Steady State Analysis of a Brushless DC Motor with Half ...
Transcript of Steady State Analysis of a Brushless DC Motor with Half ...
Purdue UniversityPurdue e-PubsDepartment of Electrical and ComputerEngineering Technical Reports
Department of Electrical and ComputerEngineering
5-1-1987
Steady State Analysis of a Brushless DC Motor withHalf Bridge InverterTodd J. KazmirskiPurdue University
Follow this and additional works at: https://docs.lib.purdue.edu/ecetr
This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] foradditional information.
Kazmirski, Todd J., "Steady State Analysis of a Brushless DC Motor with Half Bridge Inverter" (1987). Department of Electrical andComputer Engineering Technical Reports. Paper 564.https://docs.lib.purdue.edu/ecetr/564
Todd J. Kazmirski
TR-EE 87-17 May 1987
School of Electrical EngineeringPurdue UniversityWest Lafayette, Indiana 47907
Steady State Analysis of a Brushless DC Motor with Half Bridge Inverter
STEADY STATE ANALYSIS OF A BRUSFILESS DC MOTOR
WITH HALF BRIDGE INVERTER
A Thesis
Submitted to the Faculty
of
Purdue University
by
Todd J. Kazmirski
In Partial Fulfillment of the
Requirements for the Degree
of
Master of Science in Electrical Engineering
Mav 1987
This is dedicated to my family
\
Ill
ACKNOWLEDGMENTS
I would like to thank Professor Paul Krause for his support and gui
dance throughout this work and my education and, especially, for his sincere
interest in me as a person.
IV
TABLE OF CONTENTS
Page
LIST OF TABLES............................................................. ...................................vi
LIST OF FIGURES............ ..... ................................. .................... ................. ... vii
NOMENCLATURE............ ix
ABSTRACT.......................................................................................... .............xi
CHAPTER 1 - INTRODUCTION........................................................... .............!
CHAPTER 2 - DESCRIPTION OF THE BRUSHLESS DC MOTOR SYSTEM............ ...... ..................................... ....... ........................ 3
2.1 INTRODUCTION........................ 32.2 THE BRUSHLESS DC MOTOR SYSTEM............ ...........2.3 DESCRIPTION OF THE HALF BRIDGE INVERTER................... 42.4 DESCRIPTION OF THE SNUBBER NETWORK.......... ................. ...62.5 DESCRIPTION OF THE BRUSHLESS DC MOTOR...... .............. 72.6 CONCLUSION............................................................................... 8
CHAPTER 3 - DEVELOPMENT OF THE STEADY STATE .MODEL FOR THE SYSTEM........................................... ........ .......................u
3.1 INTRODUCTION......................................................................... .u3.2 OPERATING MODES AND EQUIVALENT CIRCUITS.................. ..123.3 STATE EQUATIONS FOR THE A CONDUCTION PERIOD....... 133.4 OPERATING MODES, EQUIVALENT CIRCUITS, AND
STATE EQUATIONS FOR THE B CONDUCTION PERIOD........143.5 OPERATING MODES, EQUIVALENT CIRCUITS, AND
STATE EQUATIONS FOR THE C CONDUCTION PERIOD 153.6 CONCLUSION........................................ ,«
V
Page
CHAPTER 4 - STEADY STATE DIGITAL COMPUTER SIMULATION ...22
4.1 INTRODUCTION................................... .............. ................. -4.2 DESCRIPTION OF THE COMPUTER PROGRAM..........4.3 MACHINE PARAMETERS................ ....... ............................4.4 DIGITAL SIMULATION RESULTS..............4.5 ANALOG SIMULATION RESULTS.......... ............... ...... .....4.6 TORQUE PULSATIONS AND ZERO TORQUE MINIMA4.7 ELIMINATING THE ZERO TORQUE MINIMA.... ....... ...4.8 CONCLUSION................... .................................. ...... ••••........ .
CHAPTER 5 - STEADY STATE SIMULATION WITH 120 ° < Cp < 180 °'............................................................ .....
5.1 INTRODUCTION ........... ...........................................................................365.2 OPERATING MODES AND STATE EQUATIONS .........................375.3 STEADY STATE DIGITAL COMPUTER SIMULATION................385.4 TORQUE AND EFFICIENCY OPTIMIZATION........ ...... ...............395.5 CONCLUSION.......................... ..............................................................40
CHAPTER 6 - CONCLUDING COMMENTS..............................................60
...22
.24'
.24
.25
.25
28
LIST OF REFERENCES.................................................—•••••••••<>2
APPENDIX............................................................................. ..........................64
VI
LIST OF TABLES
Table Page
3.1 Conduction period A operating modes................... .............. ........ .............19
3.2 Conduction period B operating modes............................... .20
3.3 Conduction period C operating modes.................. ..................... ............21
4.1 Machine parameters................... 35
5.1 Transistor gating sequence for 120 ° < Cp < 180 ° ..................................41
5.2 Conduction period A1 operating modes, 120° < Cp < 180°................... 54
5.3 Conduction period A2 operating modes, 120 < Cp < 180 °................. .55
5.4 Conduction period Bl operating modes, 120° ' < Cp < 180 ° ......... ..56
5.5 Conduction period B2 operating modes, 120 < Cp < 180 ° 57
5.6 Conduction period Cl operating modes, 120 ° < Cp < 180 °.......... 58
5.7 Conduction period C2 operating modes, 120° < Cp < 180 °.......... .......59
Vll
LIST OF FIGURES
Figure Page
2.1 Brushless dc motor with half bridge inverter......... ............................. ......9
2.2 Transistor gating sequence for the half bridge inverter.............................. . 10
3.1 Brushless dc motor system equivalent circuit............................—.........17
3.2 Operating mode 4 equivalent circuit................ 18
4.1 Steady state digital simulation, ias, Vc, 120 ° conduction, ■ > = 0 .......29
4.2 Steady state digital simulation, te, 120 ° conduction,_o — 0 ° ..... .30
4,3; Steady state digital simulation, ias, Vc, 120 ° conduction, o = 30 —31
4.4 Steady state digital simulation, tR, 120 ° conduction,.<;> — 30 32
4.5 Steady state analog simulation, 120 ° conduction, <;> = 0 33
4.6 Transistor gating sequence for the half bridge inverter....... ..................34
5.1 Transistor gating sequence for 120 J < Cp < 180 -41
5.2 Steady state performance, ias, Vc, Cp = 150 3 , ■, = () 3 ........... -42
5.3 Steady state performance, te, Cp = 150° , <i> = 0 3 .......... — .....,...,.,..43
5.4 Steady state performance, ias, V(., Cp = 150° , ■ > = 30 ,.,..,..44
5.5 Steady state performance, te, Cp = 150 J , ■/> = 30 J ....... .,.........,,.,...,,.,.45
Figure
5.6 Steady state performance, ias, Vc, Cp = 180 ° , <!> = 0
