Tapiola Parish Model Railway Club

Why not use ordinary variable transformers?

Most of European model railway suppliers bundle variable transformers with their starter kits. It's heart is a double insulated transformer with secondary windings so arranged that it may be tapped with a slider attached to the speed knob. The latest versions with special slow speed controls (like Fleischmann MSF) have two sliders attached to the knob to enable half wave rectification at slow speed.

If these are compared with rheostat-controllers/throttles, the variable transformer has much lower internal resistance, thus the output voltage is not so current dependent. With rheostats the less current the motor takes, the more voltage is applied. With conventional motors, having permanent magnet field and iron core armature, the current needed to to start the motor is higher than the current needed to keep it in motion. With rheostat controller, the speed control must be advanced quite far to get the loco into motion. Once the loco starts moving, one should adjust the speed setting back, in order to prevent the loco from speeding. If you have such a controller and have ammeter attached to it, you can easily prove this to be true! With rheostat we don’t actually control the voltage the motor gets or sees, but the current passing through it. As the rheostat is in series with the motor, the current passing through the both components is roughly the resistance of the motor and rheostat divided by the power source voltage (yes -- this slightly simplified). As the resistance of the dc-motor reduces as the motor starts to rotate, the resistance of the circuits becomes lower. Now due to fallen resistance in the circuit the current rises and the voltage accross the motor becomes higher and thus speeds the motor even more.

And now, why not use the variable transformer? Firstly, we wish to be able to follow the train, and that is quite impossible with a huge transformer. The second reason was money! We can make the hand held electronic controllers from scrap box parts, and only buy a cheap car battery charger at less that one third of the price of a toy train transformer! The third reason is "bells and whitsles dept." — the desire to have all kinds of nice things like inertia/momentum.

The basics of a dissipative controller

Potentiometer P1 is a linear 10 kilo-ohm potentiometer. It is connected as a voltage divider. As the graphical representation shows it has a resistor which is tapped with a slider. The pot can only deliver 1/8 W of power, so we need an amplifier.

Enter: power transistor T1.

And now the Darlington connecton!

Inertia/momentum

By adding a capacitor and an extra resitor (or adjustable resistor, like the speed control) we can achieve this. The voltage we just a few lines ago got from the speed pot could be taken from an electronic reservoir, capacitor. We can feed or drain it with our speed pot, and the rate of drain or feed is adjusted with P1, the inertia/momentum control. If we got sufficient amplification from our darlington amplifier, the capacitor could be left without a connection to the speed pot, and our train would coast at only slightly dropping speed. (yes, this is the so called the “memory function”). We could replace the speed pot with push buttons, the other would charge the capacitor trough a resistor and the other would drain it. With a connection like the one below we could get two acceleration rates and two braking rates quite easily.

It would be a good idea to add yet another button as an emergency brake, that would short the capacitor almost instantly.

The problem assocciated with this kind of brake operation is that as the speed becomes lower the rate of deceleration becomes less too. This is often referred as the capacitor discharge curve. If we compare this with the full scale world it should be quite the opposite. Why does this rate of deceleration become less as speeds goes down? It is the fact that as the voltage in capacitor becomes less, the voltage across the brake resistor becomes less, hence the current through it - -- the capacitor discharge current -- will become less too. As a remedy one could connect the other end of the brake resistance to a point that is constantly certain volts lower than the voltage in the capacitor. There are numerous methods for overcoming this effect in model railway electronics publications.

What about the overload trip!

If we place a resistor in the current path, and measure the voltage across it, we get the current that flows in the circuit. By placing a resistor R4 in the ground path like in the picture below, we may monitor the voltage across it with a transistor.

As we learned from above, the transistor will only work if it is allowed to tax 0.7 volts across base-emitter. If the current in the motor circuit rises over one amp, the voltage across the resistor R4 will be over 0.7 volts, and the transistor valve opens and starts to drain the voltage going to the darlington pair. For a one ampere overload trip a 0.33 ohm resistor is used. A diode D2 is added as we also want to get a LED illuminated at overload. The diode prevents the sneak path trough the LED to amplifier base.

The works!

What is added? Firstly the capacitor accross power transistor T2. That will help emphasing the valleys of the output voltage and aid slow speed running. The capacitor should be of 1000 mmf or more. We also added a diode across the output. This is to kill the back-EMF of the motor. Normal DC-motors act as generators when the power source is disconnected. This may be due to dirt in rail or dead crossing of a turnout. The green LED L1 is a power indicator. I’d suggest using a center off switch as direction switch as that may act as a fail safe device if the electronics get out of control.