Electrical Machines II

Dynamic Behavior, Converter Supply and Control

based on a lecture of
Univ.-Prof. Dr.-Ing. Dr. h.c. Gerhard Henneberger

at
Aachen University


Preface
These notes represent the state of the lecture Electrical Machines II at Aachen University
for the summer term effective from 2003. Some extensions of the subject matter, which are
beyond the scope of the lecture, are included for self-studies at suitable location.
In the lecture the dynamic behavior of the DC-, induction- and synchronous- machine are
discussed. Also voltage and frequency variable excitation with power converters and control
methods needed for power generation and electrical drives are treated. In addition to the
analytical solution of the differential equations under simplifying assumptions, numerical
solution by means of computers will be demonstrated. Also these simulations take the power
converter and control system into account, which is shown on practical applications.
The areas basics, mode of operation, structure and steady state operating behavior are topics
of the lecture Electrical Machines I, which is subject to the lecture held in winter terms.
Focus is put on provision of clear understanding of the physical context. Despite plain
description required accuracy is not reduced.
This script is supposed to provide an all-embracing knowledge as a basis for both the
continuation of their studies and later in practice to deal with electrical machines in detail.
The lecture electrical machines II was especially elaborated for students in main course of
the major Electro-Technique and Electronics. The knowledge of contents of the lecture
Electrical Machines I is presumed.

Please note: This script represents a translation of the lecture notes composed in German.
Most subscriptions to appear in equations are not subject to translation for conformity
purposes.







Aachen, April 2002 Gerhard Henneberger













Revision 01: Busch, Schulte, Lwenstein, Dezember. 02
Revision 02: Schulte, July 2003



Content
1 Direct and quadrature axis theory 1
1.1 Introduction........................................................................................................... 1
1.2 General rotating field machine.............................................................................. 3
1.3 Requirements and approach.................................................................................. 5
1.4 Transformation from 3 to 2 phases ....................................................................... 6
1.5 Transformation of the 2 phase rotor and stator to an arbitrary revolving
coordinate system.................................................................................................. 9
1.6 Voltage equations in the arbitrary system........................................................... 12
1.7 Balance of power and torque .............................................................................. 13
1.8 Compilation of the equations of the direct and quadrature axis theory.............. 15
1.9 Space vectors ...................................................................................................... 16
2 DC machine 19
2.1 Basics .................................................................................................................. 19
2.2 Dynamic set of equations.................................................................................... 21
2.3 Separately excited DC machine .......................................................................... 25
2.4 Coarse-step connection of DC shunt machines .................................................. 31
2.5 Cascade control of converter-fed PM DC machines........................................... 34
2.6 DC series-wound-machine as traction drive in pulse control operation............. 36
3 Induction machine 39
3.1 Dynamic equation set.......................................................................................... 39
3.2 Steady state operation ......................................................................................... 41
3.3 Rapid acceleration, sudden load change ............................................................. 43
3.4 Induction machine in field-oriented coordinate system...................................... 48
3.5 Field-oriented control of induction machines with injected currents ................. 52

ii


3.6 Steady-state operation using variable frequency and voltage converter............ 55
3.7 Field-oriented control of induction machines with applied voltages.................. 62
4 Synchronous machine 65
4.1 Dynamic system of equations ............................................................................. 65
4.2 Steady-state operation of salient-pole machines at mains power supply............ 68
4.3 Determination of X
d
and X
q
................................................................................ 73
4.4 Sudden short circuit of the cylindrical-rotor machine ........................................ 74
4.4.1 Physical explication of the sudden short circuit .................... 82
4.4.2 Torque at sudden short circuit ............................................... 83
4.5 Sudden short circuit of salient-pole machines .................................................... 85
4.5.1 Analytical calculation ............................................................ 85
4.5.2 Numerical solution................................................................. 88
4.6 Transient operation of salient-pole machines ..................................................... 96
4.6.1 Power supply operation ......................................................... 100
4.6.2 Solitary operation................................................................... 103
4.6.3 Summary and conclusion of transient operation.................... 104
4.6.4 Scheme for transient operation .............................................. 105
5 Servo-motor 107
5.1 General design and function ............................................................................. 107
5.2 Dynamic set of equations.................................................................................. 108
5.3 Steady state operation ....................................................................................... 110
5.4 Dynamic behavior ............................................................................................. 114
5.5 Voltage- and current waveforms of servo-motors with rotor position encoder 118
6 Appendix 121
6.1 Formular symbols ............................................................................................. 121
6.2 Units.................................................................................................................. 124
7 Literature reference list 127


1 Direct and quadrature axis theory
1.1 Introduction
Up to now always steady-state operation was presumed, i.e. operation on power system with
constant voltage (DC-, AC or three-phase system), constant speed and also without electrical
switching operation and mechanical load variation. Such mode of operation nearly never
occurs in real world applications, neither at drive engineering nor at central electricity supply.
The electrical machines are the main parts in drive engineering, which are the controlled
system in the actuating system. Final controlling element - today mostly a static converter -
and analog or digital control complete the drive system with speed control. DC motors are of
minor importance nowadays, since recent advances in power semiconductor and
microprocessor technology, which increased the relevance of induction- and EC-motors for
electrical drives. High dynamic responses at acceleration or braking and a short setting time in
case of mechanical load changes are demanded.
In central power supply the fundamental basis is provided by synchronous generators. They
are used in thermal power stations in conjunction with steam turbines in form of a high speed
cylindrical-rotor generator and in hydro-electric power plants as low speed salient-pole
generators. The voltage is preferably sinusoidal and the frequency should be constant. The
speed of the driving machine and the terminal voltage must be controlled in a kind of manner,
that sudden load change provoked by short circuit fault or cut-off do not result in large
changes of the frequencies and voltages. Also operation in parallel with other generators must
be possible without the occurrence of oscillations.
The knowledge of the response characteristic of the controlled system electrical machine is
therefore an important precondition for the design and the prediction of the operational
performance of an electrical machine.
While in dynamical operation the transfer function of the electrical and mechanical quantities
can always be described by a set of differential equations. The number of these equations
depends on the number of energy storage mechanism, i.e. a voltage equation for each coil and
for the rotor mass an equation of motion.
To limit the complexity of the calculation for three-phase machine with non constant mutual
inductances, some restrictive conditions must be made regarding electrical and magnetic
symmetry and fundamental wave. Additionally two kinds of transformation are required. The
three-phase winding will be transformed to a two-phase system, which magnetically
decouples the rotor and stator windings and also reduces the number of equations per
windings from three to two. The other transforms from a stationary to an arbitrary rotating
coordinate system. Performing this transformation will result in constant mutual inductances
and the feasibility to take magnetic asymmetry of some machine parts into account for
instance the different reactances
q d
X X of salient pole machines.

2 Direct and quadrature axis theory


The transformations have to be power invariant and so that resistances and inductivities stay
unchanged. The torque differential equation will be derived from the power balance and the
torque equilibrium.
The description of polyphase machines requires at least five differential equations, i.e. two for
the rotor, two for the stator and one for the rotating masses. Also the dynamic equations for
the DC machine arise from the quadrature-axis theory. In that case only three equations are
needed for field, armature and mass.
The derivation of a dynamic equation system of the polyphase machine will be accomplished
in common by the consequential use of the quadrature-axis theory. The often used
representative space vector method results in identical equations but loses clearness if
invariance of the power is taken into account. The correlation between quadrature-axis theory
and space vector diagram will be pointed out later.

Direct and quadrature axis theory 3


1.2 General rotating field machine

Fig. 1: rotating field

voltages:
[ ] [ ]
T
w u
u u u U
v S 1 1
, ,
1
[ ] [ ]
T
w v u R
u u u U
2 2 2
, , (1.1)

currents:
[ ] [ ]
T
w v u S
i i i I
1 1 1
, , [ ] [ ]
T
w v u R
i i i I
2 2 2
, , (1.2)

flux linkages:
[ ] [ ]
T
w v u S 1 1 1
, , [ ] [ ]
T
w v u R 2 2 2
, , (1.3)

resistances:
[ ]
1
1
1
]
1

1
1
1
0 0
0 0
0 0
R
R
R
R
S
[ ]
1
1
1
]
1

2
2
2
0 0
0 0
0 0
R
R
R
R
R
(1.4)

voltage equations:
[ ] [ ] [ ] [ ]
S S S S
dt
d
I R U + [ ] [ ] [ ] [ ]
R R R R
dt
d
I R U + (1.5)

flux linkage equations:
[ ] [ ] [ ] [ ] [ ]
R S S S S
I M I L + [ ] [ ] [ ] [ ] [ ]
S R R R R
I M I L + (1.6)
A generalized rotating field machine consists
of 3 phase rotor and stator winding (Fig.1).
Because the rotor revolves with ) (t , the
inductances are depending on the rotor
position.
The complete set of equations will be
presented in matrix form.

4 Direct and quadrature axis theory


inductances:
[ ]
1
1
1
1
1
1
1
]
1

,
_

,
_

,
_

,
_

,
_

,
_

1 1 1 1
1 1 1 1
1 1 1 1
3
2
cos
3
2
cos
3
2
cos
3
2
cos
3
2
cos
3
2
cos
L L L L
L L L L
L L L L
L
hw hw hw
hw hw hw
hw hw hw
S
(1.7)
[ ]
1
1
1
1
1
1
]
1

,
_

,
_

,
_

,
_

,
_

,
_

2 2 2 2
2 2 2 2
2 2 2 2
3
2
cos
3
2
cos
3
2
cos
3
2
cos
3
2
cos
3
2
cos
L L L L
L L L L
L L L L
L
hw hw hw
hw hw hw
hw hw hw
R
(1.8)

mutual inductances:
[ ]
( )
( )
( ) 1
1
1
1
1
1
]
1

,
_

,
_

,
_

,
_

,
_

,
_

cos
3
2
cos
3
2
cos
3
2
cos cos
3
2
cos
3
2
cos
3
2
cos cos
W W W
W W W
W W W
S
M M M
M M M
M M M
M (1.9)

[ ] [ ]
T
S R
M M (1.10)

W
L and
W
M are alternating inductances. The number of turns of stator and rotor windings
differ.
In the form presented above the dynamic set of equations is physically complex and it is quite
complicated to deal with it without using computers.

Direct and quadrature axis theory 5


1.3 Requirements and approach
The following requirements are be made for the quadrature axis theory for simplification
purposes:
1. Rotor and stator winding only excite spatial sinusoidal current linkages. That means,
that only the fundamental wave of the current linkages is taken into account and
winding factors of all harmonics are supposed to be zero.
2. There is no saturation, i.e. the magnetic conductance is independent of the current
linkage. The magnetic voltage drop is negligible.
3. The machine is fully symmetric, i.e. constant air gap around the whole circumference
and the influence of the slotting is negligible. Within the rotor or stator the self-and
mutual inductances are independent from the rotor position This precondition may be
partly ignored. If the stator respectively the rotor have two magnetic or electric
preferred perpendicular axes, the quadrature axis theory can still be applied through
the choice of an asymmetric part fixed coordinate system.
4. The neutral point is not connected. Therewith the number of voltage equations in rotor
and stator will be reduced from three to two.

v u w w v u
i i i i i i + + 0 (1.11)

Also this precondition can be bypassed by separation and extra handling of the zero phase-
sequence system.

The further approach is as follows:
1. Power invariant transformation of both three-phase systems (rotor and stator) to two-
phase systems, whose axes are perpendicular to each other and no interaction takes
place. Additionally the rotor will be referred to the stator winding.
2. Transformation of the steady stator winding and rotating rotor winding to an arbitrary
system, rotating with angular velocity. Thereby the mutual inductances no longer
depend on the rotor position.
3. Setting up the voltage equations for rotor and stator in the transformed system,
rotating with arbitrary angular speed.
4. Determination of the torque from balance of power.

Afterwards the dynamic, power converter and the transient behavior of DC, induction and
synchronous machine will be investigated both analytically as well as numerically.

6 Direct and quadrature axis theory


1.4 Transformation from 3 to 2 phases
W
1
V
1
i
U1
i
W1
+
U
1
+
A1
B1
i
A1
i
B1
i
V1
U
2
W
2
V
2
A2
B2
i'
A2
i'
B2
[T
32
]
[T
32
]
[T
32
]
-1
[T
32
]
-1
w
1

(t)

(t)
+
i
U2
i
W2
i
V1
+
w
2

2
Stator
Rotor

Fig. 2: three-to-two phase transformation

Conditions for rotating fields:
three-phase winding:
3
2

el mech
p
two-phase winding:
2

el mech
p

For a power invariant transformation, equation 1.12 must apply:

Ph Ph Ph Ph
I U I U S
2 2 3 3
2 3 (1.12)

According to this, there are different opportunities:

Ph Ph
U U
3 2
2
3
,
Ph Ph
I I
3 2
(1.13)

Ph Ph
U U
3 2
,
Ph Ph
I I
3 2
2
3
(1.14)

Ph Ph
U U
3 2
2
3
,
Ph Ph
I I
3 2
2
3
(symmetrical) (1.15)

Which of these possibilities is best suitable?
1 1
2
3
w
1 1
2
3
w

Direct and quadrature axis theory 7


If invariant transformation of resistances and inductances is demanded besides invariant
transformation of power, only the symmetric option is possible, which will be shown by the
following considerations:


Cu
m
L
m
A
m l w
q
l w
R

2
2
with
w m
A
q
Cu
L

2
(1.16)

D l
p
w m
L

,
_

2
2
2
1 1
0
(1.17)

To conserve previous resistances and inductances, the number of windings results in:


( )
( ) ( )
( ) ( )
Ph Ph
Ph Ph
w w
w w
const w m
3 2
2
2
2
3
2
2
3
2 3

(1.18)

Therefore follows from the equality of the current linkage for the currents:


( ) ( )
Ph Ph
Ph
Ph
Ph
Ph
I I
I
p
w
I
p
w
3 2
2
2
3
3
2
3
2
4
2
2
2
4
2
3

(1.19)

This means equal air-gap induction and for the case of same geometric data: equal air-gap
flux.

Therefore applies for the voltages:


( )
( )
2
3
2
2
3
2
3
2
3
2

Ph
Ph
Ph
Ph
Ph
Ph
U
U
w
w
U
U

(1.20)
The symmetrical transformation to a system with ( ) w
2
3
is therefore power-, resistance-
and inductance invariant and will be used below.

The conversion of the stator will be performed as follows:

If a three-phase system is intended to be replaced by a two phase system, it is advisable to set
the axis of winding A so that it coincides with the orientation of axis U. The axis B is
perpendicular to A. The turn ratio is:

( ) ( )
Ph Ph
w w
3 1 1 2 1 1
2
3
. (1.21)

8 Direct and quadrature axis theory


Both systems must have the same ampere turns across the air gap. Then the currents must be

for A-axis:

1 1 1 1 1 1 1 1
2
3
3
4
cos
3
2
cos
A w v u
i w i i i w

,
_

+ +

(1.22)
( )
1 1 1
1
1
2
3
2
1
2
A v u
v
u
i i i
i
i
,
_

(1.23)

1 1
2
3
u A
i i (1.24)
for B-axis:

1 1 1 1 1 1 1
2
3
6
5
cos
6
cos
B w v
i w i i w

,
_

+

(1.25)

1 1 1
2
3
2
3
2
3
B w v
i i i (1.26)
( )
2 2
1
2
1
1 1
1 B
v u
v
i
i i
i
(1.27)

2
2
1
1
1 v
u
B
i
i
i +
(1.28)
The transformation form of the three phase system to the equivalent two phase system can
also be written in matrix form:
[ ]
1
]
1

1
]
1

1
1
32
1
1
v
u
B
A
i
i
T
i
i
with [ ]
1
1
1
1
]
1

2
2
1
0
2
3
32
T . (1.29)

The conversion of the voltages will be performed in a similar way, because the symmetric
transformation was chosen:

[ ]
1
]
1

1
]
1

1
1
32
1
1
v
u
B
A
u
u
T
u
u
. (1.30)

Just as well a two phase system can be transformed to a three phase system by reverting the
equations:
[ ]
1
]
1

1
]
1

1
1 1
32
1
1
B
A
v
u
i
i
T
i
i
with [ ]
1
1
1
1
]
1

2
1
6
1
0
3
2
1
32
T (1.31)
[ ]
1
]
1

1
]
1

1
1 1
32
1
1
B
A
v
u
u
u
T
u
u
(1.32)

whereas [ ] [ ] [ ] 1
1
32 32


T T . (1.33)

Direct and quadrature axis theory 9


The conversion of the rotor from three to two phases can be performed in analogy.
Additionally the rotor will be translated to the number of stator turns per unit length, which is
marked in the following with a dash "
,
".


2 2
1 1

w
w
(1.34)

[ ]
1
]
1

1
]
1

v2
u2
32
,
2
,
2
u
u
T
u
u
B
A
(1.35)

[ ]
1
]
1

1
]
1

v2
u2
32
,
2
,
2
i
i

1
T
i
i
B
A
(1.36)

The conversion of the flux linkages of the stator and rotor winding takes place in the manner
from three to two phases and vice versa.

[ ]
1
]
1

1
]
1

v1
u1
32
1
1
T
B
A
(1.37)

[ ]
1
]
1

1
]
1

v2
u2
32
'
2
'
2
T
B
A
(1.38)


1.5 Transformation of the 2 phase rotor and stator to an
arbitrary revolving coordinate system
To obtain constant mutual inductances, e.g. for salient pole synchronous machines or for the
application of specific control algorithms (e.g. field oriented control), rotor or stator of three
phase machines must be transformed to stator or rotor. Sometimes it might be useful to
transform the rotor or stator to a coordinate system to rotate with the air gap flux.
In the following a generic transformation from the inactive stator and the with
dt
d
rotating
rotor to a coordinate system, rotating with arbitrary angular speed
dt
d
is presented.
The equation system is then applicable to various machine types or can be chosen freely:

0 inactive coordinate system,
t coordinate system rotating with rotor speed (
dt
d
),
t
1
coordinate system rotating with synchronous speed (line-frequency),
t

coordinate system rotating with air-gap flux (variable frequency).

