A motor controller for solar car

The Department of Computer Science and Electrical Engineering

A Motor Controller

For the Solar Car

Project

Andrew James Reghenzani

Supervisor : Mr. Geoffrey Walker

Submitted for the degree of

Bachelor of Engineering (Electrical And Electronic)

16th October 1998.

Union College,

Upland Road,

St. Lucia QLD 4067.

Ph : (07) 33771500

Fax : (07) 33713826

16 October 1998

The Dean,

Faculty of Engineering,

The University of Queensland,

St. Lucia QLD 4072

Dear Professor Simmons,

In accordance with the requirement of the degree of Bachelor of Engineering in

the division of Electrical and Electronic Engineering, I present the following thesis

entitled :

“A Motor Controller

For the Solar Car

Project”

This work was performed under the supervision of Mr. Geoffrey Walker. I

declare that the work submitted in this thesis is my own, except as acknowledged in the

text and footnotes, and has not been previously submitted for a degree at The University

of Queensland or any other institution.

Yours Sincerely,

Andrew J. Reghenzani.

A Motor Controller For The Solar Car Project

- iii -

ACKNOWLEDGEDGMENTS

The following people deserve special recognition for their contributions to my

thesis project throughout the year:

My family : who have always supported me throughout University, and have given me

the extra motivation to succeed during difficult times.

My friends : for understanding how important my thesis was and always seeming to ask

the all too familiar question “How’s your thesis going?”.

Members of the Solar Car Team : especially Charles for organizing use of a digital

camera and Anthoney for assistance with writing code. I have thoroughly enjoyed

being in the solar racing team, as it has given me the opportunity to gain valuable work

experience and gain some practical skills which complement my University studies.

My supervisor, Mr. Geoffrey Walker : for all his time, invaluable advice and

encouragement throughout the thesis project.

Keith Aldworth and the electronics workshop personnel : for the manufacture of my

PCB’s and all the labor intensive hand tinning that had to be done for both boards,

supply of components, use of the surface mount soldering station and all the technical

tips regarding PCB design and manufacture.

Keith Lane, Wayne Jenkins and Bill Slack from the electronics workshop : for building

my heatsinks and other hardware from my plans which usually consisted of a page of

dimensions, use of the tools and machines in the workshop at any time and all the

technical advice regarding manufacturing.


- iv -

ABSTRACT

The transport needs of our ever growing and evolving society is becoming

increasingly stringent and more demanding. In order to combat this, more efficient

transportation vehicles need to be developed which are faster and cleaner. As the

human race starts to realize the real extent to which the internal combustion engine has

gradually polluted the atmosphere, more research is being concentrated on alternative

forms of propulsion. A number of propulsion systems and energy sources have

undergone feasibility studies to investigate potential commercial and industrial

applications. Some projects have been shown to work successfully, while other

technologies are still well in their infancy stage of development. A handful of examples

of the technologies under consideration include nuclear energy, fuel cells, steam power,

solar power, wind power and tidal power.

Electric and hybrid powered cars are emerging as a popular transport alternative.

These type of vehicles emit far less pollutants to the atmosphere than the single internal

combustion engine, and have been proven to display moderate driving range (up to

300km). An electrically powered vehicle has essentially three major electrical

components. These are an energy source (usually a rechargeable battery bank), an

inverter or motor controller and an electric motor. In the case of a solar car, the energy

source is typically a bank of batteries, which may be recharged by photovoltaic solar

panels. The motor controller is typically a power electronics device which when

supplied with the driver’s input commands, controls the torque in the electric motor.

The electric motor converts the electrical energy supplied by the motor controller to

mechanical energy used to propel the vehicle, usually through a type of transmission.

A motor controller is custom designed for a new hub mounted Brushless DC

Permanent Magnet (BLDC PM) motor, as part of the solar car project. Efficiency and

reliability have been two of the key factors considered when designing the controller.

Due to careful selection of quality components and use of high efficiency control

algorithms, a marketable increase in efficiency over the existing system is expected with

the new controller and motor.


- v -

CONTENTS

ACKNOWLEDGEDGMENTS......................................................................................................IABSTRACT..................................................................................................................................IVCONTENTS...................................................................................................................................VLIST OF FIGURES ............................................................................................................VIILIST OF TABLES .............................................................................................................VIII1. INTRODUCTION................................................................................................................11.1 Introduction ............................................................................................................ 11.2 Problem Specification ............................................................................................ 21.2.1 Thesis Goal .......................................................................................................... 41.2.2 Motivation behind the Motor Controller and Motion Control............................. 41.3 Organization of the Thesis Document.................................................................... 52. THE UNIVERSITY OF QUEENSLAND SOLAR CAR..............................72.1 Solar Car Racing and the Races ............................................................................. 72.2 A Brief History of the UQ Solar Racing Car ......................................................... 92.3 The Nuts and Volts of a Solar Car.......................................................................... 92.3.1 Batteries............................................................................................................. 102.3.2 Solar Array ........................................................................................................ 122.3.3 Maximum Peak Power Trackers (MPPT’s)....................................................... 132.3.4 Motor Controller................................................................................................ 132.3.5 Motor ................................................................................................................. 142.3.6 Telemetry Functions and Power Supply............................................................ 152.4 Necessity for Efficient Systems............................................................................ 152.5 The Existing Drive System................................................................................... 162.5.1 Controller Type.................................................................................................. 162.5.2 Performance Characteristics .............................................................................. 172.6 The New Drive System ........................................................................................ 172.6.1 Additional Features............................................................................................ 182.6.2 Performance Requirements................................................................................ 193. MOTOR CONTROL LITERATURE......................................................................204. THEORY ...............................................................................................................................254.1 The Permanent Magnet Brushless DC Motor ...................................................... 254.1.1 Electrical and Mechanical Parameters............................................................... 284.2 Controlling a Permanent Magnet Brushless DC Motor........................................ 304.2.1 Commutation ..................................................................................................... 304.2.2 Current Regulation ............................................................................................ 354.2.3 Trapezoidal Current Excitation.......................................................................... 354.2.4 Sinusoidal Current Excitation............................................................................ 374.3 Power MOSFET Device Characteristics .............................................................. 384.4 Heatsink Considerations....................................................................................... 415. HARDWARE DESIGN STAGE...............................................................................435.1 Design of Power Stage ......................................................................................... 435.1.1 Circuit Design.................................................................................................... 445.1.2 Sensors............................................................................................................... 455.1.2.1 Bus Voltage Measurement.............................................................................. 455.1.2.2 MOSFET Heatsink Temperature Measurement ............................................. 465.1.2.3 Phase Current Measurement ........................................................................... 465.1.3 Manufacture and Construction .......................................................................... 485.2 Design of Control Stage ....................................................................................... 50


- vi -

5.2.1 Circuit Design.................................................................................................... 515.2.1.1 Auxiliary Components and Power .................................................................. 525.2.1.2 Memory Board................................................................................................ 525.2.1.3 Input/Output Ports........................................................................................... 525.2.2 Manufacture and Construction .......................................................................... 546. SOFTWARE DESIGN STAGE................................................................................566.1 System Description............................................................................................... 566.2 Main Program....................................................................................................... 586.3 Torque Control ..................................................................................................... 596.3.1 Regeneration...................................................................................................... 596.3.2 Brake.................................................................................................................. 606.4 MOSFET Heatsink Temperature.......................................................................... 616.5 Motor Temperature............................................................................................... 616.6 Speed and Direction ............................................................................................. 616.7 Commutation........................................................................................................ 616.8 Bus Voltage .......................................................................................................... 617. DISCUSSION .....................................................................................................................637.1 Discussion ............................................................................................................ 638. CONCLUSIONS ................................................................................................................648.1 Thesis Conclusions............................................................................................... 648.2 Possible Future Work ........................................................................................... 648.3 The Future of Solar Car Racing : The Big Picture ............................................... 66APPENDICES ..............................................................................................................................67APPENDIX A: SCHEMATIC AND PCB DESIGNS.................................................68APPENDIX B: MOSFET DATA SHEETS.............................................................69APPENDIX C: CSIRO/UTS MOTOR SPECIFICATIONS........................................70APPENDIX D: MICROCOMPUTER PROGRAM LISTINGS...............................71APPENDIX E: ACCOMPANYING COMPUTER DISK .............................................72MAIN PROGRAM ......................................................................................................................72SCHEMATIC FILES ..................................................................................................................72PCB FILES ...................................................................................................................................72BIBLIOGRAPHY ........................................................................................................................73

BOOKS ...................................................................................................................................... 73

JOURNAL ARTICLES ............................................................................................................ 73

INTERNET RESOURCES ...................................................................................................... 77


- vii -

LIST OF FIGURES

FIGURE 1 : BLOCK ELECTRICAL DIAGRAM OF A SOLAR CAR......................................................................10FIGURE 8 : HALL EFFECT POSITIONING SENSORS........................................................................................28FIGURE 9: NUMBERING PATTERN FOR MOSFET’S IN THE H-BRIDGE.........................................................30FIGURE 10 : 120 DEGREES COMMUTATION MODE......................................................................................32FIGURE 11 : 180 DEGREES CONDUCTION MODE.........................................................................................34FIGURE 12 : CURRENT FEEDBACK IN A BLDC MOTOR...............................................................................35FIGURE 13 : TORQUE RIPPLE IN A TRAPEZOIDAL MACHINE ........................................................................36FIGURE 14:NON-CONDUCTRING MOSFET[34] ..............................................................................................FIGURE 15:CONDUCTING MOSFET[34] ............................................................................................38FIGURE 16:WAVEFORMS AT TURN-ON[38].....................................................................................................FIGURE 17:WAVEFORMS AT TURN-OFF[38]................................................................................................39FIGURE 20 : THERMISTOR RESPONSE..........................................................................................................54FIGURE 22 : BLOCK DIAGRAM OF CONTROL ALGORITHM...........................................................................57FIGURE 23 : A FOUR QUADRANT DRIVE.......................................................................................................58FIGURE 24:ONE SWITCH ACTIVE TOPOLOGY..................................................................................................FIGURE 25:TWO SWITCH ACTIVE TOPOLOGY.............................................................................................60


- viii -

LIST OF TABLES

TABLE 1 : 120 DEGREES COMMUTATION TRUTH TABLE ............................................................................31TABLE 2 : 180 DEGREES COMMUTATION TRUTH TABLE ............................................................................33

- 1 -

1. INTRODUCTION

1.1 Introduction

The development of the internal combustion engine was certainly considered a

milestone for mankind. The focus back in the time of the Industrial Revolution was to

design machines which could fulfill time consuming, labor intensive jobs in a fraction

of the time that it took humans alone using conventional methods. Cars were developed

as a fast means of transport, and internal combustion engines soon found themselves in

many applications ranging from cane harvesters to outback generator sets. As time

progressed, most people had realized that although the internal combustion engine had

provided a much easier lifestyle, there were a number of major drawbacks. Petrol,

when combusted, forms a number of gaseous byproducts, consisting mainly of carbon

dioxide, but also containing traces of other gases such as carbon monoxide and

compounds containing lead. The potency and increasing levels of these gases and

compounds are causing gradual damage to the ozone layer in the Earth’s atmosphere.

Such gases are commonly referred to as greenhouse gases.

1. Introduction

- 2 -

Soon people began looking for alternatives to the internal combustion engine.

Quite recently, hybrid electric vehicles (EV) have been met with much success, and

commercial versions are being made today. A typical hybrid EV is driven by an electric

motor and usually contains a rechargeable battery bank and a small internal combustion

engine. The internal combustion engine still emits greenhouse gases, however only at a

fraction of the amount. In some of the latest hybrid vehicles, four wheel motors are

used (one for each wheel), and four motor controllers are used to control the torque of

each individual motor for optimal vehicle performance and control.

An alternative energy source which is very appealing is solar energy. Solar

energy is a continually advancing technology, and as photovoltaic (PV) solar cells are

being made more efficient, solar power is finding widespread use in applications such

as outback power supplies and grid connected PV arrays. A large contributor to the

increasing level of pollution is the household car, so solar cars were developed with the

vision that an ideal car could be built which could run solely from the sun for the

lifetime of the car, and never require fueling up. This indeed is a futuristic dream,

however the technology is fast approaching this stage.

1.2 Problem Specification

Design of a motor controller for the University solar car project has not been

attempted before. The new controller has incorporated a multitude of features which are

designed to make the drive system highly efficient and safer while providing a more

intuitive driver control. The new motor controller consists of a Hitachi SH1 7032 RISC

microprocessor operating at a clock speed of 20MHz accompanied by an array of

sensors and a high voltage inverter stage. The work performed in this thesis project

incorporates a number of different fields of work:

• Electronic Commutation : the switching of currents to the correct phase windings

in order to make the motor rotate and produce torque. This basic operation is

common for most types of motors. The brushless DC motor used for the solar car

1. Introduction

- 3 -

uses hall effect elements embedded in the motor to provide rotor position feedback

information (discussed in chapter 4).

