffi-
cient is 2 lb-ft/rad/s and whose angular inertia is 17.5 lb-ft-s*. The field winding is
separately excited and maintained fixed. The corresponding motor parameters are
then as follows:
R, = 0.3 }
K, = 1.0 V/rad/s
K, = 1.5 lb-ft/arm. amp

The armature winding inductance is negligible.

(a) Compute the steady-state speed corresponding to a step applied armature volt-
age of 210 V.
(b) How long does the motor take to reach within 95% of the steady-state speed of
part (a)?
(c) What is the value of the total effective viscous damping coefficient of this
motor-load configuration?
13-21. The motor of Prob. 13-20 is now used as a field-controlled dc machine in the manner
illustrated in Fig. 13-23. The armature is assumed to be energized from a constant
current source. The field winding parameters are: Rp = 50 Q and Ly = 20 H.
Moreover, the motor torque constant now expressed in lb-ft per field amperes (K,,)
is 60.
(a) Draw the block diagram.
(b) Calculate the value of steady-state speed for a step applied field voltage of
100 V.
(c) Approximately how long does it take the motor to reach within 95% of the final
speed of part (b)?
(d) Write the state equations in matrix form, using numerical values.
13-22. A two-phase servomotor has a standstill torque at full control winding voltage of 4
oz-in. Also, at full control voltage, the no-load speed is known to be 100 rad/s. The
Chap. 13 Problems 559

motor torque constant is 0.1 oz-in. per V, or 4 oz-in. per control ampere. The
control winding inductance is 6 H. Assume linear torque-speed curves and a rotor
inertia of 0.5 oz-in.?
(a) Determine the complete numerical expressions relating motor displacement to
control winding voltage.
(b) If the result of part (a) is assumed driven from an amplifier having a voltage gain
of 200, write the total expression for the direct transmission function.
(c) Place a unity feedback loop around the direct transmission function of part (b),
and find the closed loop transfer function.
(d) What is the characteristic equation of the system of part (c)?
(e) What is the characteristic equation of the system of part (b)?
13-23. An amplidyne generator is used as a power stage that drives the dc motor of Prob.
13-20. The amplidyne has a control winding resistance of 400 0, a quadrature-axis
circuit resistance of 0.1 0, and a direct-axis circuit resistance of 0.5 2. Moreover,
is
the voltage induced in the cross-axis for each milliampere of control current
direct axis for each ampere in the quadratur e
0.1 V, and the voltage induced in the
quadrature
axis is 5 V. The inductance of the control winding is 50 H and that of the
axis is 2 mH.
to amplidyne
(a) Determine the transfer function that relates motor output speed
control winding voltage.
of the motor.
(b) For a control voltage of 5 V compute the steady-state speed
the predomin ant factor in establish ing the dynamic response of this
(c) Identify
amplidyne-motor system.
tion of Prob.
A unity feedback loop is placed around the amplidyne-motor configura
that has a transducer
13-23. The feedback element is a position-sensitive device
constant of 1 volt per radian of motor displace ment.
(a) Find the closed-loop transfer function.
(b) Identify the characteristic equation.
(c) How many transient modes are there?
(d) Compute the time constant of each transient mode.
are those listed on
13-25. Refer to the system of Fig. 13-33, and assume that the parameters
p. 547.
to a reference input volt-
(a) Calculate the no-load terminal voltage corresponding
age of 30.1 V.
(b) What is the value of the closed loop gain?
corresponding to a step
(c) Determine the complete solution for the output voltage
input of 30.1 V. Assume no-load.
(d) What is the time constant of this closed loop system?
complete solution for the output
13-26. In the system configuration of Fig. 13-33 find the
the simultaneous load current
voltage for a step reference input of 30.1 V and
demand of 25 A.
shown in Fig. P13-27.
13-27. A speed control system is designed to have the configuration
ication :
The system parameters have the following identif
K, l|= amplifier gain, V/V
Ry = generator field winding resistance,

Ly = generator field winding inductance, H

560 Dynamics of Electric Machines Chap. 13

Ky = emf constant of generator, V/field amp

R = combined armature circuit resistance of generator and motor,

K, = motor torque constant, units of torque/arm. amp

J = equivalent inertia of motor and load

F = viscous friction coefficient of the load, assume negligible

is ll motor speed-voltage constant, V/rad/s

Ky = dc tachometer speed-voltage constant, V/rad/s

Develop the complete block diagram for this system, and clearly show the transfer
function for each block.

R
Reference
voltage
source Amplifier w
= e=— -_—
Load
/
| ar eT
“,

/ oe
/ !
~ n
“ constant I
'
|
|
|
I
|
eee a J
DC tach.
generator

Figure P13-27

13-28. Refer to the system of Prob. 13-27, and find the expression for the no-load steady-
state speed in terms of the system parameters for a given value of reference voltage.
13-29. Refer to the system of Prob. 13-27, and determine the expression for the final
change in speed in terms of the system parameters corresponding to an applied load
torque of 7; units of torque.
13-30. Determine the state equations in matrix form for the system of Fig. P13-27. Assume
that the system is operating in response to an applied reference voltage Er and an
applied load torque of T,;.
 Appendix
to Chapter 3

CASE
3A DEVELOPED TORQUE FOR SINUSOIDAL B AND NI: THE
OF AC MACHINES
in Eq. (3-9) has limited
The expression for electromagnetic torque as it appears
assume d constan t and all conduct ors are assumed to
application because B is
which rarely occurs in
produce equal torques in the same direction—a situation
winding is intentionally
practical machines. In fact in most ac machines the field
density, and the arma-
designed to produce an almost sinusoidally distributed flux
-conductor distribution which
ture winding is similarly arranged to yield an ampere
of the rotor structure.
is also almost sinusoidally distributed along the periphery
of an electromagnetic
For this reason we shall now consider the development
torque equation that applies for such cases.
refers to a point-by-
The armature winding ampere-conductor distribution
conductors of the armature
point plot of the current flowing through the individual of the
the instantaneous value
as depicted in Figs. 3-12(a) and (b). Note that
be sinusoidally distributed. This
current existing in the conductors is assumed to
For the 12 conductors shown
situation is readily achieved in practical machines.
the paper and the other 6 carry
in Fig. 3-12(a), 6 carry current into the plane of
ance of two poles. Actually the
current out of the paper, thus giving the appear
r distribution of the arma-
number of poles associated with the ampere-conducto

561
 562 Appendix to Chapter 3

Flux path enclosing

maximum mmf

Amp
conductor

J,amp /rad
J=uJm sina
@
LLL
ZL
a,rad
© Current sheet

(c)
mmf, A-t

fi Figure 3-12 Development of the cur-

a,rad rent sheet of sinusoidal density: (a) and
(b) show finite ampere-conductor distri-
bution; (c) current sheet representation;
(d) (d) corresponding mmf wave.

ture winding is always equal to the number of field poles produced by the field
winding.
Note too that the greater the number of conductors the more nearly sinusoi-
dal will be the ampere-conductor distribution. In fact, with a truly large number of
conductors a current sheet of sinusoidal density as shown in Fig. 3-12(c) can be
used to represent the ampere-conductor distribution. Of course this means that
the latter now takes on a purely sinusoidal variation. Moreover, the armature
winding can be considered a single sheet of one turn having the units of ampere-
turns per radian. A per-unit radian notation is used because in Fig. 3-12(c) the
distribution is no longer discrete.
A study of Fig. 3-12(a) reveals that for the ampere-conductor distribution
shown, the armature winding taken by itself looks like a solenoid, the mmf of
which is directed towards the right. For any specified armature-winding distribu-
tion the total value of this mmf is found by applying Ampere’s circuital law. The
Sec. 3A Developed Torque for Sinusoidal B and Ni 563

essence of Ampere’s circuital law is embodied in the statement that any closed
flux path can be said to exist if it encloses an mmf that is not equal to zero. Thus,
if a closed flux path is taken so that it encloses just conductor | in Fig. 3-12(a) as it
crosses the air gap twice, the associated mmf is equal to the magnitude of the
current in this single conductor. If the path is extended to embrace conductors 1
and 2, the total enclosed mmf is now the sum of the currents flowing in these two
conductors. It follows then that Ampere’s circuital law involves an integration (or
summation) of the current sheet in order to obtain the mmf characteristic of the
machine. In fact, for the closed flux path depicted in Fig. 3-12(a) showing all
conductors carrying current in the same direction (which is the same as enclosing
the current sheet over the displacement range from 0 to z radians), the maximum
mmf associated with this current sheet is given by

Ge ikimi Se we (3-49)
the
where J,, denotes the peak value of the current sheet density and # denotes
upper limit in Eq. (3-12) is left unspecifi ed, it follows that
mmf per pole pair. If the
Fig.
the ensuing integration yields the negative cosine function that is illustrated in
a plot is generated from Ampere’s circuital law, the
3-12(d). Whenever such
placed so that the areas of the mmf curve above and below
abscissa axis is always
the stator
this axis are the same, in recognition of the fact that the flux that leaves
must also be the flux that returns to the stator.
two
Because the flux path used in obtaining the result of Eq. (3-49) involves
the mmf per pole (or per air-gap crossing ) is
air gaps, it follows that

(3-50)

magnetic torque
This result will be useful later in converting the basic electro
equation to its practical forms for ac machine s.

Example 3-4
in Fig. 3-13(a). Conductors 1,
A two-pole machine has the stator construction shown
of 15 A directed out of the page. Correspondingly,
2, 3, and 4 carry a current
the page.
conductors 5 to 8 carry the same current directed into
displacement along the
(a) Draw the magnetomotive force as a function of the
clockwi se. Use Ampere ’s circuital law.
air gap starting at point A and progressing
(b) Find the maximum value of the mmf per pole.
ed by the mmf wave of
(c) Identify the location of the axis of the flux produc
part (a).
with this mmf distribution.
(d) Draw the equivalent current sheet associated
that crosses the air gap twice
Solution (a) Starting at point A, draw a closed path
d in Fig. 3-13(a). By Ampere’s
and encircles conductor 1 in the manner depicte
exist becaus e the net mmf enclosed by this
circuital law such a flux line path does
564 Appendix to Chapter 3

(d)

Figure 3-13 Determination of mmf distribution: (a) two-pole machine configura-

tion; (b) plan view of the ampere-conductor distribution; (c) mmf distribution;
(d) the equivalent current sheet.

closed path is not zero. In a practical machine conductor 1| is connected to a match-

ing conductor, which in the case of a full-pitch coil would be conductor 5. Hence
encirclement of conductor | means encirclement of the turn represented by the series
connection of conductors 1 and 5. Accordingly, the number of ampere-turns that acts
to produce such a flux line is here equal to 15 A-t. This situation is represented
graphically in Figs. 3-13(b) and (c). Observe that until a flux path starting at point A
is taken that reaches to enclose conductor 1, the enclosed mmf is zero. Just beyond
 é

Sec. 3A. Developed Torque for Sinusoidal B and Ni 565

the location of conductor 1, a step increase in mmf is assumed to occur on the

assumption that thickness of the conductor is negligible. As additional flux lines,
based at A are drawn, threading the space between conductors 1 and 2, no additional
mmf is enclosed. Hence in Fig. 3-13(c), the mmf wave stays constant at 15 A-t until
conductor 2 is reached. By drawing a closed flux path from A that just encloses
conductors 1 and 2, there occurs an additional increase in the mmf enclosed by such
apath. This is shown as the second step of the curve in Fig. 3-13(c). A continuation
of this procedure with A as the starting point, yields the resultant mmf wave of Fig.
3-13(c). Several points are worth noting. When a flux path is drawn to enclose
conductors 1 to 4, the enclosed mmf is a maximum because for such a flux path, all
the conductors carry current in the same direction. Hence this spatial integration
process carried on up to conductor 4 makes all positive contributions. As the pro-
cess is extended beyond conductor 4, the additional contributions become negative
since the newly added conductors now involve current flowing in the opposite direc-
tion. Thus if a flux path is drawn from A to include conductors 1 to 5, the net
encircled mmf is now the same as when the path is made to enclose conductors | to
3. Of course, a flux line that encircles conductors | to 8 cannot in fact exist because
the net mmf is zero.
(b) Keep in mind that the closed paths to find the mmf wave by spatial integra-
tion in part (a) were based on two crossings of the air gap, which is the situation that
prevails in practical machines by design. The double crossing, therefore, implies
that the mmf associated with the closed line integrals involves a pair of poles.
Accordingly, the mmf curve of Fig. 3-13(c) is for two poles with the mmf equally
divided. This point can be stressed by drawing a line CC’ in Fig. 3-13(c) that divides
the area between the mmf curve and line CC’ into equal positive and negative
halves. In this case, the negative half represents north-pole flux and the positive half
south-pole flux.
It is worthwhile to note that the spatial integration process could start at any
point on the stator surface. All that is required in the end is that the resulting curve
be divided equally between north- and south-pole fluxes as represented by the mmf’s
that produce these fluxes.
Once line CC’ is properly established, it follows in this case that the maximum
height of the mmf wave above (or below) line CC ‘is 30 A-t. Hence
F, = 30 A-t/pole

(c) On the assumption that the presence of the iron in the closed flux paths
produces negligible effects, the axis of the flux produced by the mmf wave is coinci-
dent with the axis of the mmf wave itself, which is along line DD’ in Fig. 3-13(c).
(d) The equivalent current sheet is depicted in Fig. 3-13(d). The distribution is
rectangular because the current carried by the conductors is the same and fixed in
magnitude. Take note of the quadrature relationship that the ampere-conductor dis-
tribution (represented by the current sheet here) bears to the mmf distribution along
the air gap of the machine.

