BATTERY ELECTRIC SPEED CONTROLBATTERY ELECTRIC SPEED CONTROL Ralf Bagnall-WildRalf Bagnall-Wild

Reliable slow running is the ideal, particularly

when working loose coupled goods trains.

IBEGINBYASKING , what do we want when we control our electric locomotives? My answer is that it should achieve the particular speed that we want, from a choice of speeds. Yet nearly all so called ‘speed controllers’ are actually voltage controllers. The mean voltage, ‘mean’ because it is often the full battery voltage pulsed on and off, determines the train

speed in a crude manner but not to my satisfaction. Voltage control is analogous to a 1930s car hand throttle. Notwithstanding the claims by some, ‘cruise control’ it is not. My simple system automatically varies the voltage continuously so that the perfect speed regulation ideal is more closely approached. It is simple, cheap and applicable to both radio controlled locos and to those controlled by the driver’s hand in the cab. Consider in the first case a theoretical constant speed moving train without friction or wind resistance, on a straight level track. Such a train requires no force to keep it moving. The point is illustrated by a curling stone sliding on swept ice. On the other hand, a train accelerating from the halt requires force as acceleration equals force divided by mass. So more motor torque (power to turn) is needed for acceleration than is needed for steady state running. In the next paragraph, the characteristics of the electric motors commonly used are discussed in the context of the needs of both acceleration and constant speed running. Most model locomotive motors have a permanent magnet stator ‘field’ in which a wire wound armature rotor operates. So all the motor current flows through the armature windings. Motor developed torque in such motors is proportional to motor current. If the current is doubled, the torque will also double. The amount of current depends in part on the voltage ‘pushing the electrons along’ but there are two voltages to consider. There is the supply voltage. (From the battery.) When the motor is turning, opposing this is a voltage generated by the motor because our motor is a generator also. The faster the motor turns, the greater the back motor generated voltage, the smaller the net voltage, the less the current through the motor, the less the motor generated torque. Equally, the slower the motor speed, the greater the torque. This ‘stiff’ characteristic is exactly what we want but is it good enough to meet the disparate needs for torque, for both acceleration and steady speed running, without varying the supply voltage?

BATTERYELECTRICTRAINS

The answer to the above question is ‘maybe’. The point turns on another question; does the train reliably self-start from halt and then go not too fast, when a constant voltage is applied to the motor? Does it do this when the train is heavy and the desired speed is slow? If the answer is ‘yes’, then voltage control is satisfactory. The closer the model is to the toy end of the scope of battery electric trains, the more likely the ‘yes’ answer.

16

GARDEN Rail
Right:This pannier tank locomotive is under the control of young

Jack who, after some practice could competently drive a locomotive

and train suing this system. This young man will go far.

Lower:Speed control fitted to a battery powered locomotive.

Photos: Ralf Bagnall-Wild

However, my minimum speed requirement is a scale walking pace. I want full battery volts to start from halt. I could reduce motor voltage after the start to give the required slow speed but this is not the better way to control. The better way is to adopt speed control instead of voltage control. To have speed control, it is necessary to measure speed. There are many established ways of doing this but most are either bulky or wasteful of power. I fitted a twenty-two bladed circular shutter on the motor shaft, and this cuts a light beam as it rotates. The pulse rate this generates is a measure of speed. This speed measurement mechanism is simple, cheap, reliable and free from friction. There is a windage loss but this is not significant. The diameter of the shutter is one inch. The following system would be the conventional way to achieve speed control but it is unnecessarily complex in my view. The pulse rate mentioned above would be converted into a voltage proportional to speed. This voltage would be compared with a second voltage proportional to the desired speed. Any discrepancy between the two voltages would act to modify motor volts, such as to null the speed discrepancy. But, my system described below is much simpler than this and its performance, whilst not matching what could be achieved conventionally, is satisfactory. Two electronic devices are used and I will describe what they do. The first is a ‘monostable’, which is simply a timer. Each time it receives a trigger pulse it switches on (becomes ‘active’) for a set period of time. It does not matter how short the trigger pulse is provided it lasts for at least a few millionths of a second, the monostable switches for the set time period. It does not matter how long the trigger pulse is, only one set time period is given per trigger pulse. If a second trigger pulse occurs during the active period, the monostable remains active and the timing restarts from zero. In this application, the trigger pulses come directly from the speed sensor. The monostable time period in this application is about half a millisecond but this time may be adjusted depending on how fast the train is to go. The second electronic device is a MOSFET. This is a highly efficient solid-state switch used in this application to switch the motor circuit on and off. The connection is made between monostable and MOSFET such that when the monostable is active the MOSFET is switched off. Power to the motor is only connected when the monostable is quiescent. The system is depicted in the upper part of the diagram.

KEY-FOBRADIO

It follows that the faster the train speed the more frequent the trigger pulses. The more frequent the trigger pulses the more the time the motor supply is switched off when compared to the time switched on. Thus, the faster the train, the less will be the mean motor voltage. Conversely, the slower the train, the greater will be the mean voltage. The variation of mean voltage with speed acts with the ‘stiff’ characteristics of the motor to

give good speed regulation. The point is illustrated in the lower part of the diagram. It should be noted that when the train is stationary, no matter what the monostable time period set to give the desired speed, full battery volts are applied to the motor. Also, should the train go much too fast, there is a speed above which no volts go to the motor. Obviously, the longer the time period set, the slower the train will run. The monostable time period is set by an external resistor that can readily be adjusted, for example by some switches on the locomotive or by the radio control. Is the speed control system driver friendly? The accompanying photograph shows my 0-6-0T locomotive being driven by my grandson Jack, aged three years and six months. He has the keyfob radio transmitter in his hand with which he can select a speed. The key-fob sequences the locomotive speed from fastforward through halt to fast reverse. There are three speeds in each direction, the slowest being scale walking pace. So it is a sequence of seven commands. This was no contrived picture; he was at it in the cold for thirty minutes or more. After he had some practice, some rolling stock was attached as a trainload. If a three and a half year old can manage the system, it cannot be so difficult. The big advantage is that irrespective of the initial train speed, and, within limits, the incline, the curve, the weight of the train and the charge state of the battery, when the driver demands a speed he gets it.

GARDEN Rail

17