• Waveform Shaping : by changing the pulse width modulation (PWM) ratio of the

output drive signals, two functions can be implemented simultaneously. Current

limiting is the process of regulating the phase currents in the motor to reflect the

torque commanded by the driver. Efficiency of the drive may be improved by

applying a weighted PWM signal to produce e.g. a sinusoidal output waveform

(PWM techniques are discussed in chapter 4).

• Sensor Technology : the motor controller has a number of sensors which provide

feedback to the software control loops. The sensors used in the motor controller

include current transducers for measuring individual phase currents, bus voltage

measurement, an integrated circuit temperature sensor for measuring heatsink

temperature and a thermistor for measuring temperature of phase windings (sensors

are discussed in chapter 5).

• Smart Control : the microprocessor is programmed to perform a number of

auxiliary functions so that the vehicle performs optimally and safely under all driver

input commands and environmental conditions. The following features will be

designed into the motor controller, and are discussed in greater detail in chapter 2:

� Regenerative braking capability

� Speed and direction of wheel output

� Cruise control function (performed by telemetry)

� Four quadrant operation

� Reverse at low speed only

� Soft start operation

� Low torque ripple operation

� Sinusoidal PWM phase current excitation

� Temperature monitoring of stator

1. Introduction

- 4 -

� Temperature monitoring of MOSFET heatsink

� Fault indicator

� Wide input voltage range

� Transient protection

� Fuse protection

� Diagnostic capability

� Cooling fan mounted to heatsink

1.2.1 Thesis Goal

The primary and most important goal of my thesis was:

“To design and construct a Brushless DC motor controller for the University of

Queensland solar car that performs motoring and regeneration at a very high efficiency.

The motor controller should also perform auxiliary functions that make the drive system

more robust, safer and easier to control.”

The controller should operate the motor with the highest possible efficiency

under steady-state operating conditions. Under abnormal conditions, the controller

should respond quickly to resolve the problem and resume normal operation to maintain

a high level of energy efficiency. On completion of the project, the motor controller

will be mounted in the solar car and be interfaced to the other electronic systems.

1.2.2 Motivation behind the Motor Controller and Motion Control

Many applications in today’s technologically advancing world require systems

with greater efficiency and more stringent operating specifications. An area in which

efficiency and reliability is an absolute must is motors and their control. Motors are

used in a vast variety of applications ranging from huge crushing mills to pinpoint

accuracy mechanisms in space applications. Some applications require motors to

operate in harsh environmental conditions, e.g. flammable gas leaks, where

1. Introduction

- 5 -

conventional DC brush motors cannot be used due to the risk of sparks forming between

the brushes and commutator. There are many types of motors available today, however

a discussion on each type is beyond the scope of this thesis.

One type of motor that boasts a very high efficiency and is very reliable is the

brushless DC (BLDC) motor. Unlike conventional DC brush motors, the brushless

motor, as it’s name suggests, has no brushes and requires extra electronic circuitry to

perform the job of commutation. The BLDC motor can be constructed in many sizes

and power ratings, and finds widespread application in many motor drives. The primary

motivation behind the thesis was to improve the efficiency and technology of the solar

car. The secondary motivation was related to the popularity of the BLDC motor and it’s

future applications. Factors such as high power to weight ratio and reliability will

definitely see BLDC motor technology improve in years to come. By studying how

such a motor is controlled, the capabilities of this motor are better understood.

1.3 Organization of the Thesis Document

The remainder of the thesis describes all work completed, problems encountered

and how these problems were overcome. Detailed descriptions including theory are

presented to support practical design choices. The following chapters form the body of

the thesis document, and may be summarized as follows:

Chapter 2, The University of Queensland Solar Car, presents first an introduction to

solar racing and how the event was first initiated, followed by a brief history of the UQ

solar racing car. The chapter then presents an electrical system overview in a typical

solar car, and how the main electrical components are interfaced. A short discussion

follows which outlines the importance of efficient systems on a solar car. The chapter

concludes by summarizing the existing drive system, then describing some of the

performance parameters of the new drive system.

Chapter 3, Motor Control Literature, presents a literature review of all relevant work

in the field of BLDC motor control. Useful formulas and control algorithms are

1. Introduction

- 6 -

extracted from the text and hi-lighted in this chapter. There is a complete list of all

references used in the bibliography section at the very back of the thesis report.

Chapter 4, Theory, provides the background material necessary to understand how a

brushless DC motor operates, and gives an insight of how to control such a motor.

Chapter 5, Hardware Design Stage, analyses the circuits designed and describes their

operation down to component level. Design formulas indicate how component values

were obtained. Mechanical factors are presented for construction of the motor

controller and when mounting into the car.

Chapter 6, Software Design Stage, describes the control algorithms implemented in

software which control the motor. There is a full listing of the code completed to date

in Appendix E.

Chapter 7, Results and Discussion, presents a discussion of the motor controller

project and the issues that emerged from such a project.

Chapter 8, Conclusions, concludes the document with a short summary of the findings

throughout the thesis project. Some possible future work is given as suggestions to

improving the motor controller. A final note is then given to the overall picture of solar

racing and where the future of such a technology is headed.

The author hopes the thesis document provides excellent reading and a useful

reference for any future work in motor control.

- 7 -

2. THE UNIVERSITY OF QUEENSLAND

SOLAR CAR

2.1 Solar Car Racing and the Races

Solar car racing first started out as a novel idea to investigate the limitations of

solar energy as a possible alternative to non-renewable energy sources. From that point

forward, solar car racing has grown in popularity and can be considered a sport, with

annual and biannual racing events being held all around the World. One of the more

prominent races is the World Solar Challenge, which covers some 3100 km from

Darwin to Adelaide along the Stuart Highway. Australian adventurer Hans Tholstrup

organized the first WSC in 1987, and it is now a bi-annual event held in October. The

Sydney City Power SunRace traverses the eastern coast of Australia from Melbourne to

Sydney and is the equivalent of the American SunRace. The American SunRace is the

largest solar event held in the United States. The World Solar Rallye in Akita, Japan is

held every year in July on a purpose-built solar racing track named the Ogata Mura

2.The University of Queensland Solar Car

- 8 -

Solar Sports Line. Many other countries hold solar related activities to promote solar

energy as a new energy alternative to existing fossil-based energy.

The solar car racing event is the most exciting part of solar car development.

Not only do competing teams have the opportunity to showcase to the world the ability

of solar energy, but have a lot of fun simply making the car perform optimally

regardless of impeding conditions. There is great satisfaction when seeing months of

hard work finally being paid off, as the solar car races through the finish line. The idea

of a solar car race is to reach the finish as fast as possible, obeying the race regulations

at all times to avoid time penalties.

For long endurance races such as the WSC, a convoy of cars accompanies the

solar car. One support vehicle usually has onboard computers and radio equipment for

data and voice interchange with the solar cars’ driver and telemetry system. Team

members ride in a scout car and place wooden boards over cattle grids so that the solar

cars’ tuned suspension is not put under great mechanical stress. In the 96 WSC,

SunShark had an RACQ representative who was able to lend assistance in mechanical

breakdowns. In races such as the World Solar Rallye in Akita, the racing track

consisted of a 30km round circuit, allowing no room for support vehicles. Telemetry

data, which was logged for an entire lap had to be transmitted in a short burst when the

car was in range of the receiving base station antenna. During the normal course of a

race, the drivers must be changed at regular intervals and a number of media stops are

usually anticipated.

There are two aspects that are essential for a highly competitive entry. A major

aspect of succeeding in a solar car race is to have a highly efficient and reliable system.

This can be accomplished by designing an aerodynamic structure made from

lightweight materials and choosing efficient electrical components. The other aspect,

which is equally important, is to have an effective race strategy. In a race situation, a

race strategy team determines an optimal speed to run the car at, depending on current

weather conditions (e.g. solar insolation, cloud cover, rain), past weather/race data (e.g.

rain patterns, road profiles) and vehicle parameters (e.g. battery state of charge, rolling


- 9 -

resistance). Most often, an unexpected weather pattern emerges or a critical breakdown

occurs. The strategy team must take into account these factors, and make a crucial “on

the spot” decision. Decisions such as these can decide the ultimate outcome of a race.

2.2 A Brief History of the UQ Solar Racing Car

The University of Queensland Solar Racing Car, commonly known is the

“SunShark”, was first conceived by a number of engineering students early in 1995.

Being only a concept and a few rough sketches at that early stage, a team decision was

finally made to build a solar car and enter it in the 1996 World Solar Challenge (WSC).

After 10 months of design and 8 months of intense construction work, the $140,000 car

was ready to roll. The WSC took Sunshark six days of racing in some of Australia’s

harshest outback conditions. The car finished in fifth place, won the silicon cell/lead-

acid battery class, and was presented with the award for technical innovation and

achievement from General Motors (GM) Holden.

A decision was made by the newly formed team early next year to participate in

the 1997 World Solar Rallye (WSR) in Akita, Japan. With only minor electrical and

mechanical modifications being made to the car in order to comply with race

regulations, the team and car were ready to compete at the Ogata Mura Solar Sports

Line in Akita. After 5 days of racing in sweltering heat, the car finished in identical

form as the WSC : ranked fifth overall and class winner of the silicon cell/lead-acid

battery category. Major electrical enhancements and some mechanical improvements

are currently underway in preparation for a large testing run near the end of 1998 and

the Sydney CitiPower Sunrace in January. The next WSC has been scheduled for

October 1999 and the team hopes to have a greatly superior car than in previous years

for this major solar event.

2.3 The Nuts and Volts of a Solar Car

A typical electrical system for a solar car is presented in Fig. 1.


- 10 -

Figure 1 : Block Electrical Diagram of a Solar Car

The central node of the electrical system is the high voltage (HV) bus. Physically it

may simply consist of a connection point or short strip of copper, however it is at this

point that the flow of current is distributed to all components. The main electrical

components are described in the next section.

2.3.1 Batteries

The primary energy source for the vehicle is the battery bank. The battery bank

usually consists of a number of individual batteries connected in series or parallel. Each

battery in the bank is typically 6 or 12V, and multiple batteries are connected in series

or parallel to obtain the desired system voltage. A single battery is actually made from

multiple “cells” contained within the battery housing. A sealed lead acid type showing

PhotovoltaicSolarArray

Maximum PeakPower Trackers

(MPPT’s)

BatteryBank

(120V DC)

HIGH VOLTAGE BUS

Telemetry andSupport Circuitary

PowerSupply

MotorController

BLDCMotor

Driver Controls andDriver Display

RadioModem


- 11 -

Figure 2 : A Sealed Lead Acid Battery

the internal structure is shown in

Figure 2. The overall battery

voltage is chosen depending on

the motor’s EMF constant and the

desired nominal cruising speed.

For the most efficient operation of

the drive system, the battery

voltage is chosen so that the

motor controller can operate with

minimal PWM (i.e. reduced

switching losses), at the

maximum desirable speed of the

car. In practice however, the

battery voltage, especially for

lead-acid batteries, fluctuates considerably around the nominal battery voltage, from full

charge to maximum discharge. For this reason, the nominal battery voltage is usually

chosen so that the lowest possible battery voltage is able to sustain a reasonably

competitive speed. An alternative solution to this problem is to implement a boost/buck

converter in the motor controller so that an optimal speed can be obtained for any

battery voltage. There are many types of commercial batteries available today. Some

examples particularly applicable for solar racing vehicles are sealed (maintenance free)

lead-acid, silver-zinc, lithium-iron and zinc-air. The SunShark solar car team chose to

obtain sealed lead-acid batteries due to ease of availability and relatively cheap cost.

One major drawback however is a relatively large weight/energy density ratio, and a full

set of batteries typically weighed in at 96kg. Each type of battery has different

characteristics (e.g. energy density/kg, charge/discharge rate) and uses, however a

comprehensive study of batteries is beyond the scope of this thesis.


- 12 -

Figure 3 : A Screen Printed Solar Cell

2.3.2 Solar Array

The capacity of batteries set out by race rules and regulations is too small for a

solar car to fully depend on during a race. Energy must be obtained from the sun by a

solar array to supplement the energy taken from the batteries. Under maximum

insolation levels, the solar array can sometimes supply ample energy, and the excess

simply flows back into the batteries. The solar array consists of a configuration of solar

photovoltaic cells, usually encapsulated to protect against the elements and damage.