We are now in a position to develop a basic torque formula as it applies to a

device that has a sinusoidally distributed flux density along the periphery of the air
ed
gap and an ampere-conductor distribution which for convenience is represent
 566 Appendix to Chapter 3

Figure 3-14 Configuration for develop-

ment of the torque equation, showing
the sinusoidally distributed flux density
produced by the field winding and the
current sheet, which represents the
Um Sin (oe+y) armature winding. The dashed arrows
denote relative magnitude and direction
of the electromagnetic torque devel-
oped per pole. The magnetic field and
the current sheet are stationary relative
to each other.

by a current-sheet density. Also, for generality, it is assumed that the current-

sheet density is out of phase with the B-distribution by # degrees as depicted in
Fig. 3-14. The procedure is really not much different from that which leads to
Eq. (3-9), but several modifications are needed. We can again start with the
fundamental force expression Eq. (3-7) and replace B with its sinusoidal variation
B,, sin a. Furthermore, in place of the conductor and the current flowing through
it, we now deal with a current element which is a differential strip of width da
of the current sheet. Accordingly, for i in Eq. (3-7) we may use
i= J da = J, sin (a + ) da (3-51)
The force appearing on the elemental current strip is then expressed as
F, = Bli = (By, sin a)/[Jm sin (a + wb) da] (3-52)
where the subscript e denotes elemental current strip. Of course the correspond-
ing torque is obtained by multiplying by the rotor radius r. Hence

T, = ByJmlr sin a sin (a + bh) da (3-53)

To determine the total torque developed on the current sheet beneath one
pole of the field distribution, we must take the sum of the elemental torques
produced by each element of the current sheet lying beneath the pole. In other
words, it is necessary to integrate Eq. (3-53) over z electrical radians, which is the
span of one pole. Hence
I, = yi torques on current elements beneath one pole

= ByJnlr |”sin a sin (a + p) da (3-54)

where T, denotes the torque developed per pole. Clearly for a field distribution
which has p poles the total resultant electromagnetic torque is

1a me frsin a sin (a + ) da (3-55)

Sec. 3A Developed Torque for Sinusoidal B and Ni 567

By introducing the trigonometric identity

sin a sin (a + #) = 3 cos & — 3 cos (2a + Ww) (3-56)

into Eq. (3-55) we obtain

T= pBaJalr [7§.c0s y doo — [7 cos Qa + W) da| (3-57)

The second term in brackets has a value of zero, so the final expression for the
developed electromagnetic torque is

T=7 EImBmlr COS (3-58)

Keep in mind that this equation is valid for sinusoidal distributions of the flux
density and ampere conductors (or current sheet). Also note that for the indicated
directions for the current sheet and magnetic flux, the developed torque is to the
left and that its direction can be reversed when w > 90°.
An inspection of Eq. (3-58) reveals that the torque is dependent upon the
space-displacement angle . Thus when # = 0° the best field pattern for develop-
ing torque prevails because the entire current sheet beneath a pole has a unidirec-
tional torque. When y = 45°, as depicted in Fig. 3-14, note that a portion of the
current sheet beneath a pole will produce an opposing torque, which results in a
reduced torque if all other things are equal. It should not be inferred, however,
that because yw # 0° the developed torque cannot be a maximum. In some ac
machines, operation at % # 0° causes a manifold increase in J,,, which gives a
considerable increase in developed torque in spite of a nonoptimum field pattern.
Often it is convenient to express Eq. (3-58) in terms of the total flux per pole,
such
which is the flux associated with any one pole of the flux-density curve. One
pole is shown in Fig. 3-15. But before proceeding further it is necessary to distin-
pointed out
guish between mechanical and electrical degrees. It has already been
ity curve, each pole spans 7
that, regardless of the number of poles in the flux-dens
mechanic al radians in a config-
electrical radians. Now, since there are always 27
that when the flux-dens ity curve
uration such as appears in Fig. 3-12(a), it follows
are the same. However,
has two poles the mechanical and electrical degrees
of mechanical
when the flux-density curve has four poles, then clearly the number
For a p-pole flux-density
degrees is one-half the number of electrical degrees.

“takes
Tate,
| Figure 3-15 Relationships for comput-
fees Tp =Lmt = oar ing flux per pole.
568 Appendix to Chapter 3

curve the relationship is

Qn = —a (3-59)
Pp

where a,, denotes mechanical degrees and a denotes electrical degrees. There-
fore, in Fig. 3-15 the pole pitch expressed in meters is
2 (3-60)
Tp = pole pitch = a,r = : ar

where r is the radius of the rotor in meters. Moreover, for a rotor axial length of /
the area associated with a pole of the flux-density curve is

A, tts = =al (3-61)

A further simplification in the computation leading to the expression for the

flux per pole results upon recalling that the average value of the positive (or
negative) half of a sine wave is

By =e Bn (3-62)

Thus the sine wave of Fig. 3-15 may be replaced with a rectangular wave having a
height of Bz, and spanning the entire pole pitch as indicated by the broken line in
Fig. 3-15. The expression for the flux per pole then follows from
® = BA, (3-63)
Inserting Eqs. (3-61) and (3-62) into Eq. (3-63) yields
4
== Bult (3-64)
Again keep in mind that this result is valid for sinusoidal variations of flux density.
By means of Eq. (3-64) the equation for the developed electromagnetic
torque may be expressed in terms of the flux per pole. Insertion of Eq. (3-64) into
Eq. (3-58) yields
T
T= 8 p?®J, Cos W N-m (3-65)

A study of Eq. (3-65) points out more clearly the three conditions that must
be satisfied for the development of torque in conventional electromechanical en-
ergy-conversion devices. There must be a field distribution represented by ®.
There must be an ampere-conductor distribution represented by J,,. Finally,
there must exist a favorable spa¢e-displacement angle between the two distribu-
 Sec. 3B Torque for Nonsinusoidal B and Uniform Ampere-Conductor 569

tions. Note that it is possible to have both a field and an ampere-conductor

distribution and still the torque can be zero if the field pattern is such that equal
and opposite torques are developed by the conductors, i.e., p = 90°.
It is also worthwhile to note that no really new concepts have been used to
derive the torque formula of Eq. (3-65). It is merely an extension of Ampere’s
law.

3B TORQUE FOR NONSINUSOIDAL B AND UNIFORM

AMPERE-CONDUCTOR DISTRIBUTION: THE CASE OF
DC MACHINES

The typical distributions are depicted in Fig. 3-16. Again for generality the cur-
rent sheet that is representative of the ampere-conductor distribution is shown
displaced in phase by the angle . These distributions are commonly found in dc
machines, as is described in Sec. 7-3. The starting point of the derivation is
Ampere’s law, again as expressed in Eq. (3-7). Of course modifications are
winding
needed consistent with the assumed distributions. Because the armature
with an elemental strip of the current
is represented by a current sheet, we deal
and the current it carries. Thus in
sheet rather than with a particular conductor
Eq. (3-7) we introduce
i =.J da (3-66)

and
Ba= B, (3-67)
the air gap. An
where B, denotes the nonsinusoidal variation of flux density along
to obtain, but this is not disturbing
exact analytical expression for B, is difficult
which is readily availabl e. Hence
because one deals rather with the flux per pole,
of the current sheet is
the force developed by an elemental strip
F, = B,|J da (3-68)

The corresponding torque for a p-pole machine is then

T, = pJ Balrda = N-m (3-69)

Figure 3-16 Current-sheet and flux-

density curves characteristic of de
machines.
570 Appendix to Chapter 3

The total resultant torque is obtained upon integrating Eq. (3-69) over one pole
pitch. Thus

T= pi|{"" Bulrda — [7 Balr da (3-70)

The second term in brackets carries the minus sign to account for the fact that J is
negative beneath the positive pole of the B, curve in the region from 7 — yw to 7.
Clearly, without an analytical expression for B,, a closed-form solution for the
torque is not readily available. However, in electromechanical energy-conversion
devices where these distributions apply, the designer by intention makes w = 0.
Unlike the preceding case this ensures that no conductors contribute negative
torques beneath the poles. Introducing & = 0 simplifies the expression for the
electromagnetic torque to

Tony ivBos Rn (3-71)

Now in the interest of avoiding B, we can express the torque in terms of the
flux per pole—a quantity which is available from other considerations. In terms
of B, the flux per pole ® is

= |"B, dA, (3-72)

But the area per pole can be expressed as

A, = Iran = Irea (3-73)

Hence Eq. (3-72) becomes

7 2 2 [a
ta) Balr=da = = [°Balr da (3-74)
Therefore

[°Bulr da =%2 @
a AS
(3-75)
Substituting Eq. (3-75) into Eq. (3-71) yields the final form for the electromagnetic
torque developed when the distributions are as depicted in Fig. 3-16 with # = 0.
Thus

T == JOo N-m (3-76)

It is worthwhile to note again that electromagnetic torque is dependent upon

a field distribution represented by ®, an ampere-conductor distribution repre-
Sec. 3B Torque for Nonsinusoidal B and Uniform Ampere-Conductor 571

sented by J, and the angular displacement between the two distributions. Of

course ws does not appear in Eq. (3-76) because for simplicity the current sheet was
assumed in phase with the flux-density curve, thereby ensuring that all parts of the
current sheet experience a unidirectional torque. A glance at Fig. 3-16 should
make it apparent that for # = 90° the resultant electromagnetic torque is zero.
Example 3-5
A p-pole machine has the sinusoidal field distribution B, sin (a — #), where a and ws
are in electrical degrees. The armature consists of a uniform current sheet of value J,
which has associated with it a triangular mmf having an amplitude per pole of Fy.
Refer to Fig. 3-17(a). The machine has an axial length / and a radius r.
(a) Determine the expression for the maximum flux/pole.
(b) Obtain the expression for the developed electromagnetic torque in terms of
B,, J, and A.
(c) Express the torque found in part (b) in terms of F, and the flux per pole ®.
Assume that A = 77/2.
(d) Is the torque found in part (b) constant at every point in the air gap?
Explain.

B=B, sin (a-y)

(b)
Figure 3-17 Configuration for Example 3-5: (a) relationship existing between flux
one pole and the
density, mmf, and current sheet; (b) a view of the rotor showing
differential arc length r da,,.
572 Appendix to Chapter 3

Solution (a) The expression for the differential flux is given by

db =BdA

where dA denotes a differential area beneath the flux density curve. This area is
given by the product of the effective axial length / and the arc length, which is
denoted by r da,,. Here use is made of the mechanical angle a,, because we are
dealing with a physical differential arc length. See Fig. 3-17(b). By using Eq. (3-59),
we can express the differential arc length in terms of electrical degrees as (2r/p) da.
Accordingly, the differential flux can then be written as (for y = 0 for convenience in
this part)

d bp
_ 2IrB
3 daa =
2IrB
——*
5 sinaa d da

The total flux per pole can now be determined by integrating the last expression over
one pole pitch which is equal to z electrical radians. Thus

@, = Bel I sina do = “Bot

Note that this result entirely agrees with Eq. (3-64).
(b) Let i denote the differential current element, which for a uniform current
sheet of value J can be written as

t= Jida

Associated with this differential current element under the influence of the pole flux
density is a differential force Bli = BJI da. When acting on a torque arm of the
machine radius r, the differential torque expression becomes
BJlr da

By integrating this quantity over one pole pitch (i.e., a electrical radians), the net
torque produced by the interaction of the field distribution and the current sheet is
determined for one pole. The total torque results upon multiplying by the number of
poles. Thus the expression for the total differential torque is
aT = p(BJlr da)

Or the total torque produced becomes

T = plrJ |"B da = plrJB, |"sin (a — W) da

Upon integrating, inserting the limits and simplifying, the result is

T = 2B, Jr cos w
Because w& and A are complementary angles, we get

w= 90°—A
which when introduced into the preceding equation and simplified yields

‘ "T = 2plrB,J sin A

 Sec. 3C The General Torque Equation 573

(c) For A = 7/2 the last equation simplifies to

T = p2B,IrJ

Algebraic manipulation of this last expression leads to expressions which can be

identified as the flux per pole and the amplitude of the mmf per pole. The flux per
pole is readily identified by rearranging the last expression to read as follows:

Pe Dpnt (-
; B,lr )deg
4 df Aj
> =P ®, 5

Clearly, the quantity in parentheses is the flux per pole.

To get F, into this formulation, it is necessary to recognize that when A = 7/2,
the total mmf per pair of poles is given by [See Fig. 3-11(a)]
F=T7J

Hence the amplitude per pole is then

Fy, =

Accordingly, by rewriting the torque equation in the further manipulated form

f
hit
Pio ale)
ye cag ES
2

We get finally

T hp:
= *~ ®,F

is
(d) No. Although the current sheet is of constant value, the flux density
, the instantan eous value of torque varies with the value of B.
not. Therefore

3C THE GENERAL TORQUE EQUATION

in Secs. 3-1, 3-4, and
The formulas for the electromagnetic torque that are derived
part of an electric ma-
3A apply in situations where the field distribution in one
the other part. Torque
chine interacts with the ampere-conductor distribution in
existing between the
is produced in the machine because of the mutual action
As long as cur-
windings to bring about an alignment of their associated fields.
exists a torque
rents exist in each winding and the field pattern is proper, there
the assumpt ion was
owing to this mutual coupling. Throughout the development
are machine s in which
made that currents exist in each winding. However, there
to produce torque when only one part of the machine is excited.
it is possible
of Sec. 1-10, where
Surely this statement is not so surprising in light of the results
in a geometrical configu-
it was shown that a force can exist whenever a flux flows
position. Refer to Eq.
ration characterized by a reluctance that varies with
recognize that the torque
(1-55). What is important to us here, however, is to
574 Appendix to Chapter 3

associated with reluctance variation cannot be computed from the results derived
in Secs. 3-1 and 3A. It seems worthwhile therefore at this point in our study to
obtain a general expression for the torque developed by an electromechanical
energy-conversion device—one that permits calculation of the reluctance torque
as well as the electromagnetic torque of the individually excited, mutually cou-
pled windings.
The general torque equation can be derived by analyzing the configuration
shown in Fig. 3-18. A current i, is assumed to flow in the rotor winding producing
the air-gap flux field. The current i, is assumed to flow in the stator winding. This
current may be direct current, single-phase alternating current, or the phase cur-
rent of a polyphase system applied to a polyphase stator winding; in any event,
‘irrespective of the form of i, it is assumed that the stator current is responsible for
the ampere-conductor distribution shown. Next consider that the rotor is allowed
to move by some differential angle da. In turn this causes a change in coupling
between the rotor field distribution and the stator ampere-conductor distribution.
Moreover, the induced emf’s of both windings will be affected, which will influ-
ence the energy input. If the input energy originates solely with the electrical
sources, it follows that corresponding to the displacement da the change in input
energy is described by
dw. = ely at + e217 dt (3-77)

But by Faraday’s law |

_ dy
ening; and _ dh
2 eae (3-78)

Figure 3-18 Geometrical configuration of an electric machine used in developing

the general torque equation. Current i; flows in the rotor winding and current i,
flows in the stator winding; together these produce the electromagnetic torque.
The nonuniform air gap gives rise to a reluctance torque.
Sec. 3C The General Torque Equation 575

where ), and A; denote the total flux linkages of the rotor and stator windings
respectively. Specifically, these quantities may be expressed in terms of the self-
and mutual inductances provided that nonlinearities and hysteresis are ne-
glected. In large measure this is justified by the presence of air gaps. Thus
Ny =. Ly ap M212 (3-79)

2 = Mi, AF Lyi (3-80)

Here L,, is the self-inductance of the rotor winding, Lx is the self-inductance of

the stator winding, and M,, is the mutual inductance between rotor and stator
windings. The positive sign is used for M2 because / and i, produce fluxes in the
same direction in the configuration of Fig. 3-18.
Upon substituting Eqs. (3-78) in Eq. (3-77) there results

dW. => i dy ct I> dnd (3-81)

