Triacs.

There seem to be problems with all of the triac circuits that they've both tried.

Although Flowering Elbow says he's got good results by using his second circuit to control a DIY pillar drill.

Variable Power Supplies.

I like the idea of using variable switched-mode power supplies to determine the optimal supplies for field coils and armature. And then use slightly less variable SMPSs to provide those supplies.

But I'm sadly lacking in practical knowledge in SMPS design, so there's A LOT of learning to do before I could go anywhere along this path.

Microcontroller.

The washing machines I've taken apart all use a microcontroller (which isn't surprising as they also have to deal with timing and sequencing too).

So, perhaps the microcontroller route is the way to go.

But I'm not that happy with the level of complexity of the circuit that Flowering Elbow has found and the code looks a bit untidy to me.

I think I could improve on the circuit:

1. Replacing the motor's tacho coil with a hall-effect sensor (I've already done a bit of work towards using a hall-effect sensor to determine RPM)
2. Using an ATTiny chip instead of an Arduino Nano - just to cut down on size.

I'd also like to tidy the software up a bit:

1. Instead of using the Arduino framework, use a more "traditional AVR" style and break it down into modules a bit.
2. (This will probably prove unpopular, but it is my obsession) Use assembly language instead of C/C++