The encapsulation of cells also increases the overall efficiency of the array. This is

achieved by carefully designing anti-reflective coatings and materials to maximize the

light energy captured. General categories of solar cells include amorphous, multi-

crystalline and mono-crystalline cells. Some types of solar cells include screen printed,

buried contact cells (BCC), laser-grooved cells and passive emitter reflective layer

(PERL). A screen printed mono-crystalline cell showing the fine metal fingers and

busbars which collect the energy

from the surface of the cell is shown

in Figure 3. The cell shown has a

rated efficiency of ~16.5%.

Commercially manufactured cells

are available with maximum

efficiencies in the order of 26%,

however cells have been produced

with peak efficiencies of 30-35%

under laboratory conditions. Solar

cells convert sunlight (photons) to

electricity (electrons) by the raising

of the energy level of electrons in

the crystalline lattice, and allowing them to move freely throughout the structure. Solar

cells are constructed from a semiconductor p-n junction, which allows current to flow in

one direction only, similar to the operation of a diode. The SunShark solar car team’s

first array contained 15.5% Sharp cells encapsulated in epoxy, giving a peak power


- 13 -

Figure 4 : Maximum Peak Power Tracker Module

output of 1kW. The new array should have higher efficiency cells and enhanced

encapsulation materials, with an expected power output of 1.5kW.

2.3.3 Maximum Peak Power Trackers (MPPT’s)

The output voltage of the PV array varies widely with changing sunlight

intensities, incident sunlight angles and PV cell temperature. As previously discussed,

the battery voltage may also fluctuate, and the PV array may be forced to operate at the

voltage depicted by the

battery. This can result in a

degraded power output from

the PV array, because the

voltage may not correspond

to the maximum power

point of the cells. The

maximum peak power

tracker (MPPT) modules

automatically hold the

photovoltaic (PV) panel at

it’s maximum power point voltage, while delivering the resulting maximum PV power

to the battery bank. It does this by electronically de-coupling the PV voltage from the

battery voltage by using a high frequency transformer and MOSFET’s. A MPPT

module is shown in Figure 4. The existing array had three MPPT modules

manufactured from the Australian Energy Research Laboratories (AERL).

2.3.4 Motor Controller

The motor controller is designed to convert the electrical energy obtained from

the batteries and solar array to suitable power waveforms to drive the motor. The motor

controller used in the solar car is designed to drive a Permanent Magnet Brushless DC

(PM-BLDC) motor. The driver becomes part of the speed regulation loop as the torque


- 14 -

produced in the motor can be controlled via controls in the cockpit. A more thorough

explanation of the motor controller is given in chapter 4.

2.3.5 Motor

The motors’ function is twofold: to convert the electrical energy to mechanical

energy when motoring and mechanical energy to electrical energy when regenerating.

There are a number of types of motors in use today, ranging from the induction,

switched reluctance, brushed DC and stepper motors. Each motor has a number of

advantages and disadvantages in particular applications ranging from large industrial

roller mills to accurate positioning control. The most popular choice for high efficiency

applications such as solar cars, is the permanent magnet brushless DC motor, or

sometimes known as a synchronous DC motor. The advantages of the BLDC motor

include:

• Very high efficiency characteristics over a large power range (98.2% recorded for

an optimized Halbach magnet arrangement).

• Require minimal maintenance, due to elimination of mechanical commutator and

brushes.

• Long operating life and higher reliability.

• No brushes means no arcing which can be paramount when working in flammable

gas locations.

• Number of motor geometry’s possible (e.g. interior permanent magnet or surface

magnet arrangements).

• High power density and torque to inertia ratio give a fast dynamic response.

• No brushes eliminates need for a high rotor inertia.

• Speed restrictions due to the traditional mechanical commutator are eliminated.

The construction and theory of the brushless DC motor is presented in greater

detail in chapter 4.


- 15 -

2.3.6 Telemetry Functions and Power Supply

The basic electrical system shown in Fig. 1 can be enhanced with the addition of

telemetry systems and support circuitry. The main aim of the telemetry system is to

calculate an optimal speed and/or power to run the car. One of the tasks performed is to

record data such as bus voltages and motor currents. The existing telemetry in the

SunShark solar car consisted of a signal conditioning board and telemetry board able to

transmit sampled data to the support vehicle via radio modems. It is envisaged the

support vehicle computers will be able to determine an optimal speed of operation, and

even take control of the car, by factoring in all relevant aspects which directly influence

the systems’ performance.

The power supply is responsible for converting the bus voltage down to supply

voltages for the circuitry. The current system converts 120V to +/-15V, 8V and 5V.

The power supply is usually a switch mode type to keep losses to a minimum.

2.4 Necessity for Efficient Systems

The photovoltaic array for solar cars is very dependent on weather conditions.

Although the sun has as much energy as a million hydrogen bombs, a fractional amount

of that energy actually reaches the Earth’s surface. Furthermore, the amount of energy

received from the sun by a photovoltaic solar array depends upon multiple factors such

as cloud cover, angle of incident sunlight, cell temperature and cell efficiency. Due to

the obvious difficulties in obtaining energy from the sun, any wasted energy (i.e. energy

that is not contributing to the forward motion of the car) is regarded as a limiting factor

on the maximum speed obtainable from the system. For the SunShark solar car,

approximately every kilogram of vehicle weight relates to rolling friction power loss

increasing by 1W. Mechanical systems and frames can be made lighter by using

different materials in an attempt to reduce rolling friction power loss. There are a

number of methods in which electrical systems can be made more efficient. Through

careful circuit design with energy efficient components, substantial power savings can

be made. The heating losses due to current flow in conductors can become substantial


- 16 -

in high power parts of the circuitry. It is advisable in this case to use over-rated cables

to help bring the conductor resistance and hence power loss down. Layout of

components is also important to reduce conductor lengths and parasitic inductive

elements. In a solar car race, the maximum velocity of the solar car is limited by the

efficiency of the system and race weather conditions. Since the weather conditions on a

race are at best highly unpredictable, in some instances a solar car may be fully reliant

on the batteries for power. At the end of the day, the performance of a solar car is

heavily determined by the overall efficiency of the system.

2.5 The Existing Drive System

The existing motor consisted of a PM-BLDC motor with toroidal flux. The

motor had no iron in either its rotor or stator and consisted of a number of cylindrical

magnets with poles opposing one another, fixed around the circumference of the rotor.

The winding were arranged so as to enclose the magnets of the rotor in a “C” or “U”

shape. The windings and the coils formed a toriodal shape, thus the name toriodal flux

(or T-Flux) motor. The back EMF waveform was of a sinusoidal shape, due to the

nature of its construction. The motor required a transmission system consisting of a

toothed drive belt. The motor was supplied from Lillington Manufacturing.

2.5.1 Controller Type

The nominal input voltage to the motor controller was 120VDC. The controller

used trapezoidal phase current excitation waveforms. A PWM chip (NE5568) was used

together with a ROM (N82S123AN) programmed with a commutation truth table to

decode the hall effect signals from the motor, and provide excitation to the correct

phases. All logic circuitry was supplied power using a linear 5V regulator, which has

an efficiency of ~50%. The inverter stage was a common three phase H bridge design,

using three paralleled MOSFET's (IRFP260) in one switch, i.e. a total of 18 MOSFET’s.

The MOSFET’s had transient suppressing metal oxide varistors (MOV) to clamp the

voltage over each MOSFET switch to a safe level. DC link capacitors (12 X 220uF


- 17 -

standard electrolytic) were used in the DC link. The MOSFET gates were driven by an

IR2130 3-phase bridge driver chip. All three lower inverter switches had a 20k ohm

resistor connected in parallel, which meant each time the upper switches were activated,

0.72 W power dissipation occurred. A simple shunt resistor was used to measure the

constant current in the DC bus, instead of in the DC link. Driver controls consisted of

two potentiometers: one to adjust speed and the other to adjust the current limit value.

A direction switch was also available however care had to be exercised when moving at

fast speeds to not bump the switch in the opposite direction, otherwise excess currents

would flow and destroy the controller and possibly the motor.

2.5.2 Performance Characteristics

The most undesirable aspect of the previous controller was the characteristic of

the driver control. The controller was speed controlled, which meant the driver had to

basically guess where to position the potentiometer for a certain desired speed. This

caused a lot of concentration by the driver as the speedometer had to be constantly

monitored and potentiometer adjusted to obtain the desired speed. Moreover the speed

ramp was not a linear function of potentiometer position, but had a slow response at low

speeds and a fast, uneven response at moderate to high speeds. This made fine

adjustment of speed a large problem. It was discovered that potentiometers are not

always fully reliable devices, and a number had to be replaced during the course of the

race. The controller experienced a number of IC faults during the 96 WSC race,

probably due to the high temperature levels. Care had to be taken if the hall effect plug

was to come out, because the controller would set the speed to maximum.

2.6 The New Drive System

The new motor is made by CSIRO/UTS and is of the permanent magnet type. The

motor features two rotors, has no iron loss and is of an axial field construction. The

motor is specifically designed to fit inside the wheel of a solar car which has a number

of distinct advantages over the original reduction belt system:


- 18 -

• All drive transmissions (e.g. indirect shaft coupling, chain, belt) are eliminated.

This can result in savings of up to 15%(dependent on drive train configuration) of

the total motor output energy in a conventional drive train arrangement, which

would have usually been lost as heat and noise.

• No need to replace broken belts/chains or dust entering transmission system.

• Better aerodynamic performance due to streamlined design.

• Motor can be sealed against dust and water

The technical specifications for the motor can be found in appendix C.

2.6.1 Additional Features

A number of improved features are to be designed into the new motor controller

to increase overall efficiency, reliability and safety:

• Torque Control Input : torque is directly controlled instead of a speed control,

which will make the driver control more intuitive. A handgrip will be used which

may be rotated in one direction for motoring and rotated in the other direction for

regeneration.

• Regenerative Braking : allows electrical braking whereby the solar car’s kinetic

energy can be reclaimed. Mechanical friction brakes will still be present for fast

stopping ability.

• Cruise Control Function : a feature which allows the driver constant speed or

torque operating modes. (performed by the telemetry unit)

• Four Quadrant Operation : meaning the motor can be driven throughout the entire

torque-speed plane, i.e. forward and reverse motoring/regeneration.

• Reverse Speed Limited : provides a safe reversing speed for better control.

• Soft Start : limits starting jerk which will improve handling and reduce tyre wear

due to wheel slip.

• Low Torque Ripple : advanced PWM modulation algorithms reduce torque ripple

to ensure smooth rotation at high and low speeds.


- 19 -

• Sinusoidal Phase Current Excitation : improves efficiency when interfacing to a

motor containing sinusoidally varying back emf and develops maximum torque

production.

• Fault Indicators : faults identified immediately by displaying fault codes when:

1. Temperature exceeded in motor stator windings (thermistor used).

2. Temperature exceeded in MOSFET heatsink (IC temp sensor used).

3. Overvoltage detected on HV bus.

4. Overcurrent detected (e.g. shorting power components)

• Diagnostic Capability : faulty components can be identified by running a number

of tests on different parts of the circuit using a microprocessor.

• Wide Input Voltage Range : 0-200V capability for different battery configurations.

• Transient Protection and Safety Devices : peripheral device for limiting inrush

current when connecting batteries and protection for power devices and

microprocessor devices.

• Fused Inputs : Protects circuitry from continued current draw.

• Cooling Fan : small fan mounted on the MOSFET heat sink to ensure extended

operation in extra hot conditions.

2.6.2 Performance Requirements

The new motor controller is designed for a more intuitive control interface and

safer operation. The controller will contain robust features and be fully self contained.

It is envisaged the overall efficiency of the system will be improved, and the average

speed of the car can be increased.

- 20 -

3. MOTOR CONTROL LITERATURE

An extensive literature search was carried out to review work completed

previously. A list of keywords relating to the topic for searching databases: e.g. motor

controller, electric drive, motion control was drawn up. A general WWW search

resulted in a number of results however I found many of the web sites were usually a

company trying to sell their product, and offer little or no technical information. The

WWW is a very convenient way of obtaining product data sheets. The main source of

information was books from the Physical Sciences and Engineering (PSE) library.

There is a reasonable selection of books in the library ranging from Power Electronics

to books specifically on motor drives and their controls. A comprehensive search using

the networked databases Inspec, Compendex, Engineering & Applied Science, National

Technology Information Service (NTIS), Current Contents and Computer ASAP was

also undertaken. This search resulted in some 32 journal and magazine articles relevant

to aspects on motor controllers.

Most articles contained an example of a motor and motor controller designed to

demonstrate a particular feature. The experimental setup was commonly explained by

3. Motor Control Literature

- 21 -

the use of diagrams. The circuit in most cases was put under a simulation and results

were compared with the actual measured values. Many of the articles obtained are from

the IEEE and IEE publications.