10 Direct and quadrature axis theory


+
A1
B1
i
A1
i
B1
A2
B2
i'
A2
i'
B2
(t)
+
d1
q1
i
d1
i
q1

(t)
+
d2
q2
i'
d2
i'
q2

(t)
+
[T

]
[T

]
[T

]
-1
[T

]
-1
stator
rotor

Fig. 3: transformation to revolving coordinate system

While transforming the rotor to the rotating system care has to be taken since both system are
moving relative to each other with the difference angel .
[ ]
1
]
1

1
1
]
1

'
2
'
2
, '
2
'
2
B
A
q
d
i
i
T
i
i

(1.44)

[ ]
1
]
1

1
1
]
1

, '
2
'
2
, '
2
'
2
B
A
q
d
u
u
T
u
u

(1.45)

with the transformation matrix of the rotor to a revolving system

[ ]
( ) ( )
( ) ( )
1
]
1

cos sin
sin cos
T . (1.46)

The reverse transformation is performed with the inverse matrices:
[ ]
1
]
1

+ +
+

cos sin
sin cos
1
T (1.47)

[ ]
( ) ( )
( ) ( )
1
]
1

+
+ +

cos sin
sin cos
1
T (1.48)

[ ]
1
]
1

1
]
1

1
1 1
1
1
q
d
B
A
i
i
T
i
i

(1.49)

[ ]
1
]
1

1
]
1

1
1 1
1
1
q
d
B
A
u
u
T
u
u

(1.50)
Therefore equations for rotating
Cartesian coordinate systems in a plane
are used:

sin cos
*
+ y x x (1.43)
cos sin
*
+ y x y (1.43)

Carried forward currents and voltages
may be written in matrix form:

[ ]
1
]
1

1
]
1

1
1
1
1
B
A
q
d
i
i
T
i
i

(1.43)
[ ]
1
]
1

1
]
1

1
1
1
1
B
A
q
d
u
u
T
u
u

, (1.43)

with transformation matrix of stator to a
revolving system

[ ]
1
]
1

cos sin
sin cos
T . (1.43)

Direct and quadrature axis theory 11



[ ]
1
1
]
1

1
]
1

, '
2
2 1
'
2
, '
2
q
d
B
A
i
i
T
i
i

(1.51)

[ ]
1
1
]
1

1
]
1

'
2
'
2 1
, '
2
'
2
q
d
B
A
u
u
T
u
u

(1.52)

This transformation is also invariant of power, because the matrices are orthogonal, i.e.
[ ] [ ]
1
cos sin
sin cos

1
]
1

+ +
+

T T
T
(1.53)

[ ]
( ) ( )
( ) ( )
[ ]
1
cos sin
sin cos



1
]
1

+
+ +

T T
T
(1.54)

Since the same transformation is applied to voltages and currents, the quotient is constant, i.e.
the transformation is invariant to inductances and resistances.
The flux linkages in the arbitrary rotating system can now be written as

1
1
]
1

+
1
]
1

1
]
1

'
2
'
2
1
1
1
1
1
q
d
h
q
d
q
d
i
i
L
i
i
L (1.55)


1
]
1

+
1
1
]
1

1
1
]
1

1
1
, '
2
'
2 '
2 '
2
'
2
q
d
h
q
d
q
d
i
i
L
i
i
L (1.56)

After conversion to the same number of turns, rotor resistance and inductance are given by:

2
2 '
2
R R (1.57)


W

2
3
2
3
M L L
hw h
(1.58)

The total inductance is

h
L L L +
1 1
(1.59)


2
2 '
2
'
2
L L L L
h
+

(1.60)

and the leakage factors ensue to

( ) ( )
2 1
'
2 1
2
1 1
1
1 1

+ +

L L
L
h
(1.61)

h
L
L

1
1
,
h
L
L
'
2
2

. (1.62 a, b)
It is remarkable, that mutual inductances do not appear in the arbitrary rotating system any
longer. This is caused by the fact that transformed rotor and stator windings rotate with the
same angular speed. Only the axes d1-d2 and q1-q2 are magnetically coupled.

12 Direct and quadrature axis theory


1.6 Voltage equations in the arbitrary system
The transformation is to be performed in two steps:
1. Transformation from the three phase to the two phase system - non connected star point:

[ ] [ ] [ ]
1
]
1

+
1
]
1

1
]
1

+
1
]
1

1
]
1

1
]
1

1
1
1
1
1
1
1
32
1
1
32 1
1
1
32
1
1
B
A
B
A
v
u
v
u
v
u
B
A
dt
d
i
i
R T
dt
d
i
i
T R
u
u
T
u
u
(1.63)
[ ] [ ] [ ]
1
]
1

+
1
]
1

1
]
1

+
1
]
1

1
]
1

1
]
1

'
2
'
2
'
2
'
2
2
'
2
'
2
32
'
2
'
2
32
'
2
,
2
,
2
32
,
2
,
2
B
A
B
A
v
u
v
u
v
u
B
A
dt
d
i
i
R T
dt
d
i
i
T R
u
u
T
u
u
(1.64)
with [ ]
32
T being constant and independent of t.
2. Transformation of the inactive stator and the rotating rotor to an arbitrary rotating
coordinate system:

[ ] [ ] [ ] [ ]

'

1
]
1

+
1
]
1

1
]
1

1
]
1

1
1 1
1
1
1
1
1
1
1
q
d
B
A
B
A
q
d
T
dt
d
T
i
i
T R
u
u
T
u
u


[ ]
[ ]
[ ]

'

1
]
1

+
1
]
1

+
1
]
1

1
1 1
1
1
1
1
1
1
q
d
q
d
q
d
dt
d
T
dt
T d
T
i
i
R

1
]
1

+

+
1
]
1

+
1
]
1

1
1
1
1
1
1
1
d
q
q
d
q
d
dt
d
dt
d
i
i
R

(1.65)

Keep in mind, that in this case [ ]

T as well as [ ]

T are not independent of t.

[ ] [ ] [ ] [ ]

'

1
]
1

+
1
]
1

1
]
1

1
]
1

'
2
'
2 1
'
2
'
2 '
2
,
2
,
2
,
2
,
2
q
d
B
A
B
A
q
d
T
dt
d
T
i
i
T R
u
u
T
u
u


[ ]
[ ]
[ ]

'

1
]
1

+
1
]
1

+
1
]
1

'
2
'
2 1
'
2
'
2
1
'
2
'
2 '
2
q
d
q
d
q
d
dt
d
T
dt
T d
T
i
i
R





( )
1
]
1

+
1
]
1

+
1
]
1

'
2
'
2
'
2
'
2
'
2
'
2 '
2
d
q
q
d
q
d
dt
d
dt
d
i
i
R

(1.66)


The voltage equations consist of the following aspects:

voltage drop at the ohmic resistance
temporal alteration of the flux linkage
rotary voltages caused by
dt
d
or.
( )
dt
d



Direct and quadrature axis theory 13


It is required to take a closer look at the following special cases and distinguish between:
( ) 0 t : transformation to a stationary system: stator

1
]
1

+
1
]
1

1
]
1

1
1
1
1
1
1
1
q
d
q
d
q
d
dt
d
i
i
R
u
u
(1.67)

1
]
1

+
+
1
]
1

+
1
]
1

1
]
1

'
2
'
2
'
2
'
2
'
2
'
2 '
2 ,
2
,
2
d
q
q
d
q
d
q
d
dt
d
dt
d
i
i
R
u
u

(1.68)

( ) ( ) t t : transformation to a system, rotating with ( ) t : rotor


1
]
1

+

+
1
]
1

+
1
]
1

1
]
1

1
1
1
1
1
1
1
1
1
d
q
q
d
q
d
q
d
dt
d
dt
d
i
i
R
u
u

(1.69)

1
]
1

+
1
]
1

1
]
1

'
2
'
2
'
2
'
2 '
2 ,
2
,
2
q
d
q
d
q
d
dt
d
i
i
R
u
u
(1.70)

1.7 Balance of power and torque
To determine the torque in an arbitrary rotating system, the balance of power is to be set up.
The following matrices are therefore defined:
[ ]
1
1
1
1
1
]
1

'
2
'
2
1
1
q
d
q
d
i
i
i
i
i [ ]
1
1
1
1
1
]
1

'
2
'
2
1
1
q
d
q
d
u
u
u
u
u [ ]
1
1
1
1
1
]
1

'
2
'
2
1
1
q
d
q
d
[ ]
( )
( )
1
1
1
1
1
1
1
1
]
1

+
+


dt
d
dt
d
dt
d
dt
d
d
q
d
q

'
2
'
2
1
1
(1.71)

So that the power input appears as:
[ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ]
43 42 1
43 42 1
43 42 1
mech
P
T
dt
m
dW
T
V
P
T T
auf
i
dt
d
i i R i u i P + + , (1.72)
where
V
P represents the ohmic losses in the winding resistances,
dt
dW
m
the change of the
magnetic energy and
mech
P the mechanical power of the total system:


p
dt
d
M
p
M
M P
el
el
el mech

. (1.73)


14 Direct and quadrature axis theory


Torque derives from

( )
( ) ( )
( )

'

+ +
'
2
'
2
'
2
'
2 1 1 1 1 d q q d d q q d el
dt
d
i
dt
d
i
dt
d
i
dt
d
i
dt
d
p
M

( ) ( )
( )

'

+
dt
d
i i
dt
d
i i
dt
d
p
d q q d q d d q

'
2
'
2
'
2
'
2 1 1 1 1

( )
'
2 1
'
2 1 1 q d d q h
i i i i L p . (1.74)

The equilibrium of torque is valid:


2
2
dt
d
p
J
M
dt
d
J M M
W W el

+

+ (1.75)

dt
d
p p
n


1
2 (1.76)

Two different cases need to be distinguished:

( ) 0 t : steady coordinate system
( )
'
2
'
2
'
2
'
2 d q q d el
i i p M (1.77)

( ) ( ) t t : with the rotor rotating coordinate system
( )
1 1 1 1 q d d q el
i i p M (1.78)

Direct and quadrature axis theory 15


1.8 Compilation of the equations of the direct and
quadrature axis theory
W
1
V
1
+
U
1
+
A1
B1
U
2
W
2
V
2
A2
B2
[T
32
]
[T
32
]
[T
32
]
-1
[T
32
]
-1
w
1

(t)

(t)
+
+
w
2

2
d1
q1

(t)
+
[T

]
[T

]
-1
d2
q2

(t)
+
[T

]
[T

]
-1

Fig. 4: transformation of revolving 2-axis coordinate system

[ ]
1
]
1

1
]
1

1
1
32
1
1
v
u
B
A
i
i
T
i
i
[ ]
1
]
1

1
]
1

'
v2
'
u2
32
, '
2
'
2
i
i
T
i
i
B
A
(1.79)

[ ]
1
]
1

1
]
1

1
1
1
1
B
A
q
d
i
i
T
i
i

[ ]
1
]
1

1
1
]
1

'
B2
'
A2
, '
2
'
2
i
i

T
i
i
q
d
(1.80)

[ ]
1
1
1
1
]
1

2
2
1
0
2
3
32
T [ ]
1
]
1

cos sin
sin cos
T [ ]
( ) ( )
( ) ( )
1
]
1

cos sin
sin cos
T (1.81)

1
1
]
1

+
1
]
1

1
]
1

, '
2
'
2
1
1
1
1
1
q
d
h
q
d
q
d
i
i
L
i
i
L
1
]
1

+
1
1
]
1

1
1
]
1

1
1
, '
2
'
2 '
2 '
2
'
2
q
d
h
q
d
q
d
i
i
L
i
i
L (1.82)

1 1
2
3
w

1 1
2
3
w

1 1
2
3
w

1 1
2
3
w

16 Direct and quadrature axis theory



1
]
1

+

+
1
]
1

+
1
]
1

1
]
1

1
1
1
1
1
1
1
1
1
d
q
q
d
q
d
q
d
dt
d
dt
d
i
i
R
u
u

(1.83)


( )
1
]
1

+
1
1
]
1

+
1
1
]
1

1
1
]
1

'
2
'
2
'
2
'
2
'
2
'
2 '
2 , '
2
'
2
d
q
q
d
q
d
q
d
dt
d
dt
d
i
i
R
u
u

(1.84)

( )
W q d d q h el
M
dt
d
p
J
i i i i L p M +
2
2
'
2 1
'
2 1 1

(1.85)

1.9 Space vectors
Complex space vectors are often used in common in the literature. In the complex area a
rotating field can be represented by a rotating space vector with the angular velocity . The
position of the space vector describes the instantaneous maximum and the quantity of the
space vector describes the amplitude of the rotating field. The complex space vector is
constructed using the instantaneous values of the three phases, for example of the currents:

( ) ( ) ( ) ( ) ( ) t i a t i a t i t i
w v u
+ +
2
3
2
(1.86)
The operator
3
2

j
e a represents a spatial displacement of 120. With the factor
3
2
the
amplitude of the space vector is adjusted to the physical magnitude. The projection on the
particular winding axis results the instantaneous value of the particular phase current.

+Re
-Im
U
V W
1

Fig. 5: space vector illustration

As well as the currents, the
voltages and the flux
linkages can be defined
using space vectors.
e.g. 0 t
1
u
i ,
2
1

w v
i i

Direct and quadrature axis theory 17


If the dynamic equations of a rotating field machine are represented with complex space
vectors in a coordinate system, which is rotating with angular speed ( ) t , the following set of
equations is achieved, if the rotor is converted to the number of the stator windings:


S
S
S S
dt
d
j
dt
d
i R u +

+

1
(1.87)

( )
'
'
' '
2
'
R
R
R R
dt
d
j
dt
d
i R u

+

(1.88)


'
1 1 R h S S
i L i L + (1.89)

S h R R
i L i L +
1
'
2
'
(1.90)

( )
dt
d
p
J
i p M
R R el

+
'* '
Im
2
3
(* conjugate complex) (1.91)

If the complex space vectors are split up into their components


1 1 d q S
u j u u (1.92)
'
2
'
2
'
d q R
u j u u (1.93)

1 1 d q S
i j i i (1.94)
'
2
'
2
'
d q R
i j i i (1.95)

1 1 d q S
j (1.96)
'
2
'
2
'
d q R
j (1.97)

with d representing the negative imaginary axis and q representing the positive real axis,
a set of equations is gained, which is, except of the factor in the torque equation, similar to the
set of equations of the two-axis theory.


1
]
1

+

+
1
]
1

+
1
]
1

1
]
1

1
1
1
1
1
1
1
1
1
d
q
q
d
q
d
q
d
dt
d
dt
d
i
i
R
u
u

(1.98)


( )
1
]
1

+
1
]
1

+
1
]
1

1
]
1

'
2
'
2
'
2
'
2
'
2
'
2 '
2 ,
2
,
2
d
q
q
d
q
d
q
d
dt
d
dt
d
i
i
R
u
u

(1.99)

( )
dt
d
p
J
i i p M
q d d q el

+
'
2
'
2
'
2
'
2
2
3
(1.100)

The factor
2
3
in the torque equation shows, that the transformation is not invariant of power.
Instead of that the space vectors should be defined as follows (for example for the current):

( )
w v u
i a i a i i + +
2
3
2
(1.101)

Thus the according space vectors would loose clearness.



2 DC machine
2.1 Basics
Because of their easy control and because of economic reasons DC machines in combination
with thyristor converters or transistor amplifiers are today still often used as high dynamic
speed-variable drives in the small and middle power range.

The essential disadvantage is the commutator. Size and dynamic are limited by the
mechanical commutation. The current transfer with brushes has a high rate of wear and
requires maintenance.
But because of the mechanical commutation with the commutator, the armature current
linkage and the excitation field are always oriented in an optimal way for the generation of the
torque. With constant excitation there is a linear correlation between torque and current. Thus
a rapid and exact speed- and position control is possible.

The following pictures show the basic structure of controlled DC machines (For the
simplification machines with only two poles are shown).

n

A

Fig. 6: DC machine

Fig. 7: DC machine, servo motor
Fig. 5 shows a large machine in medium
power range with some 100 kW for
handling applications. The machine is
equipped with commutation poles and a
compensating-field winding, to improve
the commutation and to suppress the
armature reaction. The rotor and the field
frame are made of laminated steel, to
enable rapid changes of the magnetic
field.
Fig. 6 sketches a cylindrical servomotor
with radial field, which is typically
used in the range of some kW for rapid
positioning drives. The usage of rare-
earth permanent magnets and a slim-
line type rotor are benefiting because of
dynamic reasons. With permanent
magnets in the stator, the machine can
only be controlled in the armature
circuit.

20 DC machine

Fig. 8, 9: DC machine, disc type rotor (high dynamic)

Fig. 7 also illustrates a high-dynamic DC servomotor in the power range of some kW for
robotic device applications. The motor has an axial field and a double-side stator with AlNiCo
magnets. The ironless disc-type rotor is made of distributed wires or punched segments.


DC machine 21


2.2 Dynamic set of equations
The dynamic set of equations of DC machines can be directly deduced from the two axis
theory. Variables need to be set to 0 , i.e. 0
k
, regarding:

I
A
U
A
q
1
d
1
i
d1
= -I
F
q
2
d
2
i'
q2
= I
A
U
F I
F

Fig. 10: two-axis theory applied on DC machine

The mechanical angular speed ensues to:

p
dt
d

(2.1)

It has to be considered, that the rotor is converted to the number of the stator windings, which
is not usual for DC machines. Therefore it has to be cancelled.

The following substitutions apply:

F d
U u
1
,
F d
I i
1
,
F
R R
1
,
F
L L
1
(2.2 a-d)

0
1

q
u , 0
1

q
i (2.3)

0
'
2

d
u , 0
'
2

d
i (2.4)

A q
U u
'
2
,

I
i
A
q

'
2
,
A
R R
2 ,
2
,
A
L L
2 ,
2
(2.5 a-d)

F F d
I L
1
(2.6)

I
L
A
h q

1
(2.7)

F h d
I L
'
2
(2.8)

I
L
A
A q

2 '
2
(2.9)

The d1-axis corresponds to the
excitation field.

The q2-axis is the axis of the armature
quadrature-axis field.

The polarity of the excitation field is
reversed, to turn the anti-clockwise
rotation positive. There are no windings
q1 and d2.

22 DC machine


Therefore the following equations are achieved:

dt
dI
L I R U
F
F F F F
(2.10)
( )
F h
A
A
A
A A
I L p
dt
dI
L

I
R U +
2

1
(2.11)
( )
W
A
F h el
M
dt
d
p
p
J

I
I L p M +

,
_

0 (2.12)

With the mutual inductance between excitation- and armature winding

L
M
h
three
differential equations for DC machines are gained, which describe the dynamic behaviour:

dt
dI
L I R U
F
F F F F
+ (2.13)
+ +
F
A
A A A A
I M p
dt
dI
L I R U (2.14)
W A F el
M
dt
d
J I I M p M +

(2.15)

The saturation of the excitation circuit is neglected in the considered case.
In analogy to the lecture Electrical Machines I:

c I M p
F
(2.16)
2
k
c (2.17)
i F
U n k n
k
c I M p

2
2
(2.18)
el A A F
M I c I I M p (2.19)

It is obvious from the dynamic set of equations, that not only the variables for the steady-state
operation but additionally the energy stores L
F
, L
A
and J need to be taken into consideration.
Besides the saturation, the mechanical friction and the voltage drop at the brushes in the
following paragraphs is neglected.
Fig. 11 shows the equivalent circuit diagram of the DC machine in dynamic operation.
I
F
U
F
R
F
L
F
J M
W
M,

U
i
R
A
L
A
I
A
U
A

Fig. 11: DC machine, equivalent circuit diagram for dynamic operation

DC machine 23


The set of equations needs to be converted into the state formulation. To simplify the
following considerations, it is normalized on rated (=nominal) values:

F F F
F
F
I R U
dt
dI
L
FN F
I R
1 1
(2.20)

F A A A
A
A
I M p I R U
dt
dI
L
AN A
I R
1 1
(2.21)
W A F
M I I M p
dt
d
J

N
M
1
(2.22)

Normalized values are also called per-unit values. For this purpose lower case letters are
used:

A
AN
A
u
U
U
,
A
AN
A
i
I
I
,
A
AN
AN A
r
U
I R

F
FN
F
u
U
U
,
F
FN
F
i
I
I
,
F
FN
FN F
r
U
I R

(2.23 a-h)

n

0
,
w
N
W
m
M
M


nominal values:

0

FN N
I M p U (2.24)

AN FN N
I I M p M (2.25)

time constants:

F
F
F
R
L
T (2.26)
A
A
A
R
L
T (2.27)
N
J
M
J
T
0

(2.28)

Referenced set of equations of the DC machine in state formulation:

F
F
F F
F
i
r
u
dt
di
T (2.29)
{
A
u
F A
A
A
A
i n i u
r dt
di
T
i

,
_

1
(2.30)
w
m
A F J
m i i
dt
dn
T
el

3 2 1
(2.31)

24 DC machine


Input- or excitation variables: u
F
, u
A
, m
W

State variables: i
F
, i
A
, n

Machine parameters: T
F
, T
A
, T
J
, r
A
, r
F


The according structure diagram for the three differential equations, describing the state of the
energy stores is depicted in Fig. 9:

+
-
+
-
+
-
+
-
i
A
i
A
i
F
i
F
1/r
A
1/r
F
u
A
u
i
u
F
X
X
n
m
W
m
el T
A
T
J
T
F
i
F
n
i
F

Fig. 12: equation set, structure diagram

Because of the products
F A
i i and
F
i n the set of equations is not linear and can only be
solved with numerical methods using computers.

DC machine 25


2.3 Separately excited DC machine
In many cases speed-variable DC machines are operated with constant excitation or with
permanent-magnet excitation. Thus torque and speed can only be influenced by adjusting the
armature voltage. In this case const I I
FN F
and the structure diagram of the motor
simplifies a lot using 1
F
i , as shown in Fig. 13.