5.7 Steady state performance, te, Cp = 180 ° , <b = 0
5.8 Steady state performance, ias, Vc, Cp = 180 , <b = 30
5.9 Steady state performance, te, Cp = 180 ° , <i> = 30
5.10 Steady state average torque output, Cp = 120° , 150
5.11 Steady state average torque output, Cp = 120 , 150
5.12 Steady state system efficiency, Cp = 120 ° , 150
5.13 Steady state system efficiency, Cp = 120 , 150 , 180
IX
NOMENCLATURE
ECM :
PWM
MMF
IMSL
u
R
C
P
T,
I),
electronically commutated motor
pulse width modulation or modulated
magnetomotive force
International Mathematical and Statistical Library
inverter firing angle
inverter conduction period
supply voltage in the brushless dc motor system
machine winding resistance
machine winding self inductance
machine winding leakage inductance
machine winding mutual inductance
permanent magnet flux Linkage
snubber network resistance
snubber network capacitance
number of poles in the machine
transistor i, i = A. B, C
diode L i = A. B, C
rotor mechanical angle
rotor electrical angle
X
• -v'r electrical angular frequency
• >is current in phase I of the machine, i=a, b, c
* vis voltage across phase I of the machine, i=a, b, c
• ve ; voltage across the snubber capacitor
• Te electric torque output of the machine
* Vce collector to emitter voltage of an npn bipolar transistor
• cr.or gating signal for transistor I, i = a, b, c
• on state conducting state of a solid state device
• off state non conducting state of a solid state device
* Yti . voltage across transistor I, i=a,b.c
• vdi voltage across diode I, i=a,b,c
• vei voltage induced in phase I by permanent magnet, i=a,b,
XI
ABSTRACT
Kazmirski, Todd J. M.S., Purdue University. May 1987. Steady State Analysis of a Brushless DC Motor with Half Bridge Inverter. Major Professor: Paul C. Krause.
Brushless do motors have received significant attention in the fractional
horsepower motor industry in recent years. The inverter drive used with this
system has traditionally been of the full bridge configuration. However, it is
possible to use a half bridge inverter to drive the machine. In this paper a
method to analyze the brushless dc motor with half bridge inverter is
developed. The method is integrated into a digital computer program to
simulate the steady state response of the system. The program allows both
the inverter conduction angle and the inverter firing angle to be changed.
This provides an efficient means of evaluating the effects of system parame
ters as well as inverter operating modes on the performance of the system.
The results of the simulation will demonstrate reductions in efficiency and
torque pulsations as well as an increase in average torque as the conduction
angle of the inverter is increased from 120 degrees to 180 degrees.
1
CHAPTER 1\
INTRODUCTION
The brushless dc motor is a permanent magnet synchronous machine
driven by an inverter drive. This machine also referred to as an
electronically commutated motor ■ (E,CM) has been the subject of growing
interest over the last several years. The reasons for this interest are the
advantages introduced by the brushless dc motor. Among the advantages
are:
- greater efficiency
- variable speed control
- smaller size and weight
- no brushes
- wider speed bandwidth
- bidirectional capability
- durability
- quick acceleration
- reduction of fire and toxicity hazards to the surrounding environment
The main disadvantage is the initial cost of the inverter drive and
inverter controller. The full bridge inverter operating in a discontinuous or
continuous mode has traditionally been used to drive the motor. However, it
2
is also possible to drive the motor using a half bridge inverter. This would
clearly have the advantage of reducing the number of inverter transistors
and diodes by one half. It might also reduce the complexity of the controller
used to switch the inverter.
The work contained herein is a systematic approach to analyzing the
brushless dc motor driven by a half bridge inverter. A description of the
motor, drive and snubber network is given. Next, equations for the machine
are written in machine variables. A description of the various operating
modes of the system is developed. From here, state variable equations are
written which describe the system in each of the operating modes. These
equations are used in a digital computer simulation. The simulation is then
used to study the steady state response of the system and determine the
effects of changing the firing angle of the inverter as well as conduction
period variations. The term steady state implies that the rotor speed of the
machine is held constant and that the phase currents in the machine are
periodic. ■
3
CHAPTER 2
DESCRIPTION OF THE BRUSHLESS DC MOTOR SYSTEM
2.1 INTRODUCTION
The initial step in the solution of any problem is a description of the
problem. From this description^ a readily solvable model can be developed to
obtain a solution. This chapter will begin with a description of the brushless
dc motor system. A description of the major subsystems of the system is
made followed by assumptions made for that component or subsystem. The
description of each subsystem is concluded by describing how that system
will be modeled.
2.2 THE BRUSHLESS DC MOTOR SYSTEM
The brushless do motor system consists of a permanent magnet
synchronous machine, a dc voltage source, a half bridge inverter, and a
snubber network. The system is given in figures 2.1.
2.3 DESCRIPTION OF THE HALF BRIDGE INVERTER
The inverter is of the half bridge type and is made up of three
transistors. In general, these can be any solid state switching device whose
conducting and nonconducting state can be independently controlled.
Throughout the analysis, the conducting state of a transistor will be referred
to as the "on state". This state occurs when the gating signal to the device is
on and the device is properly biased for conduction. The nonconducting
state of a transistor will be referred to as the "off state". This state occurs
when the gating signal is off or when the device is not properly biased for
conduction.
The inverter operates in the 120 degree conduction mode at all times
with the .gating sequence diagram as given in figure 2.2. Each transistor
conducts for 120 degrees and remains in the otf state for 240 degrees. As
shown in the diagram, Ta (transistor A) conducts or 120 degrees followed by
5
T0 conducting for 120 degrees and turning off and finally Tc conducting for
120 degrees and turning off. All angular displacements are in terms of the
electrical angular displacement, This sequence continually repeats itself. The
one variable in the sequence is the firing angle, <:>. The firing angle relates
this sequence to the rotor position.