This expression emphasizes that any change in electrical input energy comes
condi-
about through changes in flux linkages. Furthermore, the energy balance
in electrical input energy be equal to the change in
tion requires that this change
field energy or the mechanic al work done by the field or both. Mathemat ically
expressed,
dW, = dW; + T da (3-82)
of the rotor dis-
where T denotes the total torque developed as a consequence
the currents %
placement da and W, represents the field energy associated with
excited configur ation of Fig. 3-18 the express ion for this
and i). For the doubly
field energy is given by
W, = 4Lnii + $Lmi3 + Misin (3-83)
nature of the self-
A word is in order at this point concerning the variational
cement angle a. In accor-
and mutual inductances as a function of the rotor displa
express the self-in ductanc e of the rotor as
dance with Eq. (1-40) we can

=Ni
Lu =% (3-84)
3-84

rotor winding and %, denotes the

where N, denotes the number of turns of the
windin g. It is interesting to note here
reluctance of the flux produced by the rotor
the rotor the reluctance R, remains
that irrespective of the position taken by for
located rotor structure. Hence
invariant for a symmetrical and concentrically
this geometry Lj; is a constant.
is different. Although a simi-
However, for the stator winding, the situation
may be written, i.e.,
lar expression for the stator self-inductance
N3 (3-85)
Ly = Ry
576 Appendix to Chapter 3

examination of the geometry of Fig. 3-18 indicates that the reluctance 2%, varies
with the rotor position. This is readily understood by considering for a moment
that i, is made to flow alone (i; assumed zero) in the stator winding to produce the
ampere-conductor distribution depicted in Fig. 3-18. The stator thus behaves like
a solenoid producing flux on an axis coincident with the horizontal. Clearly, then,
when the rotor axis lies along this horizontal axis (i.e., a = 0 degrees), the
reluctance seen by this flux will be a minimum because the magnetic path consists
of two small air gaps and a large amount of high-permeability iron. This path is
called the direct axis of the machine and leads to a minimum reluctance Ry. As
the rotor iron is displaced towards a = 77/2, the reluctance increases because the
stator flux is still horizontally directed but the rotor iron is advancing towards the
vertical position as a increases which increases the air gap on the horizontal axis.
At a = 77/2 the reluctance seen by the stator-produced flux is a maximum because
it now consists of two very large air gaps. The rotor alignment with respect to the
stator-produced flux field is now said to be in its quadrature axis. Call this maxi-
mum value of reluctance &,. As a result of the shaping of the pole pieces and the
distributed nature of the stator winding,t the reluctance variation %, is a cosin-
usoidal function of a. Since for each full revolution of the rotor structure there are
two positions each of minimum and maximum reluctance, it follows that Ly. is a
double frequency function of a. The complete variation of the reluctance is de-
picted graphically in Fig. 3-19. The analytical expression for this reluctance is
given by

sa ala! Na at Si cos 2a (3-86)

a 2 2
Insertion of this expression into Eq. (3-85) yields the corresponding equation that
describes the variation of the stator self-inductance Ly) with rotor position.
There now finally remains to be determined the manner in which the mutual
inductance M,, varies with rotor displacement, if at all. A simple and direct way
to obtain this information is to perform a straightforward test on the machine of
Fig. 3-18. Apply a fixed sinusoidal voltage to the rotor winding. Remove the
source V> from the stator winding, and replace it with a suitable rms-reading
voltmeter. The voltmeter reads the voltage induced through the mutual coupling
with the changing current in the rotor winding. Specifically, this induced emf is
related to the rotor current by
di,
eo = My (3-87)
dt
For a sinusoidal variation of i; the magnitude of emf induced in the stator winding,
E,, by the constant rms rotor current J; is

Fy = oMyI; (3-88)

+The stator winding is not concentrated as a coil but rather is distributed in slots along the entire
inner periphery of the stator structure.
 Sec. 3C The General Torque Equation 577

Reluctance

Figure 3-19 Variations of reluctance

as a function of rotor displacement, as
viewed from the stator winding. To a
O z 7 37 27
different scale this plot also represents
o, electrical degrees _ the variation of Ly, versus a.

or

Mp = = iE ainax
= “aeCOS:= (3-89)
The value of E, in this equation is directly a cosinusoidal function of the coupling
between the rotor flux field and the stator winding. A glance at Fig. 3-18 shows
that at a = 0 degrees the rotor flux has maximum linkage with the stator winding,
which registers a maximum reading on the voltmeter, Ermax - At a = w/2 the net
flux linkage is zero (ideally). Hence E,is zero. At a = 7 the flux linking the stator
winding is once again a maximum, but because the flux field is reversed the stator-
induced emf undergoes a reversed polarity. Hence the mutual inductance is at its
negative maximum value. The variation of M,7 between its extreme values occurs
in accordance with a cosinusoidal function owing to the inherent construction
features of the machine. Appearing in Fig. 3-20 is the complete variation of the
mutual inductance for one full revolution of the rotor. Note that, unlike the stator
self-inductance, the mutual inductance has a fundamental variation with the rotor
displacement. If the maximum value of the mutual inductance is called M, then

Mutual
inductance

Mi2

Figure 3-20 Variation of mutual induc-

tance between the rotor and stator
windings of the machine depicted in
Fig. 3-18.
578 Appendix to Chapter 3

The foregoing results demonstrate that, for the geometrical configuration of

Fig. 3-18, 1, is independent of rotor displacement while Ly and Mj). are both
functions of a. Keeping these facts in mind, we can now return to Eq. (3-81) and
compute the change in the electrical energy input in terms of the self- and mutual
inductances. Inserting Eqs. (3-79) and (3-80) into Eq. (3-81) leads to

dW, = i, d(Liiy + Myi2) + bp (Mii + Lxir) (3-91)

Since Lj; is a constant, its differential is zero. This is not so for L2 and M).
Accordingly, Eq. (3-91) becomes
dw. = Ly di, tr 11M) dir = Ilo dM ap bM)2 di, as ih dM)

+ inLy diz + re dLy (3-92)

Also, the change that takes place in the field energy is readily found by differenti-
ating Eq. (3-83). Thus

dW, = Ly di, + bL» dip aU 113 dLy at Mi, diz = Mir2i\ dir

+ Myir di, + ijiz2 M12 (3-93)

When Eas. (3-92) and (3-93) are introduced into Eq. (3-82)—the energy balance
equation—all the terms of dW, that involve differential currents have matching
terms in the expression for dW; and so cancel out. This is by way of stating that
differential changes in the currents have no bearing on the development of me-
chanical forces. The remaining terms in the energy balance equation are then as
follows:
3 dLy + iin My = T da + 413 dLy + nin Mr (3-94)
The left side of Eq. (3-94) represents the change in electrical input energy corre-
sponding to the rotor displacement da. The last two terms on the right side denote
the change in the field energy. The remainder of the input energy goes to do
mechanical work as represented by T da. It follows from Eq. (3-94) that the
desired general expression for the developed torque in the machine of Fig. 3-18 is

(3-95)

This is the expression for the instantaneous total developed torque.{ It depends

+A little thought should make it apparent that, if the machine configuration were such that Lj
were also a function of a, the general torque equation would become

1 ,dIy;', 1 . dL . dM
‘iredibada
1 <2 eda eaae 628)
Sec. 3C The General Torque Equation 579

upon the instantaneous value of i; and i, as well as on the change of the stator self-
inductance and the mutual inductance with rotor position.
A study of Eq. (3-94) leads to a useful alternative form of the general torque
equation. Note that, if i, and i, are held constant, the change in field energy is
exactly equal to half the change in electrical energy input. Therefore the mechani-
cal work done is equal to the change in field energy. Expressed mathematically,
this is

0a i,,2 const

The partial derivative is employed here to emphasize that i, and i, are held con-
stant in this operation. Equation (3-96) can be useful in several ways. Once par-
ticular application is to reveal that the positions of maximum or minimum field
energy corresponding to zero developed torque.
Finally, using Eq. (3-95), it can be shown that the torque of mutual coupling
can be expressed in the following two alternative forms

T= 5 POF, sin A = 5p?®Jm cos (3-97)

where

A=i-y (3-98)
2

Moreover, the associated reluctance torque becomes

(3-99)
2 ee
Fein
4 Ww X4gXq
usinirh
where
Nj N3
= Lg =o R, and tq = why = oR (3-100)

Example 3-6
illustrated
Two coils are mounted in a machine with a uniform air gap in the manner
in Fig. 3-21. The self- and mutual inductanc es are given by

L,;=5H M=3 cosa

L,=2H

at an angle a from
The winding resistances are negligible. Coil 2-2’ is initially placed
1-1’ is made to carry a current of 10 sin wt, where w is the frequency of
coil 1-1’. Coil
the voltage source that energizes coil 1-1’.
with a direct
(a) Find the expression for the torque when coil 2-2’ is supplied
ment of the rotor
current of —5 A from an ideal current source. Will a displace
occur? Explain.
580 Appendix to Chapter 3

Figure 3-21 Configuration for Exam-

ple 3-6.

(b) Now assume that the current that. flows through coil 2-2’ is —S sin ot.
Repeat part (a).
(c) At what value of @ will the torque be a maximum?

Solution (a) Equation (3-95a) may be applied here. However, note that dL,/dt and
dL,/dt are zero because L, and L, are given as constants. Hence the electromagnetic
torque is attributable solely to the change in mutual inductance. Thus

T = hb vy ll (10 sin wt)(—S) a (3 cos a)

150 sin @ sin wt

Clearly, the average torque over one cycle of the source is zero and so no rotation
occurs in this case. This statement is based on the assumption that the source
frequency is rapid enough so that the rotor cannot make an instantaneous response.
(b) In this case we get
T = (10 sin wt)(—5 sin wt)(—3 sin a)

= 150 sin @ sin? ot

Introducing the trigonometric substitution
sin? wt = 4 — 4. cos 2wt
then yields
T = 75 sin a — 75 cos 2ot

In this case, there does exist an average torque which is different from zero. In fact,
specifically it is equal to the first term on the right side of the last equation. Of
course, the second term is zero over one cycle of the source frequency.
(c) A glance at the last equation makes it obvious that the maximum torque
occurs at a = 90°. Accordingly, when coil 2-2’ is energized while at position a@ in Fig.
3-21, the rotor will rotate in a:clockwise position until it reaches a position where
a = 90°, provided that a restraining force in excess of 75 sin a does not appear on the
rotor shaft.
Problems 581

PROBLEMS (cont.)

3-19. A two-pole machine is equipped with a total of 12 conductors uniformly distributed in

slots along the inner surface of the stator periphery. The conductors are paired so
that conductor 1 is series connected to conductor 7 by an appropriate end connec-
tion. Similarly, conductor 2 is joined to conductor 8 and so on until all conductors
are series connected and energized by a dc source which forces the flow of a 10-A
current.
(a) Draw the mmf distribution as a function of displacement along the air gap.
(b) Determine the number of A-t per pole.
(c) Draw the equivalent current sheet of the ampere-conductor distribution.
3-20. An electric motor is designed to have a triangular field distribution as shown in Fig.
P3-20. The ampere-conductor distribution is uniform at value J. Assume that the
machine has p poles. A rotor radius of r meters and air axial length of / meters.
(a) Find the expression for the flux per pole.
(b) Derive the equation for the developed torque in terms of the flux per pole, , and
the peak mmf per pole %,.

a,electrical
degrees

Figure P3-20

3-21. A p-pole machine has the sinusoidal current sheet variation and constant field distri-
bution depicted in Fig. P3-21. The machine has an axial length / and radius r.

Jm sin (a + w)

Figure P3-21
 Appendix to Chapter 3
582

(a) Find an expression for the maximum flux per pole.

torque in terms of
(b) Determine the expression for the developed electromagnetic
B,J,, and A.
(c) Is the torque constant at every point along the air gap? Explain.
gap as shown in Fig.
3-22. Two windings are mounted in a machine with a uniform air
P3-22. The self- and mutual inductances are

Li = 2.2H My, = V2 cos a H

Ly = 1.0H

uited with
The winding resistances are negligible. When winding 2-2’ is short-circ
applied to winding 1-1’, the current in winding 1-1’ is known to be i; = 10 V2
voltage
sin wt, where w is the frequency of the source.
torque, in
(a) If the rotor is stationary, derive an expression for the instantaneous
newton-meters, in terms of the space angle a.
(b) Compute the average torque in newton-meters when a = 45°.
tend to
(c) If the rotor is permitted to move, will it rotate continuously or will it
come to rest? If the latter, at what value of a?

Figure P3-22

3-23. In the machine configuration of Fig. P3-23, the self- and mutual induciances have the
following identification:
Ni 2NiN.
iat = Mn = re ee

2
ane
where N;, denotes the stator turns and N2 denotes the rotor turns. The machine is
equipped with a uniform air gap. Through use of the rate of change of flux linkage
with angular displacement, the torque is found to be
A dd 12
is
~ ele
Problems 583

Figure P3-23

Moreover, through the use of the Biri relationship, the expression for developed
torque is found to be T = 2N,N>B2Iri,. Finally, from the conservation of energy, the
torque equation can be written as
1 ,dL 1 ,dL .. dM
NE Pease mb ais hatte
. 2! da 22 da

Show that the torques computed by all three methods are equal.
. In the configuration of Fig. P3-23 let a = wot. By means of the flux linkage relation-
ship the emf induced in winding 1-1’ can be expressed by e; = ddj2/dt. Moreover,
through the use of the Biv relationship the emf equation becomes e; = 2N,N2Bzlv.
Finally, the expression for the speed voltage can be written as
uses dL\, : dM2 is (idL i s oe)
61s Ny, Hin Sip des + dae

Show that the speed voltages computed by all three methods are equal.
7;
3-25. In the machine configuration of Fig. P3-23 the positive directions for the currents
and i, are defined by the dots and crosses appearing in the diagram. If both currents
50°, as
are of constant magnitude with i,(t) = J and i,(t) = —(N,/N>)I, and if a =
same
depicted, in what direction is the torque on the rotor? Explain. With the
energy to be stored in the
currents, what should be the value of a for maximum
magnetic field? Explain.
3-26. In the idealized uniform air-gap machine of Fig. P3-26, excitation of either winding
Coil 2 lies at
produces a radial field. Coil 1 lies at radius r, from the axis of rotation.
radius r> from the axis of rotation. Given that

T = 2NBalryij

T = 2N2B Iroin

density
where B, is the flux density at coil 2 produced by ( in coil 1, and By; is the flux
at coil 1 produced by i, in coil 2.
(a) Show that 7; and T> have equal magnitudes.
(b) Show that 7; and 7, act in opposite directions.
584 Appendix to Chapter 3

Figure P3-26

3-27. (a) Prove the validity of Eq. (3-83) for the doubly excited configuration.
(b) Demonstrate how Eq. (3-88) follows from Eq. (3-87).
3-28. In the machine configuration of Fig. 3-18 the rotor winding is unexcited and the stator
winding is energized with the direct current Ig.
(a) Can a reluctance torque exist? Why?
(b) At what rotor speed, if any?
(c) Derive an expression for the reluctance torque in terms of the direct-axis self-
inductance L, and the quadrature-axis self-inductance L,.
3-29. The machine configuration of Prob. P3-23 is doubly excited with currents i; and i.
Are the positions in space at which the maximum torque occurs affected by the
values of i, and i,? Explain.
3-30. A reluctance torque motor has the configuration of Fig. 3-18 with a sinusoidal reluc-
tance variation along the periphery of the air gap. The motor is energized from a
120-V, 60-Hz source. With the rotor fixed in the minimum reluctance position, a line
ammeter reads 0.5 A rms. In the maximum reluctance position it reads 2 A.
(a) Determine the power developed in watts when the motor is operating with a
power angle of 10 electrical degrees.
(b) What is the speed of rotation of the rotor?
(c) What is the maximum developed power in watts for this motor?
3-31. A doubly excited system with a nonuniform air gap has inductances described by
Ly, = 6 + 1.5 cos 2a Ly = 4 + cos 2a M=6cosa

The coils are excited by dc currents with 7, = 0.5 A and i, = 0.4 A.