Reference articles [7] and [9] discuss a controller using MOSFET switches with

in-built current sensing (used IRC644, 14A cont. 250V). These MOSFET’s are referred

to by International Rectifier (IR) as HEXSense TM devices, as they contain integrated

shunt resistors, which can detect the current passing from the drain to the source. This

results in a more compact design and eliminates the need for external shunt resistors or

hall effect current transducers which results in an immediate weight saving. This type

of MOSFET was researched into, however none were found with the required voltage

rating. The only HEXSense TM devices that were found were of the 3-pin type (TO-220

case style). If 3 such devices were placed together in parallel to reduce on-state losses,

a total of 18 current readings would need to be converted using an analog to digital

converter (ADC), which would quickly clutter the available ADC channels on a

microprocessor. The controller mentioned in articles [7] and [9] can operate the motor

in all four quadrants of the torque-speed plane, i.e. forward and reverse motoring and

forward and reverse regeneration. Trapezoidal phase current excitation with 120 degree

switch conduction intervals are used so that current only flows in two of the phase

currents at any one time.

An important comment in [7] as to the position of the current sensors for current

feedback and regulation is made. The most simple method is a resistive shunt on the

DC bus. Although a simple and relatively cheap method of current detection, it cannot

detect dangerous circulating currents which may be developed in the phase windings

and power switches. This current build up can result in switch failure or

demagnetization of the rotor magnets. The only solution to this problem is to have

current transducers mounted in the phase windings so that the current may be monitored

and evasive action taken.

The controller in [7] and [9] contains separate commutation logic/control and

current regulation blocks. The commutation logic/control block was implemented with


- 22 -

the Motorola MC33034 brushless motor controller chip. The MC33034 has inputs from

the rotor position sensors and driver control (start/stop and forward/reverse), and has

commutation signal outputs which feed into the Harris GS601 HVIC half-bridge gate

drive chip. Current regulation is achieved by difference summing a current command

signal (from the driver), and the current feedback signal (from one of the lower

switches). This difference voltage represents the current error, and the GS601 driver

chip minimizes the error by varying the switches’ PWM duty cycle, effectively

regulating the current to the desired value. The current control algorithm is simple in

principle. In this case a fixed off time, TL is used. The PWM frequency is determined

by the following formula:

switches. theof timeoffT

supply, of tagesupply vol V

motor, theof emfback E

frequency, PWM where

11

L

S

==

==

−=

PWM

LSPWM

f

TV

Ef

The accuracy of the relationship described by the formula starts to deteriorate at

low speeds when the motor phase resistive drop approaches the magnitude of the back-

EMF. Another current regulation algorithm is briefly mentioned, namely holding the

total PWM frequency constant, so that the current ripple varies with speed.

The one and two switch active regeneration schemes are presented in article [9].

The two switch active scheme is preferred over the one switch active scheme at low

speeds as it is not as sensitive to the back EMF amplitude. Both methods take

advantage of the energy stored in the motor windings and transfer this energy back to

the supply. A simple speed detector circuit which works on the principle of providing a

pulse for every transition of the hall effect sensors is described. The frequency of the

pulses is proportional to the motor speed according to the following formula:


- 23 -

minute.per srevolutionin speedmotor v

6), (typically cycle electricper nscommutatio ofnumber n

polesrotor no.p

pulses, theoffrequency where

)( 120

===

==

=

rotation

rotation

f

Hznvp

f

Some articles such as [15], [19] and [32], discussed the developments in

brushless DC motors and described how a particular motor was built for the “Desert

Rose” solar racing car. The articles discuss an axial flux permanent magnet brushless

DC motor designed for an in wheel drive on a solar car. The axial flux geometry was

found to have advantages over the common radial flux geometry by reducing volume

limits and having the ability to change the air-gap between stator and rotor. Increasing

the air-gap increases the copper loss as the torque constant decreases, but decreases the

iron loss as the flux density reduces. The main author of these articles was Dean

Patterson of the Northern Territory University. Dean Patterson also has written a

journal article on the electrical system design for a solar powered vehicle [29], which

made interesting reading material as the system could be compared with our own

system and comparisons made.

Article [4] describes some common dc drive failures and how to design a control

system which can sense the failure and continue to operate normally. The results are

presented both using a simulation and measured results.

Articles [28] and [30] describe how to model electronically commutated

machines using the P-Spice simulation program. This will be very useful information

when experimenting with the inverter stage, and comparing measured results with

simulated results.

Article [24] describes the application of soft switching inverters in electric

drives. A soft switching, or sometimes known as resonant converters, eliminate

switching losses by causing the inverter switches to switch at zero voltage instants.

There are many different resonant converters available, however they all require extra


- 24 -

switches, inductors and capacitors to be arranged on the DC bus. To design a converter

of type is by itself a full thesis, so will not be further investigated. The current design

however, is flexible enough to allow future people to add a resonant converter if

desired.

Article [31] presents a bi-directional dc/dc converter which can control the DC

link voltage and control regenerative braking of an electric vehicle. An explanation

follows that describes how the converter can switch currents in both forward and

reverse directions. Motor current ripple is claimed to be reduced by constantly

changing the DC link voltage under different operating conditions.

- 25 -

4. T HEORY

4.1 The Permanent Magnet Brushless DC Motor

There are a number of configurations for the brushless DC motor, however all

operate on the same principal. There are three main components that make up such a

motor:

Stator Winding : The stator is usually wound in a three phase wye (or star)

connection. Three phase windings are usually sufficient to control most motors,

however more than three phase windings are common, and simply require additional H-

bridges and commutation circuitry. There is the option with the CSIRO motor to use

more than three phases as each phase is broken up into multiple sections. There is also

the option to connect the windings in a delta configuration, however this may introduce

unwanted circulating currents flowing around the windings. The stator of the CSIRO

motor is shown in Figure 5. Each of the three phase windings are distributed in a

4. Theory

- 26 -

Figure 5 : Stator Winding of the CSIRO Motor

Figure 6 : Magnet Ring of the CSIRO Motor

sinusoidal pattern around the

circumference of the stator

and are encapsulated in a

fiberglass resin. By winding

the phases in a sinusoidal

pattern, a sinusoidal back

emf voltage waveform is

produced between two

phases when the motor is

turned by hand. To obtain

maximum efficiency,

sinusoidal phase current

excitation must be applied to the motor.

Rotor Magnets : In conventional DC motors, electromagnets are used to create

a magnetic field. The rotor in a BLDC motor consists of rare earth magnets which

produce a constant flux (hence the name permanent magnet). One of the rotor magnet

rings of the CSIRO motor is

shown in Figure 6. The

NdFeB magnets

(neodymium-iron-boron) are

glued to the backing iron,

and are arranged in a circle

comprising 40 magnet

pieces (i.e. 40 pole motor),

in an alternating N – S – N

configuration. The backing

iron forms part of the

magnetic circuit. There are

two identical magnet rings

4. Theory

- 27 -

Figure 7 : Hall Effect PCB of the CSIRO Motor

which are placed on either side of the stator and are kept separated by special rims. The

stator will be held stationary and fixed to the trailing arm. Both rotor magnet rings are

fixed to the wheel rim, and rotate with the movement of the tyre.

Hall Effect Sensors : Hall sensors are a popular choice for rotor position

feedback in brushless DC drives, reasons being they are cheap and do not require

complex processing algorithms. Hall sensors are more suited for use with trapezoidally

controlled motors, as sinusoidal machines usually require a higher resolution sensor

such as a shaft encoder or transducer. The actual sensor is usually a N-doped InSb

semiconductor, which in the presence of a magnetic flux, an electromotive force causes

free flowing electrons to move to one side of the semiconductor which causes a

potential to form on the output terminals. In most hall elements manufactured, a

voltage regulator, amplifier

and schmitt trigger are all

integrated inside the one

device. The hall effect

sensors are glued to a PCB

which is located inside the

motor. The PCB can be

adjusted manually to align

the stator coil position with

the hall effect position. The

PCB with the hall effect

sensors mounted is shown in

Figure 7.

Three hall effects give output six different states for one full electrical cycle,

which is usually sufficient for most motor control applications. There are two possible

ways of positioning the hall effect sensors around the axis. The hall elements can either

placed at 60 or 120 electrical degree intervals (.i.e. the hall code changes every 60 or

120 electrical degrees). The hall effects to be used are configured to change every 60

electrical degrees. One electrical cycle is equal to 360 electrical degrees, and is defined

4. Theory

- 28 -

as when the hall sequence starts to repeat. The hall effect sequence can be represented

in Figure 8. Since the motor has 40 poles, for one revolution of the motor, each hall

sensor will experience 20 north's and 20 south's (i.e. 20 high and 20 low level outputs).

The mechanical separation of the magnets and hall effect sensors can be calculated

easily from knowing the number of magnets and poles. One mechanical cycle is equal

to one entire revolution of the motor or 360 mechanical degrees. One electrical cycle

repeats every 1820

360

poles/2 no.

cycle onein degrees electrical no. == mechanical degrees.

Figure 8 : Hall Effect Positioning Sensors

4.1.1 Electrical and Mechanical Parameters

The main electrical parameters of the CSIRO motor are presented in Appendix C.

Speed of Motor Calculation

Diameter of rear wheel = 510 mm diameter (approx.)

Circumference = (π)x(Diameter of rear wheel) = 1602.21mm.

Nominal Speed of Motor = 111 rad/s = 111x60/2π = 1059.97 rpm.

At the nominal speed, velocity of solar car is thus:

Speed of Car = (Circumference)x(Nominal Speed of Motor)x(60/1000000)

4. Theory

- 29 -

= 101.898 km/hr.

Electrical Parameter Calculation

The surface motor is described by the following formula:

T=kTI where

T = torque developed by motor (max. torque = 50.2 Nm, nom. torque = 16.2 Nm)

kT = torque constant per phase (0.39 Nm/A)

I = current through DC link (A)

i.e. for maximum torque, I = T/3kT = 50.2/(3x0.39) = 42.91 A

and for nominal torque, I = T/3kT = 16.2/(3x0.39) = 13.85 A

the motor can also be described by:

E = kE ωm where

E = back emf of motor (V)

kE = back emf constant (0.39 Vs/rad)

ωm = angular velocity (max. angular vel. = 300 rad/s, nom. torque = 111 rad/s)

i.e. for maximum angular velocity, E = kEωm = 0.39x300 = 117 V

and for nominal angular velocity, E = kEωm = 0.39x111 = 43.29 V

Battery Voltage Calculation

The battery voltage has to be chosen so that the motor controller may operate at near to

full PWM when running at nominal speed. The motor has a line-neutral RMS emf at

111 rad/s. Thus battery voltage required can be given as:

Vbattery = (L-N RMS EMF(peak))x(2)/(kmodulation)

= )15.1()2432( x = 105.74V, where

kmodulation = PWM factor (=1.15) due to modulation of the MOSFET switches.

Thus a battery bank of 120V should be sufficient and will leave a small amount for

overtaking.

4. Theory

- 30 -

4.2 Controlling a Permanent Magnet Brushless DC Motor

Throughout the thesis document, the following numbering pattern for

MOSFET’s in the H-Bridge will be as follows:

Figure 9: Numbering pattern for MOSFET’s in the H-Bridge

Each of the MOSFET’s contain an intrinsic diode which has a reverse recovery

time comparable to that of a discrete diode placed in parallel with the MOSFET. The

diodes will be referenced with the same numbering as the MOSFET’s, i.e. SW1 has a

corresponding diode D1, and so on.

4.2.1 Commutation

Commutation is the process of reading the hall effect sensor code, which gives

an indication of the position of the rotor. If the position of the rotor is known, then the

positions of the magnets are also known. To create a continuous rotation of the motor,

the correct phases must be switched on and off in the correct sequence so that the

applied voltage is in synchronism with the rotor position. Depending on the magnitude

of the current command, different magnitude torque can be applied to the motor. There

are two basic schemes of commutating a BLDC motor.

4. Theory

- 31 -

120 Degree Conduction

The 120 degree conduction mode switches MOSFET’s on for a length of 120

electrical degrees and off for 240 degrees. The relation between the MOSFET

switching states and hall effect codes is shown in Table 1. When the MOSFET’s are

turned on, they are not simply switched on and left on, rather they are modulated by a

PWM signal. The PWM signal varies in duty cycle depending on what current

regulation algorithm is being used. When the PWM signal is high, only two

MOSFET’s turn on at any one time, one from the high side and one from the low side of

alternate phases1. When the PWM toggles low, the low switch is turned off and the

corresponding high switch is turned on. This method is called synchronous

rectification, as it allows the current to flow through the paralleled high switch and

freewheeling diode, thus reducing conduction losses. A basic two pole motor is

presented in Figure 10 showing the rotation of the rotor magnets and the corresponding

flow of current in the motor windings and hall effect codes for 120 degree commutation.

