Electronics and Control Systems

An electric vehicle, like any other vehicle, needs some sort of control so that the speed can be regulated.

The simplest type of control is an off/on controller, but this doesn't really give much in the way of speed control - either nothing or everything! You could call it a digital controller if you like! However under certain circumstances, it does work. For instance, the original control system for the Stealth I and II both used this system.

The Stealth I simply used a heavy duty momentary push button switch located on the handlebar. It was connected between the battery and the motor. Although it worked well - for a short time - it soon became apparent that there were serious problems. Firstly, as there was no gentle take-off, the chain could snap if taking off from a standing start and the sprockets and other components could also be damaged. More importantly, because of the very high current drawn when the motor started up, the switch contacts could weld themselves together. This this meant that the Stealth I kept on going when the button was released!

The very simple control circuit of the Stealth I (click here for the full size image)

This type of controller is quite unsatisfactory and very unsafe. If anything like this is used in the Murdoch University eV Challenge the vehicle will be disqualified for safety reasons!

Someone better is a similar system but in which the switch activates a heavy duty relay or solenoid switch. In this system, the switch only carries the amount of current needed to activate the relay or solenoid. If the relay or solenoid is designed to take the current and voltage, then there is little chance of the system sticking into the 'on' position and so the system is reasonably safe.

The Stealth II had a similar control but used a heavy duty solenoid switch. This is much safer to use than the above Stealth I circuit. (click here for full size image)

This still has the disadvantage that it is only an off/on circuit and causes severe stresses in the motor, chain, sprockets and other drive-train components.

For proper control, an infinity variable control system is needed that can smoothly and gradually change from zero up to the maximum. This is achieved by using an electronic controller. The other main advantage of such a system is that it is much more efficient as there is no need to always go at the maximum speed (which can waste energy due to the higher wind resistance and rolling resistance) and this can give the eV a greater range.

Electronic controllers need to be specially designed to suit the particular type of motor - a brushed dc motor will need a different type of controller than a brushless dc motor uses. Have a look at the Motors page for more info about these.

The most common type of dc motor controller uses a technique called 'pulse-width modulation' and this can be used with both brushed and brushless dc motors, although the circuit design will be different with the different types of motors. Have a look at the Building an E-Bike article which has a brief explanation of pulse width modulation control.

To control a brushed dc motor, there are many circuit designs available and one is given in the Mini eV with plenty of go! article. Alternatively, kits are available from Dick Smith Electronics or Jaycar Electronics for $15 - $35. These circuits work well (see below for more info) - but if you want to really win the competition, you may need to find and build an even better controller!

Click here for full size image

Above: The Dick Smith Electronics Motor Controller Kit (K-3072) is based upon the design given in the June 1997 edition of Silicon Chip Magazine. It uses a TL494 PWM chip to switch a MOSFET transistor. This in turn controls the motor.

Below: The Jaycar kit is also based upon the Silicon Chip design.

IMPORTANT!!!

I made one of the Dick Smith kits and put it in the Stealth II. It worked fine - for about 10 seconds......

The problem was that the two MOSFET transistors are not rated at a high enough voltage and current. In addition, the way that they are mounted on the heatsink is very poor. They quickly overheat and cook themselves.

Firstly - the rating problems....

The MOSFETs that come with the kit are rated at 60 volts and 60 amps. Although the system may only be running at 24 volts, under certain conditions high voltage 'spikes' may be generated that exceed 60 volts. What happens is that the pulses of electric current generated by the controller go to the motor. The motor is mainly a large coil of wire - just like a transformer. As the pulse goes through the motor coil, it builds up a magnetic field (which makes the motor turn) but when the current pulse ends, the stored magnetic energy in the motor 'collapses' and generated a high voltage spike.

This spike only lasts for a few millionths of a second - but as it can exceed 100 volts can easily damage the 60 volt MOSFET.

The remedy to this problem is to use a better MOSFET - I ended up using 200 volt MOSFETS.

The second problem was over-current conditions. The motor in the Stealth II is a 350 watt electric wheelchair motor. Using a constant 24 volts dc, this means that it will only draw 15 amps from the battery. But when the battery is under heavy load - for example when it first starts accelerating - it actually draws much more than this, and if the energy is supplied in pulses, then each pulse must be even higher. It was found that the motor could easily draw over 100 amps.

Now this in itself is not a big problem - the two MOSFETS are each rated at 60 amps giving a total of 120 amps capacity. The problem arises that the silly little heatsink tabs that come with the kit just can't carry enough heat away and so the MOSFETS overheat and cook.

Using Ohm's Law, we can work out how much heat they must carry.

The MOSFETS have an 'on' resistance of 0.028 ohms. Now this may sound very low, but lets plug the values into Ohm's Law.

Ohm's Law states that:

Voltage = Current x Resistance

Now the power equation is:

Power = Voltage x Current

Combining these equations we get:

Power = (Current x Resistance) x Current

or

Power = I2R (in watts)

Now lets put in our values of 50 amps (50 amps per MOSFET) and 0.028 ohms and find out how much power is dissipated in the each MOSFETS.

Power = 50 x 50 x 0.028

= 70 Watts

or for the two MOSFETS this is 140 Watts! This is 2/5 of what is actually going to the motor! With this amount of heat being generated, if there is not sufficient cooling, the MOSFETS will heat up to hundreds of degrees and fail in only a few seconds!

How do we get around this problem? One way is to put a BIG heatsink on the MOSFETS to help carry the heat away. But because of the design of the circuit board, this is not really possible if the MOSFETS are mounted on the board itself. In the end I had to mount the MOSFETS directly on a big heatsink and then run wires from the board to the MOSFETS. It was messy, but it worked.

The other way is to choose better MOSFETS. I used MOSFETS with an 'on' resistance of 0.01 ohms. This dropped the heat produced from 140 watts to 50 watts.

If you decide to make an electronic controller, YOU WILL HAVE PROBLEMS UNLESS YOU TAKE THIS HEAT PRODUCTION INTO ACCOUNT!!!

Try getting the best MOSFETS you can afford and find! If you are at school, ask your school science technician. They probably have a 'RS Catalogue' or a 'Farnell Electronic Components' Catalogue. Other places to try are World Wide Electronics in George Street in Kensington. Have a look in the telephone book for these places.

For more information about speed controllers, have a look at the WWW pages at http://homepages.which.net/~paul.hills/SpeedControl/SpeedControllersBody.html

Thanks to Andrew Martlew, a student at Murdoch University for pointing out this one.