(a) Determine the developed torque as a function of a.
(b) Find the stored energy as a function of a.
(c) For a = 30° find the magnitude and direction of the torque on the rotor.
3-32. The dynamometer (moving-coil) instrument is distinguished from other measuring
instruments by its linear scale. In construction the moving coil is restrained by a
linear spring to have limited movement in flux field of constant value produced by a
permanent magnet. Derive an expression to show that the deflection of an indicator
needle attached to the moving coil is directly proportional to the coil current.
3-33. A singly excited reluctance motor has a variation of reluctance over the periphery of
the air gap that is as depicted in Fig. P3-33(a) for an assumed power angle displace-
Problems 585

(b)

Figure P3-33

P3-33(b).
ment A. At the same time the variation of flux with time is described in Fig.
the cooperatio n of the
(a) Can a net torque different from zero be produced through
given flux and reluctance variation? Explain.
(b) Sketch the variations of the instantaneous torque for A = 30°.
eous varia-
(c) Compute the average value of torque associated with this instantan
tion.
ctances can
A machine configuration is such that the variation of mutual and self-indu
be described by
My =1- sine H

om |=1+ sina H

Ly» = 2(1 + sin a) H

the current in coil 1 is
(a) Find the value and direction of the developed torque when
that in coil 2 is fixed at —20 A. Take
maintained constant at 10 amperes and
a = 60°.
energy supplied by each
(b) For the condition of part (a), compute the amount of
electrical source.
al current of 10 amperes
(c) If the current of coil 1 of part (a) is changed to a sinusoid
find the rms value of coil 2 current. Assume a
rms and coil 2 is short-circuited,
60-Hz, 120-V source.
produced in part (c).
(d) Determine the expression for the instantaneous torque
in part (c)?
(e) What is the average value of the torque
 Appendices

A. UNITS AND CONVERSION FACTORS

TABLE A-1 UNITS

Quantity Symbol Name of unit Dimension

Fundamental
Length IE meter ib
Mass m, M kilogram M
Time t second T
Current bal ampere I
Mechanical
Force F newton METS
Torque ip newton-meter ML>T~?
Angular displacement 0 radian =
Velocity v meter/second Lig!
Angular velocity @ radian/second 7
Acceleration a meter/second? LT
Angular acceleration a radian/second? 1
Spring constant (translation) K newton/meter MT~2
Spring constant (rotational) K newton-meter ML?T~?

586
Sec. A Units and Conversion Factors 587

TABLE A-1 CONTINUED

Damping coefficient (transla- D, F newton-second/meter MT"!

tional)
Damping coefficient (rotational) D newton-meter-second ML*T~!
Moment of inertia J kilogram-meter? ML?
Energy Ww joule (watt-second) ML*T~?
Power [2 watt ML’T3
Electrical
Charge q, Q coulomb TI
Electric potential v, V, E volt ML?>T =I"!
Electric field intensity £ volt/meter (or MLT I"!
newton/coulomb)
Electric flux density D coulomb/meter? pet |
Electric flux Q coulomb TI
Resistance R ohm METI
Resistivity p ohm-meter ML>T3] ~*
Capacitance C farad Mie saa ole
Permittivity € farad/meter WP YE ANSE
Magnetic :
Magnetomotive force F ampere(-turn) I
Magnetic field intensity H ampere(-turn)/meter Gayl
Magnetic flux 0) weber ML?T~]-!
Magnetic flux density B tesla MT"!
Magnetic flux linkages nN weber-turn MEAT Al
Inductance ip henry MAGA
Permeability 1 henry/meter MD 202
Reluctance R ampere/weber MIL TI2

TABLE A-2 CONVERSION FACTORS

I

Quantity Multiply number of by to obtain

e
_ EE a ee S
Length meters 100 centimeters
meters 39.37 inches
meters 3.281 feet
inches 0.0254 meters
inches 2.54 centimeters
feet 0.3048 meters

Force newtons 0.2248 pounds

newtons 10° dynes
pounds 4.45 newtons
pounds 4.45 x 10° dynes
dynes 10 newtons
dynes 2.248 x 10-6 pounds
newton-meters 0.7376 pound-feet
Torque
newton-meters 10’ dyne-centimeters
pound-feet 1.356 newton-meters
dyne-centimeters 10-7 newton-meters
 588 Appendices

TABLE A-2 CONTINUED

a eee LEUUEEENEnEEEEESSEESSE EE

Energy joules (watt-seconds) 0.7376 foot-pounds

joules 2.778 < 10-7 kilowatt-hours
joules 10’ ergs
joules 9.480 x 10-4 British thermal units
foot-pounds 1.356 joules
joules 1.6 x 10°” electron-volts
Power watts 0.7376 foot-pounds/second
watts 1.341 x 10-3 horsepower
horsepower 745.7 watts
horsepower 0.7457 kilowatts
foot-pounds/second 1.356 watts

B. PITCH AND DISTRIBUTION FACTORS FOR MACHINE WINDINGS

In the interest of making more effective use of the copper and iron as well as to
discriminate severely against the presence of space harmonics in the air-gap field
distribution, the windings of ac machines are intentionally distributed and pitched
so that all the phase turns do not fully link the total air-gap flux. In a three-phase
machine, the distribution of the winding is such that each phase is made to occupy
a region of 60 electrical degrees beneath each pole. Moreover, each phase coil is
made to span less than 180 electrical degrees; i.e., the span is less than a pole
pitch. Such a coil is referred to as a fractional-pitch coil. Although the use of the
fractional-pitch coil means some reduction in the value of emf induced by the
fundamental component of the field, the attenuation of the space harmonics is so
much greater that it easily justifies the cost.

The Pitch Factor

This factor serves as a measure of the effectiveness of a selected coil pitch in

reducing the influence of field harmonics on the emf induced in the armature
winding of the machine. To understand how this comes about, attention is di-
rected first to the reduction caused in the fundamental value of the induced emf
compared to the case of a full-pitch coil. Refer to Fig. B-1, and assume that the
coil consists of the two coil sides 1 and 2 which are moving toward the left relative
to the stationary north- and south-pole fundamental field component. The voltage
induced in coil side 1 is therefore out of the paper, and that induced in 2 is into the
paper. The instantaneous induced emf in these coil sides can conveniently be
represented graphically by means of Fig. B-2. The voltages E; and E, denote the
maximum values of emf associated with each coil side. Accordingly, the instanta-
neous values are found by obtaining the projection of these peak voltages on the
Sec. B Pitch and Distribution Factors for Machine Windings 589

he
Figure B-1_ Fractional-pitch coil in the
Voom fundamental field.

Reference line

Figure B-2 Phasor diagram for the

fractional-pitch coil.

each coil side has the same peak voltage.

vertical reference line. Of course,
1 is instanta-
Placing E; coincident with the reference line implies that coil side
Figure B-2 is the phasor
neously under the influence of the maximum flux density.
Fig. B-1. The projectio n of Ey
diagram representation of the situation depicted in
t with the fact that coil side 2
onto the reference line is negative, which is consisten
from that of coil side 1.
is under the influence of a pole of reversed polarity
2 provides addition of the
Because the end connection that joins coil sides 1 and
590 Appendices

coil voltages, the resultant coil voltage is found in Fig. B-2 by adding —E, to £,. It
is worthwhile to keep in mind here that the maximum value of this resultant coil
voltage remains fixed. The instantaneous value, which varies with the position of
the coil relative to the flux field, is readily found as the vertical projection of the Er
phasor. It is important to note that, because of the use of a fractional-pitch coil,
the resultant voltage can never be equal to the algebraic sum of the coil side
voltages. As a matter of fact, the pitch factor is simply defined as the ratio of the
resultant coil voltage to the absolute sum of the voltages induced per coil side.
A convenient analytical expression for the pitch factor for the fundamental
induced voltage can be derived easily from the geometry of Fig. B-2. Let W
denote the coil width in suitable units, and let 7 denote the pole pitch in the same
units. The angle between EF, and E, is then Wz/7. Drawing the perpendicular
bisector from the tip of E, onto Ep reveals two right triangles. The resultant coil
voltage can then be expressed in terms of the coil side voltage by

Er =: 2E, COS a:=) a

(1 gst (B-1)

The expression for the pitch factor for the fundamental is therefore

Kp _ Fre
aa cos ({1 es=)
7/2Une?Sil,Lipa
sin 5

_ Wo
= sin = > (B-2)

For a machine that has nine slots per pole, the use of a coil width of seven slots
leads to a value of K,; = 0.940. Thus there is just a 6% reduction in fundamental
induced emf. However, the corresponding reduction in the harmonics is much
greater. This is demonstrated next.
If n denotes the order of a space field harmonic, a little thought discloses that
the phasor diagram differs by the amount of rotation associated with one coil side
voltage relative to the other. Thus in Fig. B-2 the angle that —E, makes with the
vertical is increased by n times. Accordingly, the expression for the nth harmonic
pitch factor becomes

Pl ha i ( =) Toe eT. ee
Komagge ~ cosh 1 7) 2 ~ Sinns
sin n— > (B-3)

For the same machine with nine slots per pole, the value of fifth harmonic pitch
factor becomes
K,s = —0.174 (B-4)
Note that the emf induced by the fifth harmonic field component is reduced by
approximately 83% compared to 6% for the fundamental. The minus sign in
Eq. (B-4) indicates a phase reversal for the seventh harmonic emf, brought about
by a negative flux linkage caused by the chording of the winding.
Sec. B Pitch and Distribution Factors for Machine Windings 591

The Distribution, or Breadth, Factor

In order to focus solely upon the effect of distributing the winding of an ac

machine over sixty electrical degrees, it is assumed that only full-pitch coils are
used. Once again the derivation is carried out first for the fundamental field
wave. In this connection refer to Fig. B-3, which shows the four coils that are
assumed to constitute all of phase I of a single-layer, two-pole, three-phase ma-
chine. It is helpful to note that the instant is shown which corresponds to a net
phase voltage of zero. This means that in the phasor diagram of Fig. B-4 the
phasor denoting the phase voltage (AE) must have a zero projection on the refer-
ence line. In Fig. B-4 the quantity AB denotes the peak value of the voltage
induced in coil 1-13. Note that this coil is displaced from the position of maximum
voltage by the angle a. At the same time, because of the distribution pattern, the
voltage of coil 2-14 (denoted by BC) is away from its peak value by the angle (a +
B), where B denotes the angle between slots. It is interesting to observe that the
projection of BC onto the reference line is smaller than that of AB. A glance at
Fig. B-3 indicates consistency with the physical relationship these coils bear to the
field distribution. Coil 3-15 is represented by CD in Fig. B-4. Its projection on the

W
\ {

Direction of rotation

iF
Phase I

al field flux at
Figure B-3 Showing a distributed winding relative to the fundament
the instant of zero net phase voltage.
592 Appendices

AB = BC =CD = DE = peak coil voltage = E,

AE =Ep =peok phase voltage

Figure B-4 Derivation of the distribu-

tion factor. Illustration is for four slots
per pole per phase.

time line is negative, thus indicating that coil sides 3 and 15 are under the influence
of poles of reversed respective polarity than those for coils 1-13 and 2-14. Coil
4-16 is represented by DE in Fig. B-4. Each successive coil voltage is displaced
by an additional amount (equal to the slot angle 8) from the preceding coil. The
effect is that a resultant phase voltage is obtained that is Jess than the absolute sum
of the individual coil voltage. In fact, the distribution factor is defined as the ratio
of the resultant voltage to the absolute sum of the coil voltages comprising the
resultant.
An examination of Fig. B-4 leads to a convenient analytical expression for
the distribution factor. By drawing the perpendicular bisectors of each of the coil
voltages represented by AB, BC, and so on, the center of a circle passing through
points A, B, C, D, and E can be found. The quantities AB, BC, CD, and DE are
then also equal chords of this circle. Accordingly, it follows that
+ AOE = QB (B-5)
where Q denotes the number of slots per pole per phase. Moreover,
+ AOB = 8 (B-6)

+ AOG 2B
= 5 (B-7)

and

« AOF = £F (B-8)
 Sec. C Derivation of the Peak MMF per Pole of a Polyphase Winding 593

The expression for the peak value of the resultant phase voltage becomes

Er savor
= AE = 2(AF) L= 2(OA) sin
in SP5 (B-9)

By definition, the distribution factor is then

E 2(OA) sin a OB y (GA) sin £2

Roe eeee es bas clin
d- OEL Q2(AG) Q2(OA) sin 8/2

sin cE

~ QO sin B/2 i
For the machine shown in Fig. B-4 the values of the quantities in Eq. (B-10) are
Q = 4, B = 15°. Hence the value of the distribution factor for the fundamental is
0.958, thus representing less than a 5% reduction. The value of the distribution
factor for the fifth harmonic for the same machine is found to be 0.205, thus
representing an 80% reduction.
The net winding factor for a winding that is pitched, as well as distributed, is
equal to the product of the pitch and distribution factors. Expressed mathemati-
cally,
Kun i Kon Kan (B-11)

where the second subscript n denotes the nth harmonic. When K,,, and Kg, are
to
low for a given harmonic, the net winding factor assumes such small values as
render that particular harmonic inconsequential.