Input Output

PWM H1 H2 H3 SW1 SW2 SW3 SW4 SW5 SW6

1 0 1 1 1 0 0 0 0 1

1 0 0 1 0 0 1 0 0 1

1 1 0 1 0 1 1 0 0 0

1 1 0 0 0 1 0 0 1 0

1 1 1 0 0 0 0 1 1 0

1 0 1 0 1 0 0 1 0 0

0 0 1 1 1 0 0 0 1 0

0 0 0 1 0 0 1 0 1 0

0 1 0 1 1 0 1 0 0 0

0 1 0 0 1 0 0 0 1 0

0 1 1 0 0 0 1 0 1 0

0 0 1 0 1 0 1 0 0 0

Table 1 : 120 Degrees Commutation Truth Table

1 Note : Both high and low MOSFET’s of the same phase are never switched on at the same time.

4. Theory

- 32 -

Figure 10 : 120 Degrees Commutation Mode

4. Theory

- 33 -

180 Degree Conduction

The 180 degree conduction mode switches MOSFET’s on for a length of 180

electrical degrees and off for 180 degrees. The relation between the MOSFET

switching states and hall effect codes is shown in Table 2. Similar to the 120 degree

commutation, a PWM signal varies in duty cycle depending on what current regulation

algorithm is being used. When the PWM signal is high, three MOSFET’s turn on at any

one time, either two from the high side and one from the low side, or two from the low

side and one from the high side.2. When the PWM signal toggles low, the side with

only a single switch active toggles and it’s corresponding switch is turned on. Once

again, synchronous rectification takes place as the current flows through the paralleled

high switch and freewheeling diode, thus reducing conduction losses. A basic two pole

motor is presented in Figure 11 showing the rotation of the rotor magnets and the

corresponding flow of current in the motor windings and hall effect codes for 180

degree commutation.

Input Output

PWM H1 H2 H3 SW1 SW2 SW3 SW4 SW5 SW6

1 0 1 0 1 0 0 1 1 0

1 0 1 1 1 0 0 1 0 1

1 0 0 1 1 0 1 0 0 1

1 1 0 1 0 1 1 0 0 1

1 1 0 0 0 1 1 0 1 0

1 1 1 0 0 1 0 1 1 0

0 0 1 0 1 0 1 0 1 0

0 0 1 1 0 1 0 1 0 1

0 0 0 1 1 0 1 0 1 0

0 1 0 1 0 1 0 1 0 1

0 1 0 0 1 0 1 0 1 0

0 1 1 0 0 1 0 1 0 1

Table 2 : 180 Degrees Commutation Truth Table

2 Note : Both high and low MOSFET’s of the same phase are never switched on at the same time.

4. Theory

- 34 -

Figure 11 : 180 Degrees Conduction Mode

4. Theory

- 35 -

4.2.2 Current Regulation

Since torque is proportional to the fundamental frequency of the current, by

controlling the current, torque is also controlled. All other frequency components

contribute to losses in the motor, inductors and controller. To form a closed loop

system, there must be current feedback from the motor as indicated in Figure 12.

Figure 12 : Current Feedback in a BLDC Motor

This feedback signal is subtracted from the desired current input from the driver,

and a current error then propagates to the controller. This provides a mechanism for the

controller to accurately current limit by trying to keep the current error as close to zero

as possible. The actual current limiting is achieved using a PWM scheme for the

switching MOSFET’s. The audible range for humans is approximately between 6kHz

and 20kHz, so a PWM frequency above ~20kHz is sufficient to avoid an annoying

whine when switching.

4.2.3 Trapezoidal Current Excitation

Trapezoidal phase current excitation is a basic way to control a BLDC motor.

The MOSFET switches are activated and use a constant PWM frequency when turned

on which produces a phase current of trapezoidal shape (hence the name) as shown in

the lower trace of Figure 13. One major disadvantage of a driving a motor with

BLDCMotor

∑ Controller

Current Feedback

CurrentError

CurrentCommand

-

+

4. Theory

- 36 -

trapezoidal current, is that there are many frequency components which make up a

trapezoidal waveform, and these components only contribute to losses in the motor.

Figure 13 : Torque Ripple in a Trapezoidal Machine

Figure 13 also indicates a waveform describing torque ripple which is

characteristic for a trapezoidal motor. The torque ripple can be attributed to two major

sources:

Motor Related Torque Ripple : causes the torque waveform to be rounded during the

commutation intervals. This is caused mainly by magnetic flux leakage paths between

adjacent rotor magnet poles. This torque ripple can be minimized by careful motor

design. See label 1 in Figure 13.

Inverter Related Torque Ripple : The first of this type of ripple is caused by a current

imbalance when current is being switched between active phases. Sharp torque spikes

can be produced and are experienced every 60 electrical degrees. Special PWM

switching techniques can be used to reduce this ripple. See label 2 in Figure 13.

4. Theory

- 37 -

The second inverter related torque ripple is directly proportional to the high

frequency PWM ripple in the phase currents and produces the fast torque oscillation

(see label 3 in Figure 13.). This ripple is not usually a problem because the inertia of

the solar car usually filters out the ripple.

At high speeds, phase current and motor torque can decrease abruptly when the

supply voltage equals the combined back emf of the two conducting phases. Continued

high speed operation is possible by gradually extending the conduction time from 120

electrical degrees to 180 electrical degree conduction.

4.2.4 Sinusoidal Current Excitation

Sinusoidal current excitation is an advanced method of driving a sinusoidally

varying back emf producing motor. By driving a motor with sinusoidally weighted

PWM phase current waveforms, less frequency harmonics are present in the phase

current waveform, thus an immediate reduction in losses occurs. As a result, larger

torque is produced for the same RMS current. Sinusoidally driven motors also

experience reduced torque ripple, the principal reason being that sinusoidal machines do

not experience the abrupt phase to phase current commutations that characterize the

trapezoidal machine’s excitation waveforms.

When controlling a motor using sinusoidal excitation, the input current

command must be split into two different currents:

Id or “direct” current is aligned with the permanent magnet flux linkage phasor λm.

Iq or “quadrature” current is aligned with the back emf phasor Ef.

These currents may be related by the following formula:

Ef = (p)x( ωr)x(λm) where p = no. pole pairs.

ωr = angular velocity of motor (rad/s).

and λm = PM flux linkage amplitude.

4. Theory

- 38 -

The torque developed in a sinusoidal motor can be expressed as :

[ ])(.2

3qdqdqme LLiii

pT −+= λ

where Ld, Lq are the stator phase inductance’s.

Under normal operation, Id is set to zero and Iq is varied proportionally to input

torque. In an interior PM motor, flux weakening can be used at high speeds. Flux

weakening is the process of increasing Id to oppose the magnet flux. This results in an

extended driving range at high speeds. In a surface magnet motor, flux weakening is

not feasible, as it would not have any effect on weakening the magnetic flux, and would

only reduce efficiency and increase current drawn by the motor.

4.3 Power MOSFET Device Characteristics

A MOSFET (or metal oxide semiconductor field effect transistor) is a voltage

controlled device as opposed to a transistor which is a current controlled device.

Diagrammatically, the MOSFET can be represented in the off and on state as depicted

in Figures 14 and 15 respectively.

Figure 14:Non-Conductring MOSFET[34] Figure 15:Conducting MOSFET[34]

When the device is in the off state, the drain is insulated from the source by the

p-type region, however when a potential voltage is applied to the gate terminal, current

is allowed to flow freely between drain and source. A MOSFET’s characteristic

waveforms at turn-on and turn-off is shown in Figures 16 and 17 respectively.

4. Theory

- 39 -

Figure 16:Waveforms at Turn-On[38] Figure 17:Waveforms at Turn-Off[38]

The device’s switching speed is largely effected by the size of the gate-to-source

capacitance. This capacitor has to be charged and discharged in one switch on and

switch off cycle. A summary of the turn-on sequence is as follows:

1) The MOSFET is initially turned off with no gate voltage present. At time 0, the

gate voltage reaches the threshold gate voltage, Vth, and drain current starts to rise.

2) Between time 1 and time 2 : Gate-Source voltage waveform deviates from it’s

original trajectory due to :

a) Series source inductance develops a voltage due to the rising drain current and

causes the gate-source voltage to decrease, and

b) The decreasing drain-source voltage is reflected across the drain-gate

capacitance. A discharge current flows through this capacitor causing an increase in the

capacitive load as seen by the driver. The voltage across the source impedance

increases thus causes a retardation of gate-source voltage.

3) Time 2-Time 3 : Drain current increases further due to the reverse recovery of

the free-wheeling diode.

4) Time 3 : The free-wheeling diode starts to support the drain-source voltage and

the rate of fall of the drain-source voltage is mostly dependent on the Miller effect. It is

at this point that the MOSFET has a maximum power loss due to a large current passing

through the device and a large voltage present across the device’s terminals. Due to the

falling drain-source voltage, the drain current settles out to the current determined by

the load and this causes the gate-source voltage to drop.

4. Theory

- 40 -

5) Time 4 : The MOSFET is completely turned on and the gate-source voltage rises

rapidly to the “open circuit” value.

Similarly, the turn-off sequence can be summarized as follows:

1) The MOSFET is initially turned on with a gate-source voltage present at time 0.

At time 1, the gate-source voltage reaches a level that just sustains drain current. The

drain-source voltage increases at a rate governed by the miller effect and the gate-source

voltage is kept at a constant level which reflects the drain current.

2) The MOSFET experiences a maximum power loss again at approximately time

2, where both a large drain current and drain-source voltage are present. When the rise

of drain-source voltage is complete, both drain current and gate-source voltage decrease

until drain current reaches it’s minimum value at time 3.

3) Gate-source voltage decreases past the threshold voltage to zero between time 3

and 4. The device is fully turned off at time 4.

There are two main power losses to consider when looking at the total power

loss in a power MOSFET. Refer to the MOSFET data sheet in Appendix B:

Static On Loss : The power loss due to the resistance between the drain and source.

This power loss decreases as more MOSFET’s are paralleled together.

Pon = I2xRx(duty cycle) = (13.9)2x0.022x50% = 2.13 W.

Dynamic (Switching) Loss : The power dissipated when the MOSFET is changing

conduction state. This loss stays ~ constant when paralleling other MOSFETs.

Psw = (Vds)x(Id)x(∆t)x(fsw) = 120x13.9x200x10-9x20k = 6.67 W.

Gate Drive Requirements : the gate drivers must supply enough charge to the

MOSFET gate to enable the gate-source capacitance to charge and the device to turn on.

The power requirements increase when more MOSFET’s are paralleled.

P = VxI = VxQxf = (10V)x(190nC)x(20k) = 38 mW.

4. Theory

- 41 -

4.4 Heatsink Considerations

It is extremely important that the electronic systems run as efficiently as

possible. This is particularly relevant for the high power systems, as usually a drop in

efficiency of only a couple of percent typically relates to an increased power

consumption of tens of Watts. A potentially large power sinking device is the MOSFET

H-bridge. The MOSFET has losses as described previously, and these losses are in

most cases directly formed into heating the device’s junction. As the MOSFET is made

to switch faster, the switching losses become the most significant form of heat

generation. There is also heat caused by increased conduction losses at higher output

powers. As the junction temperature of a MOSFET increases, Rds increases, Ids

decreases, however the overall loss will increase.

To this end, to keep the losses to a minimum, a heat-sinking system has to be

made. The package of the MOSFET is designed especially to conduct heat away from

the junction and to the ambient atmosphere surrounding the device. A simple heatsink

design was calculated to obtain a feel for the correct size heatsink required:

From the MOSFET data sheet in Appendix B,

TJmax = maximum junction temp. = 150 degrees C.

TAmax = maximum ambient temp. = 70 degrees C.

Rds = drain-source resistance = 0.022 ohms.

Imax = maximum switching current = 13.9 A.

td(on) = on conduction time = 30 ns.

tr = rise time = 12 ns.

td(off) = off conduction time = 55 ns.

tf = fall time = 12 ns.

fsw = switching frequency = 20 kHz.

RthJC = thermal resistance from junction-case of MOSFET = 0.26 degrees C/W.

RthCS = thermal resistance from case of MOSFET-heatsink = 1.5 degrees C/W.

RthJA = thermal resistance from junction of MOSFET-ambient = 60 degrees C/W.

4. Theory

- 42 -

Vmax = voltage to be switched at nominal speed = 105 V.

Vs = battery voltage = 120 V.

Pon = MOSFET on power loss = WxxRI ds 25.4022.0)9.13()( 22max ==

C/W. degrees 82.1825.4

70150maxmax =−=−=on

AJthJA P

TTR

RthSAmax = RthJAmax – RthJC – RthCS = 18.82 – 0.26 – 1.5 = 17.06 degrees C/W.

kmax = maximum duty cycle = 105/120 = 0.875.

ton = total on time = td(on) + tr = 60 ns.

toff = total off time = td(off) + tf = 67 ns.

