EEE373 Electric Motor Drive

Lecture 10

Asst. Prof. Dr. Mongkol Konghirun

EE, KMUTT

Overview of Three-Phase
Synchronous Machines
{ The synchronous machine is one in which
alternating current flows in the armature
winding at stator and DC excitation is supplied
by an exciter to the field winding through slip
rings.

{ The armature (stator) winding is almost

invariably one of three-phase induction
machine, producing the sinusoidal back emfs.

{ There are two types depending on the rotor

structure: cylindrical-rotor and salient-pole.

1
Overview of Three-Phase
Synchronous Machines
{ The cylindrical-rotor type has a small number
of poles such as two or four and is usually
designed to operate at high speeds. This type
would provide the uniform air-gap.

Overview of Three-Phase
Synchronous Machines
{ The salient-rotor type has a high number of
poles and is usually designed to operate at
low speeds. This type would provide the non-
uniform air-gap.

2
 Overview of Three-Phase
Synchronous Machines
{ Stator windings are identical, sinusoidally
distributed, displaced 120o, with Ns equivalent
turns and resistance rs.

{ However, at rotor the field winding with Nfd

turns is supplied by DC voltage from an exciter
to the field winding through slip rings.

Induction VS Permanent
Magnet Synchronous Motors
phase-a phase-a

ia ia
stator stator

rotor
N S
ic ic
ωr ωr
S N

phase-c phase-c
phase-b phase-b

ib ib
permanent-magnet (PM)
rotor

PM synchronous motor
Induction motor
λm = constant permanent-magnet flux linkage

3
 Induction VS Permanent
Magnet Synchronous Motors
PMSM Induction motor

• High efficiency • Inexpensive

• High power factor
• Compact and light weight
• Easy to cooling
• No slip
• Low inertia due to magnet
(allowing fast acceleration)
• High speed operation (>
10,000 rpm)
• Popular (easy to find)
• Simple speed control
methods available (V, Rr,
or V/f)
• No rotor position sensor
used (sensor is
expensive)

Voltage and Flux Linkage Equations

in abc Reference Frame

{ The voltage equations for the PMSM are

dλ
v as = rs i as + as
dt
dλ bs
v bs = rs i bs +
dt
dλ cs
v cs = rs i cs +
dt
where rs is the resistance of the stator
winding (Ohm).

4
Voltage and Flux Linkage Equations
in abc Reference Frame

{ For a magnetically linear system, stator flux

linkages may be expressed as

λ as = Lasasi as + Lasbsi bs + Lascsi cs + λ am

λ bs = L bsas ias + L bsbs i bs + L bscs ics + λ bm

λ cs = Lcsas i as + Lcsbs i bs + Lcscsics + λ cm

Voltage and Flux Linkage Equations

in abc Reference Frame

where the machine inductances can be

expressed as
Self inductances: L asas = Lls + L A
L bsbs = Lls + L A

L cscs = L ls + L A

Mutual inductances: 1
L asbs = L bsas = − L A
2
1
Lascs = Lcsas = − L A
2
1
L bscs = L csbs = − L A
2

5
 Using Park Transformation to
Convert abc to dq Stator Variables
Permanent-magnet synchronous motor
Phase-a
winding i
phase-a ia, ib, and ic (AC)
a
stator
Park Trans.
⎡i ⎤
rotor ⎡i d ⎤ 2 ⎡ cos θλr cos(θλr − 2π / 3) cos(θλr + 2π / 3) ⎤ ⎢ a ⎥
⎢i ⎥ = ⎢ ⎥ i
N S
⎣ q ⎦ 3 ⎣ − sin θλr − sin(θλr − 2π / 3) − sin(θλr + 2π / 3) ⎦ b ⎥
⎢
ic ⎣⎢ i c ⎦⎥
ωr
S N
phase-c
phase-b
ib Phase-c ‰ id (DC) is controlled zero.
Phase-b
permanent-magnet (PM) winding
‰ iq (DC) controls torque.
winding