The choice of the zero reference for the firing angle of the inverter is a
matter of convention. In this analysis, the zero firing angle will be defined to
coincide with A equal to zero. From figure 2.1, the conduction period for
phase A is centered around B equal to zero degrees. Thus, for a firing angle
of ' ’ equal to 0 degrees, ga would be on. for ^ ranging from -60 to 60 degrees.
For a firing angle of O equal to 60 degrees, ga would be on for B ranging from
-120 degrees to 0 degrees. For O equal to -30, ga would be on over the range
of " equal to -30 to 00 degrees. Thus the gating signals given in figure 2.2
can be written as a function of B and o., ' V.;; ' r
zM = gi(^r + '■■'’).■ i=a,b,c (2.1)
For the solution of the problem, the transistors are considered to be
ideal switches. These switches have zero resistance when in the on state and
infinite resistance when in the off state. When the gating to the transistors is
off, or when Vce < 0T the transistors will function as open circuits. When the
gating to the transistors is oil, the transistors will function as short circuits
provided VGe > 0. It is also assumed that there is an accurate wav to sense
tlie pp^ition of the rotor of this machine and a suitable con trol 4yai!ab!e
which will allow for the switching of the inverter at the desired rotor
positions,
2.4 DESCRIPTION OF THE SNUBBER NETWORK
The configuration of the motor inverter system creates problems due to
the abrupt switching of the inverter when windings are conducting. During
typical operation of the system, a transistor will turn on, allow conduction of
its respective phase, and turn off at the end of its 120 degree conduction
period. The switching of the inverter takes place regardless of the currents in
the motor. As one might expect, this will cause a very large voltage to be
induced in the windings due to abruptly forcing the current in the windings
to zero. To reduce these large transient voltages, a snubber network is added
to the the system to absorb the commutating current of the windings and
reduce the large voltages which would occur during the commutation period.
The network consists of a free wheeling diode connecting each phase of
the motor to a common resistor capacitor network as given in figure 2.1.
Throughout the analysis, the diodes will be considered to be ideal. That is.
the diodes are perfect conductors when forward biased and open circuits
when not conducting. These two states will be referred to as the "on state"
and the "off state” respectively.
6
7
2.5 DESCRIPTION OF THE BRUSHLESS DC MOTOR
A 2 pole, brushless dc motor is given in figure 2.1. The windings of the
machine are Y connected with the neutral connected to the positive side of
the dc source. Before writing the equations for this machine several
assumptions will be stated. First of all, the rotor will not conduct eddy
currents[4]. Second, the windings of the machine are sinusoidally distributed.
Finally, the rotor is non-salient[4l. The voltage equations of the machine can
be written in machine variables as
^abcs ^s'abcs d* P'Abcs (2.2)
where
(2.3)
r, 0 0r, 0 r, 0 (2.4)
0 0 r,
(2.5)
8
Ls =
Lis Ls Lm j-,mLrUL|S "t" Lg *--m
Lfn L|g + Lg
(2.6)
V = X'„
sin#r
sin(^r - -y)
sM8r ~ ~)
VeV,eb■v.
(2.7)
and p is the derivative operator d/dt. The equation for torque for this
machine is
Te = (~)xm'[('as ~ '^bs —Xaijcos^ + 'W3(ibs - ics)sin0rj (2.8)
2.6 CONCLUSION
A description of the brushless dc motor system has been made. The
gating sequence for switching the inverter has been presented along with the
zero convention for the tiring angle of the inverter. The models to be used in
the analysis have also been presented. Having completed these preliminary
tasks, the method for developing suitable equations whose solution will
describe the operation of the system will be presented in the next chapter.
9
bs-ixis
Figure 2.1 Brushless dc motor with half bridge inverter
10
g.(0r>4O
gb(^r»^)
g.(M)
A A '
B
a
C
I—<j>
I27T — (j>
9r
Figure 2.2 Transistor gating sequence for the half bridge inverter
11
CHAPTER 3
DEVELOPMENT OF THE STEADY
STATE MODEL FOR THE SYSTEM
3.1 INTRODUCTION
The models developed in the previous chapter will now be used in the
solution of the brushless dc motor system. In this chapter, various modes of
system operation will be examined. For each mode, an equivalent circui t will
be developed. A system of differential equations will then be written for each
circuit, and placed in state variable form. These equations will be integrated
into a digital computer program in the following chapter to simulate the
performance of the ECM system.
. 12
3.2 OPERATING MODES AND EQUIVALENT CIRCUITS
The equivalent circuit for the brushless dc motor system incorporating
the models of Chapter 2 is given in figure 3.1. In this circuit, the conducting
state of each transistor is a function of its respective gating signal and the
collector to emitter voltage. The state of each diode is a function of the
voltage across it. For each conduction period, this circuit can be further
split up into eight equivalent circuits dependent upon the states of the
transistors; and diodes. In other words, the equivalent circuit which models
the system is dependent upon whether the transistors and diodes are
conducting or not conducting. A total of 24 equivalent circuits is needed to
represent the system over all three conduction periods. From the models of
Chapter 2, a transistor will be in the on state if its gating signal is on and
VGe>0. Otherwise, the transistor will be in the off state. A diode will be in
the on state if it is forward biased. It will be in the off state if it is reverse
biased. ' : ■
For each of the three conduction periods, there are eight distinct
operating modes. The eight modes of operation will be outlined for the
conduction period When the gating signal for transistor A is on. The: modes
for the other two conduction periods can be easily derived from the A
conduction period modes. The eight conduction modes for the A conduction
period are listed in table 3.1. Let us examine mode 4. In this mode, is
greater than or equal to zero, and both V(lb and Vdc are less than zero.
Under these conditions, phases B and C are effectively eliminated from the
' 13
circuit. Thus the equivalent circuit of figure 3.2 can be used to represent the
3.3
Circuit equations for each mode can be easily written and placed in the
state variable form
x' = Ax -r Bu
For mode 4 of table 3.1 and the equivalent circuit of figure 3.2 the following
two independent equations can be written.
bsi rs*as ^ea
CV,;'
In matrix form, these equations become
Rx' — Sx -H Tu
where
Ls 0 -rs 0 -l 10 C s = 1 T — 1L J 0 0 —
R R
14
las
i
(0 S»__
____
_j
o>
___
1
U = >
Solving for R 1 and multiplying both sides by R 1
r ' 1 1-— 0 . _ --- . —*--'Ls Ls Ls
o 1x +
0 ——RC o
1---
(3.5)
The set of equations are now in the familiar state variable form of equation
3.1. This form can be easily integrated using a variety of numerical methods.
The A and B matrices for the eight operating modes of conduction period A
are given in the Appendix.