-
+
-
+ u
A
n
n
i
A
=m
i 1
r
A
m
W
T
A
T
J

Fig. 13: DC machine, structure diagram

The set of equation now consists of only two equations being linear. Therefore it can be
solved analytically:

A
A
A A
A
i
r
n u
dt
di
T

(2.32)

W A J
m i
dt
dn
T (2.33)

The structure diagram of DC machines is obtained using Laplace-transformation. This is the
common display format in control engineering. In order to differ from the appearance of
previous equations, upper case letters are used for Laplace-transformation.

( ) ( )
( ) ( )
A
A
A A A
r
s N s U
s I s sI T

+ (2.34)

( ) ( ) ( ) s M s I s sN T
W A J
(2.35)

( )
( ) ( )
( )
A A
A
A
sT r
s N s U
s I
+

1
(2.36)

( )
( ) ( )
J
W A
sT
s M s I
s N

(2.37)


26 DC machine


+
- +
-
W(s)=U
A
(s)
-Z(s)=M
W
(s)
Y(s)=N(s)
G
1
(s) G
2
(s)
M
el
(s)
=I
A
(s)
1/r
A
1+sT
A
1
sT
J

Fig. 14: DC machine, structure diagram due to Laplace transformation

The speed of the DC machine states the output variable, which is controlled by the armature
voltage. So the armature voltage is the reference input variable. The load torque is the
disturbance. The following correlation applies:

Z
G G
G
W
G G
G G
Y
+
+
+

2 1
2
2 1
2 1
1 1
(2.38)

1.) Response to setpoint changes:

The response to set-point changes (this is the speed variation of the DC machine if the
armature voltage changes) is obtained for 0
W
M Z :

+
-
G
1
(s)
1/r
A
1+sT
A
G
2
(s)
1
sT
J
Y(s)=N(s) W(s)=U
A
(s) M
el
(s)=I
A
(s)

Fig. 15: DC machine, response to set-point changes for M
W
=0

( )
( )
( )
( ) ( )
A J A
J A
A
J A
A
A
r sT sT
sT sT
r
sT sT
r
s U
s N
s W
s Y
+ +

+
+

+

1 1
1
1
1
1
1
1
1
1
(2.39)

with the mechanical time constant:


K
N
A
N
N
N N
A
N
A J m
M
J
I
R
U
M
J
I U
R
M
J
r T T
0 0 0

(2.40)

follows:

( )
( )

,
_

+ +

+ +

m A A
A m
A m m A
T T T
s
s T T
T T s sT s U
s N
1
1
1
1
2
2
(2.41)

DC machine 27


At the time t = 0 a step change of the reference input variable is applied. For example the
stationary machine is connected to rated voltage (coarse start up).
0
1
U
A
t
U
A
(s) = 1/s

Fig. 16: DC machine, coarse start up

It is to be set:


2
0
1

m A
T T
(2.42)


A
m
m A A
T
T
D
T T
D
D
T 4
2
2
1
0

(2.43)

Then the speed characteristic is achieved by inverse transformation of

( )

'

+ +


2
0 0
2
2
0 1
2
1

sD s s
L t n (2.44)

Regarding the periodic case

1
4
<
A
m
T
T
D , (2.45)

the following solution is derived:

( ) ( )
2 2
0
2
1 arcsin 1 sin
1
1
0
D t D
D
e
t n
t D
+

(2.46)

The current is:

( ) ( ) s N sT s I
J A
(2.47)

This means a differentiation in the inverse transformation:

( ) ( ) { }
dt
dn
T
r dt
dn
T s N sT L t i
m
A
J J A
1
1


(2.48)

28 DC machine



Fig. 17: time characteristic of rotational speed for different
A
m
T
T


Fig. 18: time characteristic of the armature current for different
A
m
T
T

( ) 1 n
oscillation for 1
4
<
A
m
T
T


aperiodic limit for 1
4

A
m
T
T

( ) 0
A
i

DC machine 29


If the machine operates at steady-state no-load condition and the armature voltages is
changed, then the current increases, to achieve the new speed. If the speed is achieved, the
induced voltage has achieved its new steady-state value and the armature current decreases to
zero again. Depending on the magnitude of the damping this process causes oscillation or is
aperiodic.

2.) Response to disturbances:

The influence of the disturbance, i.e. the speed variation if the machine is loaded, can be
obtained for 0
A
U W :
The physical interpretation is a stationary machine (n = 0), which is loaded with a certain
torque. The reaction is a falling-off in speed n . By linear superposition any initial condition
can be added, i.e. n = 1.
+
-
1
sTj
1+sT
A
G
1
(s)
G
2
(s) Y(s) = N(s)
T
m
= T
j
r
A
Z(s) = M
w
(s)
1/r
A

Fig. 19: DC machine, structural diagram due to response to disturbances

( )
( )
( )
( )
( )
( )
( )
A m m
A A
A J A
A A
J A
A
j
W
T T s sT
sT r
r sT sT
sT r
sT sT
r
sT
s M
s N
s Z
s Y
2
1
1
1 1
1
1
1
1
1
1
+ +
+

+ +
+

+
+

(2.49)

For example the machine is loaded with rated torque ( )
s
s M
w
1
:
0
1
M
W
t
M
W
(s) = 1/s

Fig. 20: load with rated torque

The speed variation is:


( )
( )
( ) ( ) { } ( )
( )

,
_

+ +

'

+ +
+


dt
t dn
T t n r s N sT s N L r
T T s sT s
sT r
L n
A A A A
A m m
A A 1
2
1
1
1
(2.50)

30 DC machine



Fig. 21: time characteristic of the speed variation for different
A
m
T
T


The diagram shows the related falling-off in speed if the machine is loaded with the rated
torque. Depending on the magnitude of the damping this process causes oscillation or is
aperiodic. In steady-state operation with rated values it is:

F
A A A
i
i r u
n

(2.51)

1
A
u , 1
A
i , 1
F
i (2.52)

A
r n 1 (2.53)

A
r n n 1 (2.54)

The falling-off in speed if the machine is loaded corresponds to the referred armature
resistance r
A
.

DC machine 31


3.) Results of the analysis:

Separately excited DC machines with energy stores L
A
and J are

oscillating systems for

1
4
<
A
m
T
T
D (2.55)

aperiodic characteristics are obtained if the armature voltage or the load changes for

1
4
>
A
m
T
T
D (2.56)

If a change of state appears to separately excited or permanent excited DC machines, the
reaction is a transient phenomenon with an alternately exchange of kinetic energy of the rotor
and magnetic energy of the armature winding in the form of periodic oscillations. This can be
compared to the reaction of spring-mass system to a change of state. The damping is
responsible for the declination of the oscillation. The choice of a suitable current and speed
sensor is discussed later on.
2.4 Coarse-step connection of DC shunt machines
If the excitation current changes, the set of differential equations becomes non-linear and can
not be solved analytically anymore. Therefore computational algorithms are utilized for
analyzing transient phenomena. One phenomenon is the coarse-step connection of DC shunt-
wound machines: a stationary machine is connected to the power supply at t = 0.
The time characteristic of speed, armature current, excitation current and torque is intended to
be calculated in the following.
I
F
R
F
L
F
J
R
A
L
A
I
U
N
I
A
n

Fig. 22: DC shunt machine, equivalent circuit diagram (ecd)

32 DC machine


The general based set of equations in state formulation with 1
F A
u u is applied. The iron
saturation is neglected. The machine should be not loaded 0
W
m .

( )
A F
A
A
A
i i n
r dt
di
T 1
1
(2.57)

F
F
F
F
i
r dt
di
T
1
(2.58)

F A J
i i
dt
dn
T (2.59)

A stepwise numerical integration has to be performed on the computer.

t
k
t
k+1
y(t
k
+

t)
y(t
k
)

t

Fig. 23: numerical solution (Euler-Cauchy method)

With

( ) ( )
( )
k
k k
t y
t
t y t t y
,

+
(2.60)
follows

( ) ( ) ( )
k k k
t ty t y t y
,
1
+
+
. (2.61)

With that presupposition, a function ( ) t y at a time t t t
k k
+
+1
can be calculated, if the
function and its derivation are known at a time
k
t .

It is practical to choose the same discretization for each t . With ( ) ( ) k y t y
k
the following
set of equations is obtained:

( ) ( )
( ) ( ) ( ) ( ) k i k i k n
r t
k i k i
T
A F
A
A A
A

+
1
1 1
(2.62)

( ) ( )
( ) k i
r t
k i k i
T
F
F
F F
F

+ 1 1
(2.63)

( ) ( )
( ) ( ) k i k i
t
k n k n
T
A F J

+1
(2.64)
A simple method is for example the
Euler-Cauchy method with a given
initial value problem in the form of
( ) ( ) t y t f y ,
,
with the initial
condition ( )
0
0 y y .

DC machine 33


A recursive set of equations is formed, by moving all the variables, which are known at the
instant
k
t , on the right side and then determining the new variables at the instant
1 + k
t . Starting
at t = 0 it is integrated stepwise. As a matter of course the scanning intervals has to be much
smaller than the least machine time constant T t << .

Applied to the DC shunt-wound machine the following recursive set of equations is achieved:

( ) ( )
( ) ( ) ( )
A
F
A
A
A
A
r
k i k n
T
t
k i
T
t
k i

,
_

+
1
1 1 (2.65)

( ) ( ) ( )

,
_

+ + k i
r T
t
k i k i
F
F F
F F
1
1 (2.66)

( ) ( ) ( ) ( ) k i k i
T
t
k n k n
A F
J

+ +1 (2.67)

The following numerical values were used in the example shown in Fig. 24:

initial conditions: ( ) ( ) ( ) 0 0 0 0 n i i
F A


scanning time: ms t 1 . 0
time constants: ms T
A
13 , s T
F
42 . 0 , s T
J
57 . 0

resistances: 11 . 0
A
r , 1
F
r

A stepwise calculation results:

Fig. 24: time characteristics of m, n, i
A
, i
f

A sharp increase of the armature current and a slow increase of the excitation current lead to a
reduced torque during the acceleration. Even though an aperiodic characteristic is expected,
because of 1 4 >
A m
T T D , the speed and the armature current overshoot, because the field
is generated with a delay.

34 DC machine


2.5 Cascade control of converter-fed PM DC machines
The block diagram (Fig. 25) shows the common control system of a converter-fed drive
system, consisting of a permanent-field (= PM) DC machine, a line-commutated converter
and a cascade control. The speed control loop, which for example is realized as a PI-
controller, has a lower-level current control loop, which is also realized as a PI-controller.
+
-
+
-

Fig. 25: DC machine, control system diagram

The general set of equations of the permanent-field DC machine in based values with
const I I
FN F
, i.e. 1
F
i is:

( )
A A
A
A
A
i n u
r dt
di
T
1
(2.68)

W A J
m i
dt
dn
T (2.69)

The result of the numeric solution is the following set of equations:

( ) ( )
( ) ( )
A
A
A
A
A
A
r
k n k u
T
t
k i
T
t
k i

,
_

+ 1 1 (2.70)

( ) ( ) ( ) ( ) ( ) k m k i
T
t
k n k n
W A
J

+ +1 (2.71)

DC machine 35


The step-by-step solution of the set of differential equations is done using numeric methods:
Then the PI-controllers have to be discretized as follows (illustration due to Fig. 26):

+
-
1+sT
R

sT
R
K
R
X
1
X
2
X Y


Fig. 26: PI controller


( )
( )
( )
R
R
R
sT
sT
K s G
s X
s Y +

1
(2.72)

( ) ( ) ( ) s X sT K s Y sT
R R R
+ 1 (2.73)

,
_

+
dt
dx
T x K
dt
dy
T
R R R
(2.74)

( ) ( )

,
_

x dt t x
T
K t y
R
R
1
(2.75)

( ) ( ) ( ) ( )
{

,
_

Anteil P
Anteil I
k
i R
R
k x i t i x
T
K k y
4 4 3 4 4 2 1
1
1
1
(2.76)


The dimensioning of the current and speed controllers is not discussed in detail here. Suitable
values for the first attempt are:


J A m Ri
T r T T

1
Ri
K


Ri Rn
T T 10

10 5
Rn
K

It is important, to switch off the integral-action component if the current is limited. The figure
shows the time characteristic of speed and armature current of a permanent-field DC machine
during a setpoint step-change of the speed from 0 to 1 and then to -1, i.e. start-up and
reversing. The machine has no load. The current is limited to: 5 , 2
AN
A
I
I
.


36 DC machine



Fig. 27: actual value and desired value (dashed line) of speed and current in based values

Those drives are often used as speed-variable servo drives for machine tools and robots. They
are high dynamic, i.e. featuring a slim-line rotor to achieve low moment of inertia l D J
4
~
and a small armature time constant, because
M A
A
A
h R
L
T
+

1
~ (because of the permanent-
field, the mechanical air-gap has to be replaced by the magnetic air-gap
M
h + ).
Depending on the type of magnetic material, the ratio

M
h
amounts round about 2 ... 10.
If a position control is applied, the cascade has an additional upper-level control-loop.


2.6 DC series-wound-machine as traction drive in pulse
control operation
Speed-variable drives of ground conveyors and electric vehicles with cycle operation are
often realized using DC series-wound-machines with DC chopper controllers. Chopper
controllers are also used for voltage control of DC-fed train drives.

Fig. 28 shows the general circuit diagram of a clocked DC chopper controller, consisting of a
self-commutated thyristor switch or GTO and a freewheeling diode, connected in parallel to
the motor.

DC machine 37


J
n
M
I
M
I
G
Th
U
B
braking
driving

Fig. 28: electric vehicle drive

The method of operation is as follows:
If the thyristor is switched on (conducting), the motor is connected to the battery voltage, the
thyristor carries the current and the machine current increases with the time constant
( ) ( )
F A F A AF
R R L L T + + to the final value ( ) ( )
F A i
R R U U + . After opening the thyristor
switch, the motor is disconnected from the battery, the diode carries the current and the
machine current decreases with the same time constant to zero. Then a new cycle begins.

U
A
U
Batt
I
th
I
D
t
T
puls
= 1/f
puls
T
ein
T
aus

Fig. 29: step-up, step-down converter, duty cycles

Besides the losses in the power semiconductor devices, the chopper controllers operates non-
dissipative. With the assumption
AF m
T T >> , i.e. const n during a cycle and with
negligence of the voltage drop on the windings, it is as a first approximation:

( )
ein Batt
Puls
T
Puls
A
T U
T
dt t u
T
U n
Puls
1 1
~
0

(2.77)

The mean value of the motor voltage and so at least the speed is controlled by the ratio of
conducting time (T
ein
= T
on
) and dead time (T
aus
= T
off
) of the thyristor. The pulse frequencies
are some 100 Hz up to some kHz. The chopper controller allows only one-quadrant
operation. For braking, the thyristor and the freewheeling diode has to be interchanged in the
circuit using contactors.


3 Induction machine
3.1 Dynamic equation set
To calculate the dynamic behavior of induction machines, the general system of equations for
rotating field machines can be utilized. Because of the constant air gap the choice of the angle
( ) t is arbitrary.

An arbitrary coordinate system is to be used, whose rotational speed and initial value can be
chosen freely

( )
0
+ t t
K
,
K
dt
d

. (3.1 a, b)

The mechanical speed ensues to:

n p p
dt
d

2 . (3.2)

After converting the rotor quantities on the number of the stator windings, the voltage
equations of the induction machine are:


1
]
1

+

+
1
]
1

+
1
]
1

1
]
1

1
1
1
1
1
1
1
1
1
d
q
K
q
d
q
d
q
d
dt
d
i
i
R
u
u
(3.3)

( )
1
]
1

+
+
1
]
1

+
1
]
1

1
]
1

'
2
'
2
'
2
'
2
'
2
'
2 '
2 '
2
'
2
d
q
K
q
d
q
d
q
d
dt
d
i
i
R
u
u
(3.4)

flux linkages:

1
1
]
1

+
1
]
1

1
]
1

'
2
'
2
1
1
1
1
1
q
d
h
q
d
q
d
i
i
L
i
i
L (3.5)

1
]
1

+
1
1
]
1

1
1
]
1

1
1
'
2
'
2 '
2 '
2
'
2
q
d
h
q
d
q
d
i
i
L
i
i
L (3.6)

torque equations:

( )
dt
d
p
J
M i i i i L p M
W q d d q h el

+
'
2 1
'
2 1
(3.7)

( ) ( )
'
2
'
2
'
2
'
2 1 1 1 1 q d d q d q q d
i i p i i p (3.8)

40 Induction machine


At first a suitable coordinate system is to be chosen, with assumed synchronous rotation with
the rotating field in the stator:


1

K
dt
d
(3.9)

( )
0 1
+ t t . (3.10)

For the inverse transformation into the complex notation, the choice of the constant of
integration
0
is still arbitrary. A reasonable choice appears due to Fig. 30:

+
d
q
i
d
i
q

(t)
+Re
-Im
I
0
U
1
+
coordinate system complex plane

inverse transformation
d: direct axis
q: quadrature axis
magnetization
torque

Fig. 30: coordinate transformation

( ) ( ) ( ) [ ]
0 1 1 0 1 1 1 1 1
sin cos
3
2
cos 2 + + t u t u t U u
q d u
(3.11)

{ } { } { }
( ) { }
1
1
]
1

4 43 4 42 1
0 1
Re
0 1
1
0 1
1
1
1
Im Re
3
2
Re 2


j
e
t j
e j
j t j
q
j t j
d
t j
e e u e e u e U (3.12)


0
1
0 1
1
3 3
j q j
d
e
u
j e
u
U + (3.13)

Definition for induction machines:
2
0

, j e e
j
j


2 0

. Then follows:
quadrature axis (q-axis), in which the torque is generated, corresponds to the real axis.

direct axis (d-axis), in which the machine is magnetized, corresponds to the imaginary
axis.


1 1
1
1
1
3 3
d q
d
q
U U
u
j
u
U + (3.14)

Induction machine 41


Currents are defined in the same way:


1 1
1
1
1
3 3
d q
d
q
I I
i
j
i
I + (3.15)

The rotor is considered in the same manner:

'
2
'
2
'
2
'
2 '
2
3 3
d q
d
q
U U
u
j
u
U + (3.16)


'
2
'
2
'
2
'
2 '
2
3 3
d q
d
q
I I
i
j
i
I + (3.17)

As the currents
1 q
i and
1 d
i in the rotating two-phase system are direct currents in the steady-
state operation, the currents
1 q
I and
1 d
I in the complex notation are currents with system
frequency, whose r.m.s. value equals
3
1
times the DC value and whose phase angle is
determined by
2
0

.


1 q
I and
'
2 q
I are active currents in the complex notation.


1 d
I and
'
2 d
I are reactive currents in the complex notation.

3.2 Steady state operation
In steady-state operation the flux linkages in the rotating system are constant, i.e.

0

dt
d
(3.18)

and the speed is constant, i.e.

0
dt
d
(3.19)

Torque equation and voltage equation are therefore decoupled and independent


1 1 1 1 1 q d d
i R u
1 1 q id
u (3.20)


1 1 1 1 1 d q q
i R u +
1 1 d iq
u (3.21)

( )
'
2 1
'
2
'
2
'
2 q d d
i R u + (3.22)

( )
'
2 1
'
2
'
2
'
2 d q q
i R u (3.23)

( )
1 1 1 1 q d q d el
i i p M (3.24)

42 Induction machine


The inverse transformation of the voltage equations provides:

( ) ( )
3 3
'
2 1 1 1 1 1
'
2 1 1 1 1 1
1
q h q d d h d q
i L i L i R
j
i L i L i R
U
+

+ +

(3.25)
( ) ( )
3 3
1
'
2
'
2 1
1
1 '
2
'
2 1
'
2
'
2 1
1
1 '
2
'
2
'
2
q h q d d h d q
i L i L i R
j
i L i L i R
U
+

(3.26)

'
2 1 1 1 1
'
2 1 1 1 1 1 q h q d d h d q
I X j I X j I R I X j I X j I R U + + + + + (3.27)

1
'
2
'
2
'
2
'
2 1
'
2
'
2
'
2
'
2
'
2 q h q d d h d q
I X s j I X s j I R I X s j I X s j I R U + + + + + (3.28)

'
2 1 1 1 1 1
I X j I X j I R U
h
+ + (3.29)

1
'
2
'
2
'
2
'
2
'
2
I X s j I X s j I R U
h
+ + (3.30)

Finally the well-known voltage equations of the induction machine is achieved, corresponding
to the symmetric equivalent circuit diagram (ecd) as depicted in Fig. 31.