Voltage and Flux Linkage Equations

in dq0 Reference Frame

{ When rotor is PM and surface mounted,

producing uniform air gap, then the stator flux
linkages are simplified as

λrds = L s i ds
r
+ λm
λrqs = L s i qs
r

λr0s = L ls i 0r s
where
Ls = stator self inductance
L s = L mq + L ls = L md + L ls

6
Voltage and Flux Linkage Equations
in dq0 Reference Frame

{ Also, the voltage and torque equations

are expressed as r
dλ ds
r
v ds = rsi ds
r
− ωr λ qs
r
+
dt
dλ qs
r
r
v qs = rsi qs
r
+ ωr λ ds
r
+
dt
dλr
r
v 0s = rs i 0s
r
+ 0s

dt
⎛ 3 ⎞⎛ p ⎞ r r
Te = ⎜ ⎟ ⎜ ⎟ ( λ ds i qs − λ qs
r r
i ds )
⎝ 2 ⎠⎝ 2 ⎠
where p is number of poles

Voltage and Flux Linkage Equations

in dq0 Reference Frame

{ If three-phase stator winding are Y-

connected, then
i as + i bs + i cs = 0

therefore,
r
v 0s r
, i 0s , λ 0s
r
=0

because
2⎛1⎞
r
i 0s = ⎜ ⎟ ( i as + i bs + i cs ) = 0
3⎝2⎠

7
Vector Control of Permanent-Magnet
Synchronous Motor (PMSM)

{ We are trying to control the PMSM just

like separately excited dc motor.
{ To accomplish that, the d-axis current is
forced to be zero ( i dsr = 0 ), the stator
flux and torque equations become
λ ds
r
= Ls i ds
r
+ λm = λm
λrqs = L s i qs
r

Te =
22
( λ dsiqs − λ qsr idsr )
3p r r

=
22
( λ ds i qs ) =
3p r r 3p
22
( λ m i qs
r
)

Vector Control of Permanent-Magnet

Synchronous Motor (PMSM)

{ Finally, the torque equation of PMSM

looks similar to one of the separately
excited dc motor.

{ Therefore the vector control of PMSM is

accomplished where total flux
appearing in torque equation is the
constant magnet flux linkage and
independently controls the torque.

8
 Vector Control of Permanent-Magnet
Synchronous Motor (PMSM)
Permanent-magnet synchronous motor
Phase-a
winding i
phase-a ia, ib, and ic (AC)
a
stator
Park Trans.
⎡i ⎤
rotor ⎡i d ⎤ 2 ⎡ cos θλr cos(θλr − 2π / 3) cos(θλr + 2π / 3) ⎤ ⎢ a ⎥
⎢i ⎥ = ⎢ ⎥ i
N S
⎣ q ⎦ 3 ⎣ − sin θλr − sin(θλr − 2π / 3) − sin(θλr + 2π / 3) ⎦ b ⎥
⎢
ic ⎣⎢ i c ⎦⎥
ωr
S N
phase-c
phase-b
ib Phase-c ‰ id (DC) is controlled zero.
Phase-b
permanent-magnet (PM) winding
‰ iq (DC) controls torque.
winding

Vector Control of Permanent-Magnet

Synchronous Motor (PMSM)

{ In summary, to establish the vector

control of PMSM, not only the rotor flux
angle (i.e., PM angular position) must
be known, but also the d-axis current
must be controlled to zero.

{ In fact, the vector control concept of

PMSM is to feed the three-phase stator
currents in phase with the
corresponding three-phase back emfs.