3.4 OPERATING MODES, EQUIVALENT CIRCUITS,
AND STATE EQUATIONS FOR THE B CONDUCTION PERIOD
The operating modes, equivalent circuits, and state equations ot
conduction period B can be easily derived from those of conduction period A.
The derivation is accomplished by a simple change of variables. The change
of variables consists of changing As variables for conduction period A to Bs
variables, Bs variables to C:s, and Gs to As. When .this transformation is
applied to the state equations only the state vector and the input vector are
changed as given in equation 3,6.
15
■ ■
4*bs veb^cs Veclas
u =
Vea;vc :y,
(3-6)
The A and B: matrices remain unchanged which is very convenient for the
digital computer simulation. The operating modes for the B conduction
period are given in table 3.2. A similar transformation can be used to
transform the operating modes, equivalent circuits, and state equations to
the C conduction period.
3.5 OPERATING MODES, EQUIVALENT CIRCUITS,
AND STATE EQUATIONS FOR THE C CONDUCTION PERIOD
The following transformation is used to transform the operating modes,
state equations, and equivalent circuits from the A conduction period to the
C conduction period. Change all A^ variables to Cs variables. Bs to As, and
C, to Bs. Again the A and B matrices remain unchanged with only the state
and input vectors changing. The operating modes for the C conduction
peripd are given in table 3.3 and the state and input vectors are givon in
equation 3.7.
16
V.s v„clas Vea
#1bsU —
Veb
vc V,
(3tf)
3.6 CONCLUSION
The eight operating modes for each conduction period have been defined.
For each state in the A conduction period, an equivalent circuit has been
developed. For each circuit, a set of state equations has been derived whose
solutions will describe the operation of the brushless dc motor system. Two
transformations have been introduced which transform the operating modes,
equivalent circuits, and state equations for the A conduction period to the B
and C conduction periods. The operating modes and the state equations can
now be used as the basis for a digital computer program to simulate the
operation of the brushless dc motor system.
17
db -
Figure 3.1 Brushless dc motor system equivalent circuit
18
Figure 3.2 Operating mode 4 equivalent circuit
19
Table 3.1 Conduction period A operating modes
CONDUCTION PERIOD A OPERATING MODES
Vu- Vdb Vdc operating mode
< 0 < 0 < 0 0
<0 < 0 > 0 1
<0 > 0 < 0 2
< 0 > 0 > 0 3
>o <0 <0 4
>0 < 0 > 0 5
> 0 > 0 < 0 6
> 0 >o > 0 7
20
Table 3.2 Conduction period B operating modes
CONDUCTION PERIOD B OPERATING MODES
vtb ' VdcVda j operating mode
< 0 < 0 <0 0
< 0 < 0 > 0 1
< o > 0 < o 2
<0 > 0 > 0 3
>0 <0 < o 4
>0 <0 > 0 5
>0 >0 < o 6
>0 > 0 > 0 7
Table 3.3 Conduction period C operating modes
CONDUCTION PERIOD C OPERATING MODES
V'.: Vtc ■■■■ Vda V operating mode
< 0 < 0 < 0 0
< 0 < 0 > 0 ; i ^ ^
<0 > 0 <0 2
< 0©A
l > 0 . 3
> 0 <0 < 0 4
> 0 < 0 > 0 : 5 a
> 0 > 0 <0 6 ' ■
> 0■
> 0 >0 7 ■
22
CHAPTER 4
STEADY STATE DIGITAL COMPUTER SIMULATION
4.1 INTRODUCTION
The state equations developed in Chapter 3 are now used in a digital
computer program to simulate the steady state operation of the brushless dc
motor system. Two digital simulations are presented. The difference between
the two is the inverter firing angle used. An analog simulation is also
presented for comparison to the digital simulation. The simulation results
show a large torque pulsation with instantaneous minima torque values
nearly equal to zero. These zero minima are inherent to the ECM with a half
bridge inverter operating in the 120 degree conduction mode. The cause of
the zero torque minima will be discussed and a easily implemented solution
will be presented.
23
4.2 DESCRIPTION OF THE COMPUTER PROGRAM
The method used in the digital computer program is straightforward.
The state equations developed in the preceding chapter are integrated using
the IMSL routine DGEAR. The algorithm is based on the Adams Method of
solving a system of differential equations.
The program can be summarized as follows. Throughout the simulation,
the rotor speed is held constant. Before each integration step, the transistor
and diode voltages are calculated to determine the operating mode of the
system. The proper A and B matrices are then supplied to the integration
routine and the integration is performed. These two steps are repeated until
one complete electrical cycle has been simulated. The final state vector is
then compared to the initial state vector. If the relative difference between
the two is less than the user supplied value, the simulation ends. Otherwise,
a new initial state vector is calculated and the the simulation is repeated.
This process continues until the user specified relative difference is satisfied.
Thus, the initial state vector supplied by the user need not be the steady
state state vector. The results of the integration at each step, namely the
state vector, can be used to calculate any system parameter desired by the
user.
24
4.3 MACHINE PARAMETERS
Results of analog and digital computer simulations of the system are
given in the following sections. The machine parameters used in the
simulation are given in table 4.1. All simulations which follow are performed
with a constant rotor speed of 3600 revolutions per minute.
4.4 DIGITAL SIMULATION RESULTS
Figures 4.1 through 4.4 show the results of two steady state digital
computer simulations of the ECM system with o = 0 degrees and' S == 30
degrees. These results show that this system operating in the 120 degree
conduction mode has a very large pulsating torque. For the case of o — 0,
the torque equals zero at discrete points. For the case of = 30, the torque
nearly equals zero at discrete points. In both cases, these points coincide
with the end of commutation for the winding which is most recently switched
off.