I
1
U
1
X
h
I
0
I
2
'
U
2
'/s
R
1
R
2
'/s X
1
X
2
'

Fig. 31: induction machine, equivalent circuit diagram (ecd)

Torque ensues to:

( )
d id q iq
d
q q
d
el
I U I U
p i
i
p
M +

,
_

1
1
1 1 1
1 1
1
3
3 3 3 3
3

(3.31)


Induction machine 43


3.3 Rapid acceleration, sudden load change
Using numerical integration methods on the computer, now the rapid acceleration of an
induction machine with squirrel-cage rotor is calculated. At the time 0 t the machine in
standstill is connected to the supply voltage. It is assumed, that the supplying system is a stiff
system and the machine is loaded only with its moment of inertia. Afterwards a sudden
change load change with rated torque occurs.
The parameter ( ) t is chosen as ( )
2
1

t t and the system of equation is to be
transformed in state form. For the squirrel-cage rotor is considered: 0
'
2
'
2

q d
u u


1 1 1 1 1
1
q d d
d
R i u
dt
d
+

(3.32)


1 1 1 1 1
1
d q q
q
R i u
dt
d

(3.33)

( )
'
2 1
'
2
'
2
'
2
q d
d
R i
dt
d

(3.34)

( )
'
2 1
'
2
'
2
'
2
d q
q
R i
dt
d
+

(3.35)

( ) [ ]
W d q q d
M i i p
J
p
dt
d

1 1 1 1

(3.36)

Currents result from the inverse inductance matrix.


( )
( )
( )
( )
[ ]
1
1
1
1
]
1

1
1
1
1
]
1

+
+
+
+

1
1
1
1
]
1

'
2
'
2
1
1
1
1
1
2
2
1
'
2
'
2
1
1
1 0 1 0
0 1 0 1
1 0 1 0
0 1 0 1
1
q
d
q
d
L
h
q
d
q
d
L
i
i
i
i
4 4 4 4 4 4 4 4 3 4 4 4 4 4 4 4 4 2 1

(3.37)

The following structure diagram for induction machines, shown in Fig. 32, accords to the
discussed behavior:

Values are defined as:

excitation values:
1 d
u ,
1 q
u ,
W
M

state values:
1 d
,
1 q
,
'
2 d
,
'
2 q
,

default:
1

K


44 Induction machine




Fig. 32: induction machine, structural diagram

Induction machine 45


Initial conditions for 0 < t in case of machine being turned-off appear as:

0
'
2
'
2 1 1

q d q d
i i i i (3.38)

0
1 1

W q d
M u u (3.39)

Excitation values of the rapid acceleration case of induction machines are (for 0 > t ):

0
1

d
u ,
1 1
3 U u
q
, 0
W
M (3.40 a-c)

After rapidly starting up, the induction machine is operating at no-load and the initial
conditions before the sudden load change are considered:


0 1
3 I i
d
,
1
(3.41 a,b)

0
'
2
'
2 1

q d q
i i i (3.42)

For
L
t t > the considered induction machine is to be suddenly loaded with rated torque. The
excitation values ensue to:

0
1

d
u ,
1 1
3 U u
q
,
N W
M M (3.43 a-c)

The simulation is based on the following machine data:

o kW 400
N
P

o
1
0
min 1000

n

o V U
N
380
1
(verk.)
4
o A 160
0
I

o A 715
1

N
I

o 05 , 0

o 90 , 0 cos
N


o Nm 3906
N
M

o s
M
J
T
kipp
m
33 , 0
0

In both cases a dynamic transient reaction takes place.
When it is switched on, the induction machine generates heavy oscillating torques at
standstill because of the DC components in combination with high symmetrical short-
circuit currents. After the declination the machine runs up (more or less fast,
depending on the coupled masses) and adjusts itself with overshoots at no load.
The sudden load change is braking the machine at first until the electrical torque is
built up. Afterwards the machine adjusts itself to a steady state.

46 Induction machine


The according diagrams (Fig. 33-37) show:

o stator current ( ) t i
u
, rotor speed ( ) t n and electrical torque ( ) t f M
el
as time
dependent functions
o dynamic torque-speed characteristic ( ) n f M
el
and the dynamic circle diagram
( )
Blind Wirk
I f I

The deviations from the steady-state characteristics are notably.

For that case the induction machine is not practical as a dynamic actuator in a drive system. In
the following it will be examined, if it is possible, to let the induction machine have the same
dynamic performance as the DC machine.


Fig. 33: induction machine, stator current ( ) t i
u 1
vs. time



Fig. 34: induction machine, rotational speed ( ) t n vs. time

Induction machine 47



Fig. 35: induction machine, torque ( ) t M vs. Time


Fig. 36 : induction machine, dynamic speed-torque characteristic ( ) n M


Fig. 37: induction machine, dynamic circle diagram ( )
Blind Wirk
I I

48 Induction machine


3.4 Induction machine in field-oriented coordinate system
Because of the effect of the collector, the excitation flux is always perpendicular to the
armature current-linkage and their spatial position is stationary.
If the machine has a commutating-field winding and a compensating-field winding, then the
armature quadrature-axis field is completely compensated 0 + +
K W A
. Thus the
armature flux-linkage in the armature quadrature-axis equals zero 0
qA
and the excitation
flux is not influenced by the armature current. Therefore the armature flux-linkage in the
direct axis depends only on the field excitation current
F dA
I ~ and the torque is
F A el
I I M ~ .

x
x
x
x
x


Advantage can be taken of this for the induction machine by choosing a rotor-flux-oriented
coordinate system, which is rotating with the speed of the rotor flux

( )
0


+ t t (3.44)

whereas the instantaneous value of the angular speed


R

+ (3.45)

does not necessarily need to correspond with the stationary value
1
of the stator field at
rated frequency.
Fig. 38: separately excited,
compensated DC machine (VZS)

Induction machine 49



d1
q1
q2
d2

Fig. 39: coordinate system, revolving at rotor flux rotational speed (d=direct, q=quadrature)


A direction convention of the axis (direct/quadrature) is used due to:
d: direction of the rotor flux
q: perpendicular to the rotor flux

This is called field-oriented operation. An observer, rotating with the system with (t), detects
the same field distribution and the same torque generation as in comparable DC machines.

The result are simple relations for the controlled variables rotor flux and stator active current.
Both variables can be adjusted independently ( compare to: DC machine).

For the rotor flux-linkages it is postulated:
( )

i L i L i L
h d h d h d
+ +
!
1
'
2 2
'
2
1 (direct reference axis) (3.46)

( ) 0 1
!
1
'
2 2
'
2
+ +
q h q h q
i L i L (quadrature axis) (3.47)

Thereby the variable

i is a user-defined magnetizing current, which is proportional to the

rotor flux linkage.

From this follows for the rotor currents:

( )
1
2
'
2
1
1
d d
i i i
+

(3.48)

( )
1
2
'
2
1
1
q q
i i
+

(3.49)
A decomposition of the stator and rotor
current-linkages direct ( )
'
2 1
,
d d
i i and
quadrature ( )
'
2 1
,
q q
i i components regard-
ing the rotor flux leads to a clear
decoupling and permits a suitable control,
to inject currents in such a manner, that
the rotor flux-linkage in the quadrature
axis becomes zero 0
'
2

q
and the that
the rotor flux-linkage in the direct axis
only depends on the magnetizing current

i
d
~
'
2
.

Then the torque is only generated by the
perpendicular components of rotor flux
and stator active current
1
'
2
~
q d el
i M .

50 Induction machine


The angular speeds ensues to:

p
dt
d

(3.50)


R K
dt
d

+ (3.51)

whereas
R
is the angular frequency of the rotor currents.

The rotor voltage equations are:


( )
'
2
'
2 '
2
'
2
0
q
d
d
dt
d
dt
d
i R

+

(3.52)


( )
'
2
'
2 '
2
'
2
0
d
q
q
dt
d
dt
d
i R

+

(3.53)

After pasting the above relations follows:

( )
dt
di
L i i R
h d

+
+

1
2
'
2
1
1
0 (3.54)

( )

i L i R
h R q
+
+

1
2
'
2
1
1
0 (3.55)

With the rotor time constant
2
T due to


( )
'
2
'
2
'
2
2
2
1
R
L
R
L
T
h

, (3.56)

the rotor equations in field-oriented coordinates are obtained:


1 d
i controls

i :
1 2 d
i i
dt
di
T +

(3.57)


R
is proportional to
1 q
i :

i T
i
q
R
2
1
(3.58)

The torque equation in field-oriented coordinates is:

( )
1
2
'
2 '
2
'
2
'
2
'
2
1
q
d
q d d q el
i p i i p M
+

(3.59)

el
M is proportional to
1 q
i

and

i :

dt
d
p
J
M i i
L p
M
W q
h
el

+
+

1
2
1
(3.60)

The stator voltage equations do not have to be regarded further on, if the stator currents are
injected by power converters with high switching frequencies and short sampling times
(servo-converter).

Induction machine 51


The equation for

i shows, that the direct component of the stator current controls the rotor
flux. The large rotor time constant is the controlling time constant ( compare with the field
winding of DC machines). Therefore the magnitude of the rotor flux is not suitable for rapid
control processes.
The equation for
R
shows, how the angular speed of the rotor flux is composed of the
mechanical angular speed of the rotor and the angular speed of the slip, which results from the
active current component of the stator current and the magnitude of the rotor flux.

The torque equation describes the mechanical dynamic response and the torque generation,
which now results from the direct flux and the quadrature current ( compare with DC
machines). If const
d

'
2
applies, torque
el
M und the angular speed
R
are proportional to
1 q
i . These three equations describe the model of the induction machine in field-oriented
coordinates.

In analogy to DC machines, an equivalent structure diagram for induction machines in field-
oriented coordinates with injected stator currents can be found. This means, that under certain
conditions induction machines behave like separately excited DC machines with vanishing
time constant. At constant rotor flux the torque generation follows the quadrature current
instantaneously and the rotor flux can be adjusted solely with the direct current component.


1 2 2 q
i i T i T

(3.61)


1 2 d
i i
dt
di
T +

(3.62)


dt
d
p
J
M i i
L p
M
W q
h
el

+
+

1
2
1
(3.63)


Fig. 40: induction machine, control strategy diagram

Same control strategy as used in DC machine applications:
Speed setpoint input with

, flux input with

1 d
i , instantaneous torque generation with
1 q
i :
high-dynamic drive

52 Induction machine


3.5 Field-oriented control of induction machines with
injected currents
As a matter of principle the intention is achieved. If the induction machine is controlled in
field-oriented coordinates, it is required to know the angular position of the rotor flux.

It is not possible to measure any of the electric values in the rotor of squirrel cage machines -
neither rotor currents nor rotor voltages. The measurement of the air-gap flux describes an
approximation, which furthermore is expensive and susceptible to faults.
However magnitude and angular position of the rotor flux can be calculated from measured
values of the stator currents and the speed, using the rotor equations of induction machines.
This is called flux model, which is implemented on a microcontroller running in online
operation.


1 2 d
i i
dt
di
T +

(3.64)


dt
d
i T
i
R
q

3 2 1
2
1
(3.65)


1
2
1
q
h
el
i i
L p
M
+

(3.66)

[ ]
1
]
1

1
]
1

1
1
1
1
B
A
q
d
i
i
T
i
i

(3.67)

[ ]
1
]
1

1
]
1

1
1
32
1
1
v
u
B
A
i
i
T
i
i
(3.68)

searched
Rotor-
flux
mea-
sured
value

Fig. 41: induction machine, flux model in field-oriented coordinates

Induction machine 53


For this reason induction machines can be controlled in field-oriented coordinates with
injected currents. The following picture shows the structure diagram of a complete drive
system with speed-, torque- and flux controller.
c
u
r
r
e
n
t

c
o
n
t
r
o
l
l
e
r
s
p
e
e
d

c
o
n
t
r
o
l
l
e
r
P
W
M
%
X
f
i
e
l
d

w
e
a
k
e
n
i
n
g
f
l
u
x

c
o
n
t
r
o
l
l
e
r
T
2

Fig. 42: entire drive system, structural diagram

The correct determination of

i and requires the knowledge of

2
2
2
R
L
T .
Problem:
2
R depends on temperature and
2
L depends on saturation effects.

54 Induction machine


The setpoint values for the flux and the torque are calculated from the speed setpoint input. In
base speed operation the machine is driven with full flux and above the synchronous speed
with field weakening (field control). Using a flux model with appropriate controller, the
setpoint values for the transformed stator direct and quadrature currents are calculated. The
three-phase current setpoint values for the pulse-controlled inverter result from an inverse
transformation. Current

i and position are calculated online with the flux model, with
knowledge of the measured stator current and speed values.

If the machine is in standstill at the instant 0 t and a setpoint step-change 1
1

is
enforced, the control unit injects the currents
1 u
i ,
1 v
i and
1 w
i (
1 d
i and
1 q
i in the flux model
after the transformation) with the pulse-controlled inverter in such a manner, that the
magnetizing current

i is built-up on its rated value with the rotor time constant

2
2
2
R
L
T . The
machine accelerates nearly linear according to the adjusted quadrature current
max 1 q
i during
the acceleration time
max
1
M p
J
T
J

.

There is no overshoot and there are no oscillations any more. high dynamic drive


Fig. 43: simulation of an acceleration process and an afterwards braking
N
q
i
i
3
1
1

N
i
i
3

Induction machine 55


3.6 Steady-state operation using variable frequency and
voltage converter
There are two possible fundamental operational performances:

1.) Operation with constant stator flux-linkage

The stator flux-linkages as well as the currents and voltages in steady-state operation are
obtained by inverse transformation from rotating systems with ( )
2
1

t t :


3 3 3 3
'
2 1 1
'
2 1 1
1
1
1
d h d
q h q
d
q
i L i L
j
i L i L
j
+

(3.69)


'
2 1 1
'
2 1 1 d h d q h q
I L I L I L I L + + + (3.70)

,
_

+
+ +
1
'
2
1 1
'
2 1 1
1
I
I L I L I L
h
(3.71)

( )
*
0 1
*
2 1 1
I L I I L + (3.72)

For this purpose the equivalent circuit diagram (see script EM I, chapter 7.2) with
( )
1
2 2
1 1
1
w
w

is suitable in particular.
*
1
1

Fig. 44: induction machine, short-circuit ecd

In the equivalent circuit diagram (ecd) as shown in Fig. 44, with 0
1
R the current
*
0
I in the
shunt arm has to be kept constant to achieve const
1
.

const
j
U
X j
U
L I L

1
1
1
1
1
*
0 1 1

(3.76)

This means operation on a supply with const
f
U

1
1
. The neglect of
1
R is only valid for high
frequencies. For lower frequencies the supply voltage must be increased to compensate the
resistance voltage drop at the stator resistance.
1
'
2
*
2
1 +

I
I (3.73)

( )
2
1
'
2
*
2
1 + R R (3.74)

1
*
2
1
X X

(3.73)

56 Induction machine


The locus diagram of the stator current is the well-known Heyland diagram of induction
machines. The circle is practically parameterized with the rotor frequency.

( )
2 1 *
2
*
2 *
2
~ tan s s
R
X
(3.77)


Fig. 45: induction machine, locus diagram (Heyland circuit)


The speed is adjusted with supply frequency and supply voltage const
f
U

1
1
.
The peak value of the current is
1
I . The maximum torque is the breakdown torque
kipp
M .

The relation between the breakdown slip and the rotor time constant is described by:

( )

+

1
1
1 1
1
2
1
'
2
1
*
2
*
2
1
L
R
X
R
s
kipp kipp
(3.78)


( )
( ) ( ) ( )
2
'
2
'
2
2 1 1
2
1
'
2
1
1 1 1
1
T L
R
L
R
h

+ + +
+

(3.79)


The Kloss equation describes the torque as a function of rotor frequency.

2
2
2 2

kipp
kipp
kipp
kipp
kipp
s
s
s
s M
M
+

+
(3.80)

For different values of the supply frequency a family of characteristics is obtained, which
appears quite similar to the characteristic of DC machines for
kipp
<
2
.

Induction machine 57


For
10 1
(no-load flux-linkage) is
1
, 1
0 1
X
U
I I
Str N
N
(no-load current).
n
M
P

=

c
o
n
s
t
M

=

c
o
n
s
t

Fig. 46: induction machine, torque-speed diagram


2.) Operation with constant rotor flux-linkage

Again the rotor flux-linkage as well as the stator flux-linkage are obtained by inverse
transformation from the rotating system with
2
1

t .
Practically the rotor flux-linkage for field-oriented operation is to be chosen. With that the
induction machine is expected to behave like a separately excited DC machine.


3 3
0
3 3
'
2
'
2 '
2

i L
j j
h
d
q

(3.81)

We define:


3
0

i
j I
+
(3.82)

and divide
'
2
by ( )
2
1 +

( ) ( )
( )
( ) ( )
( )
+
+ +

+ +
+

0 1
2 1
0 1
2
0
2
'
2
1
1 1
1
1 1
I L
I L I L
h h

(3.83)
field weakening range
1
1
1
~
f

const U
1

2
1
1
~
f
M
kipp

base speed range
const
1

const
f
U

1
1

const M
kipp


58 Induction machine


For this purpose the equivalent circuit diagram (ecd) with
( )
2 2 2
1 1
1
1
w
w

is reasonable.

Fig. 47: induction machine, ecd for constant rotor flux linkage

In the equivalent network with 0
1
R the current
+
0
I in the shunt arm has to be kept
constantly to achieve const
'
2
. This is field-oriented operation. The voltage drop at
1
L
has to be compensated load-dependent:

( )
+
+
0 1 1 1
1
1
1 I L j I L j
U

(3.87)

For this purpose the stator has to be dimensioned adequate, to avoid saturation phenomena.

Again the neglect of
1
R is only valid for high frequencies. In the case of low frequencies the
supply voltage must be increased to compensate the resistance voltage drop at the stator
resistance
1
R .

The locus diagram of the stator current
1
I

dependent on the rotor frequency
2
is now a line
for const
'
2
and the stator voltage
1
U is load-dependent.

( )
+
+
0 1 1 1 1
1 I X j I X j U (3.88)


+ +
+
2 1 0
I I I (3.89)

( ) 0 1
1 0
2
2
+
+
+
+
X j I
s
R
I (3.90)

For
1
I follows:

( ) ( )

,
_

+

+
+
+
+
+
+ + +
1 2
2
1 1
0
2
1 0
0 2 0 1
1
1
1


R
L
j I
s R
X I
j I I I I (3.91)


( )
( ) ( )
( )
( )

,
_

+ +

,
_

+
+ +
+
+
+ + +
kipp
h
j I T j I
R
L
j I

2
0 2 2 0
2
2
'
2
2 1
2 1
0
1 1
1
1 1
1
1 (3.92)
( )
2
'
2 2
1 +
+
I I (3.84)

( )
2
2
'
2
2
1 +

+
R
R (3.85)

'
2 1 0
I I I +
+
(3.84)

Induction machine 59




-

Fig. 48: induction machine, phasor diagram


Specific points:
: 0
2

1
, 1
0 1
X j
U
I I
Str N
N

+
no-load flux linkage (3.96)

:
2

1
I stator current (unlimited) (3.97)

i) The active current is:

{ }
kipp kipp
N
W
I
I
I I

2 2 0
1 1
Re

(3.98)

which means that the rotor frequency adjusts itself load-dependent.

ii) The reactive current is constant and equal to the no-load current, the power factor cos is
is improved (=increased) compared to const
1
.