C. DERIVATION OF THE PEAK MMF PER POLE OF

A POLYPHASE WINDING

To begin simply, consider a two-pole machine with three phases and a total of six
slots. Each phase therefore consists of a single coil, which for a full-pitch winding
means that the coil spans from slot 1 to slot 4, as depicted in Fig. C-1. For a
uniform air-gap machine, any current that flows through this coil produces a
rectangular mmf distribution. The total coil mmf is distributed equally between
is
slots 1 and 4 (say for flux entering the air gap) and slots 4 and 1 (where the flux
assumed leaving the air gap for the same instant). Attention for the moment is
directed to the mmf distribution of one phase. The effect of the remaining phases
is considered presently.
If the phase coil is assumed to be carrying a sinusoidally varying current of
a
rms value J, then the height of the rectangular distribution alternates between
594 Appendices

Figure C-1 Magnetomotive force distribution of a full-pitch, single-coil phase

winding in a uniform air-gap machine.

positive maximum and a negative maximum value, in accordance with the follow-
ing expression:

F= Aezl sin wt (C-1)

where z, denotes the conductors per slot. Note that when the argument is 77/2, the
coil mmf has a positive maximum value. Inthe interest of obtaining a result that is
applicable generally, i.e., to the harmonics as well as to the fundamental con-
tained in the rectangular distribution of Fig. C-1, a Fourier analysis is applied to
this wave. The equation for the mmf distribution for phase I, expressed as a
function of time and space displacement from the indicated reference position a =
0°, then becomes

F(a) = >, a, cos na (C-2)

n=1,3

where

a, = — i's F tor na da (C-3)

1 2a

and & is given by Eq. (C-1).

Inserting Eq. (C-1) into Eq. (C-3) and noting that the given wave possesses
quarter-wave symmetry, we may write

4 a/2 Ns A : 5
An = — [. “3a Iz; sin wt cos na da = 0.9z,/ sin wt (=sin n ) (C-4)

Substituting this result into Eq. (C-2) then yields

F(a) = 0.92.1 sin wt (cos a — § cos 3a + $cos5a—4+cos7a+---) (C5)
Now, if instead of one slot per pole per phase the machine is equipped with Q slots
Sec. C Derivation of the Peak MMF per Pole of a Polyphase Winding 595

per pole per phase, the foregoing equation can be made to apply by introducing
two modifications: replacing z, by z,Q and including the distribution factor of each
harmonic. Accordingly, we get

F(a) = 0.9z,Q1 sin wt (Ka cos a — Be cos 3a + 5Kis cos Sa —:: ) (C-6)

Furthermore, if each coil is assumed to be fractional pitch, then each term of

Eq. (C-6) must be further modified by the pitch factor for each harmonic. Thus
the complete expression that describes the mmf distribution of one phase of a
fractional-pitch distributed winding is

F(a) = 0.9z,Q/ sin wt bees cos a — ne cos 3a + A cos 5a —:: :) (C-7)

where Ky, = Kan Kpn is the winding factor obtained as the product of the individual
distribution and pitch factors.
Since our interest is chiefly in the resultant mmf produced by three-balanced
windings cooperating to produce a revolving mmf in the manner described in Sec.
4-1, we need first the expressions for the mmf of phases II and III. These are
readily obtained from Eq. (C-7) by introducing the required 120° displacement in
time and space. Hence

F(a) = 0.9z,QI sin (wt — 120°) [Ku COS. (ae t2U0) = ‘e cos (3a — 360°)

+ Kys
=~ cos (Sa + 600°)
°
= - jus (C-8)

Ky :
F(a) = 0.9z,QI sin (wt — 240°) be cos (a — 240°) — ee cos 3(a — 240°)

+ ‘Sscos 5(a@ + 240°) —:- | (C-9)

It is helpful to note that the mmf of each individual phase has an alternating
character. This is true of the harmonic as well as the fundamental components.
However, the resultant fundamental or harmonic waves are found to revolve with
time. For any particular harmonic this is readily corroborated by summing the
contributions for each phase. Thus for the fundamental the result is found to be

F(a) = $(0.9)z;QIKw1 sin (wt — @) (C-10)

As already demonstrated in Sec. 4-1, Eq. (C-10) is the equation of a traveling sine
wave moving in the positive a direction. It follows therefore that the amplitude of
this resultant wave is the coefficient of the trigonometric term. Hence

Fa = 3(0.9)z,;OKwil (C-11)
596 Appendices

The factor 3 appears because we are dealing with three phases. For a q-phase
machine, Eq. (C-11) becomes

= 5 (0.9)zs OKinl (C-12)

It is often desirable to express this quantity in terms of the total number of turns
per phase (N). This is possible to do upon recognizing that
( slots ;
(turns) _ Zs (Ss (C-13)
erate = ae p (poles)
phase
xle
ce @ \no
Accordingly,

ities (C-14)
D
Insertion of the last expression into Eq. (C-12) yields the final and desired form of
the mmf amplitude of a polyphase winding, namely,

alley I (C-15)

A similar procedure can be followed to find the expression of the peak mmf
of the harmonics in the resultant revolving field. Adding the third harmonic com-
ponents of each of the three phases yields zero—a result that is obvious from an
inspection of Eqs. (C-7), (C-8), and (C-9). This conclusion also applies to all
multiples of the third harmonic. Accordingly, there is no resultant third harmonic
component in the traveling mmf wave of a balanced three-phase winding.
The addition of the fifth harmonic contributions in each of the three phases
leads to the following result:

F,5(a) = 5(0. 9) Zs or Az
> sin (wt + a) (C-16)

Note the form of the argument is such as to indicate that the fifth harmonic
component of the resultant wave moves in the negative a direction. The ampli-
tude of the fifth harmonic wave is then clearly

wkAyw5 N ‘
Fas = 3(0.9) = “—— I = 3(0. are I (C-17)

For a q-phase machine this becomes

NK,
Fas =0.99
| (C-18)

A little thought concerning the derivation of the last expression readily

reveals that the general expression for the amplitude of the resultant traveling
 Sec. D Reduction Factors for Polyphase Windings
597

wave of the nth harmonic of a g-phase winding is

NKvn I
Fan
A = 0.9q mp (C-19)
Upon inserting n = 1, Eq. (C-15) is obtained.

D. REDUCTION FACTORS FOR POLYPHASE WINDINGS

A reduction factor is a numerical quantity which, when applied to a rotor quantity

such as voltage, current, impedance, and so on, refers that quantity to the stator.
This means that the effect produced by the referred quantity in the stator is
identical to the effect caused by the original quantity in the rotor. The reduction
factors are different for the various electrical quantities, and they also differ
depending upon whether the rotor is of the wound or squirrel-cage type.

The Wound Rotor

There are three reduction factors of interest: one to handle voltage, one to handle
current, and the third to deal with resistance, reactance, and impedance. As
described in Chapter 2, a reduction factor is essentially a transformation ratio that
converts the actual rotor winding into an equivalent rotor winding having the same
number of turns as the stator winding. If we let E,; denote the standstill emf
induced per phase in the stator winding by the air-gap flux, and E, denote the
standstill emf induced per phase in the rotor winding by the same air-gap flux, then
by Eqs. (4-9) and (4-10) the ratio of these two quantities is

E, Li. Ni Kw =a, (D-1)

E, N:2Ky2

Equation (D-1) is the reduction factor for referring any voltage associated with the
rotor winding back to the stator. The convention used in this book to denote the
referred quantity is the prime notation. Accordingly, to refer the quantity E, to
the stator, we write
Ey = a,ky (D-2)
A comparison with Eq. (D-1) shows E} is identical to E,, and they should be since
both quantities are related to a winding having the same effective turns cut by the
same gap flux.
The reduction factor for current can be derived by recalling how the rotor
current is made to flow. Ideally, at no-load and for an induction motor with
negligible rotational losses, the rotor current is zero. The slip correspondingly is
also zero. As load is applied to the rotor shaft, the rotor structure slows down,
598 Appendices

thus causing an emf to be induced in the rotor winding, which in turn produces
enough rotor current to supply the required energy to the load. As rotor current
flows through the q)-phases of the rotor, it presents a total rotor mmf of magnitude
q2N2Kw2Iz that acts in a direction to oppose the original stator mmf, gi NiKwilm.
The latter quantity is the effective stator mmf that produces the air-gap flux, the
value of which is essentially determined by the applied stator voltage. To offset
this opposing action of the rotor mmf and thus preserve the value of air-gap flux
demanded by the applied stator voltage, there must occur an increase in stator
mmf equal and opposite to the total rotor mmf. Calling the increased stator cur-
rent [4, at equilibrium we can then write

MLN Kwi = q2l2N2Kw2 (D-3)

Or, the required increase in stator current becomes

aS" f= a D-4
: aN Kwi é : ( )

where

a, = PNR current reduction factor (D-5)

aN; Kw
The application of a; to any rotor phase current gives the equivalent effect of that
current in the stator winding. It is interesting to note that when the rotor and
stator phases are the same, the reduction factor for current is the inverse of the
reduction factor for voltage.
The manner of transferring the effect of a given rotor resistance to the stator
can be found from the relationship that preserves either copper loss or per-unit
resistance drop. For example, for a rotor winding resistance r; and rotor current
I, the per-unit resistance drop is J,r2/E,. When placed in the stator, the effect of
this rotor resistance (now called r5) must be the same as it produces in the rotor
itself, in spite of the fact that in the stator the r3 carries the current J; and must be
associated with the stator voltage E; (or E3). Equivalence is preserved by satisfy-
ing the following equation:

Ir, _ lar
ERIE (D-6)
It then follows that the rotor resistance as it must appear in the stator is

eee: (D-7)

ee) rr = azr2 (D-8)

Sec. D Reduction Factors for Polyphase Windings 599

where

aa (= == _ reduction factor for resistance, (D-9)

Zz
go \NoKy2 reactance and impedance ”
It is a simple matter to show that Eq. (D-9) applies with equal validity to reactance
and impedance quantities. This is left as an exercise for the reader.

The Squirrel-Cage Rotor

It is helpful to keep in mind that the squirrel-cage rotor consists of solid copper
bars that are often driven through the slots and connected at both ends by means
of copper end rings. Asa result of this construction each bar can be considered to
be a separate and individual phase of a g-phase system, reckoned in terms of the
traditional definition of what constitutes a phase. Since each bar (or conductor) is
a phase, it follows that the number of turns per phase must be 3 (it takes two
conductors to make one turn), and the winding factor must be identically unity.
The voltage reduction factor for the squirrel-cage rotor is readily determined
by inserting N, = 4 and K,,2 = 1 into Eq. (D-1), which is a general result. Thus
ay = 2Ni Ky (D-10)
When the number of rotor bars is not divisible by the pole pairs through an
integer, the number of rotor phases is equal to the number of bars. Hence, under
these circumstances for the squirrel-cage rotor, Eq. (D-3) becomes

HN Kw = Sah (D-11)
or

S»
I, = (2) ve D-12
EIGN Kay! 7 ( )
where S> denotes the number of rotor slots, and the quantity in parentheses is the
current reduction factor.
The reduction factor for resistance is found again by employing Eq. (D-6).
For the squirrel-cage rotor this leads to

n=sSah (D-13)

Introducing the appropriate current and voltage reduction factors into the last
equation yields

esa [Aan Ae Jn (D-14)

2

where the quantity in brackets identifies the desired result. As before, this reduc-
tion factor is applicable to reactance and impedance as well as resistance.
 600 Appendices

E. SATURATED SYNCHRONOUS REACTANCE METHOD BASED

ON RESULTANT FLUX

Frequently, the design of synchronous machines is such that when rated power is
delivered at rated voltage the operating point on the magnetization curve is lo-
cated somewhere along the knee of the curve. Accordingly, the machine can be
described as operating in a partially saturated state. A good indication of the
degree of saturation of the magnetic circuit is given by the resultant flux ®, or its
associated emf E,. Thus, if a synchronous generator is assumed to be delivering
rated power at rated voltage and a specified power factor, the actual armature
induced emf E, can be found as illustrated by Eq. (5-22) and then located on the
magnetization curve. This is illustrated in Fig. E-1, where E, is represented by
point a on the saturation curve and is produced by the field current R. It is
worthwhile to note here that if there were no saturation of the iron, the voltage E,
would be induced by a smaller excitation R,,. In essence, then, it can be said that
the degree of saturation prevailing under these conditions is represented by the

/
Air—gap cee

v “Linearized
aif Cc magnetization
alate ea ee Satie nT
24 Oe Rs curve corresponding
to @ (i.e. E,)

m
Line =
volts
voltage,gm
—to—neutral
Magnetization curve

m dt

Field amperes

Figure E-1 Illustrating the saturated synchronous reactance method based

on E,.
Sec. E Saturated Synchronous Reactance Method Based on Resultant Flux 601

ratio of these two quantities. That is, the reluctance of the magnetic circuit has
been increased by a saturation factor k where

k=- Re
_ Fra
E (E-1)

The second form of this equation follows from the geometry of similar triangles in
Fig. E-1.
How would this synchronous generator perform if it were now assumed that
the degree of saturation remained fixed at ®, and the armature current were
reduced to zero while the field current remained constant? A study of Fig. E-1
discloses that the operating point would move from a to c along a linearized
magnetization curve that corresponds to the resultant flux. Point c is found as the
point of intersection of a line drawn from the origin through a and the ordinate line
placed at the constant field excitation (F'). The no-load voltage that corresponds
to c is identified as E f and represents the no-load terminal voltage that would be
produced if the air-gap flux remained invariant as the armature current diminished
to zero. Since this does not actually happen, clearly E f is entirely a fictitious
quantity. On the other hand, the field current F (that is associated with E 1) is not
fictitious, but real. In fact, F is the quantity that we are seeking, and the foregoing
construction has been introduced simply to facilitate the determination of F. The
determination immediately follows from the similar triangles OaR and OcF. Thus

F= BeE, R (E-2)

where these quantities are defined as illustrated in Fig. E-1.

In applying Eq. (E-2), the quantity E, is found by means of Eq. (5-22); R is
found from the actual magnetization curve. E f is then computed by the synchro-
nous reactance method in the manner illustrated in Example 5-2 but with the
necessary modification that that value of synchronous reactance be used that
corresponds to the degree of saturation represented by the resultant flux. If x,
were entirely associated with the iron structure of the machine, the saturated
synchronous reactance could be found by dividing the unsaturated x, by the
saturation factor k. However, as indicated by Eq. (5-16), the synchronous reac-
tance also includes the effect of leakage reactance, which remains essentially
unaffected by the changes in the degree of saturation of the main iron paths of the
machine. Accordingly, the saturation factor must be applied to xg only in Eq.
(5-16). This leads to the following expression for the saturated synchronous reac-
tance X,;:
Xs — XI
Xs5 = XI a (E-3)
k
where x; denotes leakage reactance, k is the saturation factor given by Eq. (E-1),
and x, is the unsaturated synchronous reactance given by Eq. (5-18). The follow-
ing example illustrates the effectiveness of this method.
 Appendices
602

Example E-1
rated terminal
For the machine of Example 5-1, find the field current that produces
armature current is delivered to a 0.8 lagging pf three-phase
voltage when rated
balanced load. Use the saturated synchron ous reactance method.
which are
Solution The values of E, and R are given by Eqs. (5-22) and (5-23),
repeated here for convenience.
E, = 8444.4V and R=222A

The saturation factor is found to be

aS es eyR24
Hence by Eq. (E-3) the saturated synchronous reactance is
7, =x “0s Dip
dad Wee, 8
Lap mekieayy pata: 1s tine ay © 16.2 0

Accordingly, we get
Ey = V, [0° + Talixss)
= 7967.4 + j418.4(0.8 — j0.6)16.2
= 7967.4 + 4042 + j5423
= 12,009 + j5423 = 13,177 /24.3°

Finally, by Eq. (E-2) the required excitation current is found to be

13,177 222) ui= 346A
F = 3754.4
which is identical to the result found by the general method.