25
4.5 ANALOG SIMULATION RESULTS
Figure 4.5 shows the results of a steady state analog computer
simulation of the ECM system with 0=0 degrees. The model used for this
simulation is not the same as that used for the digital simulation'. The
difference is that the diodes and transistors both have a forward voltage drop
when conducting. In an attempt to compensate for this difference, the dc
source voltage was increased to 14.8 volts in the analog simulation. The
results of the analog simulation are very close to those obtained from the
digital simulation. The slight difference is attributed to the difference in the
models used for the simulations. In both the analog and digital simulations,
the torque equals zero or is very near zero at discrete points. This condition
is undesirable and may be unacceptable for certain applications. A method
to eliminate these near zero torque points will be discussed in the next
section.
4.6 TORQUE PULSATIONS AND ZERO TORQUE MINIMA
The zero torque minima associated with the ECM operating in the 120
degree conduction mode are undesirable. In order to eliminate this problem
it is necessary to understand the source of the torque pulsation. Upon close
examination of figure 4.1, one can see that at certain instants of time none of
the phases of the machine conduct current. These points correspond to the
time at which the last phase to be switched off ends commutation. To
examine this problem more closely, assume the currents which exist in the
machine are those associated with the conduction period of the transistors.
Thus, the currents which flow through the diodes after commutation are
neglected. From figure 2.2 which is repeated as figure 4.6, when ST = —o, Tc
is switched off, the current in phase G will begin commutation and at the end
of commutation, none of the windings conduct current. The resultant
magnetomotive force(MMF) is zero. Next, the A phase immediately begins to
conduct current. This results in an MMF which can be represented by a
vector in the as axis of the machine. As the current increases and decreases
in the A phase the magnitude of the MMF vector changes but its position
remains constant. Thus including the commutation period for the A phase,
the MMF vector will remain in the as axis for 120 electrical degrees. The
vector will then go to zero for a brief instant and then jump to the b3 axis
coinciding with the B conduction period. At the end of commutation of the
B phase, the MMF vector will again go to zero and then jump to the cs axis
due to the current conduction in the C phase. This current conduction
pattern gives rise to an MMF vector which is discretely positioned in one of
three axes dependent upon the phase of the machine which is conducting.
The result is a large pulsating torque which has minima which reach zero at
the end of commutation of each phase.
27
4.7 ELIMINATING THE ZERO TORQUE MINIMA
Ii ilder to eliminate the pulsating torque, the currents in the motor
must produce an MMF which is constant and rotates at the speed of the
rotor. These conditions can only be satisfied if the phase currents are
sinusoidal and displaced by 120 degrees. In order to satisfy these conditions,
a complex controller must be used. This type of controller would certainly be
expensive and may make the ECM system with half bridge inverter an
unacceptable alternative to other drive systems. However, the operating
mode of the existing system can be changed to reduce the torque pulsations
with no additional hardware.
The zero minima of the steady state torque curve can be eliminated if at
least one phase of the machine- is always conducting current. This condition
can be satisfied if the conduction period of the half bridge inverter is
increased from 120 degrees. Using this type of inverter, at least one phase of
the machine will always be conducting current. During the commutation
period, two windings will be conducting. This current conduction pattern
will result in an MMF vector which moves in a continuous fashion.
To visualize this, assume at time tt only phase A conducts current. The
resultant MMF vector is coincident with the as axis of the machine. Since
the conduction period is greater than 120 degrees, phase B will he switched
on before phase A is switched oil". At t.>, the B phase is switched on and the
increasing current in phase B results in an MMF which rotates from the as
axis towards the bs axis. At t3, the A phase is switched off. .As the A phase
28
commutates, the MMF vector continues to rotate towards the bg axis. When
the current in the A phase goes to zero the MMF vector will be coincident
with the bs axis of the machine. The MMF vector will make a similar
transition to the cs and as axes as the three phases of the machine are
switched at the desired rotor positions. It is important to note that although
the MMF vector moves in a continuous fashion, the magnitude and the
rotation speed of the vector is not constant. The result is that the zero
minima of the steady state torque curve are eliminated, but the torque
pulsations will remain although reduced in magnitude.
4.8 CONCLUSION
The results of steady state digital and analog computer simulations have
been presented in this chapter. The results of the digital simulation are in
close agreement with those of the analog simulation. Both simulations show
the undesirable near zero value points occurring in the steady state torque
curves. A solution to this problem has been proposed. The solution involves
extending the conduction period of the half bridge inverter to a value greater
than 120 degrees. This solution will be explored in Chapter 5 with the
presentation of simulation results obtained using several different conduction
periods. . . ' i .
Uc UOLTS
las AMPERES
29
20.00
16.00
12.00 -
8.00
0.0060.00•60.00 0.00 120.00 180 00 240.00 300.00
ELECTRICAL ROTOR POSITION DEGREES
27 oo
26 oo
25.50
24.50
24.00•60.00 0.00 60.00 180.00 240 00 300 00
ELEGTRIGAL ROTOR POSITION DEGREES
Figure 4.1 Steady state digital simulation, ias,' Vc, 120 ° conduction, 0 — 0
torq
ue in
n
30
240.00 300 00
ELECTRICAL ROTOR POSITION DEGREES
Figure 4.2 Steady state digital simulation, te, 120 ° conduction, •.>
UOLTS
las AM
PERES
31
20.00
16.00
12.00
8.00
H.OG
0.00-90.00 30.00 90.00 150.00 270.00210.00
ELECTRICAL ROTOR POSITION DEGREES
27.50
27 00
26.50
26.00 -
25.50
270 00150.00 210.0090.00-30.00 30.00-90.00
DEGREESELECTRICAL ROTOR POSITION
Figure 4.3 Steady state digital simulation, ias, Vc, 120 conduction, O — 30
TOR
QU
E IN N-M
32
\
0.00-90.00 -30.00 30.00 90.00 150.00 210.00 270:00
ELECTRICAL ROTOR POSITION DEGREES
Figure 4.4 Steady state digital simulation, te, 120 ° conduction, d> ==
;
volts
'a
s amps33
0.250
0.125
Figure 4.5 Steady state analog simulation, 120 ° conduction, (h = 0 °
34
g.(0r»$
gb(0r>^)
g•(*„*)
A A
B
<1
C
I—
7T3
9*
5tt
3
!— <£ 27T <P
Figure 4.6 Transistor gating sequence for the half bridge inverter
Table 4.1 Machine parameters
35
MACHINE PARAMETERS
PARAMETER VALUE UNITS
Rs 5.5 ohms
C* 1.0 millifarad
Vs- . 14.0 volts
r 0.1 ohms
:■ 4 0.3 millihenries
L.is 0.03 millihenries
-'in. ' -0.135 millihenries
.....10.38 millivolt—second
radian
poles 6 V
36
CHAPTER 5
STEADY STATE SIMULATION WITH 120 ° < Cp <180°
5.1 INTRODUCTION
In Chapter 4, increasing the conduction period of the inverter from 120
degrees was proposed as a method to reduce the torque pulsation of the ECM
system. In this chapter, the results of a digital computer simulation will be
presented which prove that extending the conduction period does in fact
decrease the torque pulsation. The results also show that larger conduction
periods result in larger average torque values for the system at the expense of
system efficiency.