{ }
N B
I I I
0 1 1
Im (3.99)

For
kipp

2
is

I I
W 1
.

Breakdown slip and current limitation do not apply anymore.

The torque is calculated with the induced voltage and the active current ( compare to DC
machines).

( )
kipp
N
N W i
I
I X
p
I U
p
M

2 0
0 1
1
1
1
1
3 3

(3.100)


kipp
kipp
kipp
N
M
X
U p

2 2
1
2
1
1
2
1
3

(3.101)
For
'
20
'
2
(no-load flux linkage) is (3.93)


1
, 1
0 0
X
U
I I
Str N
N

+
(no-load current). (3.94)

and the stator current ensues to:
2 2
1 Wirk
I I I +

. (3.93)

60 Induction machine


For
kipp

2
is
kipp
M M 2

The speed-torque characteristic is now a (declining) straight line and looks like the
characteristic of separately excited DC machines.

kipp
kipp
M
M

2
2

(3.102)

kipp
kipp
M
M
p
f
p
f
p p
f
n n n

2 2
1 2 1
2 1

(3.103)



Fig. 49: induction machine, torque-speed characteristics


The speed drop in the load case is halved for const
'
2
compared to the case of const
1
.



A summarizing comparison and conclusion of both described methods

1.) constant stator flux-linkage
2.) constant rotor flux-linkage

is shown in the table on the next page.









Induction machine 61


constant stator flux-linkage constant rotor flux-linkage
10 1
const

ecd with ( )
1
2 2
1 1
1
w
w

const I L
N

0 1


const I
R U

1
1
1
1
1


for 0
1
R : const
f
U

1
1

locus diagram: circle

,
_

+
j
j
I I
kipp
N
2
0 1
1
1

kipp
kipp
kipp
M M

2
2
2
+


kipp
M M ( )
max
M
2

1
I
I
W

0

1
2
I
I
I
B
+

may become unstable,

operation at variable voltage
and frequency supply
2
'
20
2
'
2
1 1 +

const

ecd with
( )
2 2 2
1 1
1
1
w
w

( ) const I L
N

0 1
1

const I L j I
R U

1 1 1
1
1
1
1

for 0
1
R : const I L j
f
U

1 1
1
1
2
locus diagram: line
( )
2 2 0 1
1 T j I I
N
+



kipp
kipp
M M

2
2



kipp
M M 2 ( )
max
M

I I
W 1
( ) 3
1 B
i

0 1
I I
B
( ) 3
1 A
i


may not become unstable,
better power factor cos ,

DC machine behavior,
field-oriented control
speed control with
1

2
adjusts itself due to load
current entsprechend
here:
1 1
0
3
L
U
I
N
N
N

(Does not have to be no-load flux-linkage. Rated flux-linkage is also possible.)
if
kipp

2
follows

62 Induction machine


3.7 Field-oriented control of induction machines with
applied voltages
Up to now the field-oriented control of induction machines was deduced using the simplifying
assumption, that a current injecting power converter with high switching frequency, with fast
control unit and sufficient voltage reserve is available. This is the case for servo drives with
transistorized frequency converters and switching frequencies up to 20 kHz in the kW power
range. For larger drives GTO-pulse-controlled inverters with variable-voltage link and
switching frequencies lower than 1 kHz are used. In the discussed case the assumption is no
longer fulfilled, so the stator voltage equations of the induction machine have to be taken into
consideration.
After pasting the rotor currents from chapter 3.4, the stator flux-linkages are:
( ) ( ) ( )
1
2
1
1 1 1
'
2 1 1 1 1 1
1
1 1
d
h
d h d h d h d
i i
L
i L i L i L
+
+ + + +

(3.104)

( )
( )( ) ( )
1 1
2
1
2
1
2 1
1 1 1
1 1 1 1
1
1 1
d
h h
d h
i L i
L
i
L
i L +
+

+
+

,
_

+ +
+

(3.105)

( ) ( )
1
2
1
1 1 1
'
2 1 1 1 1 1
1
1 1
q
h
q h q h q h q
i
L
i L i L i L
+
+ + +

(3.106)


1 1 q
i L (3.107)

Now stator voltage equations can be converted:


1 1
1
1
2
1
1 1 1
1
1 1 1
1
q
d h
d q
d
d d
i L
dt
di
L
dt
di
L
i R
dt
d
i R u +
+
+

(3.108)


1 1
2
1 1
1 1 1 1
1
1 1 1
1
d
h q
q d
q
q q
i L
i L
dt
di
L i R
dt
d
i R u +
+

+ + +

(3.109)

With the definition of the stator time constant
1
T


( )
1
1 1
1
1
R
L
T
h
+

(3.110)

the stator voltage equations in field-oriented coordinates are obtained:

( )

,
_

+ +
dt
di
L i L u
R
i
dt
di
T
q d d
d

1 1 1 1
1
1
1
1
1
1
(3.111)

( ) ( )

i L i L u
R
i
dt
di
T
d q q
q
+
1 1 1 1
1
1
1
1
1
1
(3.112)

The stator voltage equations complete the machine model with the co-action of the stator
voltages and the stator currents.

Induction machine 63


Concerning the components of the stator currents the induction machine behaves like a first-
order time-delay element with the time constant
1
T and the gain
1
1
R
. The components of
the stator currents are coupled by the right side of the equations.
1 1 d
i L

and
1 1 q
i L

are rotary induced voltages, caused by the current in the particular quadrature
axis. ( )
dt
di
L


1
1 is a transformer voltage, which is caused by the change of the
magnetizing current. ( )

i L
1
1 is the rotary induced voltage of the magnetizing field.

Both control systems are coupled by the stator currents and are not independent from each
other. However a decoupling is desired in such a manner, that the current controllers can be
adjusted independently. This is realized by adding compensating voltages with negative sign
to the controller output voltages
Rd
u and
Rq
u , so that the coupling voltages disappear. So the
controllers see decoupled controlled systems. Assumption for the compensation: the rotor flux
linkage needs to be constant, i.e. 0
dt
di

.

1
1
1
1
R
u
i
dt
di
T
Rd
d
d
+ (3.113)


1
1
1
1
R
u
i
dt
di
T
Rq
q
q
+ (3.114)


1 1 1 q d Rd
i L u u

(3.115)

( )

i L i L u u
d q Rq
+ +
1 1 1 1
1 (3.116)

The compensating voltages are generated in a decoupling network (Fig. 50).
X
X
X

Fig. 50: induction machine, control strategies, decoupling network

Fig. 51 shows the complete circuit of a field-oriented induction machine with voltage
injecting actuator with pulse-width modulation. A cascade control is used for the direct- and
the quadrature-current, which can be adjusted independently with the aid of a decoupling
network. The actual values of magnitude and position of the rotor flux, which are required for
the control, are determined with a flux model (see next page for figure).

64 Induction machine


2 3
23

P
W
M

[

T


]
X

[

T


]
-
1
i
A
1
i
d
1
, s
o
l l
u
A
1
u
V
, s
o
l l
u
W
, s
o
l l
u
U
, s
o
l l
u
R
q
u
q
1
i
d
1
i
V
i
W
i
U
i
B
1
u
q
1
, s
o
l l
u
R
d
u
d
1
i
q
1
i
d
1
S
i
q
1
S
i
+
-
-
-
-
+
+
+
S
+
-

m
S
m
i

S
i

B
1
i

q
1
i
d
1
i

J
T
M
3

~

f
i
e
l
d
w
e
a
k
e
n
i
n
g
s
p
e
e
d
c
o
n
t
r
o
l
f
l
u
x
c
o
n
t
r
o
l
c
u
r
r
e
n
t
c
o
n
t
r
o
l
t
o
r
q
u
e
c
o
n
t
r
o
l
i
n
v
e
r
t
e
r
i
n
d
u
c
t
i
o
n
m
a
c
h
i
n
e
t
r
a
n
s
f
o
r
m
a
t
i
o
n
d
e
c
o
u
p
l
.
.
n
e
t
w
o
r
k
i
n
v
e
r
s
e
t
r
a
n
s
f
o
r
m
a
t
i
o
n
f
l
u
x
m
o
-
d
e
l
c
u
r
r
e
n
t

c
o
n
t
r
o
l

p
L
1
h
1
+

2
E
d
u
E
q
u

Fig. 51: field-oriented induction machine, voltage injecting actuator, pulse-width modulation


4 Synchronous machine
4.1 Dynamic system of equations
In this section focus is put on salient-pole synchronous machines with distinctive magnetic
biaxiality and definitive damper winding. The following features have to be taken into
consideration:
permeances of direct- and quadrature-axis differ,
q d
X X

transformation has to be carried out on the asymmetric part of the system, which
means the rotor.
dt
d
dt
d


damper winding takes effect as well in the direct-axis as in the quadrature-axis.
Therefore a short-circuited direct-axis damper-winding and a short-circuited
quadrature-axis damper-winding has to be taken into account additionally to the
excitation winding.
the generator reference-arrow system (EZS) is usually chosen for generator operation.
The reference angle of the rotating coordinate system equals

+
2
0


Differing from the previous section we now omit the indices 1 and 2 for the stator and the
rotor. According to the literature lower case letters are used for the stator and capital letters
for the rotor. The conversion of the rotor on the number of the stator windings is retained and
indicated with a dash. For this reason the following denominations are obtained for the real
machine and for the transformed machine in the biaxial system:
i
u
u
u

u
v
i
v
u
w
i
w
i
F
u
F
i
D
,i
Q

d
u
i
d
d
u'
F
i'
F
q
i'
Q
i'
D
u
q
i
q

(t)

Fig. 52: real machine Fig. 53: machine in biaxial system
note: stator in EZS d: stator direct axis, q: stator quadrature axis
rotor in VZS D, Q according for damper windings
F index for rotor direct axis

66 Synchronous machine


In the equivalent winding system all the windings are separated in magnetizing in direct- and
quadrature axis. Five voltage equations and one torque equation apply:

q
d
d d
dt
d
i R u

+
1
(4.1)

d
q
q q
dt
d
i R u +

+
1
(4.2)

dt
d
i R u
F
F F F
'
' ' '

+ (4.3)

dt
d
i R
D
D D
'
' '
0

+ (4.4)

dt
d
i R
Q
Q Q
'
' '
0

+ (4.5)

( )
A d q q d el
M
dt
d
p
J
i i p M

(4.6)


In generator operation the torque on the shaft is not driving but braking:

W A
M M (4.7)


The following relations result for 5 accordant flux linkages (summarized in a matrix):

1
1
1
1
1
1
]
1

1
1
1
1
1
1
]
1

1
1
1
1
1
1
]
1

'
'
'
'
'
'
'
'
'
0 0 0
0 0 0
0 0
0 0
0 0
Q
q
D
F
d
Q hq
hq q
D hd hd
hd F hd
hd hd d
Q
q
D
F
d
i
i
i
i
i
L L
L L
L L L
L L L
L L L
(4.8 a-e)


Utilized resistances and inductances are denominated as follows:

o stator direct axis:
1
R
hd d
L L L +
1


o stator quadrature axis:
1
R
hq q
L L L +
1


o excitation winding:
'
F
R
hd F F
L L L +
' '

o damper winding direct-axis:
'
D
R
hd D D
L L L +
' '

o damper winding quadr. axis:
'
Q
R
hq Q Q
L L L +
' '

The matrix of the inductances is symmetric because the transformation is power-invariant.
converted to the
number of stator
windings

Synchronous machine 67


The transformation of the stator voltages and currents results in:

[ ] [ ]
1
]
1

1
]
1

v
u
q
d
u
u
T T
u
u
32
(4.9)

[ ] [ ]
1
]
1

1
]
1

v
u
q
d
i
i
T T
i
i
32
(4.10)

The rotor quantities do not need to be transformed, because they are already separated in 2
perpendicular axes rotating with
dt
d
.
So that the salient-pole synchronous-machine is completely described with

o five voltage equations,
o one torque equation and
o five flux-linkage equations

in a two-axis system rotating with synchronous angular speed.

The non-salient-pole machine (cylindrical-rotor machine) can be understood as a special case
featuring
q d
L L .

If damper windings are omitted and permanent magnets are treated as an injected magnetic
field, the permanent-field synchronous machine with rotor position encoder, the so called EC-
motor, is also described.

The system of differential equations is nonlinear and therefore it can only be solved
completely using numerical methods on a computer. Only if the speed is defined to be
constant, an analytic solution can be found.

As it was formulated in the requirements, the zero phase-sequence system is neglected. But if
the star point is connected and the load is not symmetrical, then there are a star point current
and a star point voltage:

0 3
0
+ + u u u u u
w v u St
(4.11)

If the zero phase-sequence system is separated from the unsymmetrical system, a symmetrical
system ensues, which can handled as before:

0
*
u u u
u u
,
0
*
u u u
v v
,
0
*
u u u
w w
(4.12 a-c)

For the zero phase-sequence system additional voltage- and flux-linkage equations are
obtained, which can be solved separately:

dt
d
i R u
0
0 0 0

+ (EZS) (4.13)

0 0 0
i L (4.14)

68 Synchronous machine


0
L and
0
R can be measured, if all the three winding phases are supplied in phase:

1 0
3 R R ,
1 0
3 L L (4.15 a,b)

After the inverse transformation of the solution of the symmetrical system, the solution is
added to the zero phase-sequence system solution:

0
*
i i i
u u
+ ,
0
*
i i i
v v
+ ,
0
*
i i i
w w
+ (4.16)

4.2 Steady-state operation of salient-pole machines at
mains power supply
In steady-state operation the flux-linkages in the rotating system are constant, i.e. 0

dt
d
and
the rotational speed is constant, i.e. 0
dt
d
.
Therefore the voltage equations and the torque equation are decoupled. The stator resistance
is neglected, i.e. 0
1
R . For this reason the system of equations is simplified as follows:

q d
u (4.17)

d q
u (4.18)

' ' '
F F F
i R u (4.19)

0
'

D
i (4.20)

0
'

Q
i (4.21)

( )
q d d q A
i i p M (4.22)

'
F hd d d d
i L i L + (4.23)

q q q
i L (4.24)

The other flux-linkages are not taken care of, the voltage equation for the excitation winding
is trivial and the damper currents in steady-state operation are equal zero. Pasting the flux-
linkages results in:

q q d
i L u (4.25)

'
F hd d d q
i L i L u (4.26)

( )
q q d d A
i u i u
p
M +

(4.27)

Synchronous machine 69


For the inverse transformation of the system rotating with
dt
d
dt
d
the choice of the
integration constant
0
is still free. Practically we choose for the synchronous machine:

+
2
0
(4.28)

d
q
i
q
i
d
(
t)
+

-Im
d
+Re
U
q

2
-
+

+
U
d
U
q

Fig. 54: coordinate system Fig. 55: complex plane


q d
j q j d
U U e
u
j e
u
U + +
0 0
3 3

(4.29)

Then the rotor axis, in which the machine is magnetized (direct axis) lags behind with an
angle of

+
2
and the quadrature axis (the axis of the synchronous generated voltage) is
leading with an angular displacement of . As desired the terminal voltage coincides with the
real axis.


Inverse transformation:

j q
q
j
j
q q
d
e
i
L j e e
i L
U

3 3
2
(4.30)

j
j
F hd d d
q
e e
i L i L
j U

2
'
3




j F
hd
j d
d
e
ji
L j e
ji
L j

3 3
'
(4.31)
inverse
transformation

70 Synchronous machine


We define:


j q
q
e
i
I
3
(4.32)

j d
d
e
ji
I

3
(4.33)

j F
F
e
ji
I

3
'
,
(4.34)

Then we obtain:

q q d
I L j U (4.35)

'
F hd d d q
I L j I L j U (4.36)

With appliance of reactances:


d d
L X ,
q q
L X ,
hd hd
L X (4.37 a-c)

and the synchronous generated voltage:


'
F hd P
I jX U (4.38)

follows:


p d d q q q d
U I jX I jX U U U + + (4.39)

and respectively:


d d q q p
I jX I jX U U + + (4.40)

I
d
U
q
U
q
I
q
I
d
U
d
jX
q
I
q
jX
d
I
d

I
'
F
U
P

For this reason the phasor diagram of
salient-pole machines can be drawn, if
U and I and the directions of the d-
(direct) and q-(quadrature) axes are
known respectively predetermined.









Fig. 56: SYM, phasor diagram

Synchronous machine 71


If decomposed in separate components follows:

q d
U U U + (4.41)

q d
I I I + (4.42)

+ (4.43)

sin I I
d
, cos I I
q
(4.44 a,b)

sin U I X U
q q d
, cos U I X U U
d d p q
(4.45 a,b)

For practical use mostly the following depiction of the phasor diagram is applied:
I
d
U
q
U
p
I
q
I
d
jX
q
I
jX
d
I

j(X
d
-X
q
)I
d

Fig. 57: SYM, common phasor diagram

Construction manual:

from the power supply the type of load is predetermined: U , I , are known
orientation of axes d and q is unknown
phasors of I jX
q
and I jX
d
are to be drawn
the q-axis, i.e. the direction of
p
U and , is determined by the straight line from the
origin through the pivot of I jX
q

based on that, the current I can be decomposed into
d
I and
q
I
the perpendicular from the pivot of I jX
d
on the q-axis is due to the magnitude of
p
U

The phasor diagram of salient pole machines differs from that of cylindrical-rotor machines in
the difference ( )
q d
X X , i.e. the different admittances in d- and q-axis
d q
X X 8 , 0 ... 5 , 0 .
For X X X
d q
the phasor diagram of the cylindrical-rotor machine is obtained.

72 Synchronous machine


If the direct components are substituted by the root-mean-square values, then the torque
ensues to:

,
_

+
3 3 3 3
3
q q
d d
A
i u
i u p
M

(4.46)

( )
q q d d A
I U I U
p
M +

3
(4.47)

That means, that the active power in the d- and the q-axis are adding up.
With the relations found in the phasor diagram

d
p
d
X
U U
I
cos
, (4.48)

sin U U
d
(4.49)

q
q
X
U
I
sin
, (4.50)

cos U U
q
(4.51)

the torque is converted and described as a function of the angular displacement.

1
1
]
1

+


q d
p
A
X
U
U
X
U U
U
p
M

sin
cos
cos
sin
3
(4.52)


1
1
1
1
]
1

,
_

4 4 4 4 3 4 4 4 4 2 1
4 43 4 42 1
l
Syn M
d q
M
d
P
X X
U
X
U U p
Re
2 sin
1 1
2
sin
3
2

(4.53)

M
M
kipp

kipp

2

M
Rel
M
Syn
M
A

Fig. 58: SYM, torque vs. lagging angle

Synchronous machine 73


The torque of salient-pole machines is composed of two components. The first part also
appears in cylindrical-rotor machines and is depending on the excitation. It is called
synchronous torque. The second part does not depend on the excitation and results from the
difference of the permeances
q d
X X . This is called the reluctance torque.
The magnetic unsymmetrical rotor tries to adjust itself in such a manner, that the magnetic
energy is minimized. This does not depend on the polarity of the excitation. Therefore the
reluctance torque has twice the frequency of the synchronous torque. The breakdown torque is
moved to values
2

<
Kipp
.
In low power ranges machines without excitation winding are built, only using the reluctance
torque caused by the difference of the permeances in d- and q-axis. These are the so called
reluctance machines.
For X X X
q d
the torque equation of the cylindrical-rotor machine is matched.
4.3 Determination of X
d
and X
q

d
X can be determined with a measurement in continued short circuit with no-load excitation
and pure direct axis field.
I
k
= I
d
U
p
= U
NStr
jX
d
I
d

Fig. 59: SYM, phasor diagram for continued short circuit

q
X can be determined with a measurement of the angular displacement when the machine is
loaded with pure active-power load.

