But an even number of windings 2,4,6 is easier to compensate, one in phase in series with the second in antiphase. Like a bifilar. But I don't know what to do with three windings.
I do not know yet. I keep thinking about how to do it.
Use a factor divisible by both 2 and 3, like 12, 18, 24, ...
Also consider what rakarskiy was trying to promote with the slots.
There is something interesting going on there. When we place
a bundle of wire (output coil) in a slot, the magnetic field it
produces when put under load has to go someplace, but if
wherever it goes does not significantly disrupt the magnetic
fields produced by the exciter windings, we have an opportunity
for power gain.
We have to look at this device in terms of magnetic flux paths.
You have paths for the exciter coils and paths for the output
coils. These paths can cross through each other, but should
not collide with each other.
From what I can see with the Holcomb device, the output coils
are positioned identically to how they would be found in a typical
generator stator, but the exciter windings are positioned totally
different, even though the effect they would have on an actual
spinning rotor is nearly the same. Remember, the exciter windings
are there to replace a spinning rotor, to replace the magnetic
fields that would be produced by a spinning rotor under torque.
You are trying to use three phase input to replicate a spinning rotor,
but the output need not be three phase also. It can be two phase,
four phase or whatever. Look at the flux paths and see where the
transitions happen and have your output coils catch the change in
flux across these transitions.
I wish I had a fancy machine shop and plenty of time and material
to crank out one or more of these units to test with. If I had that,
I could show you real hardware and take measurements. I'm
commenting here based on concepts I "think" may have some
importance. What the raw fundamental principal is for the HES,
we may never know and it is still a possibility it cannot physically
be put into hardware without exotic materials. I'd like to think
Holcomb is correct in his claims, but until I do it myself and convince
myself, all I can do is make suggestions based on what I think.