When it is desirable to compute voltage regulation by the saturated synchro-

nous reactance method, a glance at Fig. E-1 reveals that the no-load terminal
voltage to be used in Eq. (5-26) is to be found on the actual magnetization curve
corresponding to the value of F computed by Eq. (E-2). This is denoted by point b
in Fig. E-1.

F. SLIP TEST FOR SALIENT-POLE SYNCHRONOUS MACHINES

Information about the direct-axis synchronous reactance and the quadrature-axis

synchronous reactance of synchronous machines can be obtained from a rela-
tively simple no-load test known as the slip test. The reason for the name will be
apparent presently. Figure F-1 depicts the wiring arrangement for the test, includ-
ing the instruments that permit calculations of the reactances.
Sec.F = Slip Test for Salient-Pole Synchronous Machines 603

Figure F-1_ Schematic diagram for the slip test.

The three-phase variac is needed to apply an adjustable voltage of rated

frequency to the three-phase stator winding, which is set appreciably below the
rated machine voltage. Application of this voltage creates a revolving set of mag-
netic poles in the air gap of the synchronous generator, traveling at synchronous
speed n,. At the same time, a separate source is used to drive the salient-pole
rotor structure at a speed slightly different from synchronous speed and in the
same direction as the revolving stator mmf. A quick and easy way to ascertain
that the direction is proper is to put a voltmeter across the open-field winding
terminals. If a small reading is registered the direction is proper. (Why?) The use
of a dc motor is convenient as the separate drive source because of the ease with
which the speed may be adjusted.
By driving the rotor structure at a speed either slightly below or slightly
above synchronous speed, the rotor appears to be slipping either behind or ahead
of the rotating magnetic poles of the stator. Consequently, there will be a period
of time when the stator magnetic poles are in a position of alignment with the rotor
iron pole structure, thus presenting a minimum reluctance to the stator mmf and
resulting in a minimum magnetizing current, as indicated by the line ammeters, A,
in Fig. F-1. Ata slightly later instant the direct axis of the rotor pole structure will
be in a position of quadrature with respect to the axes of the stator magnetic
poles. Accordingly, the air gap is very large, and so a large magnetizing current
must flow to establish essentially the same air-gap flux demanded by the applied
voltage. The increase in current is quite appreciable and is easily registered by the
line ammeters. Keep in mind that this variation of the current from a minimum to
a maximum level occurs at a frequency that is dependent directly upon n, — n.
 Appendices
604

small
Obviously, one technique required in this test is to keep this frequency
true maximum and minimum values can be recorded. How-
enough so that the
ever, at the same time, this frequency must not be made so small that the rotor
structure is pulled into synchronism with the stator magnetic poles through the
action of the reluctance power. As a matter of fact, the reason for performing this
test at reduced voltage is precisely to minimize the influence of the reluctance
torque to pull the rotor into synchronism. Unfortunately, a lower limit must be
placed on this applied voltage; otherwise, the computed value of synchronous
reactance will be nearer to the unsaturated value, which we already know can be
misleading.
Because the internal impedance of the variac is not likely to be negligible,
the wide swing in the minimum and maximum values of the line ammeters will
cause some fluctuations in the voltmeter reading. These too should be recorded.
Also note that the maximum voltage reading is associated with the minimum line
current, and the minimum voltage reading goes with the maximum current.
The direct-axis synchronous reactance is found by using Vinax and Jmin Since
these are the readings that are registered when the rotor-pole axis and the mag-
netic poles of the stator are in alignment. Assuming a Y-connected stator wind-
ing, the expression for the direct-axis synchronous reactance thus becomes
ce Vmax:
(Bi)
ia
Average values of the three ammeters and two voltmeters should be used.
By a similar argument, it should be clear that the quadrature-axis synchro-
nous reactance can be computed from

x, = Vinin
(F-2)
ORV Grae
The reliability of the results obtained from Eqs. (F-1) and (F-2) can be
improved by employing oscillographic methods to record and measure the varia-
tions in line currents and line voltages. This permits a larger slip frequency to be
used, which in turn allows a higher applied voltage to be used without the danger
of pull-in. Furthermore, the maximum and minimum values can be measured
much more accurately because the inertia effects of the indicating ammeters and
voltmeters are eliminated entirely.

G. A SHORT BIBLIOGRAPHY

The following is a short selected list of books that the reader can find helpful as a
supplementary source of information in the study of this textbook.
CHAPMAN, STEPHEN, J., Electric Machinery Fundamentals (New York: McGraw-Hill
Book Company, 1985).
Sec. G A Short Bibliography 605

EL-Hawary, M.E., Principles of Electric Machines with Power Electronic Applications

(Englewood Cliffs, NJ: Prentice-Hall, Inc., 1986).
FITZGERALD, A.E., C. KINGSLEY, JR., AND S.D. UMaANs, Electric Machinery, 4th ed.
(New York: McGraw-Hill Book Company, 1983).
LiwsHITz-GARIK, M. AND C.C. WHIPPLE, Electric Machinery, vols I and II (Princeton,
NJ: D. Van Nostrand Co. Inc., 1946).
McPHERSON, GEORGE, An Introduction to Electrical Machines and Transformers (New
York: John Wiley & Sons, Inc., 1981).
NASAR, S.A. AND L.E. UNNEWEHR, Electromechanics and Electric Machines (New York:
John Wiley & Sons, Inc., 1979).
STEIN, ROBERT AND WILLIAM T. Hunt, Electric Power System Components: Trans-
formers and Rotating Machines (New York: Van Nostrand Reinhold, 1979).
 Answers to
Selected Problems

1-1. (a) 4/a A/in (b) 32 x 10-7 N/m (c) 0.32 N/m
1-3. (a) 0 (b) —0.1 N/m, repulsion (c) same as (b)
1-5. 20 A
1-7. (a) 98 A-t (b) 1.225 A
1-9. 183.75 A-t
1-11. 0.00475 Wb
1-13. 357.4 turns
1-15. 158.8
1-17. (a) 0.5 H (b) 0.0025 J
1-19. 2.41 W
1-21. 0.302 J/cycle
1-23. Hysteresis: 1400 W @ 60 Hz, 2100 W @ 90 Hz; Eddy: 400 W @ 60 Hz,
900 W @ 90 Hz
1-25. F = (B’/p)bg
1-27. (a) 0.53 A (b) 0.4975 J (c) 0.165 J (d) 4.72 H

2-1. (a) 0.00108 Wb (b) 1.44 in

2-3. ¢ decreases, J, remains constant
2-5. (a) 0.01126 Wb (b) 1.5 (c) Increase exceeds 50% (d) 90 Hz

606
Answers to Selected Problems 607

2-7. (a) 97.65% — (b) 2453.2/0.33° (©) 2.216%

2-9. 96.68%
2-11. 4.82%
2-13. 0.7055 A
2-15. 206 A
2-17. Purely capacitive
2-20. 87.8 V
2-21. 0.8566 + j0.8072
2-23. Leading load with pf angle of 28.6°
2-25. 1A
2-27. 10H
2-30. (a) 1.9H (b) 0.2 H
2-33. (a) 1.48% (b) 0.575 © (c) 55 kW, 16.5 kW (d) 22 kVA, 440/330 V
2-35. 621 /—68.75° A, high side
2-37. 48.4 /—60.46° A, 50 /—36.9°
2-39. (a) 225 kVA (b) 25 kVA (c) 200 kVA
2-42. (a) 0.0206, 0.0206 (b) 95.63%
2-43. 0.0256 pu, 0.016 pu
2-45. 2 ohms
2-47. 0.866
2-49, (a) 120.09 V (b) 126.6 V

3-1. (a) 0.09425 Wb (b) 39.27 V (c) 3.927 Wb/s

3-3. (a) 4.8 N-m (b) 50 V (c) 400 W
3-5. (a) 0.0024 Wb (b) 0.144 Wb-t (c) 0.144 cos Smt (d) 2.26
(e) 0 (f) 1.96 V
3-7. (a) & = Bir(a — 26) (b) e = 2NBIv (c) 7.54 V (d) 4.8 V
3-9. (a) A = 2.4 cos wot (b) 130.6 V (c) 130.6 V (d) 150.8 V
3-12. 532 A-t/pole
3-15. (a) 8620 A-t (b) 53.1 N-m (c) 104 V (d) 0.00122 Wb

4-2. 900 rpm, 1.8 Hz

4-5. (a) 1.50d¢,, directed CW 60° from vertical in Ist quad. (b) 1.50¢,, @ 120 CW
4th quad.
4-9. F,,, cos (wt — a)
4-10. (a) 1800 rpm (b) 3 Hz (c) 90 rpm (d) 1800 rpm (e) 0 (f) yes
4-13. (a) 700 rpm (b) 1700 rpm
4-16. 87.45%
4-17. 99.3 A
4-19. (a) 21.68 /—25.3° (b) 8.38 hp (c) 80.3% (d) 36.32 N-m
(e) 56.73 /—78.4° A
 Answers to Selected Problems
608

4-21. (a) 0.288 lag (b) 0.98 lag (c) 5.41 kW

4-23. (a) 160.5 /—27.53° (b) 90.49%
4-25. (a) 0.9 lag (b) 87.35% (c) 41 hp
4-28. (a) yes (b) 5 hp (c) slightly less (d) 2
4-30. 16.7% drop
4-33. (a) 0.096 (b) 45.55 N-m
4-35. 0.0333 ohm
4-37. (a) 40.033 A, 13.52 kW (b) 593.5 V, 71.75 kW
4-40. 1.215 ohms
4-41. 0.1174 QO, 72 = 0.13 O, x; + x3 = 0.9.0, r,; = 136.6 0, xg = 15.7 Q.
r, =
4-43. 6 A, 0.301 lag
(a) (b) 534 W
4-45. (a)
160 V
4-47. (a)
900 rpm (b) 945 rpm
4-49, (a) —0.2 (b) 1440 rpm (c) 60 kW
4-51. (a) 32 V (b) 6 V in quadrature with sE,
4-53. (a) 10 V (b) 3 kW
4-56. 1.1 ohms
4-57. 25.87 ohms

5-1. 1.8 pu
5-3. 17.8 A
5-5. (a) 340/14.75° V (b) 329 V (c) 233 kW
5-7. (a) 10.56 A (b) 12.68 A
5-9. (a) 1.58 0 (b) 1.284 0 (c) 33.06A
5-11. 11.3 t/p
5-13. 38.6 A
5-15. (a) 31.5% (b) —18.38%
5-17. 40.55%
5-19. 2 ohms
5-21. (a) 192.25 V (b) 35.9° (c) 35.9°
5-23. (a) 112.78 V (b) 8.74° (c) 45.64°
5-25. (a) 101 V (b) 3.053 A
5-27. 275 /24.42° V
5-29. (a) 6000 (b) 164A (c) 50.9°

6-1. (a) 0.605 leading (b) 1000 V/phase

6-4. (a) 293 joi
6-5. (a) 0.778 leading (b) 10,000 /—14.7° (c) 1517.2 kW
6-7. (a) 2/3 pu (b) 2.6 pu
6-8. (a) motor (b) —30° ,
6-11. (a) 24.85 A, 0.79 lagging (b) 280.3 /—25.03°
Answers to Selected Problems 609

6-13. 627 N-m

6-15. (a) 2.58 pu (b) 0.375 pu
6-18. xq = 0.7 pu, x, = 0.4 pu
6-19. 1.805 pu, 0.72 leading
6-20. (a) 72° (b) 1.926 pu
6-22. 481
6-24. 0.797 lag

7-2. (b) 3.73Em

7-3. (b) 3.86E,,
7-8. (a) 366.7 ohms (b) 12 V (c) 176.5 ohms (d) 0.068 A (e) 436 rpm

7-9. 190 V
7-11. (a) 0.01 Wb (b) 1.24 Wb (c) 1.16 Wb
7-13. (a) 366.7 ohms (b) 17.3 V (c) 1.48 A
7-15. (a) 258 V (b) 58.3 A (c) 1.54A (d) 236 V
7-17. (a) 8.125 ohms (b) 210 V (c) 84.1%, 454 N-m
7-19. (a) 11.2 ohms (b) 552 V
7-21. (a) 250 V (b) 270 V (c) 8.56 A (d) 276 V
7-24. 38.7 ohms

8-1. (a) 1192.27 rpm (b) 6.8%

8-3. 998.2 rpm
8-5. (a) 430.6 rpm (b) 672.6 N-m (c) 40.7 hp
8-7. (b) o = 2847/VT (c) 47.6 rad/s (d) 50.6 rad/s (e) 6.3%

8-9. (a) 40.16 A, 42.16 A (b) 73.2 N-m (c) 10.6 hp

8-11. (a) 1112 rpm (b) 93.4% (c) 119.6%
8-13. (a) 134 ohms (b) 880 rpm
8-15. (a) 0.009 Wb (b) 129 N-m (c) 0.58 kW (d) 15.35%
8-17. (a) 229A (b) 1413 rpm
8-19. (a) 260 A (b) 576.2 N-m (c) 86.76%
8-21. (a) 72.92 N-m (b) 1342.3 rpm (c) 81.3%
8-23. 1432.4 rpm or 150 rad/s
8-25. 69 ohms
8-27. (a) 2.045 ohms (b) 10.226 (c) 62.06%
8-29. (a) 24.355 rad/s (232.57 rpm) (b) 80.59%
8-31. 67.5 rad/s (644.6 rpm)
3. armature voltage
8-34. (a) 1. external armature resistance, 2. field control,
(b) 1. 2000 rpm, 200 hp; 2. 500 rpm, 200 hp; 3. 2000 rpm, 200 hp
control
8-36. 880 rpm, 77.5 A
8-39. 80.6%
8-42. (a) 24.4 ohms (b) 262 V (c) 175 ohms
610 Answers to Selected Problems

8-45. 0.585V,
8-47. (a) 393 rpm (b) 379 rpm (c) 3.7%
8-49, (a) 176.93 rpm (b) 191 rpm, 7.9% (c) 381.57 rpm, 3.8%
8-51. 189.44 Hz

9-1. Yes
9-5. CW: 671 A-t; CCW: 1432 A-t
9-7, 8 [90° A
9-9, (a) 95 [14.7° (b) 37/—40.6°
9-11. 123.9 /20.9°, 68.7 /93.19°
9-13. (a) 7.854 A, 0.57 lag (b) 0.411 hp (c) 62.28%
9-15. 1224 W, 20.34 A
9-17, (a) rz = 3.34.0, x; = x, = 3.211 O, (xg/2) = 49.21 0 (b) 65.28 W
(c) 546.5 0
9-19. (a) 8 A, 0.772 lag (b) 1.303 hp (c) 71.49%
9-21. (a) 49.8% rated torque (b) 213 wF (c) 266.67% rated torque

10-1. (a) 25.5 /—13.35° (b) 104.2 /25.06° (c) 9.24 /106.1° (d) 21.52 /119.4°
10-5. (a) 60/—210° (b) 60 /—120°
10-7. (a) 18.83 /—143.5° (b) No
10-9. (a) No (b) 75 /0° (ce) 25 [0°
10-11. (a) 90 /0°, 60.8 /—99.46°, 70 /—180°, 60.8 /99.46° (b) 70 /0° (c) Yes
(d) 10 /0° (e) 10 /0°
10-13. 123.9 /20.9°, 68.7 /93.19°
10-15. (a) 100 [0° (b) 20 [0°
10-16. 3.37 S-W or 0.0179 N-m
10-19. (a) 115.2 /—0.68°, 735.52 [—50.84° (b) 61.314 /—113.62° mA
10-21. (a) 5.23 S-W (0.01386 N-m) (b) 2.614 S-W (0.00693 N-m) (c) 0.000039
10-24. (a) 13.173 S-W (b) 13.7 S-W (c) 9.21 S-W
10-25. 6.964

11-1. (a) 0.8637,, CCW (b) 18°

11-3. 0, a condition of equilibrium
11-5. 48 bits/s

te Coil A-A’ Coil B-B’

ooo.ore
oc Oo
O'S
= SO
Cs
= oo
oor &
oor
 611
Answers to Selected Problems

11-8. (a) 7.2° (b) 40

12-3. Line-to-neutral voltages: 26, 26, —52; Line-to-line: 0, 78, —78.