U
NStr
jX
d
I

jX
q
I
U
P

Fig. 60: SYM; phasor diagram for pure active power load
0 U (4.54)
0 0
q q q d
I I X U (4.55)

0
d d p q
I X U U (4.56)
K
d
p
d
I
X
U
I (4.57)
0 F F
I I :
NStr p
U U (4.58)
K
NStr
d
I
U
X (4.59)
1 cos (4.60)
NStr
q
U
I X
tan (4.61)
tan
I
U
X
NStr
q
(4.62)

74 Synchronous machine


Small salient-pole machines can be driven unexcited with a very small slip on the power
supply. Depending on the rotor position concerning the stator field different currents appear
depending on the width of air gap. The slip has to be very small (a few per mill), so that a
damper winding has no influence on the measurement. Current and voltage, oscillating with
the slip frequency, are measured.


Fig. 61: SYM, current and voltage

Direct-
d
X and quadrature component
q
X of the so called synchronous reactance ensue to:


min
max
I
U
X
d
(4.63)


max
min
I
U
X
q

(4.64)


4.4 Sudden short circuit of the cylindrical-rotor machine
As an example for the use of the dynamic equations of the direct- and quadrature-axis theory
for the calculation of dynamic phenomena, we now discuss the three-phase sudden short
circuit of the cylindrical-rotor machine from the no-load operation.

The sudden short circuit is the transient phenomenon, which appears instantaneously after
short-circuiting of the stator-circuit terminal. After all phenomena faded away, the machine is
in continued (or sustained) short circuit.


Synchronous machine 75


The solution of this problem will be carried out analytically without using a computer.
Therefore some simplifications have to be made for easing purposes:

o resistances of the stator winding are neglected, 0
1
R .

o rotor speed remains constant during the dynamic phenomenon and is equal to the
synchronous speed.

o The cylindrical-rotor machine is symmetrical
h hq hd
L L L

with two identical rotor
windings with an electrical displacement of 90. The excitation winding with slip-
rings is supplied with DC current. The quadrature-axis damper-winding is short-
circuited. The machine has no direct-axis damper-winding.
o a rotating coordinate system is chosen with
dt
d
dt
d

. The stator system has the

denominators d and q with
1
R and
1
L . The rotor system has the excitation winding F,
the quadrature-axis damper-winding Q with
'
2
R and
'
2
L .

o reference axis is again the rotor direct axis. The initial condition is given by the
switching instant

+
2
t .
U
V
W
X
Y
Z
Q
F

-Im
d
+Re
U
q

2
-
+

+
U
d
U
q

Fig. 62: SYM, orientation of windings Fig. 63: SYM, revolving coordinate system

With the upper called simplifications the following set of equations is derived:

q
d
d
dt
d
u

(4.65)

d
q
q
dt
d
u +

(4.66)

dt
d
i R u
F
F F
'
' '
2
'

+ (4.67)

dt
d
i R
Q
Q
'
' '
2
0

+ (4.68)

76 Synchronous machine


( )
q d d q A
i i p M (4.69)

'
1 F h d d
i L i L + (4.70)

'
1 Q h q q
i L i L + (4.71)

d h F F
i L i L +
' '
2
'
(4.72)

q h Q Q
i L i L +
' '
2
'
(4.73)

The torque equation is decoupled and can be handled separately. Before starting the solution,
initial conditions need to be pre-defined.

The initial state before switching is no-load operation and rated voltage:

( ) 0 0
d
i , ( ) 0 0
q
i , ( ) 0 0
'

Q
i , ( )
'
2
'
'
0
'
3 0
R
u
I i
F
F F
(4.74 a-d)

'
0 F
I is the root-mean-square value of the no-load excitation current based on the stator, which
causes the induction of rated voltage at rated speed. For this case the flux-linkages for the no-
load operation are:

( ) ( ) 0 0
'
F h d
i L (4.75)

( ) 0 0
q
(4.76)

( ) ( ) 0 0
' '
2
'
F F
i L (4.77)

( ) 0 0
'

Q
. (4.78)

Out of this follows the stator voltage for the no-load operation:

( ) 0 0
d
u (4.79)

( ) ( ) ( ) 3 3 0 0 0
'
0
'

NStr F h F h d q
U I X i L u (4.80)

The inverse transformation with

+
2
0
results in:
( )

j
j
F h j q j d
e e
i L
j e
u
j e
u
U

+ +

2
'
3
0
0
3 3
0 0


( )
j
NStr p F h
j F
h
e U U I jX e
ji
jX

0
'
0
'
3
0
(4.81)

with being the switching angle

o if 0 , then at the time 0 t the voltage in phase U is in the real axis, i.e. switching
at maximum voltage (flux-linkage equal zero).
o if
2

t , then at the time 0 t the voltage in phase U is in the plus/minus

imaginary axis, i.e. switching at voltage zero crossing (flux-linkage is maximum).

Synchronous machine 77


After defining the initial conditions the solution of the system of equations for 0 t can be
started. First the voltage equations of the stator are considered. For a three-phase short circuit
follows:

q
d
d
dt
d
u

0 (4.82)

d
q
q
dt
d
u +

0 (4.83)

The solution of these two differential equations can be easily done.
q
and
d
are harmonic oscillations:

t B t A
q
sin cos + (4.84)

t A t B
d
sin cos + (4.85)

The integration constants are determined by pasting the initial conditions:

( ) 0 0 A
q
(4.86)

( ) ( ) 0 0
'
F h d
i L B (4.87)

For this case the according stator flux-linkages are obtained:

( ) t i L
F h d
cos 0
'
(4.88)

( ) t i L
F h q
sin 0
'
(4.89)

Stator currents in the flux-linkage equations are eliminated by solving the stator flux-linkages
for stator currents and pasting them into the rotor flux-linkages:

1
'
1
1 +

F d
d
i
L
i (4.90)

1
'
1
1 +

Q q
q
i
L
i (4.91)

( )

,
_

+
+

,
_

+
+

' '
1
'
1
'
1
cos 0
1
1
1 1
1
F h F h F h d F
i L t i L i L

(4.92)

( )
,
_

+
+

,
_

+
+

' '
1
'
1
'
1
sin 0
1
1
1 1
1
Q h F h Q h q Q
i L t i L i L

(4.93)

Besides the known stator flux-linkages, the rotor flux-linkages still include the unknown rotor
currents. By pasting the rotor flux-linkages into the rotor voltage equations the differential
equations for the rotor currents are achieved:

( ) ( )

,
_

+
+
+
dt
di
L t i L i R i R u
F
h F h F F F
'
'
1
' '
2
' '
2
'
1
sin 0
1
1
0

(4.94)

( )

,
_

+
+
+
dt
di
L t i L i R
Q
h F h Q
'
'
1
' '
2
1
cos 0
1
1
0

(4.95)

78 Synchronous machine


It is to be divided by
'
2
R and needs to be defined ( compare to transformers)

open-circuit time constant of the rotor winding, open-coil stator:

( )
'
2
2
'
2
'
2
0
1
R
L
R
L
T
h
F
+


(4.96)

short-circuit time constant of the rotor winding, short-circuited stator:

0 F FK
T T (4.97)

After conversion follows:

( )
,
_

+ + t T i i
dt
di
T
FK F F
F
FK

sin
1
1 0
' '
'
(4.98)

( ) t T i i
dt
di
T
FK F Q
Q
FK

cos
1
0
'
'

+ (4.99)

These two inhomogeneous differential equations can be solved using Laplace transformation.
Considering that 1 >>
FK
T , i.e. 0
1

FK
T
, solutions for the rotor currents follow as:

( )

,
_

FK
T
t
F F
e t i i

1
cos
1
1 0
' '
(4.100)

( ) t i i
F Q

sin
1
0
' '

(4.101)

An inverse transformation is not required, because the excitation winding and the damper
winding are both arranged perpendicular and mounted on the rotating rotor. Now we know
the stator flux-linkages and the rotor currents.
Rotor currents in the transformed system can be determined. Pasting known relations leads to:

( )
( )
1
'
1
'
1
1
cos
1
1 0
cos 0

,
_

FK
T
t
F
F h
d
e t i
L
t i L
i (4.102)

( )
1
'
1
'
1
sin
1
) 0 (
sin 0

t i
L
t i L
i
F
F h
q
(4.103)

( )
1
1
1
1
]
1

,
_

1
]
1

t
e t
i
i
i
FK
T
t
F
q
d

sin
1
1
1 cos
1
1
0
1
'
(4.104)

Synchronous machine 79


Now the inverse transformation to the stationary three-phase system can be executed. Focus is
put on the current in phase U:

,
_

,
_

,
_

2
sin
2
cos
3
2
3
2
t i t i i i
q d A u
(4.105)

After pasting, converting and using ( )
h
Nstr
F
L
U
i

3
0
'
the stator current ensues to:

( )

,
_

+ +

,
_

sin
1
sin
1
1
2
1
t e
L
U
i
FK
T
t
NStr
u
(4.106)

The stator current
u
i now consists of an alternating component, which decreases from high to
lower values, and a DC component, which does not fall off.
The bahavior of the DC component is physically not correct. The reason for this purpose is
the neglect of the stator resistance, to make the system of equations solvable analytically. To
obtain a universal valid solution for the time characteristic of the stator current, the
declination of the DC component has to be added by the factor
dK
T t
e

.
dK
T is the short-circuit time constant of the stator winding.

1
1
0
R
L
T
d
(4.107)

0 d dK
T T (4.108)

Then follows:

( )
,
_

+ +

,
_

dK FK
T t T t NStr
u
e t e
L
U
i

sin
1
sin
1
1
2
1
(4.109)

By introducing the synchronous reactance
1 1
L X and the short-circuit reactance
1 1
X X
K
the stator current finally results in:

( )

'

+ +
1
]
1

,
_

+

dK FK
T t
K
T t
K
NStr u
e
X
t e
X X X
U i sin
1
sin
1 1 1
2
1 1 1 1
(4.110)

As well as for the DC component of the stator current, the declination has to be physically
added in the corresponding alternating component of the rotor currents:

( )

'

+

dK FK
T t T t
F F
e t e i i

cos
1 1
1 0
' '
(4.111)

( )

'

dK
T t
F Q
e t i i

sin
1
0
' '
(4.112)

80 Synchronous machine


Now the results can be illustrated and interpreted:

switching at maximum voltage: 0

t U u cos 2 (4.113)

1
1
]
1

'

,
_

+

t e
X X X
U i
FK
T
t
K
N u
sin
1 1 1
2
1 1 1
(4.114)


Fig. 64: voltage, current vs. time


switching at voltage zero crossing:
2

t U u sin 2 (4.115)

1
1
]
1

'

,
_

+

dK FK
T
t
K
T
t
K
N u
e
X
t e
X X X
U i
1 1 1 1
1
cos
1 1 1
2 (4.116)


Fig. 65: voltage, current vs. time

Synchronous machine 81


( )
1
1
]
1

+

dK FK
T
t
T
t
F F
e t e i i

cos
1 1
1 0
' '
(4.117)


Fig. 66: exciter current vs. time


( )
dK
T
t
F Q
e t i i

sin
1
0
' '
(4.118)


Fig. 67: damper current vs. time


The time characteristic of the stator currents in the stator windings depends on the
instant of switching, because of the sinusoidal changing flux-linkage before switching.
Switching at maximum voltage: There is no DC component. The short-circuit current
has a time-delay of
2

and starts at zero crossing

,
_

K
NStr
u
X
U
I
1
max
2
.

Switching at voltage zero crossing: There is a DC component in full, because the
short-circuit current with a time-delay of
2

has to be compensated to zero. After one

period the DC component and the alternating component add up to twice their value.
The DC component declines with
dK
T .

,
_

K
NStr
u
X
U
I
1
max
2
2

Instantaneously after switching the magnitude of the short-circuit current relevantly
depends on the short-circuit reactance
1 1
X X
K
. The initial symmetrical short-
circuit current
K
NStr
X
U
1
2
flows. The initial symmetrical short-circuit current declines
with the short-circuit time constant
FK
T to the sustained short-circuit current
1
2
X
U
NStr

, which depends on the synchronous reactance.

82 Synchronous machine


The time characteristic of the rotor currents does not depend on the instant of
switching, because the flux-linkage before switching was constant.
The DC component of the stator current corresponds to the alternating component of
the rotor. The alternating component of the stator current corresponds to the DC
component of the rotor.
After switching the flux linkage can not change in the short-circuited stator winding.
Therefore the currents in the excitation winding and in the damper winding have to
magnetize oppositely and counteracting. Because of the resistance of the stator
winding, the alternating component declines with
dK
T

. The DC component is
increased as well as the stator current with a factor of

1
and declines as well as the
alternating component of the stator current with
FK
T
1
.

In practice the following values are typical for cylindrical-rotor generator:
o 2 2 , 1
0
1
1
L
K
N
N N
I
I
I U
X
x

o 25 , 0 15 , 0
1 1
L x x
K


o ms T
dK
250 60L

o s T
FK
2 5 , 0 L

4.4.1 Physical explication of the sudden short circuit
If synchronous machines are driven in no-load operation with rated voltage or if driven with
rated load with sudden short-circuit of the terminals, short-circuit currents occur, which are a
multiple of the sustained short-circuit current. Current peaks also occurring in the rotor are
multiple times higher than the excitation current in operation at rated values.

Figures 68 and 69, showing a snapshot of the flux distribution in the synchronous machine
before the sudden short circuit and half a period after, give a simple physical explication.

U
V
W Y
Z
E F
X

h

U
V
W Y
Z
E F
X

F

s

s

h
2
h

Fig. 68: 0 t Fig. 69: t

Synchronous machine 83


For 0 < t the flux-linkage in the stator winding U is maximal. Therefore the voltage in
phase U is:

t U u sin 2 (4.119)

And the flux-linkages are:


'
1
2
Fo h
I L (4.120)

( )
'
0
'
2
'
0 2 2
2 2 1
F F h
I L I L + (4.121)

The synchronous machine is three-phase short circuited, when phase U has its voltage zero
crossing. With neglect of the resistances of stator winding and rotor winding we obtain in
stator and rotor (for 0 t ):

0

dt
d
u , i.e. const (4.122)

i.e. a constant flux-linkage is forced. This means, that the flux
h
, whose magnetic loop was
closed along the stator at no-load operation, keeps its magnitude but is now displaced to the
magnetically worse conductive leakage path. If the rotor has turned half a revolution, the
stator winding has to generate a current-linkage, which is able to drive a flux with the double
magnitude of the pole-flux through the leakage path, to retain the original flux-linkage.

For t follows:

( )
'
0
!
1 1
'
0 1
2 2 1 2
F h h F h
I L I L I L + + (4.123)

and then

1 1
'
0
1
2 2 2 2
2
X
U
L
L I
I
N h F

(4.124)

The sustained short-circuit current is increased with a factor of

1
. The relevant reactance is
1 1
X X
K
. When switching at voltage zero crossing, the DC component occurs entirely,
which doubles the amplitude of the sudden short-circuit current.

4.4.2 Torque at sudden short circuit
The dynamic calculation shows, that the occurring currents at sudden short circuit are much
higher than the sustained short-circuit current. Consequentially the forces in the machine are
much higher and have to be taken into consideration for the construction of synchronous
machines. Therefore the torque at sudden short circuit is to be estimated. We want to consider
the maximal mechanical stresses so the declination of the currents is neglected, leading to:

o 1

FK
T t
e

o 1

dK
T t
e

84 Synchronous machine


Then the torque and the currents ensue to:

( )
'
1
'
1 Q d F q h el
i i i i L p M (4.125)

( ) 1 cos
3 1
1 cos
1 3
1 1

,
_

,
_

t
L
U
t
L
U
i
NStr NStr
d

(4.126)

( ) t
L
U
i
NStr
q


sin
3
1

(4.127)

( ) ( ) t
L
U
t
L
U
i
h
NStr
h
NStr
F

cos 1 1
3 1
cos
1
1
3
'

,
_

(4.128)

( ) ( ) t
L
U
i
h
NStr
Q


sin 1
3
'

(4.129)

Pasting and converting leads to:

t
X
U p
M
NStr
elK


sin
3
1
2

(4.130)

The peak transient torque oscillates at system frequency. Based on the rated torque results:

t
x
I U
p
t
X
U p
M
M
m
N K
N N N
NStr
N
elK
el

sin
cos
1
cos
3
sin
3
1
2

(4.131)

If the numerical values are calculated, it can be seen, that the peak transient torque is
increased by a factor

5 5 , 9
8 , 0 25 , 0
1
7 , 0 15 , 0
1
L L



related to the according rated torque. This is important for the dimensioning of the mechanical
system. The torque results from the co-action of the DC component of the stator current and
the rotating field of the rotor.


Synchronous machine 85


4.5 Sudden short circuit of salient-pole machines
4.5.1 Analytical calculation
Subject of discussion will be the sudden short circuit at no-load operation and rated voltage of
the salient-pole machine again using simplifying assumptions. The complexity of the
calculation is considerable a result can be given directly.
Stator voltage

( ) + t U u
N u
sin 2 (4.132)

Stator current:

{
( ) ... cos
1 1 1 1 1
2
...
...
'
...
' ' '
,
' ' '
+

'

1
1
1
1
]
1

,
_

,
_

t
X
e
X X
e
X X
U i
state steady
d
transient
T
t
d d
nt subtransie
T
t
d d
Str N u
d d
4 4 4 3 4 4 4 2 1 4 4 4 3 4 4 4 2 1


( )

,
_

,
_

4 4 4 4 4 4 3 4 4 4 4 4 4 2 1 4 4 4 4 4 3 4 4 4 4 4 2 1
frequent double
T
t
q d
asymmetric
T
t
q d
A A
e
X X
t e
X X
' ' ' '
...
' ' ' '
1 1
2
1
2 cos
1 1
2
1
cos ... (4.133)

The single components of the short-circuit current are determined by the magnetizing
reactances and the leakage reactances of the concerned windings. For the occurring couplings,
the equivalent magnitudes
' '
d
X ,
' '
q
X and
'
d
X are defined.

The subtransient reactance
' '
d
X is the effective reactance in the d-axis at first point of time.
Here the stator winding d, the damper winding D and the excitation winding F feature a
magnetic coupling at first. The reactance at sudden short circuit results from the magnetizing
reactance and the leakage reactances ( compare with transformers). The according transient
time constant
' '
d
T is obtained as well.
X
hd
X
D
X
F
X
1
X''
d
d
D
F

Fig. 70: exciter-, damper windings
F D hd d
X X X X X +
1
' '


1
' '
' '
R
L
T
d
d


' '
' '
d
NStr
K
X
U
I

86 Synchronous machine


Besides the stator winding q only the quadrature-axis damper-winding Q is active in the q-
axis. For the calculation of the quadrature-axis subtransient reactance
' '
q
X , the simple parallel
connection of the magnetizing reactance and the damper winding in the q-axis are gained.
X
hq
X
Q
X
1
X''
q
q
Q

Fig. 71: quadrature components

If the influence of the damper winding is declined rapidly, still the excitation winding and the
short circuited stator winding in the d-axis are coupled. The transient reactance
'
d
X and the
transient time constant
'
d
T are calculated as follows:
X
hd
X
F
X
1
X'
d
d
F

Fig. 72: direct components

A transient reactance
'
q
X does not exist, because there is no other winding in the rotor q-axis
besides the damper winding.
After the declination of the longer lasting transient process, the stator current fades to the
sustained short-circuit current, which is determined by the direct reactance
d
X .

d
NStr
K
X
U
I (4.134)

The appearing DC component depends on the instant of switching. Switching at zero voltage
crossing results in an entire DC component. If the switching is executed at maximum voltage,
no DC component comes up. The DC component falls off with the asymmetric time constant
A
T .

' ' ' '
' ' ' '
1
2
1
d q
d q
A
L L
L L
R
T
+

(4.135)
Q hq q
X X X X +
1
' '

X X X X +
' '
F hd d
X X X X +
1
'
1
'
'
R
L
T
d
d

'
'
d
NStr
K
X
U
I

Synchronous machine 87


The double-frequency component, as well dropping with the asymmetric time constant
A
T ,
results from the magnetic asymmetry
' ' ' '
q d
X X at first instant.
Standard machines with direct-axis and quadrature-axis damper-winding are not asymmetrical
because of
' ' ' '
q d
X X .
For this reason the double-frequency component disappears and the time constant
A
T is equal
to the subtransient time constant
' '
d A
T T .