37
5.2 OPERATING MODES AND STATE EQUATIONS
The same method which is used to analyze the ECM system with 120
degree conduction mode is used to analyze the the more general case of
conduction periods ranging from 120 degrees to 180 degrees. Similar to the
120 conduction mode, there are three symmetrical 120 degree periods. These
three periods can be defined by the rotor position where they start. The three
positions are given in figure 5.1 which shows the transistor gating signal as a
function of rotor position and inverter firing angle. In figure 5.1. the 120
degree symmetrical periods have been further split into two periods and are
determined by the gating signals on at the given rotor position. The period
when Ta is switched on until Tc is switched off is defined as A2. The
subscript two is used to indicate that two transistors are switched on. The
period when Tc is switched off until Tb is switched on is defined as A,. In this
case, the subscript one is used to indicate that only one transistor is switched
on. The periods defined by B and C are symmetrical to the A conduction
period. In other words the state equations for the B and C conduction
periods can be derived from those of the A conduction period by using the
change of variables introduced in Chapter 3. The modes of operation for the
A conduction period are given in tables 5.2 and 5.3 The A and B matrices
corresponding to the state equations for the A, B, and C conduction periods
are given in the Appendix. The modes of operation for the B and C
conduction periods are listed in tables 5.4 through 5.7.
38
5.3 STEADY STATE DIGITAL COMPUTER SIMULATION
AM of the simulations which follow were obtained using a digital
computer program and the same general algorithm used in Chapter 4. The
system parameters are the same as those given in Chapter 4 with the
exception of the conduction period, Cp, which ranges from 120 degrees to 180
degrees. k,::.-
Given in figures 5.2 and 5.3 are the results of a simulation with
Cp = 150 , and 0 — 0 . Figures 5.4 and 5.5 give the results of a simulation
with Cp = 150 , and </> = 30 . Given in figures 5.6 and 5.7 are the results
of a simulation with Cp = 180 ° , and <;> = 0 °. Figures 5.8 and 5.9 give the
results of a simulation with Cp = 180 °, and 4> = 30°. Upon close
examination of these results and those of Chapter 4 several conclusions can
be drawn. First, the near zero torque minima of the 120 degree conduction
mode system are eliminated by increasing the conduction mode to 150 or 180
degrees. Second, a significant increase in the average torque output is
obtained by increasing the conduction period from 120 degrees to 180
degrees. Third, an average torque increase also results when O is advanced
from 0 degrees to 30 degrees for each of the three conduction modes studied.
Finally, the peak capacitor voltage and phase current vary widely as
Cp and <i> are changed. The results also suggest that there may be unique
values of o and Cp to optimize the average torque. Likewise, these two
parameters may be adjusted in order to optimize the efficiency of the ECM
system.
5.4 TORQUE AND EFFICIENCY OPTIMIZATION
39
In order to investigate the effects of Cp and <!) on system efficiency and
average torque, several simulations were run. These simulations consist of
measuring the system efficiency and average torque output for several steady
state operating modes of the system. The efficiency is defined as the
mechanical energy output divided by the electrical energy input into the
system, multiplied by 100%. Results are given in figures 5.10 through 5.13.
Given in figures 5.10 and 5.11 are the average torque output results for
conduction periods of 120, 150, and 180 degrees. Figures o.l2 and 5.13 give
the system efficiency results for conduction periods ol 120, 150, and 180
degrees. These results demonstrate the effects oi' Cp and <!> on the average
torque output and system efficiency. The average torque can be increased by
increasing Cp from 120 degrees. This increase however comes at the expense
of operating efficiency since the 120 conduction mode is the most efficient of
the three modes studied. The firing angle of the inverter on the other hand
can be used to optimize both the average torque output and the system
operating efficiency. As an example, examination of the 120 conduction mode
shows that as the firing angle is changed from 0 to 30 degrees the average
torque increases from 0.15 newton-meters to 0.35 newton-meters. T.he system
efficiency moves from 65% to 70%.
40
5.5 CONCLUSION
A digital computer program was developed and used to compare the
brushless dc motor system with three different conduction modes and several
different inverter firing angles. As expected, the results demonstrate that the
near zero minima of the instantaneous torque output can be eliminated by
changing the 120 degree conduction to 150 or 180 degrees. The results also
demonstrate the use of the conduction period to increase the average torque
output and decrease the system operating efficiency or vice versa. Larger
conduction modes provide for larger torque, but only at the expense of
operating efficiency as compared to the 120 conduction mode. The firing
angle however can be adjusted over various ranges to provide for increased
operating efficiency and torque output for any of the three conduction modes
studied;
41
Table 5.1 Transistor gating sequence for 120° < Cp < 180°
Si ON, gioa CENTER, gic OFF, gi0ff
Sa 1 1 |*o.
-.0. cp
-? + -
gb - (j) - + 120 °: ; 2
- 6 + 120 °CD
- <i> + — + 1202 . .
gc - 6 - — - 120 °' ' 2 .
