Fullscreen HTML5Med Res (??MB)HTML4

- ctr,c Electronic control circuitry has a big role to play in modern electric vehicles. We look at the brain that controls the drain. By GERRY NOLAN limb in, switch on and start a conventional (ICE) motor ' vehicle and it usually settles into a steady idle. This is because the engine controller is set up to make it do just that. Basically, the accelerator is sprung to the fully closed position and idle jets in the carburettor control the fuel/ air mixture ratio and amount to produce the idle. If the engine is cold, the choke (manual or auto) is used to vary the ratio and quantity. The same result is achieved in a C different way with fuel injected vehicles. By contrast, if you climb in and switch on your electric vehicle, nothing happens until you press the accelerator. When you press the accelerator in your ICE vehicle, you increase the amount of fuel/air mixture introduced to the cylinders to be burnt, thereby increasing the amount of energy released to be converted into power and VROOM! VROOM! - engage low gear and away you go. In your electric vehicle, as you accelerate abruptly away without the benefit of gears, the complete lack of VROOM! VROOM! will make you feel even more strongly that what is happening shouldn't be happening. EV motors and controllers are so closely combined that discussing one without the other is difficult so we have arbitrarily decided to discuss controllers first and then motors in the next instalment.. Theoreticaliy, the power control of an electric vehicle is as simple as controlling the volume on your radio or television. But what about the losses Virginia? Obviously, using a whacking great rheostat would result in enormous power losses, especially at low speeds. Remember, we're talking several hundreds of amperes here. Even racing model electric cars have up to 160A pulse ratings for braking. Something with a little more finesse is called for and now power control- lers are almost all based on some form of digital, solid state circuitry, many of which use MOSFET inverters with more and more using microprocessors to completely automate motor control. Much of recent electric vehicle development is focussed on the use of laptop or small built-in computers for the collection of data over the whole charging and running regime of the vehicle. This data is then used to write software that will "tell" the vehicle what to do when particular demands are made on it. When enough data has been collected, it can be used to program a microprocessor chip to carry out all of the requirements automatically and efficiently. What are the requirements? Perhaps the most important requirement is that the control system responds quickly and smoothly to the driver's signals for higher or lower speed. While it is doing this, it should provide overload protection for the battery, drive motors, drivetrain and, not the least important, for itself. It must be able to do this while using the minimum energy, with the minimum losses and, at the same time, maximise the energy available for the vehicle and reduce battery and motor losses. Because of the very rapid acceleration capabilities of electric vehicles, the power controller should also have Pt.3: motor control - the part 6 SILICON CHIP ductors while the motor is running, the main methods of control are field and armature control, or a combination of both. DRIVER ELECTRICAL INPUT, EG ACCELERATOR PEDAL DC motor control '\V SPEED OR P0SITION / CONTROL LOGIC ' i/ ' CURRENT LEVEL \ / , I/ \V DRIVE SHAFT ' / POWER CONTROLLER '/ DRIVE MOTOR I Block diagram of electric vehicle (EV) motor control Fig .1: the motor and its controller are closely interrelated. If the driver wants to increase speed, for example, he presses on the accelerator and the control logic senses this via the output from a small potentiometer controlled by the accelerator pedal. This 'increase power' signal goes to the power controller, which increases the power to the motor. As the motor speeds up, a tachometer sends a signal back to the control logic where it is compared with the original input signal. When the motor speed reaches the desired level, the input and feedback signals will balance each other and no further 'increase speed' signals are sent to the power controller. an inbuilt "high-pedal lockout." This is a time delay so that, if the accelerator pedal is accidentally knocked down hard from fully off, the vehicle will not jerk away and injure someone. Also very important is the regenerative braking capability of the controller. This is its ability to enable energy that would normally be dissipated as heat during braking to be used to recharge the batteries - a vital factor in increasing vehicle range. How the controller works Essentially, the control logic senses the speed required by the driver, from the position of the accelerator pedal, and adjusts the amount of power going to the motors to achieve this speed. The block diagram of Fig.1 will give you a clear picture. The torque, speed and regeneration characteristics of electric motors are primarily governed by the following factors: • armature current; • magnetic flux per pole; • number of armature conductors; • number of poles; • armature speed; and • field current. As it is impractical to change the number of poles and armature con- played by electronics One way to achieve this is to vary the resistance of the armature or of both the armature and field windings by switching resistances in and out of the motor circuit with contactors. This may be achieved manually or through sensing the motor speed but, either way, the discrete changes in voltage produce a jerky motion. And of course, it produces high FR losses. A more efficient but no less jerky way is to use contactors to switch the batteries into different combinations of series and parallel to provide more or less power to the armature. These methods are rarely used now, even for model electric vehicles. High power solid-state controllers offer the most practical, reliable and efficient method of motor control, through pulse width modulation (PWM), frequency modulation and a combination of these two techniques. The PWM technique uses a constant DC voltage which is "chopped" into pulses of varying widths (see Fig.2). At low speed, the ON pulses are quite short