Fig. 73: sudden short-circuit current with maximum DC component, one phase

The maximum peak of the sudden short-circuit current in phase U occurs half a period after
short-circuiting the three phases at voltage zero crossing of phase U. The amplitudes of the
subtransient and the DC component add up. Regarding the damping, the current peak is:

' '
8 , 1 2
d
NStr
S
X
U
I (4.136)

In agreement with a VDE (association of german electrical engineers), the maximum value
has to be smaller than 15 times the peak value of the rated current, because of the high
occurring forces and oscillating torques.

o 15
8 , 1 8 , 1
2
' ' ' '

d N d
NStr
N
S
X I X
U
I
I


o 12 , 0
15
8 , 1
' '

d
X


88 Synchronous machine


Typical values of the reactances and time constants of salient-pole machines with definite
damper winding are:

o 4 , 1 8 , 0 L
d
x p.u.

o 9 , 0 4 , 0 L
q
x p.u.

o 4 , 0 2 , 0
'
L
d
x p.u.

o 25 , 0 12 , 0
' ' ' '
L
q d
x x p.u.

o s T
d
5 , 2 5 , 0
'
L

o s T T
A d
1 , 0 02 , 0
' '
L

Those reactances and time constants can be determined from the sudden short-circuit.

4.5.2 Numerical solution
Using numerical integration methods on the computer, the complete system of equations of
synchronous machines can be directly solved without any simplifying assumptions.
Particularly the assumption of a constant speed is not to apply anymore.
Indeed the system of equations has to be converted into state form, which is well known from
control engineering.

q d d
d
R i u
dt
d
+

1
(4.137)

d q q
q
R i u
dt
d

1
(4.138)

' ' '
'
F F F
F
R i u
dt
d

(4.139)

' '
'
D D
D
R i
dt
d

(4.140)

, ,
,
Q Q
Q
R i
dt
d

(4.141)

( ) [ ]
q d d q A
i i p M
J
p
dt
d

(4.142)

Therefore the inductivity matrix has to be inverted. This is not explained in detail here.

[ ]
1
1
1
1
1
1
]
1

1
1
1
1
1
1
]
1

'
'
'
1
'
'
'
Q
q
D
F
d
Q
q
D
F
d
L
i
i
i
i
i
[ ]
1
1
1
1
1
1
]
1

'
'
'
0 0 0
0 0 0
0 0
0 0
0 0
Q hq
hq q
D hd hd
hd F hd
hd hd d
L L
L L
L L L
L L L
L L L
L (4.143)

Synchronous machine 89


For this reason the complete system of equations is known as interpreted in the shown
structure diagram (Fig. 74), which has to be solved using:

o excitation values:
d
u ,
q
u ,
F
u ,
A
M

o state values:
d
,
q
,
'
F
,
'
D
,
'
Q
,

P
P
i
d

d
M
A
+
+
-
R
1

P
J
d

dt
u
d

q
R
1
i
d

d
+
- -
u
q

d
R
1
i
q
-
-

q

q
R
1
i
q
u
F
R
F
i
F
+
-

F

F
R
F
i
F
R
D
i
D
-

D

D
R
D
i
D
R
Q
i
Q
-

Q

Q
R
Q
i
Q
(L)
-1
'
'
'
'
'
'
'
'
-
'
'
'
'
'

Fig. 74: structure diagram of salient-pole machines

1) Initial conditions for 0 < t need to be predefined, for example no-load operation at
rated voltage and synchronous speed.

( ) ( ) ( ) ( ) 0 0 0 0 0
' '

Q D q d
i i i i (4.144)

( ) ( ) 0 0 0 +
q d
u (4.145)

( ) ( ) ( )
NStr F hd d q
U i L u 3 0 0 0
'
(4.146)

( ) ( ) 0 0
' ' '
F F F
i R u (4.147)

and 0
A
M .

90 Synchronous machine


Practically we define

+
2
t . Then it is:
( ) ( ) ( )
,
_

,
_

+ +

2
sin 3 0
3
2
sin 0 cos 0
3
2
0 1
t U u u u
NStr q d u


( ) + t U cos 2 (4.148)

o 0 switching at maximum voltage

o
2

t switching at zero voltage crossing

2.) The following machine parameters need to be known:
h
L ,
d
c ,
q
c

1
R ,
F
R ,
Q D
R R

N
U
1
,
F
U

J ,
N
M ,
N
f , p

and the leakage factors of the single windings:


1
,
F
,
Q D


3.) Excitation values:
For 0 > t the stator voltages for the sudden short-circuit are substituted by zero.

0
q d
u u

For coarse synchronizing the stator voltages are:

0
d
u ,
NStr q
U u 3

The two other excitation states are:

0
A
M , ( ) 0
, ,
F F F
i R u

4.) For example a machine is chosen with the following data:

o apparent power: kVA S
N
100

o nominal voltage: .) ( 380 verk V U
N


o power factor: 8 , 0 cos
N


o pole pairs: 3 p

o frequency: Hz f
N
50

o nominal current: A I
N
152

o nominal torque: Nm M
N
764

Synchronous machine 91



Fig. 75: salient-pole machine, sudden short-circuit and coarse synchronizing: stator current
U
i



Fig. 76: sal. pole machine, sudden short-circuit and coarse synchronizing:excitation current
F
i

92 Synchronous machine



Fig. 77: sal. pole machine, sudden short-circuit and coarse synchr.: direct stator current
d
i




Fig. 78: sal. pole mach., sudden short-circuit and coarse synchr.: quadrature stator current
q
i

Synchronous machine 93



Fig. 79: salient pole mach., sudden short-circuit and coarse synchr.: direct damper current
D
i



Fig. 80: sal. pole mach., sudden short-circuit and coarse synchr.: quadrature damp. current
Q
i

94 Synchronous machine



Fig. 81: salient pole machine., sudden short-circuit and coarse synchronization.: torque M



Fig. 82: salient pole machine., sudden short-circuit and coarse synchronization.: speed n

Synchronous machine 95


Summarized description of the behaviour of salient pole synchronous machines in coarse
synchronization and sudden short-circuit operation:

Because the switching instant was at the zero voltage crossing of phase U, the stator
current
u
i contains an extended DC component. The time constant for the slope of the
sustained short-circuit current is
' '
d
T respectively
'
d
T .

The excitation current jumps to a value, which is

1
times the no-load value, and falls
off corresponding to the stator current. DC components of stator currents cause an
alternating component of the excitation current.

The transformed stator currents
d
i and
q
i are DC currents in steady-state operation.
d
i
represents the reactive component, which is responsible for the magnetization,
q
i
represents the active component, which generates the torque.

The damper currents
D
i and
Q
i are only effective at first instant after switching,
anytime else they are equal zero.

Even though the driving torque remains constant, the speed declines after short-
circuiting because of the ohmic losses in the resistances.

After the sudden short-circuit the machine was coarse synchronized by reconnecting to the
power supply and again transient phenomena occur with high current- and torque-peaks.


96 Synchronous machine


4.6 Transient operation of salient-pole machines
Previous sections show, how the system of differential equations of synchronous machines
can be solved to calculate the dynamic behavior using analytical or numerical methods.
Using simplifying assumptions, dynamic transient phenomena can be handled in analogy to
steady-state phenomena. In this case non-linear system of equations are not necessarily
subject of integration:

1.) It is assumed, that the speed of the machine is constant during the transient
phenomenon. The bigger the moment of inertia of the rotating machine, the better this
assumption is fulfilled.

2.) The effect of the damper winding is not taken into consideration, because the
subtransient phenomenon declines rapidly. In lots of cases there is no damper winding
at all.

3.) The induced voltage components in the voltage equations can be neglected in
contrast to the rotary induced voltage ( ) 1 >> T . Therefore the DC components of the
currents do not occur.

4.) The excitation flux-linkage during the transient phenomenon is assumed to be
constant, i.e. the excitation current changes according to the stator current. This
assumption is fulfilled, if the resistance of the excitation winding is very small
( ) 0
F
R , because the short-circuited winding keeps its flux constant 0

dt
d
F
and
const
F
. The assumption can be fulfilled too, if the machine is equipped with a
voltage controller to compensate the ohmic voltage drop if the current changes, i.e.
dt
d
i R u
F
F F F

0 .

5.) The resistance of the stator winding can be neglected: 0
1
R

6.) There is no zero phase-sequence system.

With these assumptions the system of dynamic equations is simplified as follows:

o voltages:
q d
u (4.149)

d q
u (4.150)

0
'

F
u (4.151)

o currents:
'
D
i does not appear!

'
Q
i does not appear!



Synchronous machine 97



o torque
( )
q d d q W
i i p M (4.152)

o flux linkages
'
F hd d d d
i L i L + (4.153)

' ' '
F F d hd F
i L i L + (4.154)

'
D
does not appear!

q q q
i L (4.155)

'
Q
does not appear!

Flux linkages
d
and
q
are inserted into the voltage equations and the inverse
transformation with

+
2
0
is to be executed according to section 4.2.

q q
j
j
q q j d
d
I jX e e
i L
e
u
U

2
3 3
0
(4.156)

( )
'
2
'
3 3
0
F hd d d
j
j
F hd d d j q
q
I jX I jX e e
i L i L
j e
u
j U
+

(4.157)

whereas

j q
q
e
i
I
3
(4.158)

j d
d
e
i
j I
3
(4.159)

j F
F
e
i
j I
3
'
'
(4.160)

besides


'
F hd p
I jX U (4.161)

The inverse transformation is done for the excitation flux in the direct axis in the same way:

( )
' '
' ' '
'
3 3
0
F F d hd
j F F d hd j F
F
I L I L e j
i L i L
e +
+

(4.162)

The index 0 describes the state before switching. If the excitation flux-linkage has to be
constant after switching, then it has to be:

.
'
0
'
0
' ' '
const I L I L I L I L
switching before
F F d hd
switching after
F F d hd F
+ +
4 4 4 3 4 4 4 2 1 4 4 3 4 4 2 1
(4.163)

98 Synchronous machine


This means, that the excitation current changes according to the stator current:

2
0
'
0
'
1 +

+
d d
F F
I I
I I (4.164)

Therefore the voltage equations in the quadrature-axis are:

4 4 3 4 4 2 1
'
2
0
'
1
F
I
d d
Fo hd d d q
I I
I jX I jX U

,
_

(4.165)


( )( ) ( )( )
2 1
0
2 1
'
0
1 1 1 1
1
1
+ +

,
_

+ +

d d
d d F hd
I jX
I jX I jX (4.166)

( )
0
'
0
1
d d d d p
I X j I jX U (4.167)

0 p
U is the synchronous generated voltage before switching,
d d
X X
'
is the transient
reactance.

With that the transient synchronous generated voltage
,
p
U can be defined, therefore remaining
constant after switching still after switching:

1.) ( ) const I X U I X j U U
d d q d d p p
+
'
0 0
'
1 (4.168)

The synchronous generated voltage changes according to the excitation current:

,
_

+
2
0
' '
1
d d
Fo hd F hd p
I I
I jX I jX U (4.169)


( )( ) ( )( )
2 1 2 1
0
'
0
1 1 1 1 + +

+
+ +

d d d d
F hd
I jX I jX
I jX (4.170)

( ) ( )
d d d d p
I X j I X j U + 1 1
0 0
(4.171)

2.) ( )
d d p p
I X j U U + 1
'
(4.172)

The direct-axis voltage is still:

3.)
q q d
I jX U (4.173)

Aspects 1), 2) and 3) describe the transient operation of a salient-pole machine, i.e. the
transient phenomenon after switching. The following cases need to be distinguished:

mode of operation physical synchr. gen. voltage direct axis reactance
steady state const I
F
const U
p

d
X
transient const
F
const U
p

'

'
d
X

Synchronous machine 99


Now the
p
U is not constant anymore regarding the phasor diagram for the transient operation,
but const U
p

'
. The voltage
p
U changes according to
d
I . The value of
'
p
U can be
determined using the initial conditions before switching.
I
I
d
U
I
q
U
q
U'
p
U
d
jX
q
I
q
jX'
d
I
d
jX
q
I
U
p
jX
d
I
(
1
-

)
X
d I
d

Fig. 83: phasor diagram for the transient operation

Torque in transient operation ensues to:

( )
q q d d
I U I U
p
M +

3
(4.174)

with replacements due to:

,
'
d
q p
d
X
U U
I


q
d
q
X
U
I (4.175 a,b)

sin U U
d
cos U U
q
(4.176 a,b)

Using insertion, the following equations are obtained:

1
1
]
1

cos
sin
sin
cos
3
'
'
U
X
U
U
X
U U
p
M
q d
p
(4.177)


1
1
]
1

,
_

2 sin
2
1 1
sin
3
2
' ,
'
U
X X X
U U
p
d q d
p
(4.178)
An equation for the torque in transient operation is achieved, which shows nearly the same
structure like the equation for the torque in steady-state operation but
p
U is replaced by
'
p
U
and
d
X is replaced by
'
d
X .

100 Synchronous machine


4.6.1 Power supply operation
There are two kinds of load cases, which cause a transient phenomenon, if the machine is
connected to the power supply:

1.) The electrical sudden load change, caused by a switching in the power supply
network. The supply voltage and the supply reactance are suddenly changed.
Because
'
p
U has to be constant, the currents
d
I and
q
I change stepwise, because
the supply voltage and the supply frequency are suddenly changed. Therefore the
generated torque is changed too. Instantaneously after switching, is equal
0
,
because the speed does not change immediately due to the moment of inertia. The
speed of the machine does not change fast, therefore there is a surplus/shortfall of
torque, which causes a motion process ( ) t of the rotor. The resulting currents and
the time characteristic of the torque in transient operation can be determined using
the precondition const U
p

'
.
2.) The mechanical sudden load change, caused by a stepwise change of the driving
torque. The supply voltage and the supply reactance do not change in this case.
Shortly after the sudden load change, the same currents flow as before. Therefore
the synchronous machine generates the same torque as before. The surplus/shortfall
of torque causes a motion process ( ) t of the rotor. Therefore the currents and the
torque of the synchronous machine change. Their time characteristic can be
determined using the precondition const U
p

'
.

After the declination of the transient phenomenon, a new steady-state operating state adjusts
itself gradually.

For easier description the following base values (so called per-unit quantities) are to be
used:

NStr
U
U
u (4.179)

NStr
I
I
i (4.180)

NStr
p
F
F
F
U
U
I
I
i
0
(4.181)

N NStr
I U
R
r (4.182)

N NSr
I U
X
x (4.183)

N SN
N N NStr
N
M
M
I U
p
M
M
M
m

cos
cos
3

(4.184)

Synchronous machine 101


characteristic of the excitation current:
o steady-state 1
F
i (4.185)
o transient ( )
d d p F
i x u i + 1
'
(4.186)
'
'
cos
d
p
d
x
u u
i

(4.187)

cos
1
'

u
u
i
p
F
(4.188)

i
F

N
1
stationary
transient

2

Fig. 84: current
F
i versus

characteristic of the torque
o steady-state
1
1
]
1

,
_

2 sin
1 1
2
1
sin
cos
1
d q d
p
N
x x x
u
m (4.189)

o transient
1
1
]
1

,
_

2 sin
1 1
2
1
sin
cos
1
' '
'
d q d
p
N
x x x
u
m (4.190)
1
m

2

kipp
>

kipp
<

2
stationary
transient

Fig. 85: torque m vs.

102 Synchronous machine


Because of the constant flux in the excitation winding, the machine compounds itself during
the transient phenomenon by increasing the excitation current and with it the torque on a
multiple of the rated values. Therefore the dynamic stability is increased to angular
displacements of more than 90 (electrical).

The transient phenomenon of a sudden load change in power supply operation and the self-
compounding because of the constant excitation flux can be illustrated with the phasor
diagram.

1. initial state for the steady-state operation: 1 u , 1 i ,
N
(4.191)

draw phasor diagram :
d d q p
i x u u + (4.192)

d d q p
i x u u +
' '
(4.193)

2. transient operation: const u
p

'
,
N
> , 1 u (4.194)

Draw direction of , take out (=read, measure)
d
u and
q
u , calculate:
q
d
q
x
u
i (4.195)

'
'
d
q p
d
x
u u
i

(4.196)

q d
i i i + (4.197)

Thus follows: ( )
d d dF p p
i x u u + 1
'
(4.198)

u
p
u'
p
u
q
i

Fig. 86: phasor diagram

Synchronous machine 103


4.6.2 Solitary operation
The transient theory can also be applied for isolated operation. Again applying const U
p

'


'
p
U is determined from the instant before switching. Because of the constant flux in the
excitation winding, the currents and voltages instantaneously after switching result from
const U
p

'
.

After a certain period time, if the transient phenomenon has fallen off, const U
p
is applied.
For this reason the currents and voltages for the most recent steady-state can be determined.
Between initial and final state there is a transient phenomenon with
0 d
T (
0 F
T ) if the stator
winding is open-circuited or short circuited, respectively
dK
T (
FK
T ) if the excitation winding
is open-circuited or short circuited.

Example: Synchronous machine connected to power supply; rated current and 0 cos
(inductive) assumed.
i = 1
u = 1
jx'
d
i
u'
p
u
p jx
d
i

Fig. 87: synchronous machine, phasor diagram

o after switching applies:

0 i and .
'
const u
p


o instantaneously after switching we obtain for the transient operation:

{
0
' '

+
d d p
i x u u (4.199)

o after a long period of time we obtain for the new steady-state operation:

{
0
+
d d p
i x u u (4.200)
Values:

o 1 u , 1 i , 90

o 1
q
u , 0
d
u

o 0
q
i , 1
d
i

o
F d p
i x u + 1

o
' '
1
d p
x u +

are obtained from the phasor diagram (Fig. 87) at
steady-state operation before switching.

104 Synchronous machine


If the stator winding is open-circuited, the time constant ensues to:
time characteristic = final state + (initial state final state)
. const time
t
e

(4.201)

So that the transient phenomenon can be described as:

( )
0
'
F
T
t
p p p
e u u u u

+ (4.202)

The discussed time characteristic is shown in the diagram of Fig. 88:

u=u
p
t
u=u'
p
u=1

Fig. 88: time characteristic


4.6.3 Summary and conclusion of transient operation
Three equations are basically used:

d d q p
I X U U +
' '
(4.203)

( )
d d q d d dF p p
I X U I X U U + + 1
'
(4.204)

q q d
I X U (4.205)


Synchronous machine 105

dK
d
T
T
0
stator

'

4.6.4 Scheme for transient operation
1.) Steady-state operation before switching,
phasor diagram, determination of
'
p
U

2.) Transient operation instantaneously after switching,
calculate new currents, voltages and torque using .
'
const U
p


3.) Steady-state operation after a long period of time
phasor diagram, determination of
p
U , calculate currents, voltages and torque after that

4.) Transient phenomenon in between: e-function
time constant:
case: switching appears in field.

FK
F
T
T
0
Feld

'

case: switching appears in stator

field open-circuited
field short-circuited
stator open-circuited
stator short-circuited


5 Servo-motor
Permanent-field synchronous machines with rotor position encoder are also known as servo-
motors.
5.1 General design and function
Design:
The stator of this machine type features a conventional three-phase winding. The permanent-
field rotor is equipped with rare-earth or ferrite magnets. The power inverter is controlled by
the rotor position encoder.

Fig. 89: permanent-field synchronous machine (servo motor)

Function:
The three-phase winding of the stator is supplied with a square-wave or sinusoidal three-
phase system depending on the rotor position. Thus a rotating m.m.f is caused, which rotates
with the exact rotor speed and which generates a time constant torque in co-action with the
magnetic field of the permanent-field rotor. The rotating field in the stator is commutated
depending on the rotor position in such a manner, that the rotating m.m.f in the stator and the
rotor field are perpendicular with a constant electrical angle of 90.
Thus results an operating method, which does not correspond to the operating method of
synchronous machines anymore, but to that of DC machines. Here the armature current
linkage and the excitation field are also perpendicular with a constant electrical angle of 90.
This constant angle is adjusted mechanically with the commutator of the DC machine.