- <b - 120 ° Vo + ~ 12 0 °
— A2 | '.AT
= Cp - 120
gbC^r*^) x2 = 120 °
B2 ; B1 I, i x3 = 120 °
«*—------ x3 ------- »U»,A)
*2 -| C2 ! Cl ! .* '•
■C,
I/
47T3
n
Figure 5.1 Transistor gating sequence for 120 ° < Cp < 180 °
•JO
LTS
42
HO . 00
33.00
2H 00
16.00
0.0075.00 15.00 105.00 165,00 225.00 285.00
ELECTRICAL ROTOR POSITION
uu -15.00 H5.00 105.00 165.00
electrical rotor position DEGREES
Figure 5.2 Steady state performan
TOR
QU
E IN N-
43
i
225.00 285 0075.00 15.00 45.00 165:00105.00
ELECTRICAL ROTOR POSITION DEGREES:
Figure 5.3 Steady state performance, ter.Cp = 150 ° , <> = 0
UO
LTS
las AMPERES
44
HO 00
32.0 0
24.00
16.00
0 .00255.00195.00135.0075.0015.00■45.00105.00
ELECTRICAL ROTOR POSITION DEGREES
38.00
37.00
36.00
35.00
34.00
33,00 255.00135.00 195 ,0075.. 0015.00105.00
ELECTRICAL ROTOR POSITION DEGREES
Figure 5.4 Steady state performance, ias, Vc, Cp = ISO ° , < > = 30 °
ELECTRICAL ROTOR POSITION DEGREES
Figure 5.5 Steady state performance, te, C
MOLTS
las AM
PERES
46
48.00
40.00
24.00
16.00
8.00
-30.00 30.00 210,00 270.0090.00 90.00 150.00
ELECTRICAL ROTOR POSITION
43,00
39.00
210.00 g70?9B150.0030.0030.00
ELECTRICAL ROTOR POSITION
Figure 5.6 Steady state performance, ias, Vc, Cp =» 180 " , o « () /
TORQUE IN
47
0.00-90.00 -30.00 30.00 90.00 150.00
ELECTRICAL ROTOR POSITION DEGREES
Figure 5.7 Steady state performance, te, Cp = 180 ° , <:> = 0 °
UOLTS
las AM
PERES
48
50.00
HO . 00
20.00
10.00
0.00120.00 •60 00 0.00 60.00 120.00 180.00 2H0.00
ELECTRICAL ROTOR POSITION DEGREES
HH.00
38.00
180.00 gHQ.OQ60.00 120.00120.00 ■60.00
ELECTRICAL ROTOR POSITION DEGREES
5.8 Steady state performance
TOR
QU
E IIS IS-
49
180.00120.0060.000 .00-120.00 -60.00
ELECTRICAL ROTOR POSITION DEGREES
Figure 5.9 Steady state performance, tRt Cp = 1.80 , = 30
AUER
AGE
TORQUE! N-
N
AUER
AGE
TORQ
UE
50
= 120
120.0.90.0060.0030.000 .00-30 00
PHI DEGREES
.20
Cn = 150
-1090.0030.00 60.00-30 00 0.00
PHI DEGREES
Figure 5.10 Steady state average torque output, Cp == 120 ° , 150
51
1.20
90
5 .30
Cn = 180
0.00-30.00 0.00 30.00 60 00 90 00
PHI DEGREES
1.20
Iz:.90
ZD C3 C£ O f—idID<rLdZ><1
.60
30
0 .00
CL = 150
* Cn = 120120.0060.00 90.0030.000 . 00
PHI DEGREES
Figure 5.21 Steady state average torque output, Cp = 120 ° , 150 ° , 180 °
52
100.00
80.00
60 00
h HO.00
CO 20.00
0.00120 .0090 0060.0030 000 . 0030.00
DEGREES
100 .00
70.00
40.00
10.00
20.0 030.00 0 .00 30.00 60.00 90.00 120 .00
DEGREES
Figure 5.12 Steady state system efficiency, C
EFFI
CIE
NC
Y
SYST
EM EFF
ICIE
NC
Y
53
100 .0
80.00
60.00
40 00
0.0030 00 0.00 30 00 60 00 90 00
DEGREES100 o
80.00
60.00
40.00
CO 20.00
0 . 000120 0090 0060.0030.0030.00
DEGREES
Figure 5.13 Steady state system efficiency, C
Table 5.2 Conduction period A1 operating modes, 120 < Cp < 180 °
CONDUCTION PERIOD A1 OPERATING MODES
>db vdc operating mode
< 0 ■■ < 0 < 0 0
< 0 < 0 > 0 :' 1 ■:;
< 0 > 0 < 0 2
<0 > 0 > 0
> 0 < 0 < 0 4 ■■
> 0 ' < o V 0 5
> 0 > 0 <0 6 ■'
> 0 > 0 > 0 7
' 55
Table 5.3 Conduction period A2 operating modes, 120 < Cp < 180
CONDUCTION PERIOD A2 OPERATING MODES
■ - ; Vtc Vdb operating mode
< 0 < 0 <0 8
<0 <0 > 0 9
<0 >0 <0 10
<0 >0 > 0 n
>0 <0 <0 ■ 12 . .
> 0 <0 > 0 : 13
>0 > 0 <0 14
> 0 > 0 > 0 15 ’
56-
Table 5.4 Conduction period Bl operating modes, 120 0 < Cp <180 °
CONDUCTION PERIOD Bl OPERATING MODES
;: ;;vtb : vdc Vda operating mode
< 0 CO < 0 0
< 0 <0 > 0 ■ ■ T'■
< ° . 1 0 : <0 2 ■.
< o >0 >0 3 • "
>0 <0 < 0 i.
> 0 ' <0 > 0 5 .
>0 >0 <0 6
■ . > o ■ _ 0 > 0 fy-i
. 57
Table 5.5 Conduction period B2 operating modes, 120 ° < Cp < 180 °
CONDUCTION PERIOD B2 OPERATING MODES
. ; ' : ; Vtb , . : " Vta . Vdc operating mode
<0 < 0 C 0 , s: .
<0 < o > 0 9
< 0 > 0 <0 10
o > 0 > 0 11
> 0 <0 <0 12
> 0 <0 > 0
>0 >0 <0 14
>0 >0 >0 !5
■. 58
Table 5.6 Conduction period Cl operating modes, 120 ° < Cp <180 °
CONDUCTION PERIOD Cl OPERATING MODES
Vda Vdb operating mode
o ' < 0 < 0
< 0 <0 > 0 i
<0 '■ > 0 <0 2
< o > 0 > 0 . 3
> 0 < 0 < 0 4 . . ' < : ■
> 0 < 0 > 0 5
>0 >0 <0 . ''6
> 0 > 0 > 0 4
; 59
Table 5.7 Conduction period C2 operating modes, 120 ° < Cp < 180 °
CONDUCTION PERIOD C2 OPERATING MODES
Vtb Vda operating mode
A, ' . < 0 <0 8
<0 < 0 > 0 • 9,.
<0 > 0 <0 10
<0 > 0 > 0 : 'n. 'V
> 0 <0 <0 • 12 ...
