I have also tried the switched winding path to attempt to match the alternator to the wind energy. I had tried series parallel arrangements of coils in the same phase, as well as star/delta switching. All these require some form of rpm sensing and relay type controlled switching of the coils, with some hysterysis and timing delays to extend the life of switching components, if using mechanical relays. The relays tend to make a racket in gusty conditions.
At the end of the day, the lowest possible winding resistance and the alternator wound for a normal cutin at approx 2/3 of max rpm gives the highest practical efficiency. The trouble is no output to a battery load at the bottom 2/3 of the mill rpm.
For the windmills that are not blade stall limited.
If you normally wind coils for cutin at 100 rpm, with a 300-400rpm top end. The alternator may have had an efficiency of 50% at the maximum power.
This same alternator could be wound for cutin at 200rpm. The alternator would now have an efficiency of probably 75% at maximum power. This would translate for most mills of additional power to the load, as less is converted to heat in the stator, as the resistance is half for the same load current. For a typical dual rotor AxFx alternator, say nom 1kW. an additional 250W could be available to the load at maximum power. The power at the bottom rpm end, below this higher cutin rpm is recoverable with an additional external capacitor voltage multiplier arrangement. This is easily made for a 3phase alternator, with some additional rectifiers and cap coupling to them. The caps do not pass the full power of the alternator, but approx 20%, at maximum output power. For a typical alternator dual rotor, 12mags per rotor, 9coils, 3phases, 100rpm cutin required, approx 1000uF caps will do. AC rated caps are required, and best results are with at least 200VAC ratings, to handle the AC ripple current.
The higher alternator efficiency will reduce the liklihood of peak current coil failure.
There is no rpm sensing, or switchmode DC-DC converter, and no relays. The arrangement is just in parallel with the existing rectifiers. The inherant electrical characteristics of capacitor gives an increasing loading with mill rpm. Once the mill reaches approx the 200rpm, the normal rectifiers add to the power output. The multiplyer will vary the voltage gain automaticaally between the mill and the load. As the windmill increases rpm, the voltage gain will reduce. The AC voltage across the caps will increase until half the voltage from the windmill appears across the caps, when the normal rectifiers start to conduct as well.
The windmill may stall the blades at higher windspeeds, towards the point where normal furling would occur, due to the lower alternator impedance. This will all depend on the load impedance. The windspeed, and wind energy distribution at a location will determine the significance of potential power loss if this happens.
The capacitor multiplier arrangement is like an external electrical gearbox, with no moving parts, that can be tuned to the application.
One aspect to consider is what to do with the additional power. If eventually a cap fails, usually goes open cct, the windmill will not become dangerous, but power output will just decrease at the lower rpm. I have cap doublers on my iron cored windmill and I will incorporate them in my new 48V AxFx mill. The winding phase resistance will be under 0.5ohms, and the windmill will not have startup stall problems.
Gordon.