I was thinking about it a bit more.
What may be more significant is that the RMS current in the individual coils is higher when they're in series than when they're in parallel. (In this case I mean the true RMS, not the .707*peak approximation for sine waves.) In parallel they only conduct during their own peak - when their voltage is the highest. In series they conduct for that AND for the peaks of the other coils (except when their voltage is low enough that some or all of the current stays in the diodes, making both sides conduct).
Higher current in the coils means more heating. The heating of the coils is the limit on how much a particular machine can generate - so you have to run the genny at lower power to get the same heating.
Again I have other committments and can't spend the time right now to figure out how much this hits you. But I bet it will be large - like a factor of five (the number of phases), of the same 3-ish that we saw with the voltage change for the rewind, or some lower number due to some of the current bypassing the coil thorugh the diodes due to resistive drop in the coils and the diode forward current non-linearity. Any of these would be a disastrous derating of your genny.
I was also thinking about connections: If you connected it in a five-point Y you could get away with half the multiplier on the rewind - AND half the diodes (though the diode drop would be about the same - in fact a tad higher unless you used diodes rated for twice the current). This is close enough to the original wind that you might not have to bother rewinding the coils - just make the mill spin a bit faster. The "current from other windings" effect due to the 5-wire Y replacing 10-wire separate commections would be zero with AC resistive loads, present but quite small with rectification. (In fact your 5-phase setup should reduce it considerably compared to a similar recitfied 3-phase setup.)
This discussion is far back on the horizon. So if you don't reply to it in the next couple days (if only to say you've seen it) I'll try to send it to you directly or point you to it in a reply to some other posting you make. (Or maybe I should work out more details and post an article.)
Or if you want to correspond directly my email we can figure out some way to exchange email addresses without alerting the spambots. B-) (Note that I already get so much spam that I only muck out the box and read the correspondence occasionally.)
cheers
Rod