9
Vector Control of Permanent-Magnet
Synchronous Motor (PMSM)
The sensored vector control of PMSM using a
position sensor can be depicted as follows:
DC sup ply
Digital Signal Pr ocessor voltage
i qsr
r* − r*
v qs v∗as PWM1
i qs + i cs
ω*r,e PI PI v∗bs
+ ωr,e PWM VSI
(K )
r −1 i bs
− r* ∗
PMSM
r*
i ds =0 v ds
s
v gen. 3 − ph i as
cs
PI
+ − PWM6 position
r
i ds θ r ,e sensor
θ r ,m
θ r ,e
r
i ds i bs

r
i qs (K ) r
s
i as

ωr,e
Speed θ r ,e Pole
calc. pairs

Position Sensors

(1) Quadrature Encoder Pulse

(QEP) - Optical Encoder

(2) Resolver

10
Quadrature Encoder Pulse (QEP)
- Optical Encoder

{ It is the incremental encoder.

{ Following figure shows the structural drawing

of an incremental encoder’s relative angular
position detector of outputs A and B type,
having signal Z of the phase zero (or index
signal), which is produced only one pulse per
mechanical revolution.

Quadrature Encoder Pulse (QEP)

- Optical Encoder

{ The light projected from the LED goes

through the slits of the rotary disc and
each of the slits A, B, and Z of the fixed
board with slits.

11
Quadrature Encoder Pulse (QEP)
- Optical Encoder

{ Then, the light is detected by the light

receiving devices A, B, and Z. Slits A and
B on the fixed board with slits have a
phase difference of 90 degrees. The
electric outputs are also the rectangular
waveforms with the same 90-degree
phase difference

{ The rotational direction can be detected

by looking at which the output signal lead
another signal by 90 degree.

Quadrature Encoder Pulse (QEP)

- Optical Encoder

{ Signal outputs for both directions:

12
Resolver
{ A resolver is a detector of the absolute
position or the absolute angle of the
rotation.
{ The encoder converts the amount of
displacement into the digital format, while
the resolver converts it into the analog
format.
{ The encoder converts the amount of
displacement into the digital format, while
the resolver converts it into the analog
format.

Resolver
{ Figure shows the connection diagram
and the associated signals of the
resolver.

13
Resolver
{ The sinusoidal excitation signal with a
high frequency ranging between 2 kHz
and 10 kHz is fed into the rotor winding
(or primary winding, U0).

{ Other two stator windings (or secondary

windings, U1 and U2) are mechanically
placed orthogonal of each other on the
stator side.

Resolver
{ These windings produce two induced
voltages that contain the rotor position
information. Specifically, these resolver
output signals are actually the sinusoidal
excitation signals modulated with the
amplitude of sine and cosine of shaft
angle. The equations of these resolver
output signals can be written as follows:

u 1 (ε, t ) = U 0 .k. sin(ε). sin(ω ref t )

u 2 (ε, t ) = U 0 .k. cos(ε). sin(ω ref t )

14
Resolver
u 1 (ε, t ) = U 0 .k. sin(ε). sin(ω ref t )

u 2 (ε, t ) = U 0 .k. cos(ε). sin(ω ref t )

where - k is the transformation ratio of the

resolver.
- U0 and ωref is the peak value and the
frequency of the excitation
signal, respectively.
- ε is the rotor position

Resolver
{ As a result, the resolver can sense the
rotor position not only when the motor is
rotating, but also when the motor is
standstill.

{ The process to calculate the absolute

position and speed in the digital format
from the analog resolver signals is known
as Resolver-to-Digital (R/D) conversion.

15
Resolver
{ The simplified block diagram of
calculating the rotor position from the
resolver signals is illustrated as follow:
sin ε
u 1 (ε, t )
⎛ sin ε ⎞ ε
arctan ⎜ ⎟
De mod ulation cos ε ⎝ cos ε ⎠
u 2 (ε, t )

{ Once the rotor position is available from

the position sensor, then the speed could
be simply calculated as follow:
dε
ω=
dt

Resolver
{ Normally, the low pass filter (LPF) is
added to reduced the noise effect on the
measured angle as follow:
(LPF)

ε d
dt
1
τ s +1
ω
1
where τ = filter time constant =
2πf c
fc = cut-off frequency (Hz)

16