in relation to the OFF pulses, while at high speed they are proportionally longer, right out to top speed where they are effectively 100% DC. Frequency modulation maintains a constant pulse width but varies the rate at which the pulses occur until again, at maximum speed, there is practically no time between the pulses. A combination of these two techniques may also be used. Until recently, power transistors, thyristors, SCRs and bipolar junction transistors (BJTs) were used to achieve the high switching rates needed for both PWM and frequency modulation methods. In the last few years, the development of high powered metal oxide semiconductor field effect transistors (MOSFETs), with much faster switching times and the ability to be switched directly by logic gates, has made them the preferred option. One of the reasons David Gosden of Sydney University opted for MOSFETs, despite the availability of much higher powered and cheaper thyristors and bipolar junction transistors, was because of their superior MARCH 1991 7 Fig.2(a): a constant frequency, variable pulse width (PWM) waveform. The motor slows down as the pulses become narrower & speeds up as the pulses widen. VOLTAGE VARIABLE - k: , - Tfixed ' / - - > TIME A switching capabilities. Switching times are an order of magnitude shorter than BJTs, allowing the switching frequenci es to be increased to over 15kHz. This gets it out of the unpleasant 2-BkHz range of BJT inverter switching frequ encies and al lows much quieter vehicle operation. AC motor controllers Although new developments in DC motors and controllers may change the situation at any time, at present the trend is to us e AC motors for electric vehicles. The General Motors Impact is powered by two 3-phase , AC inductiontype motors, each driving one of the front wheels. Together they develop 85kW at 6600rpm and the full torque of 1.27Nm over the whole speed range. Inverters convert DC from the batteries to AC and the whole system, which is capable of handling up to 100kW, weighs only 28kg. Pulse width modulation (PWM) and frequency modulation inverters (FMI) , and various combinations of the two , are both used in AC motor controllers. Braking During mechanical braking of a vehicle, kinetic energy is dissipated as heat in the brakes, which of course wear out over a period of time. In an EV, the life of the mechanical brakes can be increased by using the drive motor/s as a generator supplying a resistive load, thus providing a braking torque to the wheels. This type of braking, in which the vehicle's kinetic energy is dissipated in resistance, is known as "dynamic braking". Both mechanical and dynamic braking dissipate the kinetic energy as heat, which is therefore wasted. When the vehicle is travelling faster than the speed required by the driver (that is, over-running), the drive motor/s may be used as generators to convert a part of the kinetic energy loss to electrical energy which is then used to recharge the batteries. This is "regenerative braking" and is used in all modern electric vehicles in the forward direction and by some in the reverse direction. Regenerative braking is not an option that is available to internal combustion engine vehicles and, in a world where non-reusable energy resources are perceived to be running out, any means by which energy wastage can be limited obviously has valuable advantages. Nevertheless, because there is a lower speed limit to the use of regenerative braking, all EVs also have a mechanical braking system similar to that used in conventional vehicles. Energy management This photograph shows the solid state control & inverter circuitry in the Sydney University Suzuki Carry Van. (Photograph by Robert MacDonnell) . 8 SILICON CHIP An important function of the EV power control system is to provide the driver with information about the amount of energy remaining. As mentioned in the previous instalment, VOLTAGE Fig.2(b): a variable frequency constant pulse width waveform. The higher the pulse frequency, the higher the motor speed. FIXED ~ T variable > TIME B this is not quite as simple as providing a battery state-of-charge indication to replace the current fuel -remaining gauge. Some of the things an energy management system (as relate d to batteries) should do are: • provide battery state-of-charge indication; • control the maximum current drawn from the battery; • control the· depth of discharge; • control charging characteristics; The Solar Star, which recently broke the world-speed record for a solar/electric vehicle (see story this issue). actually provides a range-atpresent-rate-of-discharge readout by using a small on-board computer. When all the bugs have been worked out of the software being used in this , it is hoped that it wUl be developed into a microchip that can be used in off-the-shelf instruments . As mentioned above , microprocessors are used in conjunction with computers to facilitate the testing of various motor control strategies and combinations of strategies with few or no hardware changes. For example, Sydney University uses an Intel 80C196 processor in conjunction with a laptop computer. It has operated the motor in such a way that the stator currents are determined by the torque demand and, in another series of tests, where the motor is operated under torque control. It is safe to assume that, by the time EVs are out of the experimental stages and accessible to the motoring public , every aspect of the vehicle con- trol will be under the direction of microprocessors . Even the steering signals from the driver may be modified according to speed, gradient and so on, p erhaps even to the extent that the controller will automatically slow the vehicle down if it is going too fast for the amount of steering input indicated. The Solar Star already has a 'cruise control' selection to optimise energy usage for a given speed. Once we realise that, being electric , the EV is a fertile fi eld for all the clever ideas already developed in electronics - that we can control every aspect directly and not be restricted to controlling various valves, diaphragms an d pumps - deve lopments in EV control will take a quantum leap. SC VOLTAGE I~ ton Fig.2(c): the waveform from a variable frequency/ variable pulse width controller. VARIABLE - k T variable ' --> TIME C MARCH 1991 9