12-4. Line-to-neutral voltages: 26, —52, 26; Line-to-line: 78, —78, 0.
12-11. 0.06 pu
12-13. 0.275 pu
12-14. a 0° 5° 10° ile 20° pass? 30°
Ep 0 0.023 0.31 0.022 0.007 —0.001 0

12-15. (a) —0.1736E, (b) 0.3472E; (c) Yes

z i+ -) where a = Ar,
13-1. (a) T(s)= i( ili F tRB

) 7) =3-.
13-3. (a) T(s) = =RC OTD =Faget = Ca a
;=
(c) T(s)= a ce

@ 1) =2(E8).1- eR ene
1+ sRoC
13-5. (a) T(s) = (b) Same as Fig. 13-3(b)
(tae s(R, =f R,)C

Biaee a 3
13-7. |) = Gs ip
13-9. (a) 45.72 rad/s (b) 6
‘a
(c) 5.37 X 10-4 @ + @ = 4.71e;
os) 4.71 ag:
13-11 nn} E\(s) rg 65.57 x 107% 8

13-13. L;JRo + (JRR; + RFL; + K,K,L p)o + RARF + K; K,)o = KaK, KE

where R = Rm + Re
1 K
‘f Tf ° 1) |Rey E
13-13 = .
Rely) KK, + RF
2 RET RFT m
Re
where t= L/lRy, Tm = J/F, and R = Rm +
je) AE
13-17. (a) M(S) = F524 (F + K,KnK+ o)s8 + KpKaKn
K5k,h, =U (c) provides damping of transient
(b) Js? + (F + KaKmKo)s
terms
(b) 7.58 (c) 7 lb-ft/rad/s
13-20. (a) 150 rad/s
i Fis 0 4] ie :
ne (b) 60 rad/s
13-21. d/ (©) 26.62 s d(d) Al = a ea ire ? Oalet
(c) v(t) = 300 — CME elt) (d) 1/1204 s
13-25. (a) 300 (b) 10
612 Answers to Selected Problems

Rik .
pain ood RAK,K, + RF) + K,K,KiKo ”
Appendix to Chapter 3
3-19. (b) 30 At/p

3-21. (a) ¢ = Bir(2m/p) (b) T= p? J, sin A (c) T varies

3-25. a = 0°

3-28. (a) Yes because Ry # R, (b) standstill (c) T=

3-30. (a) 15.45 W (b) 3600 rpm (c) 45 W

 Index

Armature reaction, synchronous

A generator:
cross magnetizing effect, 247
Actuating signal, 509 demagnetizing effect, 247
Air-gap characteristic, 226 direct-axis, 247
Alternating field mmf, 158 in equivalent field amperes, 230, 244
Amortisseur winding, 273 quadrature-axis, 247
as a reactance, 233
Ampere-conductor distribution, 564, 565
relation to current sheet, 565 Armature reaction reactance, 234
relation to mmf axis, 565 Armature winding, 3, 127
dc machine, 136
Ampere’s circuital law, 7, 563
Ampere’s law, 2, 9 distributed, 591
Ampere-turns per pole, evaluation, synchronous machine, 134
565 three-phase motor, 132
Amplidyne generator, 540 Armature winding circuit resistance, dc
Applications:
generator, 322
dc motors, 369 Asynchronous machine, 153
single-phase motors, 426 Autotransformer, 89
advantages, 89
synchronous motors, 287
conducted power, 94
synchros, 482
three-phase motors, 206 step-down, 91
step-up, 96
Armature mmf:
ac machines, 561, 564
transformed power, 94
dc machines, 312, 569
Auxiliary winding, 396

613
614 Index

Compound generator, 309

Compound motor:
Backward-field impedance, 400 applications, 368-369
Base values for per unit system: controllers, 370
synchronous machine, 253, 281 equivalent circuit, 345
transformers, 98
power flow diagram, 346
Bibliography, 605 speed control, 352, 373
by adjustable armature voltage, 358
Block diagrams (Signal flow graphs):
by adjustment of pole flux, 352
amplidyne generator, 544, 545
by external armature resistance, 356
dc generator, 520, 522
speed-torque characteristic, 340
de motor, 513, 515, 517, 518
Conduction angle (see Firing angle)
feedback system, 508
field-controlled dc motor, 534, 535 Conservation of energy, 127, 143
two-phase servomotor, 539 Controllers:
voltage control system, 549 dc motors, 370
Blocked rotor test: current-limit type, 371
single-phase motor, 412 definite-time limit type, 372
three-phase motor, 184
three-phase induction motor, 211
Breadth factor (see Distribution factor) Control differential, 478
Breakdown torque, 179 Control of reactive power, 261
Brushes, 137, 143, 305 of real power, 263
Brushless excitation, 294 Control transformer, 477
Brush shift motor, 194 Control transmitter, 476
Conversion factors, units, 587, 588
Core losses, 38
Core loss resistor, transformer, 65
C three-phase induction motor, 183
Core-type transformer, 61
Capacitor, synchronous, 284 Critical field resistance, 309
Capacitor-start, capacitor-run motor, 419 Cross-magnetization of armature
Capacitor-start, induction-run motor, 417 reaction, 247
Characteristic equation, 507, 517, 528, Cross-product rule, 10
534, 545 Cumulative compounding, 310, 335
Chopper drives, 381 Curie point, 13
Closed-loop transfer function, 509 Current density, 20
Coercive force, 16 Current element, 2
Coercivity, 16 Current-limit controller, 371
Commutating poles, 329 Current sheet, 562, 569
Commutation, 325 relation to mmf per pole, 563
compensating winding, 330 Cut-out switch, 396
flashover, 330 Cylindrical rotor, 224
interpoles, 329
linear, 327
overcommutation, 329
resistance, 328 D
undercommutation, 328
Commutator, 136 Damper winding (see Amortisseur
Compensating winding, 330 winding)
Index 615

DC compound motor (see Compound Delta-wye transformer connection, 108

motor) Demagnetization of armature mmf, 311
DC generator: Demagnetization of armature reaction,
armature winding circuit resistance, 247
322 Diamagnetism, 11
commutation, 143, 325 Differential compounding, 311
series field, 309 Differential synchro, 478
compound, 309 Diode, free-wheeling, 288, 383
construction features, 136 Direct-axis path, 245
critical field resistance, 309 Distribution factor, 591
demagnetization of armature mmf, 311 Diverter, series field, 316
dynamics, 519 Doubly-excited machines, 154
equivalent circuit, 321 Drive amplifiers, stepper motors, 465
external characteristic, 313 Dynamic braking, 364, 372
field resistance line, 309 Dynamics:
interpoles, 139 amplidyne generator, 540
performance calculations, 320 block diagram, 508
power flow diagram, 322 dc generator, 518
shunt, 308 dc generator-de motor system, 524
voltage build-up, 309 dc motor with load, 510, 516
voltage generation, 324 field winding current, 507
types, 308 signal flow diagram, 508
DC motor: time constant, 507
applications, 369 two-phase servomotor, 535
chopper drive, 381 voltage-control system, 546
compound, 340, 345, 346, 352
controllers, 370
dynamics, 512
equivalent circuit, 345
E
permanent magnet type, 365
power flow diagram, 346 Eddy-current loss, 37
series, 341, 345, 346, 352 Efficiency:
shunt, 336, 338, 345, 346, 363 dc generator, 324
speed control, 348-363 dc motor, 348
speed regulation, 338 single-phase motor, 411
speed reversal, 363 three-phase motor, 172
transformer, 80, 103
speed-torque curves, 336-345
state equations, 516, 578 Electrical degrees, 568
Electric field intensity, 19
steady-state performance, 347
transfer functions: Electromagnetic torque (see Torque)
armature-controlled, 529 Electromechanical energy conversion
field-controlled, 533 principles, 127
Electromotive force (see Induced
types, 335
voltages)
thyristor drives, 373
single-phase, 373
Electronic speed control, 373
three-phase, 376 Energy, conservation of, 127, 143, 147
Equivalent circuits:
Definite-time limit controller, 372
compound de generator, 321
Delta-delta transformer connection, 105
616 Index

Equivalent circuits (cont.) Forward-field impedance, 400

dc motor, 345 Four-phase system, 432
servomotor, 441 Fractional-pitch coil, 140, 588
single-phase motor, 397—406 Free-wheeling diode, 288, 383
synchronous generator, 235 Frequency response, transformer, 71
synchronous motor, 274 Ferromagnetic material, 12
three-phase motor, 164
transformer, 66, 86
Equivalent reactance: G
autotransformer, 93
three-phase motor, leakage, 186 Gauss, cgs unit of flux density, 22
transformer, 69 General method, cylindrical-pole, 226
Equivalent resistance:
salient-pole, 252
autotransformer, 93
nonlinear analysis, 238
three-phase motor, 186 Generated voltage (see Induced voltages)
transformer, 69
Generator, defined, 123
Error, harmonic displacement, 499
Gap power:
Error detector, 482
single-phase motor, 409
Excitation voltage, 231, 238, 274, 275
three-phase motor, 167
Exciting current, 63
Gilbert, cgs unit of mmf, 21
External characteristics:
compound generator, 316
derivation of, 317 H
shunt generator, 316
derivation of, 319
Half-stepping, 469
Harmonic displacement error, 499
Hysteresis, 16
F Hysteresis loss, 34
Hysteresis motor, 423
Faraday’s law: Hysteretic angle, 55
electric machines,
126
induced emf, 129
transformers, 53 l
Feedback voltage control system, 546
Field excitation by rectifiers, 286 Impedance:
Field resistance line, 309 backward-field, 400
Field winding, 3, 127 equivalent, transformer, 78
of dc machine, 136 forward-field, 400
of synchronous machine, 136 negative-sequence, 401
of three-phase motor, 134 positive-sequence, 401
Firing angle, 288 rotor-leakage, 163
Flashover, 330 stator-leakage, 185
Flat compounding, 314 Induced voltages:
Flux linkage, 575 ac machine, 125, 128, 140
Flux per pole: f dc machine, 143, 304
ac machine, 567 : transformer, 53
dc machine, 570 Inductance:
Force, magnetic, 41—42 circuit viewpoint, 29
Index 617

definition, 29 equivalent circuit, 19

energy viewpoint, 32 Ohm’s law of, 20
linear, 30 reluctance of, 8
mutual, transformer, 83 toroid, 23
nonlinear, 30 units, 22
physical viewpoint, 33 Magnetic field energy, 32, 575
units, 30 Magnetic field intensity, 6, 19
Induction motor: coercive, 16
construction features, 132 units, 6
(see also Single- and Three-phase Magnetic force, 38
induction motors) Magnetic flux, definition, 5
Inductor, 29 fringing, 27
Infinite bus, 264 units, 6
Iron-core reactor, 51 Magnetic flux density, definition, 3
Insulation, induction motors, 211 right-hand rule, 3
Interpoles, 139, 329 unit of, 3, 5
Inverter, 199 Magnetic moment, 11
Magnetic potential difference, 20
Magnetic saturation, 13
L Magnetic starters, dc motors, 370
three-phase motors, 212
definite-time type, 214
Laminations, purpose of, 37
Magnetism:
in dc machine, 138
Curie point, 13
Leakage flux, primary, 58
diamagnetic, 11
secondary, 59
domain theory, 12
Leakage reactance (see Reactance,
ferromagnetic, 12
Synchronous generator, Three-
hysteresis, 16
phase induction motor, and
paramagnetic, 11
Transformers)
saturation, 13
Linear commutation, 327
theory of, 11
Lenz’s law, 53
Magnetization curve, 13
Lines of force, 2
dc machine, 309, 323
Long-shunt connection, 316
normal, 15, 16
Loop transfer function, 509
universal plot, 14
Losses: Magnetizing current, transformer, 63
in dc generator, 322
Magnetizing reactance, transformer, 67
in dc motor, 346 Magnetomotive force, mmf, 7
rotational, 168 relation to ampere-conductor
in single-phase motor, 409 distribution, 565
in synchronous motor, 275 relation to current sheet, 563
in three-phase motor, 169 Maximum torque:
in transformer, 80 single-phase motor, 408
three-phase motor, 179
M Maxwells, lines of flux, 21
Mechanical degrees, 567
Magnetic circuit: Mmf distribution, 564
analogy with electric circuit, 18 Mnf per pole, peak value, 593
computations, 21 Motor, defined, 123
618 Index

