As my signature shows, I am an ASE certified day-dream engineer.
I'm always dreaming of making things better. Anywhere fat can be trimmed I process it in my mind. And this time, the issue of controller efficiency is on my mind. And the idea rests upon the restructuring of the motor. Although this is sprung from the hub motor topic in EC, I felt it went a different enough route to warrant a separate topic.
Imagine for me if you will, the permanent magnet motor. As I recall, it has several sets of windings, and only one or perhaps 2 windings are activated at any given point by the brushes. (I use a brushed motor for this example as I've only just as I'm writing this topic looked up how a brush-less motor works, and the concept has not yet sunk in).
My idea is, in the case of a hub motor, wherein there would not be a driveshaft, rather, the whole front face and outer circumferential face, (think of it like a brake drum), would rotate, while the coils would fixed. Now lets say that while the case materials are rated for the motor to put out 10hp, each coil only provides 1hp.
Now, in this case, each coil would consume ~770w. Which I believe is switchable at reasonably high voltages without significant arcing. So now the controller, instead of firing mosfets to control motor current draw, fires relays, to activate or deactivate windings based upon load demand.
Obviously this would require a motor re-designed from the ground up, but sometimes that's not a bad thing. Now anyone with some real credentials want to poke holes into my daydreaming?
</3am half-awake ramblings>