The constant angle of a permanent-field synchronous machine with rotor position encoder is
adjusted electrically using a power inverter. This machine type can not fall out of
synchronism anymore and behaves like a DC machine. That is why the machine is also called
electronically commutated DC motor (EC-motor).

108 Servo-motor


x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
u
v
w
q
d
x
y
z
h
M

Fig. 90: servo-motor, direction convention


5.2 Dynamic set of equations
Focus is put on salient-pole machines with magnetic asymmetry
d q
L L in the following.
The arrangement of permanent magnets can be reduced to 1 <
i
if the machine is supplied
with sinusoidal currents, to save permanent magnet material and to improve the voltage
waveform. The permanent magnets are magnetizing in the direct axis. The magnetic air-gap in
the q-axis is assumed to be equal to the mechanical air-gap. The magnetic air-gap in the d-
axis is
M
h + and therefore it is
d q
L L > . This effect can be utilized to reduce the current, as
to be seen later on. Figure 90 shows a two-pole model.
In case of no damper windings, permanent-field synchronous machines only have three
windings to be considered. This applies, if the rotor is made of laminated sheets and if the
conductive rare-earth magnets (SmCo, NdFeB) are subdivided into small laminations, to
avoid eddy currents.
The following denominations are used:
o d: stator direct axis
o q: stator quadrature axis
o F: rotor direct axis, permanent magnets

Due to
d q
L L a suitable coordinate system is to be chosen, requiring to rotate with the rotor,

dt
d
dt
d
. (5.206)

EC-motors are used in machine
tools and robotic devices
because of their excellent
dynamic behavior simple
controllability. The brushless
technique has a low rate of wear
and is maintenance-free.

Servo-motor 109


Practically the previously discussed coordinate system for salient-pole machines without
damper windings is used. The load reference arrow system for motor operation is applied on
the considered case.
There are only voltage equations and the flux-linkage of the stator winding, but no voltage
equation for the permanent magnets. The constant rotor flux
M
in the d-axis, which is
caused by the permanent magnets, is considered by a constant equivalent excitation current
'
0 F
i in the equation for the stator flux-linkage in the d-axis. The stator winding has no effect
on the permanent magnets.


q
d
d d
dt
d
i R u

+
1
(5.1)


d
q
q q
dt
d
i R u +

+
1
(5.2)

( )
W d q q d el
M
dt
d
p
J
i i p M +

(5.3)


3 2 1
M
F hd d d d
i L i L

+
'
0
(5.4)

q q q
i L (5.5)

The flux-linkages can be directly pasted into the stator voltage equation and into the torque
equation.


q q
d
d d d
i L
dt
di
L i R u +
1
(5.6)


'
0 1 F hd d d
q
q q q
i L i L
dt
di
L i R u + + + (5.7)

( )
d q q q d d q F hd el
i i L i i L i i L p M +
'
0
(5.8)

( ) [ ]
W q d q d F hd
M
dt
d
p
J
i L L i i L p +

'
0
(5.9)


110 Servo-motor


5.3 Steady state operation
In steady-state operation the flux-linkages in the rotating system and the speed are constant,
i.e.
o 0

dt
d
(5.10)

o 0
dt
d
(5.11)

For this reason the dynamic system of equation can be simplified and the torque equation is
decoupled.


q q d d
i L i R u
1
(5.12)


'
0 1 F hd d d q q
i L i L i R u + + (5.13)

( ) [ ]
q d q d F hd el
i L L i i L p M
'
0
(5.14)

For the inverse transformation of the rotating system with

dt
d
dt
d
(5.15)

the (still) arbitrary integration constant is practically chosen:


2
0

(5.16)

After the inverse transformation into the complex notation, the rotor axis, which is the
magnetizing axis of the permanent magnets, is pointing in direction of the negative imaginary
axis and the q-axis, where the torque is generated, is orientated to the real axis ( compare
with induction machine).


q d
j q j
d
U U e
u
j e
u
U + +

0 0
3 3

(5.17)

( )
q q d
q q d
d
I X j I R j
i L i R
U +

1
1
3

(5.18)

( )
'
0 1
'
0 1
3
F hd d d q
F hd d d q
q
I X j I X j I R j
i L i L i R
j U + +
+ +


(5.19)

whereas currents are defined as:

3
q
q
i
I ,
3
d
d
i
j I ,
3
'
0
'
0
F
F
i
j I (5.20 a-c)

Servo-motor 111


With:

'
0 F hd p
I X j U (5.21)

The complex voltage equation is:

( )
p d d q q q d
U I X j I X j I I R U + + + +
1
(5.22)

Torque is obtained from the root-mean-square values:

( )
3 3 3
3
'
0
q
d q
d F
hd el
i
L L
i i
L
p
M
1
]
1

(5.23)

( ) [ ]
q d q d p el
I X X I U
p
M

3
(5.24)

Based on the knowledge of equations 5.17 - 5.22 the phasor diagram can be drawn.
p
U and
q
I are in phase. Two modes need to be distinguished:

normal operation 0
d
I field-weakening 0 <
d
I
+ 0 < + 0
U
p
U
U
p
U

q
=
j X
q q

R
q

F0
j X
d d

R
d
j X
q q

R
q

F0

q
d
d
q

Fig. 91: normal operation Fig. 92: field weakening

Compared to DC machines, field weakening of permanent-field synchronous machines can
only be performed within certain limits, if a negative direct current component is injected
additionally to the torque generating current in the q-axis, 0 < . Thus the angle of stator
current linkage and rotor field is increased to an electrical angle of more than 90. The speed
increases too. The injection of a negative direct current can also be used to minimize the
current and to improve the power factor. The precondition is a rotor with a magnetic
asymmetry
d q
X X > . Then the negative direct current component and the reluctance in
conjunction with the quadrature-axis component of the current generate an additional torque
contribution.

112 Servo-motor


Thus the total current consumption is decreased. At the same time the phase displacement
between terminal voltage and stator current is reduced, so a power factor cos of nearly 1 is
obtained. This method allows an operation with lower current consumption and therefore with
lower copper losses. Thus the utilization of the machine is increased, power of the converter
output power is reduced. With increasing difference ( )
d q
X X this so called pre-control
0 < becomes more and more effective.

The characteristics of the permanent-field synchronous machine with rotor position encoder
in steady-state operation can be determined using the direct and quadrature components of the
voltage equations and the torque equation.


q q d d
I X I R U
1
(5.25)


p d d q q
U I X I R U + +
1
(5.26)

( ) ( )
q d q d p
I X X I U
p
M

3
(5.27)

The frequency-dependency of the reactances and of the synchronous generated voltage is
considered by basing on rated frequency.

0
0
d d
X X

(5.28)

0
0
q q
X X

(5.29)

0
0
p p
U U

(5.30)


q q d d
I X I R U
0
0
1

(5.31)


0
0
0
0
1 p d d q q
U I X I R U + +

(5.32)

( )
q d q d p
I X X I U
p
M
1
]
1

,
_

0 0
0
0
0
3

(5.33)

Finally results:
o speed (shunt characteristic)
d d p
q q
I X U
I R U
n
n
+


0 0
1
0 0

(5.34)

o torque (shunt characteristic)
( ) [ ]
q d q d p
I X X I U
p
M

0 0 0
0
3

(5.35)

o control instruction (electrical commutator):
q q d d
I X I R U
0
0
1

(5.36)

Servo-motor 113


M
k
M n
n
0
=

qmax
U
R
1

N Str
q
d
=
0

d
<
0
U
N
S
tr
U
q =
,

d
<
0
U
N Str
U
q <
,
d =
0
U
N
S
tr
U
q =
,

d =
0

Fig. 93: control instruction, schematic overview

The operational performance of permanent-field synchronous machines with rotor position
encoder corresponds to the behavior of separately excited DC machines.

o
A q
U U

o
A q
I I

o
F d
I I

o
0 0
n k U U
NStr p


By set-point selection of the voltage
NStr q
U U the speed can be adjusted non-dissipative in
the speed range
0
n n . By injecting a negative direct current
d
I with a voltage component
d
U the speed can be further increased in spite of the permanent field. Torque increases
simultaneously.

114 Servo-motor


5.4 Dynamic behavior
With time constants

o
1
R
L
T
d
d
(5.37)

o
1
R
L
T
q
q
(5.38)

a system of differential equations is obtained, to be able to completely describe permanent-
field synchronous machines.


q q
d
d
d
d
i T
R
u
i
dt
di
T + +
1
(5.39)


'
0
1 1
F
hd
d d
q
q
q
q
i
R
L
i T
R
u
i
dt
di
T

+

(5.40)

( ) [ ]
W q d q d hd F
M i L L i L i p
dt
d
p
J

'
0

(5.41)

The transformation of the stator voltages to a rotating system with

dt
d
is done as usual:

[ ] [ ]
1
]
1

1
]
1

v
u
q
d
u
u
T T
u
u
32
(5.42)

Based on this, the structure diagram of a complete servo drive can be drawn. Besides the
machine model a PWM-converter and the speed control are included.

Thus the machine can be simulated and computational calculated with (spice-oriented-)
simulation tools.

excitation values:
d
u ,
q
u ,
W
m
state values:
d
i ,
q
i ,
initial conditions: ( ) ( ) ( ) 0 0 0 0
q d
i i ,
2
0



Figure 94 shows the according structure diagram of a permanent-field synchronous machine
with rotor position encoder (top left: converter; top right: control; bottom: machine).

Servo-motor 115

[ T ]
2
3
2
3
q
i

PWM

[ T ]

X

-1
X
X

X
+
-
+ -
+

+ -
-
- +
+
+ -
q
i
m
W
L
q
q
i
qS
i
+
-
d
i R
1
dS
u
qS
u
AS
u
BS
u
VS
u
WS
u
US
u
V
u
W
u
U
u
A
u
B
u
d
u
q
u q
i
d
i
d
i
q
i
d
i
p
T
d
T
q
T
q
L
q
L -
d

J
p
T
d
q
i

T
q
1
R
1
1
R
1
+
-

8
i
F0
' L
hd
R
1
T
d
i
d
-
i
F0
' L
hd
I
F0
' L
hd
R
1

Fig. 94: servo-motor drive, structure diagram

The machine is controlled in such a manner, that armature current-linkage and excitation field
are perpendicular in normal operation. This means that, if the flux of the permanent magnets
is represented by
'
0 F hd
i L , the stator current has no direct component 0
d
i and the
quadrature component
q
i is the torque-generating component.

In order to keep 0
d
i , follows for
d
u :

0
1
+
q q
d
i T
R
u
(5.43)

q q d
i L u (5.44)

If a control method with 0
d
i is applied, for example field-weakening 0 <
d
i , the voltage
control in the d-axis has to be:

+
d q q
d
i i T
R
u

1
(5.45)

+
d q q d
i R i L u
1
(5.46)

Thus the time characteristic of the direct current ensues to:

,
_

d
T
t
d d
e i i 1 (5.47)

116 Servo-motor


Voltage in the quadrature axis results from the speed which is supposed to be adjusted. For
example at no-load operation and rated frequency:

0
'
0
1 1

F
hd
q
i
R
L
R
u
(5.48)

NStr F hd q
U i L u 3
'
0
(5.49)

Fig. 95: measured mechanical speed



Fig. 96: measured phase-current (setpoint and actual value)

Servo-motor 117



Fig. 97: calculated mechanical speed


Fig. 98: calculated phase current

Figures show reversing process of a servomotor: comparison of measurement and calculation.

118 Servo-motor


5.5 Voltage- and current waveforms of servo-motors with
rotor position encoder
Permanent-field (PM = permanent-magnet, permanent-field) synchronous machines can be
operated in different ways. In block-operation, the motor is supplied with rectangular (block)
currents and the distribution of the air-gap flux density is rectangular. If the motor is supplied
with sinusoidal currents and the rectangular distribution of the air-gap flux density is retained,
then we have mixed operation. In sinusoidal operation, the current and the distribution of the
air-gap flux are sinusoidal. The figure shows the characteristics of flux density, current and
voltage.
If a machine is operated in block-operation, then it is also called brushless or electrically
commutated DC machine. If a machine is operated in sinusoidal operation, it is also called
self-controlled synchronous machine. The operational performance of permanent-field
synchronous machines with rotor position encoder generally corresponds to the operational
performance of DC machines.
If the machine is supplied with sinusoidal currents, sinusoidal induced voltages are necessary.
We can obtain a nearly sinusoidal air-gap field using parallel magnetized instead of radial
magnetized permanent magnets and by designing a suitable stator winding (chording for
example).
Another possibility is to supply the machine with rectangular (block) currents. The total
supply current has a constant magnitude and is distributed cyclic to the three stator phases,
which results in current blocks with an electrical length of 120 and dead times of 60. If the
induced voltage during the length of a current block is constant, then power of the phase is
constant too. During the dead times the induced voltage has no influence on the torque
generation. The trapezoidal characteristic of the induced voltage results from 1 > q and
because of the skewing of the stator slots of one slot pitch.
The advantages of the rectangular supply in comparison to the sinusoidal supply are a 15%
higher utilization of the machine and the usage of simple position sensors (three photoelectric
barriers) instead of expensive resolvers and an easier signal processing.
The disadvantages of the rectangular supply in comparison to the sinusoidal supply are:
with increasing speed eddy-current losses arise in the conductive rare-earth permanent
magnets (in comparison with non-conductive ferrite magnets) caused by the slot
harmonics and the jumping rotating m.m.f..
because of the machine- inductances and the voltage limitation of the converter, there are
heavy deviances from the rectangular current form at high speed. The results are a reduced
torque and higher losses.
because of the non-ideal commutation of the phase currents at rectangular supply, angle-
dependent huntings occur at lower speed, which has to be compensated by the control.

In contrast the mixed operation has advantages. If the machine is supplied sinusoidal, if it has
a rectangular flux-density distribution in the air-gap and if the stator winding is chorded, to
achieve a sinusoidal induced voltage, then the best motor utilization is obtained. In this case
the fundamental wave of the flux-density in the air-gap is increased and at the same time the
losses are reduced. A 26% higher machine utilization can be achieved, compared to sinusoidal
supply, respectively 10% higher compared to rectangular supply. The operation with
sinusoidal currents requires an exact information of the rotor position, which requires an
expensive encoder-system.

Servo-motor 119



Fig. 99: table, comparing waveforms and resulting power values


6 Appendix
6.1 Formular symbols
A current coverage; area (in general)
a number of parallel conductors
B flux density (colloq. induction)
b width
C capacity
c general constant, specific heat
D diameter, dielectric flux density
d diameter; thickness
E electric field strength
e Eulers number
F force; form factor
f frequency
G electric conductance, weight
g fundamental factor, acceleration of gravity
H magnetic field strength
h height; depth
I current; I
w
active current; I
B
reactive current
i instantaneous current value
J mass moment of inertia
j unit of imaginary numbers
K cooling medium flow, general constant
k number of commutator bars; general constant
L self-inductance; mutual inductance
l length
M mutual inductance; torque
m number of phases, mass
N general number of slots
n rotational speed

122 Appendix


O surface, cooling surface
P active power
p number of pole pairs; pressure
Q reactive power; cross section; electric charge
q number of slots per pole and phase; cross section
R efficiency
r radius
S apparent power
s slip; coil width; distance
T time constant; length of period; absolute temperature; starting time
t moment (temporal); general time variable
U voltage (steady value); circumference
u voltage (instantaneous value); coil sides per slot and layer
V losses (general); volume; magnetic potential
v speed; specific losses
W energy
w number of windings; flow velocity
X reactance
x variable
Y peak value (crest value)
y variable; winding step
Z impedance
z general number of conductors

pole pitch factor; heat transfer coefficient
brushes coverage factor
constant of equivalent synchronous generated m.m.f.
air gap; layer thickness
dielectric constant
Pichelmayer-factor
efficiency; dynamic viscosity
electric current linkage
load angle; temperature; over temperature
electric conductivity

Appendix 123


power factor, thermal conductivity; wave length; ordinal number; reduced magnetic
conductivity
magnetic conductivity
permeability; ordinal number
ordinal number; kinematic viscosity
winding factor
specific resistance
leakage factor; tensile stress
general partition; tangential force
magnetic flux
phase displacement between voltage and current
flux linkage
angular frequency

124 Appendix


6.2 Units
The following table contains most important physical variables and their symbols and units to
be used. An overview of possible unit conversions is given in the right column additionally.

physical variable Symbol SI-unit abbrev. unit conversion
length L Meter m
mass M Kilogramm kg 1 t (ton) = 10
3
kg
time T Second s 1 min = 60 s
1 h (hour) = 3600 s
current intensity I Ampere A
thermodynamic
temperature
T Kelvin K temperature difference in
Kelvin
celsiustemperature

Degree
Centigrade
C = T T
0
light intensity I Candela cd
area A - m
2

volume V - m
3
1 l (Liter) = 10
-3
m
-3
force F Newton N 1 kp (Kilopond) = 9.81 N
1 N = 1 kgm/s
2
pressure P Pascal Pa 1 Pa = 1 N / m
2

1 at (techn. atm.) = 1 kp / cm
2

= 0.981 bar, 1 bar = 10
5
Pa
1 kp / m
2
= 1 mm WS
torque M - Nm 1 kpm = 9,81 Nm =
9,81 kgm
2
/ s
2


Appendix 125


physical variable Symbol SI-unit abbrev. unit conversion
mass moment of
inertia
J - Kgm
2
1 kgm
2
= 0.102 kpms
2
= 1 Ws
3

impetus moment GD
2

GD
2
= 4 J / kgm
2

frequency F Hertz Hz 1 Hz = 1 s
-1

angular frequency

- Hz = 2f
rotational speed N s
-1
1 s
-1
= 60 min
-1

speed (transl.) V - m / s 1 m / s = 3,6 km / h
power P Watt W 1 PS = 75 kpm / s = 736 W
energy W Joule J 1 J = 1 Nm = 1 Ws
1 kcal = 427 kpm = 4186,8 Ws
1 Ws = 0,102 kpm
el. voltage U Volt V
el. field strength E - V / m
el. resistance R Ohm


el. conductance G Siemens S
el. charge Q Coulomb C 1 C = 1 As
capacity C Farad F 1 F = 1 As / V
elektr. constant
0
- F / m =
0

r
= relative diel.-constant
inductance L Henry H 1 H = 1 Vs / A = 1 s
magn. flux

Weber Wb 1 Wb = 1 Vs
1 M (Maxwell) =
10
-8
Vs = 1 Gcm

126 Appendix


physical variable Symbol SI-unit abbrev. unit conversion
magn. flux density B Tesla T 1 T = 1 Vs / m
2
= 1 Wb / m
2

1 T = 10
4
G (Gau)
1 G = 10
-8
Vs / m
2

magn. field
strength
H - A / m 1 Oe (Oersted) =
10 / 4 A / cm
1 A / m = 10
-2
A / cm
magn.-motive force

- A
magn. potential V - A
magn. constant
0
- -
0
= 410
-7
H / m

0
= 1 G / Oe
permeability

- - =
0

r
= relative permeability
angle

Radiant rad 1 rad = 1 m / 1 m
= l
curve
/ r


7 Literature reference list
Books and scripts listed as follows may exceed the teaching range significantly. Nevertheless
they are recommended best for a detailed and deeper understanding of the content of this
lecture.

B. Adkins
The general Theory of electrical Machines, Chapman and Hall, London

Ch.V. Jones
The unified Theory of electrical Machines, Butterworth, London

L.E. Unnewehr, S.A. Nasar
Electromechanics and electrical Machines, John Wiley & Sons

Electro-Craft Corporation
DC Motors, Speed Controls, Servo Systems, Pergamon Press

Ch. Concordia
Synchronous Machines, John Wiley, New York

Bahram Amin
Induction Motors Analysis and Torque Control

Peter Vas
Vector Control of AC Machines, Oxford Science Publications

T.J.E. Miller
Switched Reluctance Motors and their control, Magna Pysics Publishing