Multispeed motor, 189 salient-pole synchronous motor, 281

Mutual flux, 58 transformers, 99
Mutual inductance: Permanent magnet dc motor, 365
electric machine, 577 alnico type, 365, 366
transformer, 83 ceramic type, 365, 366
Permanent-split capacitor motor, 417
Permeability, 4
N relative, 4
Permeance, 41
Nameplate rating, transformer, 70 Phasing procedure, transformer:
Negative feedback, 509 delta-delta connection, 105
Negative sequence, symmetrical delta-wye connection, 108
components, 401, 434 Phasor diagram, transformer, 58, 59
No-load characteristic, 226 salient-pole generator, 249, 251
No-load tests: single-phase motor, 396, 415
single-phase motor, 412 synchronous generator, 232
three-phase motor, 182 synchronous motor, 275
transformer, 75 unbalanced four-phase system, 437
Normal magnetization curve, 15 Pitch factor, windings, 590
Null position, 494, 496 Pole pitch, 568
Null voltage, 597 Positive sequence symmetrical
components, 401, 433
Potier reactance, 244
O Power:
cylindrical rotor, synchronous
Oersted, 21 machine, 256
Open-circuit characteristic (see No-load salient-pole synchronous machine, 258
characteristic) Power angle, 256, 274, 277
Open-circuit test: Power factor control, 283
synchronous machine, 226, 237 Power flow diagram:
transformer, 75 dc generator, 322
Overcommutation, 329 dc motor, 346
Overcompounding, 314 single-phase motor, 409
Overexcitation, 274 synchronous motor, 275
Overload protection, 212 three-phase motor, 169
Power formula, machine design, 360
Primary leakage, flux, 58
P
reactance, 58
Primary winding, transformer, 51
Parallel operation of synchronous Pull-out torque, 287
generators, 260
Paramagnetism, 11
Peak mmf per pole, polyphase winding,
593 Q
Per unit calculations:
definition of base values, 98, 253 Quadrature-axis path, 245, 576
salient-pole synchronous generator, — Quadrature-axis synchronous reactance,
252 249, 250, 604
Index 619

R Residual voltages, in synchros, 500

Resistance commutation, 328
Ratings: Resistance-start, split-phase motor, 396
dc motors, 369 Retentivity, 15
single-phase motors, 426 Revolving magnetic field, 154
stepper motors, 473 conditions for, 157, 394
synchronous motors, 287 mathematical description, 158
three-phase motors, 206 Right-hand rule, 54
Ratio of transformation, Rotational losses, 168
autotransformer, 92 Rotor, 124
squirrel-cage motor, 599
synchros, 479
three-phase motor, 161, 597 S
transformer, 57
Reactance: Salient-pole synchronous machine:
armature reaction, direct axis, 248 construction, 134
armature reaction, quadrature axis, cross-magnetizing effect of armature
249 reaction, 247
equivalent, transformer, 69 demagnetizing effect of armature
leakage, in transformers, 58, 60 reaction, 247
magnetizing, transformer, 67 general method of analysis, 252
Potier, 244 phasor diagram, 249, 251
saturated, synchronous, 239, 601 power, electromagnetic, 258
stator leakage, 164, 242 power, reluctance, 258
synchronous, 235 two-reactance method, 245
direct-axis, 250 voltage regulation, 255
quadrature-axis, 250 Saturated synchronous reactance, 239,
three-phase motor, equivalent leakage, 601
186 Saturated synchronous reactance
stator leakage, 164 method, 239
Reactive power, control of, 261 based on resultant flux, 600
Rectifier circuits: Saturation, magnetic circuit, 13
single-phase, full-wave, 290, 373 effect on magnetizing current, 62
three-phase, diodes, 197 Schrage brush-shift motor, 194
three-phase, SCR, 198, 292, 376 Secondary leakage, flux, 59
Reduction factors: reactance, 59
squirrel-cage, 599 Self inductance, 84, 575
wound-rotor, 597 Series field winding, 309
Relay, 38 Series generator, 311
Reluctance, magnetic: Series motor:
definition, 8 applications, 368-369
differential, 9 equivalent circuit, 345
motor, 471, 576 power flow diagram, 346
units, 9 speed control, 352
Reluctance power, 258, 579 by adjustable armature voltage, 359
Reluctance torque, 471, 579 by adjustment of pole flux, 352
Remote positioning, with synchros, 485 by external armature resistance, 356
Residual flux density, 15 speed-torque characteristic, 341
620 Index

Servomechanism, 484 capacitor-start, capacitor-run motor,

Servomotor (see Two-phase servomotor) 419
Shaded-pole motor, 419 capacitor-start, induction-run motor,
Shear line, PM motor, 367 418
Shell-type transformer, 61 effect of rotor resistance on, 408
Short-circuit ratio, 240 hysteresis motor, 423
Short-circuit test: permanent-split motor, 419
single-phase motor, 412 shaded-pole motor, 420
synchronous generator, 227 universal motor, 423
three-phase motor, 184 universal motor, 421
transformer, 77 Singly-excited machines, 153
Short-shunt connection, 317 Sinusoidal winding, advantages of, 487
Shunt generator, 308 SI units, 21
Shunt motor: Skin effect, deep-bar squirrel-cage, 209
applications, 368-369 Slip, three-phase induction motor, 163
controllers, 370 for maximum torque, 178
equivalent circuit, 345 single-phase motor, 399
power flow diagram, 346 two-phase motor, 443
speed control, 349, 373 Slip frequency, 163
by adjustable armature voltage, 358 Slip test, 602
by adjustment of pole flux, 349 Space displacement angle, three-phase
by external armature resistance, 354 motor, 175
speed regulation, 338 Speed control:
speed reversal, 363 compound motor, 349, 354, 358
speed-torque characteristics, 336 dc motor, 348-363
Signal flow graph, 508 (see also Block with electronics, 373
diagrams) series motor, 352, 356, 359
Silicon controlled rectifiers, 197 shunt motor, 349, 354, 358
switching sequence, 200 three-phase motor, 163
Single-phase induction motor: via electronics, 196
- applications, 426 by external rotor resistance, 191
blocked-rotor test, 412 by injected armature voltage, 192
capacitor types, 417-419 Speed regulation, 338
characteristics, 412 Speed reversal, 363
equivalent circuit: Split-phase motor, 396
double revolving field theory, 398 Squirrel-cage induction motor, 133
symmetrical components, 401 NEMA, class A, 205
method of operation, 400 NEMA, class B, 208
no-load test, 412 NEMA, class C, 209
performance calculations, 408 NEMA, class D, 210
power flow diagram, 409 skin effect in, 209
resistance-start, 396 Starters (see Magnetic starters)
rotating field, 394, 397 Starting torque:
sequence impedances, 401 single-phase motor, 395
shaded-pole type, 420 three-phase motor, 177
slip, 399 State equations:
torque-speed curves, 406 amplidyne generator, 546
Index 621

de generator, 521, 522 short-circuit characteristic, 227

de generator—dc motor, 527 short-circuit ratio, 240
de motor, 516, 578 slip test, 602
field-controlled de motor, 535 synchronizing procedure, 260
RL circuit, 510 synchronous reactance, 235
two-phase servomotor, 539-540 three-phase generated voltage, 224
Ward-Leonard system, 524, 529 two-reactance method of analysis, 245
State variable, 509 zero pf characteristic, 243
Static errors, in synchros, 499 Synchronous machine, defined, 153
Static stability, synchronous machines, Synchronous motor:
way amortisseur winding, 273
Stepper motor: applications, 285
construction features, PM type, 459 basic torque equation, 297
drive amplifiers, 465 brushless excitation, 294
half-stepping, 469 equivalent circuit, 274
method of operation, 459 field excitation by rectifiers, 286
switching sequence, 467, 470 how load is accommodated, 276
ratings, 473 performance calculation, 275
reluctance type, 471 phasor diagram, 275, 279
torque-speed curves, 472 power angle, 274, 277
translator logic, 466 power factor control, 283
Symmetrical components: power flow diagram, 275
applied to two-phase motor, 401 static stability, 277
two- and four-phase systems, 432 as a synchronous capacitor, 284
Synchronizing procedure, 260 three-phase rectifier for, 292
Synchronous capacitor, 284 V-curves, 284
Synchronous generator: Synchronous reactance:
armature reaction, 233 cylindrical-pole machine, 235
in field amperes, 230 direct-axis, 250
reactance, 233 quadrature-axis, 250
brushless excitation, 294 saturated, 239
construction features, 134 unsaturated, 235
control of reactive power, 261 Synchronous reactance method, 235
control of real power, 263 Synchronous speed, 162, 272
equivalent circuit, 235 Synchros:
excitation voltage, 231 advantage of sinusoidal winding, 487
general method of analysis, 226, 238, applications, 482
py) construction features, 476
leakage reactance, 242 control receiver and transmitter, 476
no-load characteristic, 226 control differential, 478
nonlinear analysis, 238 control transformer, 477
phasor diagram, cylindrical pole, 232 as error detector, 482
salient-pole, 249, 251 governing equations, 481—482
power, 255 harmonic displacement errors, 499
salient-pole theory, 245 mmf equation, 492
saturated synchronous reactance null voltage, 494
method, 239 remote positioning, 485
622 Index

Synchros (cont.) rotational losses, 168

residual voltages, 499 slip, at maximum torque, 178
static errors, 500 slip speed, 163
for torque transmission, 484 space displacement angle, 175
Synchroscope, 261 speed control, 188
squirrel-cage rotor, 133, 205
T starting torque, 177
stator-referred rotor quantities, 167
Temperature effects on motor life, 211 torque-speed curves, squirrel-cage
Thévenin’s theorem, applied to motor motor, 208
analysis, 180 Three-phase transformer connections,
Third harmonic voltages: 104
delta-delta connection, 108 Thyristor drives, 198, 373
single-phase transformer, 63 Time constant, 507, 513, 519, 521, 525,
wye-wye connection, 111 $26 9531,:53737 542
Three-phase induction motor: Torque:
ampere-conductor distribution, 174, breakdown, 179
176 conservation of energy, 127, 143
applications, 206, 207 general equation, 573, 578
armature mmf, 174, 176 ac machine, 568
blocked-rotor test, 184 dc machine, 570
breakdown torque, 179 of mutual coupling, 579
classes, 205 practical forms:
computation of performance, 168 ac machines, 141
construction features, 132 dc machines, 145
controllers, 211 production of, 126
deep-bar rotor, 205 reluctance, 574, 579
double squirrel-cage rotor, 205 in two-phase servomotor, 444
effect of rotor resistance on torque- see also DC motor, Single-phase
speed curve, 179 motor, Three-phase motor, and
effect of temperature rise, 211 Synchronous motor
electronic speed control, 196 Torque transmission, in synchro, 484
equivalent circuit, 164, 168 Torque-speed curves:
equivalent leakage reactance, 186 de motors, 336-345
equivalent resistance, 185 effect of rotor resistance:
equivalent resistor for mechanical single-phase motor, 397
load, 167 three-phase motor, 179
gap power, 166 single-phase motors, 406
how torque is developed, 173 stepper motor, 473
maximum torque, 179 three-phase squirrel-cage motor, 208
parameters from no-load tests, 182 two-phase servomotor, 449
phasor diagram, 184 Transfer functions:
power flow diagram, 169 amplidyne generator, 540, 542, 544
reactances, leakage, 164 armature-controlled dc motor, 529
magnetizing, 164, 183 characteristic equation, 507
ratings and classifications, 206 closed-loop system, 509
reduction factor for voltage, 161 de generator, 520, 521
revolving field, 154 de generator—de motor, 524, 528
 Index
623
dc motor, 510
performance calculations, 441—448
definition, 506
symmetrical components, 438
field-controlled dc motor, 533
torque, 444, 447
loop, 509
torque-speed curves, 449
RL circuit, 506
transfer function, 535
transformer, 71
Two-reactance method, 245
two-phase servomotor, 535, DEE wets
Two-value capacitor motor, 419
voltage control system, 549, 550
Transformer:
autotransformers, 89
U
balance current, 60
base values of, 99
conventional, 56 Undercommutation, 328
construction features, 61 Undercompounding, 314
delta-delta connection, 105 Underexcitation, 275
delta-wye connection, 108 Undervoltage protection, 212
efficiency, 80 Units, magnetic, 22
equivalent circuits, 65, 68 Universal motor, 421
coupled-circuit viewpoint, 86 Unsaturated synchronous reactance, 235
equivalent leakage reactance, 69
equivalent resistance, 69
V
exciting current, 63
frequency response, 71 e
input impedance, 71 V-curves, 284
Lenz’s law, 53 Voltage:
magnetizing current, 62 build-up, in dc generator, 309
magnetizing reactance, 67 control system, 546
mutual inductance, 83 excitation, 231
nameplate of, 70 generation of dc, 304
open-circuit test, 75 induced (see induced voltages)
per unit calculations, 98
phasor diagram: BUD ea) ;
relations in synchros, 479
residual, 500
Hahn ze three-phase, generation of, 225
;
primary leakage reactance, 158 Voltage regulation, 81
primary winding, 51 cylindrical-pole synchronous
secondary leakage reactance, 59 generator, 239
secondary winding, 51 dc generator, 324
short-circuit test, 77 salient-pole synchronous generator,
three-phase circuits with, 104 255
transfer function, 71 transformer, 81
voltage regulation of, 81
wye-delta connection, 113
wye-wye connection, 111 WwW
Translator logic, stepper motors, 466
Turbogenerator, 224 Ward-Leonard system, 524
Two-phase servomotor: Winding:
equivalent circuit, 441 amortisseur, 273
 Index
624

Winding factor, 141

Winding (cont.)
armature, 310 Wound-rotor induction motor, 132, 190
auxiliary, 396 Wye-delta connection, 113
compensating, 330 Wye-wye connection, 111
distribution factor of, 592
pitch factor of, 590
reduction factors of, 597
Z
series field, 301
shunt field, 310
sinusoidal distribution of, 487 Zero sequence symmetrical component,
432
wave, 306
Zero pf characteristic, 243
Wave winding, 306
 VLTIV.

Athens Regional Library System

3207 00094 4811

423600

REGIONAL LIBRARY SERVICE

ATHENS AREA
 VINCENT DEL TORO

_—BASIC.
_ELECTRIC
This comprehensive treatment of basic electric machines is comprised of two parts.
Part | is devoted to an exposition of those topics that can be considered as
constituting the basic subject matter of a core course in electric machinery. Part II is
concerned with a series of eclectic topics that easily builds on the foundation
established in the first part.

The first three chapters in Part | establish the UTaTol-Tanl-1e

cl er-(o1.<olcelU
alomm al(olg
makes it possible in later chapters to analyze the conventional electric machines.
namely, single-phase and three-phase induction motors, dc generators and motors,
and synchronous generators and motors. Part || includes in-depth treatment of
these specialized topics: two-phase motors such as the servomotors, stepper
motors, and the synchro. While analysis of the machines in Part | is restricted to
steady-state behavior, an entire chapter in Part I places emphasis on formulations
that allow dynamic as well as steady-state behavior to be described.

ISBN O-13-Ob014b-2 |
9000

PRENTICE HALL, Englewood Cliffs, N.J. 07632 9°780130"601469 |