τ = average on and off time = (ton + toff)/2 = 63.5 ns.

Thus Psw = MOSFET switching power loss =

.06.120105.632

9.13120.

2

. 9max Wkxxx

fIV

sws == −τ

(for an inductive load)

Thus TOTAL Power Loss = Ptotal = Pon + Psw = 5.31 W.

The maximum power the MOSFET can dissipate without a heatsink is:

Pdmax = .33.160

)70150()( maxmax WR

TT

thJA

AJ =−=−

If we use an aluminum plate with the following characteristics:

Thermal Conduction for Aluminium = λ = 2.08, Thickness of Al plate = t = 5mm.

Heatsink orientation factor = Cf = 0.43 (for a black anodized vertically mounted plate).

Area of both sides of heatsink = A.

RthSA = C/W. degrees 24.12)5.133.1(31.5

)70150()(

)(max

maxmax =+−−=+−−thCSd

total

AJ RPP

TT

The following equation can be used to describe the heatsink required:

C/W. degrees 6503.3 25.0

+

= ffthSA C

AC

tR

λ

Solving for A, we obtain A = 18.92 square cm.

Thus a 5 mm thick piece of aluminum with dimensions ~3X3 cm is required.

- 43 -

5. HARDWARE DESIGN STAGE

This section details the process in designing the hardware of the motor

controller. A similar device should be able to be constructed based on the information

given here. The motor controller consists of two PCB boards: a high voltage board and

a control board. Each board will be described separately.

5.1 Design of Power Stage

The power stage of the motor controller comprises all of the high voltage

components. It was decided to place all such components on the one PCB so that

potentially fatal high voltage kept confined to the one board, and didn’t have to be

routed across boards. The major components on this board are the MOSFET H-bridge,

driver circuits for the MOSFET’s and high voltage capacitors. There are a number of

auxiliary circuits also placed on the board. Most of them have something to do with the

high power section, therefore it was convenient to include these circuits on the same

board as the other components. These include bus voltage measurement, temperature

5. Hardware Design Stage

- 44 -

sensing of the MOSFET heatsink and two phase current sensing circuits. These circuits

are described in more detail in section 5.1.2. The PCB contained positive and negative

battery terminals on one end of the PCB and three phase connections on the other end.

The interface to the control board was through a 20 pin IDC connector. MOSFET drive

signals, power supply and sensor readings were sent through this connector, as

described in section 5.2.

5.1.1 Circuit Design

The most important issues to consider when designing the high power part of the

motor controller are, in order of priority, as follows:

1) Reduce Stray Inductance : Any conductor of finite length will possess some

form of inductance. Stray inductance, especially in a circuit switching large currents

(i.e. large di/dt), may slow down turn-off of the MOSFET’s and produce unwanted

oscillations. One method of reducing stray inductance, and in some cases stray

magnetic flux (i.e. EMI), is to minimize the effective loop in which current flows. In

other words, forward and returning current paths have to be as close to one another as

possible. The upper limit to reducing stray inductance is the use of bus bars : copper

strips, with a thin insulator sandwiched between them. The first design of the motor

controller inverter section used copper busbars with a thin insulate sheet between layers.

Power components were to bolt onto the copper, however it was later abandoned due to

difficulties in construction and insulating problems.

2) Reduce Power Loss : An envisaged source of power loss is to use PCB tracks.

A calculation waas carried out to determine the power loss in a PCB track (dimensions

4mm wide, 0.0034 mm thick and 80mm long), carrying 13.9 A of (nominal) current.

R = pL/A = 1.7x10-8x80/(0.0034x4) = 100u ohms.

Where p is the resistivity of copper = 1.7x10-8 ohm x metre.

This indicates that using a PCB with wide tracks should pose no problem with losses.


- 45 -

3) Low ESR Capacitors : The effective series resistance of a capacitor

contributes to the power loss within a capacitor as described in chapter 3. Standard

electrolytic capacitors may be used, however an alternative is “Unlytic” capacitors.

When compared to the standard electrolytic capacitors, the Unlytic capacitors feature

10X greater current density, 3X greater over-voltage protection, 10 year shelf life, dry

construction (electrolyte free), non-polar and good current ripple characteristics.

4) Heat Conduction from MOSFET’s to Heatsink : Effective heat-sinking on the

power devices will help keep the junction temperature at a controlled level, thus

reducing loss due to increased power consumption at elevated temperature.

5) Weight of Hardware : Any unnecessary weight simply adds to the weight of

the overall car and reduces performance. Although there is ample room within the

bulkheads of the solar car, the motor controller should be contained as the one unit.

6) Ease of Replacement of Components : An important issue when components

must be replaced quickly in a race situation.

5.1.2 Sensors

Refer to Schematic Listings in Appendix A for details.

5.1.2.1 Bus Voltage Measurement

The bus voltage circuit simply works from a voltage divider consisting of

resistors R37-R39 inclusive. The divider has a ratio of exactly 0.025, so that if

Vin=120V, Vout=3V and if Vin=200V, Vout=5V. Resistors R37 and R38 are not

simply replaced by a single resistor for safety reasons. If either R37 or R38 were to

become short circuit, the other resistor would still act as a voltage divider and prevent

the full bus voltage propagating through to the rest of the system and destroying

components. Capacitor C21 simply filters the input signal to remove any voltage spikes


- 46 -

which may be present on the bus. Diodes D4 and D5 are schottky type diodes and are

designed to clamp the input signal from 5V+Vd to 0V-Vd. The op-amp is configured as

a voltage follower and simply acts as a buffer. The output of this circuit under normal

operating conditions is 0-5V which corresponds to a bus voltage of 0-200V DC.

5.1.2.2 MOSFET Heatsink Temperature Measurement

The sensor consists of a LM35 precision centigrade temperature sensor. This

sensor is designed to output 10mV/degrees C. The sensor can work over the range –55

to 150 degrees C, however for the purpose of measuring the temperature of the motor

controller heatsink, a range of 0 to 150 degrees C was adequate. The RC network

consisting of resistor R39 and C22 is threefold. Resistor R39 with a value of 560 ohms

is designed to remove the effect of any cable impedance. This assumes the connection

between the actual sensor (on heatsink) and the signal conditioning circuitry behaves

similar to a transmission line. Capacitor C22 stops DC current causing self heating of

the sensor. Finally the RC network can be used as low pass filter to stop induced

pickup on the sensor wires interfering with the measurement operation. The op-amp is

configured as a non-inverting amplifier with a gain of 3.3333, so that if the temperature

of the heatsink is 150 degrees C, Vout=5V. It was felt there was no real requirement for

protection diodes, because the entire controller will probably be shut down if the

MOSFET heatsink temperature reaches 100 degrees C. Under normal operating

conditions therefore, the output of the circuit ranges from 0V to 5V which corresponds

to 0 to 150 degrees C.

5.1.2.3 Phase Current Measurement

The phase current may be measured directly in the phase windings or

alternatively measured in the DC link. The disadvantage in measuring the current in the

DC link is that circulating currents may build up in the phase windings and MOSFET

switches are not detected. It is because of this disadvantage that the current sensors

were placed in the phases. The various components considered were:


- 47 -

Resistive Shunt

Advantages : cheap, readily available, no power supply required

Disadvantages : non-isolated, relatively high power dissipation with large current draw,

becomes inaccurate if shunt becomes hot.

Hall Effect Current Transducer

Advantages : isolated, low power consumption, high current capability, accurate.

Disadvantages : expensive, require +/- 15V +/-10% typical for most models.

A hall current transducer was chosen over a traditional resistive shunt because :

• Cost was not a major issue

• Isolation was required and which meant using an isolated op-amp with a

conventional shunt resistor, and

• A power supply was already supplying +/-15V for another current transducer used.

To effectively measure the phase currents of all three phases, two identical

circuits were used to measure the phase currents of phases A and B. The current in the

third phase, C, was readily determined by applying the simple Kirchoff’s current law to

the star configuration of the phase windings (i.e. IphaseA+IphaseB=IphaseC).

The actual current transducer was an instantaneous type, capable of measuring

+/- 50A DC. By passing the current carrying conductor through the core twice, the

resolution of the sensor could be doubled, however the measuring range would also

halve (i.e. +/-25A would be measured). The supply rails for this sensor required +/-

15V to be supplied. This was not a large problem, however a current transducer that

simply ran from a single supply, say a supply already used by other devices would have

been an advantage, as it would have simplified the entire system. This could possibly

be a future thesis topic : to build a current transducer than could measure current in both

directions and run off a single supply. The sensor had two types of output: a voltage

and current output. The voltage output was used in this instance, and pins 1 and 4 had

to be linked to enable this option. The sensor’s output was a handy 100mV/A


- 48 -

Figure 18 : The completed Power Board

measured, so that the voltage output ranged from –5V(-50A measured) to +5V(+50A

measured). The analogue-to-digital converter on the SH1 only accepts an analogue

voltage of between 0-5V, thus the sensor’s output had to be scaled accordingly.

Resistors R44 and R45 are of equal value and form a voltage divider. In this

configuration, the input voltage to the op-amp still ranges from 0-5V, however 0V

denotes –50A measured, 2.5V denotes 0A measured and 5V denotes +50A measured.

Diodes D6 and D7 are schottky type diodes and are designed to clamp the input signal

from 5V+Vd to 0V-Vd. The op-amp is configured as a voltage follower and simply

acts as a buffer. The output of this circuit under normal operating conditions is 0-5V

which corresponds to a measured phase current of –50 to +50A DC.

5.1.3 Manufacture and Construction

Manufacture of this part was fairly straightforward. The PCB was designed

using Protel Advanced PCB CAD software. Due to the nature of high power on the

PCB, all high power tracks were set to 10mm wide where possible. Power supply

tracks to the driver chips were set to 1mm and other tracks set to 0.5mm. Pad sizes

were set to 2mm where possible so that components had a large area to solder to,

making the PCB more reliable. Since the final PCB was fairly simple and contained no

smaller tracks than 0.5mm,

the electronics workshop at

university was able to

manufacture the PCB. To

increase the current

capacity of the tracks, all

copper tracks were tinned.

A solder through lacquer

completed the manufacture

of the PCB.


- 49 -

Figure 19 : MOSFET Heatsink and Cooling Fan

The most important factor when designing the PCB was the layout of

components, especially the high power devices. A logical layout was used, having the

battery connection on one end of the PCB and the motor connections at the other end.

The finished design is shown in Figure 18. The position of the capacitors was not

critical, however were placed close to the battery input. Each MOSFET “switch” is

comprised of three MOSFETS placed in a row. A compact way of achieving this was to

mount them vertically in line with pins through the board.

Simple heatsinks made from aluminum were able to be manufactured by the

electronics workshop, as no

commercially available

heatsinks were suitable.

The entire heatsink was

black anodized so that heat

conduction was maximized,

and holes were drilled

through the heatsink in order

to take weight from the

heatsink. The heatsink is

insulated from the back of

the MOSFET case by using

strips of electrically insulating/heat conducting material. The cooling fan was mounted

onto an aluminum bracket and simply bolted to the PCB. Figure 19 shows the heatsinks

and cooling fan.

All high power connections to the MOSFET were placed under the board so that

no vias had to handle high current. All drive connections to the MOSFET was achieved

on the top of the board using surface mount devices(SMD’s). The final layout of

MOSFET’s resulted in a compact layout which helps to reduce the inductance due to

large current flow. The position of the Metal Oxide Varistors (MOV’s) was not critical,

however were placed close to the device they intend to protect : the MOSFETS. A


- 50 -

LM35 temperature sensor (case T0-92) was mounted to the heatsink by simply gluing

the case down. The TO-92 case style was preferred to a surface mount version

available, due to anticipated difficulties in mounting the SMD to the heatsink without

shorting the pins on the heatsink. Connections to the sensor were made through a three

wire connector which plugged into the PCB.

The driver chips were placed as close as practical to the MOSFET’s and

auxiliary components to the IR2110 were able to be placed very close to the IC. To

minimize inductance in the drive signals running between the microprocessor and the

driver chips, twisted pair cable should be used. Due to the way the PCB’s were

constructed, the two cables were made so that they ran close together. In the IDC20

plug which connects the two boards, each 2nd wire is connected to ground, so that every

signal wire runs next to a ground wire. Special attention is made on the PCB to have a

ground track running directly opposite a signal track.

5.2 Design of Control Stage

The control stage comprises the SH1 microprocessor and it’s auxiliary

components, serial communications, sockets for hall effect input and output to power

board and some signal conditioning circuitry. The serial communications port forms

two primary duties. Power is supplied from the power board through the serial link and

is then distributed throughout the controller. Voltage is supplied at +/-15, 8 and 5V.

