Astro - The amperage has to be pretty high at 4 volts. Even at 40 amps, that's only 160 watts, which a 7' windmill can do without it being too windy. Check out post #215 in this thread and the three graphs. What I am showing in the graphs is that my generator was ramping up the load too fast versus how much shaft power the blades can provide. I re-wired the alternator so it now can spin twice as fast before hitting the cut-in voltage and it pushes blades to a point where they will make more power than the generator can load it to. The other highly complicated thing is that the blades were effectively stalling all of the time, so the lift and efficiency of the blades was really low (probably less than 10% efficient). If I can get the blades to spin up faster when compared to a set wind speed (see reply #210 for a generic Cp graph), then I can easily triple my efficiency to say 30% and I'll have a lot more shaft power to play with.
To make this even more complicated, when I re-wired it, there is now less heat being lost in the windings since the current was cut in half on all of the coils (4x1 coils vs. 2x2 coils for each phase). This greatly improves the efficiency, which means I need less shaft power. And to add to this, I removed my diode in the system, so the voltage drop across that diode is now gone, which also was a huge waste of power. The original setup was about 20% efficient at taking shaft power and sending it into the battery. Now I'm hoping it's closer to 40-50%.
The alternator has 12 coils and 3 phases. Each phase has 4 coils wired in parallel and they use 17 gauge wire. The magnet rotor has 16 magnets that are 4"x1/2"x1/2" arranged in a NN-SS-NN-SS.... pattern such that there are 8 poles. I did this due to the spacing of the 12 coils and how I fit my laminated steel shims to transfer the magnetic flux (very similar to an electric motor, specifically a brushless 3 phase motor).
It runs at about 30 RPM/ volt in the 4x1 configuration. My old treadmill motor was about 35 RPM/volt. Of course, my new alternator is going to output something like 30 amps at ~600 RPM versus only 6 amps with the treadmill motor at the same speed since there is a lot less resistance in the coils. With the diode removed, the 30 amps will probably be more like 50 amps.
"I re-wired the alternator so it now can spin twice as fast before hitting the cut-in voltage and it pushes blades to a point where they will make more power than the generator can load it to."
"To make this even more complicated, when I re-wired it, there is now less heat being lost in the windings since the current was cut in half"
So in essence you raised the voltage and lowered the amperage making it perform better. OR just do not draw so many amps is what I was saying. Hit the sweet spot where the mill is turning and you are putting out a more stable and lower charging current. What you had going on was in essence a electrical breaking system. Lots of amps and it slowed down to much. Although a battery does charge faster with higher amps. So in the first stage of charging, you want to put as many amps into the batteries as they can handle without gassing, (that happens at a certain temp which I can not recall off hand and is why good charge controllers have a temp sensor) AND WITHOUT stalling the mill. As the battery voltage catches up, you do not need as many amps and then finally when almost charged and drawing very little amps, you back the voltage off to the batteries peak voltage. That is the process you are trying to achieve. So again you can raise the voltage and thus lower the amperage of the mill. You can lower the amp draw of the load. Or you can let it buck with a relay that causes it to charge hard slowing the mill from say 400 rpm down to 300 (or whatever works for your situation), then cut the load, let the mill get back up to 400 rpm and then do it again and again and again until the batteries do not need such a high amp input because they are starting to get charged. If doing that, you are going to want a mill that can recover that hypothetical 100 rpms very fast or you want a mill that is hard to slow down. That will keep you charging most of the time, or as much as the mill can handle at a high amp draw from the load.
You can accomplish this with a current sensing or a voltage sensing relay, since the two are dependant of each other. You will have x amount of voltage at 300 rpms and x amount at 400 rpms in in our hypothetical example, we know that 300 rpm is as much as we want to stall the mill and so we set a low voltage circuit into play, cutting the load until another relay set at the voltage of whatever 400rpms makes, then reconnect the load of batteries. and ta da Bob's your uncle. It is also going to do the same thing in low wind situations, because it will try to take as much as it can without stalling the mill past whatever set points you use. So in low winds and you are only getting 25v, it is still going to take as many amps as it can to try and charge the batteries, but as soon as the voltage drops to say 20v, or whatever stall point you desire, it is going to kick the load off and let the mill speed back up.
It is not complicated at all.
As I said in my situation, I am going to have starting current of compressors and motors. So I want a mill that is hard to slow down. One that just wants to spin at the same speed all the time and does not get worried if the wind goes up or down for 10 seconds. OR if the load amperage goes up or down slightly for short amounts of time.