We recently bought a train set made by a renowned company and just couldn’t resist looking inside the locomotive. Although it did have an electronic decoder, the DCM motor was already available 35 (!) years ago. It is most likely that this motor is used due to financial constraints, because Märklin (as you probably guessed) also has a modern 5- pole motor as part of its range. Incidentally, they have recently introduced a brushless model. The DCM motor used in our locomotive is still an old-fashioned 3-pole series motor with an electromagnet to provide motive power.
The new 5-pole motor has a permanent magnet. We therefore wondered if we couldn’t improve the driving characteristics if we powered the field winding separately, using a bridge rectifier and a 27 Ω current limiting resistor. This would effectively create a permanent magnet. The result was that the driving characteristics improved at lower speeds, but the initial acceleration remained the same. But a constant 0.5 A flows through the winding, which seems wasteful of the (limited) track power. A small circuit can reduce this current to less than half, making this technique more acceptable. The field winding has to be disconnected from the rest (3 wires). A freewheeling diode (D1, Schottky) is then connected across the whole winding. The centre tap of the winding is no longer used. When FET T1 turns on, the current through the winding increases from zero until it reaches about 0.5 A.
At this current the voltage drop across R4-R7 becomes greater than the reference voltage across D2 and the opamp will turn off the FET. The current through the winding continues flowing via D1, gradually reducing in strength. When the current has fallen about 10% (due to hysteresis caused by R3), IC1 will turn on T1 again. The current will increase again to 0.5 A and the FET is turned off again. This goes on continuously. The current through the field winding is fairly constant, creating a good imitation of a permanent magnet. The nice thing about this circuit is that the total current consumption is only about 0.2 A, whereas the current flow through the winding is a continuous 0.5 A. We made this modification because we wanted to convert the locomotive for use with a DCC decoder. A new controller is needed in any case, because the polarity on the rotor winding has to be reversed to change its direction of rotation. In the original motor this was done by using the other half of the winding. There is also a good non-electrical alternative: put a permanent magnet in the motor. But we didn’t have a suitable magnet, whereas all electronic parts could be picked straight from the spares box.