Since the other thread about generators that speed up under load has degenerated into insults I'm starting this new one in the hope that we can keep it clean.

This is for people who have done experiments and already know the basics.

I just wanted to share a few thoughts and findings based on my experiments with my dc motor driven two coil setup.

First of all we know that in order to see acceleration effects under loaded or shorted condition we need to attain a critical frequency. To be more precise we need a magnet to move past the coil above a certain speed, meaning faster than the current buildup in the coil in order to get the delayed Lenz effect, which is what gives us low drag or even speed up conditions.

One would think that fitting as many magnets as possible around the rotor is best as it raises the frequency for the same rpm. From my experience this is not the case. I mean bigger rotors with more magnets are better, but not too many.

Who has studied the Kromrey generator knows that compared to a standard generator, on the Kromrey the rpms can be lowered down to almost 50% of full speed while still delivering almost full output power to the load. This I have verified personally. In fact the output is based on the core demagnetisation while the magnet is leaving the core. The demagnetisation needs time to occur based on load impedance/resistance, so we need enough space between one magnet and the next. I could find no improvement adding magnets after certain amount, Thane Heins like.

The other thing is magnet - core gap. My conclusion is that as close as possible is best in order to see speed up conditions. But you need a solid structure that can handle the huge forces if you use neo magnets.
From Bedini's notes it seems that when the setup is "built right" with the right load we should see a decrease of about 50% of input power. Personally I haven't been able to see this, but what I'm thinking is this:

Let's say you have rotors who's magnet/coil gaps can be adjusted while in operation. Let's say you're so far away with the rotors that they don't even affect the coils, that's our base line speed (and input power). If we start moving the magnets closer to the coils (open) we start to see some drag and deceleration. If we then short the coils (our base line speed has to be over the critical frequency) our speed goes back to almost base line speed.
If we move the magnets to get an extremely narrow gap we see a strong deceleration, but also a much bigger acceleration when we short our coil, or add the right load. What I mean is, the smaller the gap the bigger the speed difference between the unloaded and the loaded (or shorted ) condition. This might be how Bedini could report such a hug he difference between loaded and unloaded.

Bifilar coils: franky, although they act a bit as capacitors I don't think bifilar coils have a magic role in this case other than simplifying finding the right number of turns, especially if you have many strands to play with. You need the largest wire (more amps) you can fit while allowing you to have enough turns to get the needed self inductance, which at a given speed will give you no drag or speed up.

About what I wrote at the beginning, one thing I need to think about some more is, if we get basically the same output for much lower rpms and input power, why bother getting up to the critical frequency..? I know this doesn't sound right, need to think some more........

Mario