■ 0 < 0 > 0 ■ 13
> 0 ■ > 0 <0 14
> 0 >0 >0 . 15
60
CHAPTER 6
CONCLUDING COMMENTS
The work presented herein provides mathematical equations whose
solution describe the operation of the brushless dc motor system operating in
conduction modes ranging from 120 to 180 degrees. These equations which
are in state variable form have been integrated into a digital computer
program to provide a simulation of the brushless dc motor system.
The results of a set of simulations have been used to study the use of
the conduction mode and the firing angle of the inverter to control the
operation of the system. Among the conclusions which were drawn from the
simulations is that the conduction mode of the inverter can be used to
eliminate the near zero torque minima encountered with 120 degree
conduction mode system. Also, the conduction mode can be used to increase
the average torque output of the system, but only at the expense of a
decrease in system operating efficiency. The variation of the firing angle of
the inverter was shown to increase both the average torque and the system
efficiency over a limited range.
The system studied was not a pulse width modulated system. In order
tQ study the PWM ECM system, a new set of mathematical equations
describing the system must be developed. The method used to develop the
state equations which describe the non PWM system can also be used to
develop equations for the PWM system. These equations could then be
integrated into the digital computer program by changing the A and B
matrices and changing the logic in the program. The same method could
LIST OF REFERENCES
62
LIST OF REFERENCES
[1] P. C. Krause Analysis of Electric Machinery McGraw-Hill, 1986.
[2] N. A. Demerdash and T. W. Nehl, “Dynamic Modeling of Brushless DC Motors for Aerospace Actuation,” IEEE Trans, on Aerospace arid Electronic Systems,Vol. AES-16, pp. 811-821, November 1980.
[3] S. Williams and C. Eng, DC Motor,” Proc. IEE, Vol. 132, pp 53-56, January, 1985.
4j T. W. Nehl, F. A. Fouad, N. A. Demerdash, and E. A, Maslowski, Electronically Operated Synchronous Machines with Parameters Obtained from Finite Element Field Solution,” IEEE Trans on Industry Applications, Vol, IA-18, pp 172-182, March/April 1982.
[5] F. A. Fouad, T. W. Nehl, and N. A. Demerdash, Operated Synchronous Machines Using Finite Elements,” IEEE Trans, on Power Apparatus and Systems, Vol. PAS-100, pp, 4125-4134, September 1981.
[6] R. S. Ramshaw, A. W. J. Griffin, and K. Lloyd, Control, pp 40-44, January 1966
[7] J. R. Woodbury, IEEE Trans, on Industrial Electronics and Control Instrumentation, ” Vol. IECI-21, pp. 52-80, May 1974.
[8] T. W. Nehl, F. A, Fouad, and N. A. Demerdash, Electronically Commutated DC Permanent Magnet Machines,” IEEE Trans, on Magnetics Vol. MAG-17, pp 3284-3286, November 1981.
[9] T. M. Jahns with Rectangular Current Excitation,” IEEE Trans, on Industrial Applications, Vol. IA-20 No. 4, pp 803-813, July/August 1984.
[10] E. Richter, T. J. E. Miller, T. W. Neumann, T. L. Hudson, Economical Assessment,” IEEE Trans, on Industry Applications, Vol LA-21, No. 4, pp. 644-650, May/June 1985.
[11] R. DeCarlo, Class Notes, 1987.
63
APPENDIX
The conduction modes zero through seven for conduction periods greater
than 120 degrees have the same A and B matrices as the corresponding
modes of the 120 degree conduction mode. The A and B matrices are listed
below.
64
Aq —
0 0 0 0 0 0 0 0 0
0 0 0
000
RC
A, =
0 0 0 0
0 0 —
00
oo —
00_1_'l.
RC
Ao —
0 0
0Ls0 0
— 0c
0
0RC
Aak2
0 0 0 00 —rLs rLm Lm - Ls
® rLs Lm — Lgko ko ko.
0 — —C C RC
A4.;—
~u 00
0 0 0 0 0 0
0 0 0-
0
00
1RC
..'As. _1_k2
-rLs 0 rLm 0 0 0
rLm 0 —rLs k2
Lm0
-Lsk2
RC
A6 k2
-rLg rLm 0 rLm -rLg 0
0 0 0 ko
o — 0c
“Lg0k2
”rc
>\
66
—k2r -k3r -k3r -2k3
II
.
-k3r -^k2r ■^k3r —k2 ~ k3-k3r -k3r —k2.r —k2 — k3
0 illC
k]_C
kiRC
Ag — Aq
Ag — Ao
10
0 0 0 0 0 0 0 0
0 0
0 0
Ls
0
0
1RC
An k2
0 0 0 —rLs
0 rLn k2
0 0
rLs Ljjj
0k2RC
Ai2 — A4
67
A-13 — Ag
A14k2
-rLs 0 rLm 0 0 0
rLm 0 “rLs
0 0 0
000k2
RC
Ais kl
k2r k3rk3r —k2r k3r —k3r
•k3r•k3r■kor
0
”k3
-k2-k3
kiRC
Where
k, = L,3 + 2Lm3 - 3Lm2L,
k, = L,- - Lm-
^3 ^nr^s
68
B00 0 0 0 0 0
0 0 0
0 0 0 _1_
Ls0p1
RC
Bt =
0 0 0 0
0 0
00
0 0 0
00
J_Ls
1RC
B2
0
0
0
0
0J_Ls0
0
0
0
01
0
0
Ls01
RC
b3 k2
0 0 0 0 0 —Lc Lm L, — Lms m s • m
0 Lm —Ls Ls — Lmko
0 0 0RC
GO I
69
B40 0 0 0 0 0
0 0 0
J_Ls001
RC
5 k2
—Le 0 Lm L« — Lms m s m
0 0 0 0 Lm 0 —L, L. — Lmin s s m
o ^2RC
b6 k2
-LsW
0
0
Lm
-Ls0
0
0 L3 — L 0 Ls — L 0 00 >-
RC
mm
b7 _2_ki
■k2
'^3'^3
■k3 —k3 k2 + 2k3 •k2 — k3 k2 + 2k3 •k3 —k2 k2 + 2k3
A.RC
0 0 0
70
B8 Bo
Bg — B,
Bio — Bi
Bn — B3
B12 — B4
Bis — B6
Bh — B5
Bis — B7
71
Where
k, = Ls3 + 2Lm3 - 3Lm2Ls
k2 = Ls2 - Lm2
^3 — Lm‘ — LmLs