The second is a half duplex communication link between the motor controller and the

telemetry board. The input signals from the telemetry board are as follows, and are all

represented as a digital word:

� Variable torque (motoring and regenerating) control : This signal originates

from a potentiometer mounted inside the driver handgrip control. A number of

other avenues regarding the handgrip were considered. Rotary optical encoders and

strain gauges were considered however were found to be either too bulky or

expensive. The potentiometer has a full swing of 5V, where 0V indicates full


- 51 -

regeneration, 2.5V standstill operation and 5V full motoring. There is a hysterisis

band centered at 2.5V embedded in software to allow for the handgrip to not return

to the zero center position. The diver display unit accordingly converts the full

analogue swing to a digital word scaled from 0x0000 to 0xFFFF. This data is sent

via the telemetry unit, which has the ability to change the torque signal to one which

is calculated from strategy software. The torque signal is then sent to the motor

controller for processing.

� Forward/Reverse Selector Switch : This switch forms part of the driver controls

and simply implies which way the driver would like to motor. This signal likewise

is sent as a digital word, where 0x0000 represents forward and 0xFFFF reverse.

� Brake : To indicate to the motor controller that the driver is pressing the brake

calipers, two miniature switches are inset into either caliper. When the driver starts

to place pressure on the caliper, the switches are activated via a simple strip of metal

which runs along the front of the caliper. This signal likewise is sent as a digital

word, where 0x0000 represents brakes not activated and 0xFFFF brakes activated.

A socket is also set up to accept hall effect inputs and thermistor resistance from

the motor. Power is supplied to the hall effect devices also via this socket. A 20-pin

IDC connector forms the interface to the other PCB as described earlier in the chapter.

MOSFET drive signals are output, two signal conditioned phase currents, bus voltage

and MOSFET temperature are input and power supplied to the other PCB on this

socket. A two pin connector is mounted on the board so that the fan may be

conveniently plugged and unplugged from the PCB.

5.2.1 Circuit Design

The schematic was drawn using Protel and was spread over three sheets (see

Appendix A for schematic printouts). The three sections will each be given a short

description:


- 52 -

5.2.1.1 Auxiliary Components and Power

The SH1 microprocessor requires a number of auxiliary components (Refer to

Sheet 2 of 4 in Appendix A). Capacitors C4 to C11 inclusive are simply decoupling

capacitors for the multiple Vcc pins on the microprocessor. Pins 85 and 86 are the

reference voltage pins for the analogue to digital converter and can be connected to an

accurate regulated supply if accuracy is desired, however it was not necessary in this

instance. The crystal oscillated at 20Mhz and capacitors C21 and C22 completed the

oscillator circuit. U1 is an undervoltage sensing circuit, and will drive the /RES pin low

when the supply voltage drops below 4.7V. Switch SW1 is the manual reset switch and

jumper JP1 is used to temporarily continually reset the SH1 when testing.

Power arrives at the motor controller from the serial port. The fuse and the

“trans-zorb” transient suppressor provides a degree of protection against continued

excessive current draw and overvoltage respectively. Capacitors C2 and C23 are

standard decoupling values for the 5Vsupply. Capacitors C3 and C24 are used in

conjunction with the cooling fan supply, and are designed to absorb the relatively large

current draw of the fan at startup. Decoupling capacitors for the +/- 15V supplies are

found on the other PCB.

5.2.1.2 Memory Board

The SH1 has no integrated non-volatile memory thus external memory is

necessary. The flash memory provides 128kB of memory space, and the RAM socket is

a means of downloading the code to the SH1. Capacitors C19 and C20 are simple

decoupling capacitors for the memory chips.

5.2.1.3 Input/Output Ports

The MAX483 with associated decoupling capacitor C13 forms the serial

interface chip which runs on the RS485 protocol. Two output pins are only needed to

interface the chip to the outside world. As a general rule, 100 ohm resistors are used in


- 53 -

series to the SH1 for all digital and analogue input and output lines. This simply

provides some protection for the microprocessor, as the resistors limit to some extent

any current into or out of the SH1. Resistors R1 to R10 inclusive perform this function.

The 20 pin IDC was wired exactly the same as the corresponding IDC on the other

board so that the boards could be mounted above one another and a cable could run

directly from one to the other. JP5 was included incase optoisolation was needed

between the SH1 and the driver chips. The optoisolation simply provides added

protection to the SH1, however it was felt this was not required and would consume

power unnecessarily. In the case the optoisolation is required, a patch PCB can be fitted

into the socket and optoisolation chips fitted.

The hall effect signals are a open collector output thus require pull-up resistors

to 5V. Resistors R26 to R27 inclusive provide this function, and a value of 1k ohm

resulted in clean output. Each hall signal is then filtered using a simple RC series

network before being cleaned further by a pair of schmitt triggers. Decoupling

capacitor for the schmitt trigger is C18. The motor temperature sensing circuit

consisted of a Wheatstone bridge and a LM358 op-amp configured as a differential

amplifier. The thermistor resistance formed one arm of a Wheatstone bridge along with

resistors R18 to R20 inclusive and R32. Variable resistor R32 is for fine calibration of

the output temperature. The resistor values around the op-amp were calculated as

follows:

The thermistor is a non-linear device, it’s resistance having an exponential response

approximately equal to:

Tthermistor eR 3977016.0= where Rthermistor is the resistance in ohms across the

thermistor terminals.

T is the temperature of the thermistor in Kelvin (273 +T =

temperature in degrees).

This is plotted in Figure 20, and shows the expected resistance values at key

temperature points.


- 54 -

Figure 21 : The Control PCB

NTC Thermistor Response

0.00

5000.00

10000.00

15000.00

20000.00

25000.00

30000.00

35000.00

1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97 103

109

115

121

127

133

139

145

Temperature (degrees C)

The

rmis

tor

Res

ista

nce

(ohm

s)

10k ohm @ 25 degrees C

34k ohm @ 0 degrees C

508 ohm @ 110 degrees C 194 ohm @ 150 degrees C

Figure 20 : Thermistor Response

5.2.2 Manufacture and Construction

The PCB was designed using Protel software and measured a small 137x69mm.

The PCB had to be

manufactured out of

University as 0.25mm

tracks were used and a

solder mask was necessary

for the surface mount

components. A silkscreen

was also printed onto the

PCB showing placement

of components. All

decoupling capacitors and

most resistors were surface

mount components. The cost of an IC socket for the SH1 probably would have been

comparable to the chip itself, so a surface mount version was used. The control PCB is


- 55 -

shown in Figure 21. The main issues when soldering surface mount components using

the soldering station were:

1) Hold the component down and place and sparingly place the solder paste over

the component pins.

2) Hold the hot air gun a fair distance from the PCB to dry the solder paste out to

avoid spreading the paste over the board and then move in closer and bring the paste to

melting point. The solder mask repels the melted solder and is attracted to the

component pins.

As an extra precaution, a piece of plastic was fitted under the fuses to avoid shorting to

any tracks or vias.

- 56 -

6. SOFTWARE DESIGN STAGE

Although the code has not been fully completed at the time of writing this thesis,

it should be working at the time of demo day. Software flowcharts are included where

necessary.

6.1 System Description

A block diagram of the system is shown in Figure 22. Block 1 represents the

driver controls. The torque input passes through a block with gain K before entering

block 2. The gain is simply a linear scaling factor to convert the commanded torque to

a commanded current. The other driver controls input to block 2 is the forward/reverse

switch and the brake command. Block 2 is an algoritm that determines the correct

division of current to the three phases. This algorithm receives position feedback from

the hall effects in the motor. Block 9 is a lookup table which receives the hall effect

codes and converts them to a sequence of MOSFET states using the commutation truth

6. Software Design Stage

- 57 -

table. The circles labelled 3 represent differentiators. The output is the difference

between the commanded current from the driver and the actual current sensed in the

phase windings at the ellipse marked 7. Block 8 simply calculates the current in all

three phase windings from the current feedback from two phase windings. The

difference in current from stage 3 is a current error, and is an indication as to the

discrepancy between the actual and commanded current. The three error signals are fed

into block 4 which determines the ITU register values. Block 5 produces the PWM

output signals which are generated using the ITU in the SH1. Block 6 represents the

MOSFET H-bridge which outputs a three phase waveform to the motor.

Figure 22 : Block Diagram of Control Algorithm

The controller must be able to drive the motor in all four quadrants. The quadrants can

be presented diagrammatically on the torque-speed plane as in Figure 23.

7

K

1

2

3

4 5 6

BLDCMOTOR

8

9

10


- 58 -

Figure 23 : A four quadrant drive

6.2 Main Program

The main program is the control loop which ultimately determines the switching

state of the MOSFET’s. A number of subroutines, some described below multitask with

the main program. The SH1 does not do full multitasking, but assigns a very small time

period to all subroutines and programs running, and keeps cycling through these.

The forward/reverse switch is first checked to determine the desired motoring

direction. The torque is sampled to determine whether the driver would like motoring

or regeneration. If regeneration is desired, the program skips to the regeneration

algorithm and commences regeneration. If motoring is desired, the program first checks

whether the car is moving above a certain base speed in the reverse dirction to the

motoring command. If this holds true, the car will simply coast until it is moving with a

low enough reverse speed, is stationary or is moving in the desired motoring direction.

When motoring, the program observes the hall effect code and phase currents and

determines the correct MOSFET’s to switch on.

Torque

Speed

ReverseMotoring

ForwardRegeneration(Braking)

ForwardMotoring

ReverseRegeneration(Braking)

12

3 4


- 59 -

6.3 Torque Control

As depicted in Figure 22, the driver inputs a torque control rather than a speed or

position control. The driver is effectively controlling the current into the motor. The

torque is converted to a current and the control will work with this value throughout the

control loop.

6.3.1 Regeneration

Why incorporate regeneration in the first place? Regeneration allows recapture

of most of the momentum of the car which is usually lost as heat and noise in the

friction brakes. The implementation of regeneration should result in a considerable

energy saving of up to 10%. This figure is probably optimistic and assumes all energy

generated can be stored immediately. Two major factors limiting the storage of energy

are the voltage across the capacitor must remain below 200Vand the charging rate of the

batteries. A highly efficient solution is the inclusion of a super capacitor able to store

large amounts of energy. A flat battery may be used as a temporary reservoir,

especially at the start of a race when overcharging batteries could pose a danger.

Another way is to use a resistive shunt which is modulated and simply dissipates excess

energy as heat, however this idea was knocked back in favor of the mechanical brake

calipers. If the amount of energy returned is less than expected, the rate of deceleration

may be too slow and friction brakes will have to be used in parallel with regen.

Regeneration is achieved by making the torque reverse polarity (i.e. cause to go

negative), so if the vehicle were travelling forward, the motor will want to turn

backward. The back emf of the motor is used to generate current flow back to the

batteries. Two regeneration schemes are shown in Figures 24 and 25.


- 60 -

Figure 24:One Switch Active Topology Figure 25:Two Switch Active Topology

Lab represents the combined inductance’s of phases A and B and Vab represents

the combined back emf of phases A and B. The circuit operation is described below:

One Switch Active Topology

All switches are originally open. S4 is closed, and due to the back emf of the

motor, current flows around the solid line which causes energy to build up in the phase

windings. When the upper current threshold is reached, S4 opens and assuming

Vab<Vs, current flows through D2 and D3 and back to the batteries as depicted by the

dashed line. When the lower current threshold is reached, S4 turns on once more. This

procedure repeats on the next active windings and the motor experiences regeneration.

Two Switch Active Topology

S1 and S4 are both turned on simultaneously, allowing current to build up in the

phase winding as depicted by the solid line. S1 and S4 both turn off at the upper current

threshold and current flows back through D2 and D3 to the batteries as depicted by the

dashed line. S1 and S4 turn back on once again when the lower current threshold is

reached and the cycle repeats again. The two switch active topology is preferred at low

operating speeds as the PWM frequency goes to zero at standstill using the one switch

active topology. Refer to reference [9] for more information on these regeneration

schemes.

6.3.2 Brake


- 61 -

When the brake is pressed, the program will halt and set the current input to a

large regenerative setting. A speed control loop determines whether the car has

completely slowed down, and stops the car from rolling back.

6.4 MOSFET Heatsink Temperature

A subroutine is dedicated to observing the heatsink temperature. If the

temperature goes above a set limit, the fan will be activated and greater current limiting

will be enforced. If the temperature exceeds an absolute limit, the controller will have

to be shut down.

6.5 Motor Temperature

A subroutine is dedicated to observing the motor temperature. If the

temperature goes above a set limit greater current limiting will be enforced. If the

temperature exceeds an absolute limit, the controller will have to be shut down.

6.6 Speed and Direction

A subroutine is dedicated to calculating the motor’s speed and direction of

rotation from observing the hall effect frequency and switching sequence respectively.

6.7 Commutation

A simple lookup table located in memory will be used to translate hall effect

codes into MOSFET switching states.

6.8 Bus Voltage


- 62 -

If the bus voltage exceeds a certain limit which approaches the maximum

MOSFET drain-source voltage, the MOSFET’s will have to be disabled until the over

voltage condition has subsided. This will most likely occur during regeneration.

- 63 -

7. DISCUSSION

7.1 Discussion

All hardware design was completed, however the software has not been

completed as of this hand in date. It is hoped this will be operational by the demo day.

- 64 -

8. CONCLUSIONS

8.1 Thesis ConclusionsThis thesis has described the building of a motor controller for a hub mounted

BLDC motor. The following conclusions have been attained throughout the duration of

the thesis project:

• Different types of motors available

• Hardware requirements for the controller

• Control Algorithms needed to drive a motor

8.2 Possible Future Work

Although the new motor controller is a large improvement over the previous

controller, providing a more efficient control of the motor, there are a number of

projects that may form a thesis or small project for any future people interested in the

area of motor control. The projects include:

8. Conclusions

- 65 -

A Resonant DC Link Inverter (sometimes called soft switching inverter): This

additional circuit can be added to the motor controller as it exists now, and has the

potential for a marked improvement in efficiency. A resonant converter is usually

placed between the battery or energy source and the input terminals of the controller.

The circuit usually consists of a power switch (e.g. MOSFET) and a number of passive

components (e.g. inductors and capacitors) and can be arranged in many configurations.

The resonant converter works in parallel with the switching of the MOSFETS in the H-

bridge, and it’s basic operation is to force the voltage on the bus to zero whenever the

MOSFET’s in the H-bridge switch off. As discussed in the Theory section, the most

significant power loss in MOSFET’s occurs at the instant of switch off, when there is a

large voltage across the switch and the switch is still conducting a large amount of

current. A useful place to start research on this topic is to consult references [13], [14],

[31] and [24] and the textbooks listed in the bibliography.

A Current Transducer : In the current motor controller, two current transducers are

used to measure the individual phase currents. One disadvantage with the sensors used,

was that they required dual supplies (+/- 15V DC) so that the output was able to swing

both in the positive and negative directions. This is particularly undesirable in most

cases when trying o interface with a microprocessor’s A/D port, which only accepts 0-

5V DC input. It should be possible to develop custom made transducers that work from

a single supply, (e.g. 5V DC), measure current in both directions and output voltage in

the range 0-5V. This sensor can then be used to measure all current sensing

requirements throughout the car, resulting in significant energy saving especially when

sensing large currents, as compared with the standard shunt resistor method.

Switching Algorithms and Advanced Features : The current design can be easily

modified especially in the code. New switching algorithms may be found which will

improve efficiency of switching. Other features may be added onto the board easily, as

each pin not used on the SH1 is brought out to jumper pins.

8. Conclusions

- 66 -

8.3 The Future of Solar Car Racing : The Big Picture

The development of the future’s clean vehicles (i.e. vehicles which do not emit

potentially harmful greenhouse gases) has a brighter future thanks to solar car vehicles.

Solar cars are, at the present day, simply primarily racing vehicles and have no direct

commercial applications. Their restriction is mainly from the limitations of the slowly

emerging technology, namely photovoltaic solar energy. It is envisaged however, that

the innovations and discoveries made through solar cars will spark new ideas and

products to develop the clean cars of the future. Of course the automobile industry will

not be the only organization to benefit from the spin-offs from solar cars. The work has

far reaching applications in solar houses, remote area power systems (RAPS) and space

applications to name a few examples.

Appendices

- 67 -

APPENDICES

Appendices

- 68 -

APPENDIX A: SCHEMATIC

AND PCB DESIGNS

Appendices

- 69 -

APPENDIX B: MOSFET

DATA SHEETS

Appendices

- 70 -

APPENDIX C: CSIRO/UTS

MOTOR SPECIFICATIONS

Appendices

- 71 -

APPENDIX D: MICROCOMPUTER

PROGRAM LISTINGS

Appendices

- 72 -

APPENDIX E: ACCOMPANYINGCOMPUTER DISK

MAIN PROGRAM

SCHEMATIC FILES

PCB FILES

Bibliography

- 73 -

BIBLIOGRAPHY

Books

[33] Crowder, R. M., Electric Drives and their Controls, Oxford : Oxford University

Press, 1995.

[34] Kenjo, T., Power Electronics for the Microprocessor Age, Oxford : Oxford

University Press, 1994.

[35] Mohan/Undeland/Robbins, Power Electronics : Converters, Applications, and

Design, John Wiley and Sons Inc., 1995.

[36] Williams, B. W., Power Electronics : Devices, Drivers and Applications,

[37] Rashid, M. H., Power Electronics : Circuits, Devices and Applications, 2nd Ed.,

Prentice Hall International Inc., 1993.

[38] International Rectifier, Control Integrated Circuits CIC-1, International

Rectifier, 1996.

Journal Articles

[1] Jahns, T. M., “Torque Production in Permanent-Magnet Synchronous Motor

Drives with Rectangular Current Excitation,” IEEE Transactions on Industry

Applications, vol. IA-20, no. 4, pp. 803-13, July/August 1984.

[2] Jahns, T. M., “Flux-Weakening Regime Operation of an Interior Permanent-

Magnet Synchronous Motor Drive,” IEEE Transactions on Industry Applications, vol.

IA-23, no. 4, pp. 681-9, July/August 1987.

Bibliography

- 74 -

[3] Bose, B. K., “A High Performance Inverter-Fed Drive System of an Interior

Permanent Magnet Synchronous Machine,” IEEE Transactions on Industry

Applications, vol. 24, no. 6, pp. 987-97, November/December 1988.

[4] Wallace, A. K., Spee, R., “The Simulation of Brushless DC Drive Failures,”

PESC ‘88 Record, pp. 199-206, April 1988.

[5] Alberkrack, J., “Selecting Brushless DC Motor Controllers,” Machine Design,

pp. 109-14, August 11, 1988.

[6] Morimoto, S., Takeda, Y., Hirasa, T., Taniguchi, K., “Expansion of Operating

Limits for Permanent Magnet Motor by Current Vector Control Considering Inverter

Capacity,” IEEE Transactions on Industry Applications, vol. 26, no. 5, pp. 866-71,

September/October 1990.

[7] Jahns, T. M., Becerra, R. C., Ehsani, M., “Integrated Current Regulation for a

Brushless ECM Drive,” IEEE Transactions on Power Electronics, vol. 6, no. 1, pp.

118-26, January 1991.

[8] Low, T., Lee, T., Tseng, K., Lock, K., “Servo Performance of a BLDC Drive

with Instantaneous Torque Control,” IEEE Transactions on Industry Applications, vol.

28, no. 2, pp. 455-62, March/April 1992.

[9] Becerra, R. C., Ehsani, M., Jahns, T. M., “Four-Quadrant Brushless ECM Drive

with Integrated Current Regulation,” IEEE Transactions on Industry Applications, vol.

28, no. 4, pp. 833-41, August 1992.

[10] Jahns, T. M., “Motion Control with Permanent-Magnet AC Machines,”

Proceedings of the IEEE, vol. 82, no. 8, pp. 1241-52, August 1994.

[11] Holtz, J., “Pulsewidth Modulation for Electronic Power Conversion,”

Proceedings of the IEEE, vol. 82, no. 8, pp. 1194-213, August 1994.

Bibliography

- 75 -

[12] Chan, C. C., Jiang, J. Z., Xia, W., Chau, K. T., “Novel Wide Range Speed

Control of Permanent Magnet Brushless Motor Drives,” pp. 539-46, 1995.

[13] Lia, K., Hsiao, C., “Modelling and Implementation of a Resonant DC-Link

Inverter Driving a Permanent Magnet Synchronous Servo System,” pp. 370-5, 1995.

[14] Chickamenahalli, S. A., Cathey, J. J., “A Variable Frequency Resonant

Commutated Converter,” College of Engineering Division of Engg. Technology and

Dept. of Electrical Engineering University of Kentucky, pp. 272-7, 1995.

[15] Patterson, D. J., Spee, R., “The Design and Development of an Axial Flux

Permanent Magnet Brushless DC Motor for Wheel Drive in a Solar Powered Vehicle,”

IEEE Transactions on Industry Applications, vol. 31, no. 5, pp. 1054-61, 1995.

[16] Melear, C., “Brushless DC Motor Positioning System,” Motorola Corporation,

pp. 466-72.

[17] Tuckey, A. M., Patterson, D. J., “Brushless DC Machine Controller for a Kinetic

Energy Tidal Generator – Progress in Implementation, and Results,” pp. 323-8.

[18] Gasser, S. C., Cochrane, I. C., “High Efficiency PWM-VSI for a Solar-Electric

Racing Car,”, pp. 353-8, 1995.

[19] Patterson, D. J., “Recent Advances in the Design and Construction of Axial Flux

Permanent Magnet Machines,” National Engineering Conference ‘The Darwin

Summit’, pp. 435-9, 1996.

[20] Caricchi, F., Crescimbini, F., Honorati, O., Napoli, A. D., Santini, E., “Compact

Wheel Direct Drive for EVs,” IEEE Industry Applications Magazine, pp. 25-32,

November/December 1996.

Bibliography

- 76 -

[21] Chung, S., Kim, H., Youn, M., “A New Instantaneous Torque Control of PM

Synchronous Motor for High Performance Direct Drive Applications,” IEEE

Proceedings, pp. 721-27, 1997.

[22] Gair, S., Hredzak, B., Eastham, J. F., “Design of Advanced Electric Drives and

Torque Control Methodologies for Electric Vehicles,” EVS 14 Conference, Transport

Research Institute, Napier University and Dept. Electrical and Electronic Engineering,

University of Bath, pp. 1-10.

[23] Sato, Y., Fujita, K., Yanase, T., Kinoshita, S., “New Control Methods for High

Performance PM Motor Drive Systems,” EVS 14 Conference, Fuji Electric Corporate

Research and Development Ltd., pp. 1-9.

[24] Chan, C.C., Chau, K.T., Chan, D. T. W., Yao, J. M., “Soft Switching Inverters

in Electric Vehicles,” EVS 14 Conference, Dept. Electrical and Electronic Engineering,

The University of Hong Kong, pp. 1-10.

[25] Kogure, S., Kawabata, Y., Yamada, E., Kanamori, A., “Development of Motor

Control Technology for Electric Vehicles,” EVS 14 Conference, Toyota Motor

Corporation, pp. 1-14.

[26] Bose B. K., Szczesny, P., “A Microcomputer-Based Control and Simulation of

an Advanced IPM Synchronous Machine Drive System for Electric Vehicle

Propulsion,” IEEE Transactions on Industrial Electronics, vol. 35, no. 4, pp. 547-59,

1988.

[27] Xu, C. H., Schroder, D., “Modelling and Simulation of Power MOSFET’s and

Power Diodes,” PESC ’88 Record, pp. 76-83, April 1988.

[28] Liang, Y., Gosbell, V. J., “Diode Forward and Reverse Recovery Model for

Power Electronic SPICE Simulations,” IEEE Transactions on Power Electronics, vol.

5, no. 3, pp. 346-56, July 1990.

Bibliography

- 77 -

[29] Patterson, D. J., “Electrical System for a Solar Powered Vehicle,” Recordings of

21st Annual IEEE Power Electronics Specialists Conference PESC ‘90, pp. 618-22,

1990.

[30] Fardoun, A. A., Fuchs, E. F., Huang, H., “Modeling and Simulation of an

Electronically Commutated Permanent-Magnet Machine Drive System Using SPICE,”

IEEE Transactions on Industry Applications, vol. 30, no. 4, pp. 927-37, July/August

1994.

[31] Caricchi, F., Crescimbini, F., Noia, G., Pirolo, D., “Experimental Study of a

Bidirectional DC-DC Converter for the DC Link Voltage Control and the Regenerative

Braking in PM Motor Drives Devoted to Electrical Vehicles,” IEEE Applied Power

Electronics Conference APEC 1994, pp. 381-86, 1994.

[32] Patterson, D. J., “An Efficiency Optimized Controller for a Brushless DC

Machine, and Loss Measurement using a Simple Calorimetric Technique,” Record of

26th Annual IEEE Power Electronics Specialist Conference PESC ’95 Atlanta, pp. 22-7,

June1995.

Internet Resources

[1] Component Data Sheets

The Electronics Mall

http://faraday.ee.latrobe.edu.au/~mg/ic.html

Chip Directory

http://www.ideal.net.au/chipdir/

A motor controller for solar car

Technology

internal combustion

hall effect

hall effect

queensland

switch active

reducing stray

solar powered

pwm signal