I rewound the secondary coils of a 50watt transformer with 4 sets of 5 turns of 15 gauge wire, giving me a no load output voltage of 1.2v. I expect maximum current output to be around 30 amps.
A weird thing I was not expecting is that I was able to measure 1.2v of potential between any two leads of the four sets of windings. These are windings with I believe no electrical connection bridging them. It's also independent of the winding polarity. I don't know if that is to be expected or if I made some mistake.
This is intended to be a component of a power supply in a future project.
Disassembling the core is time consuming. I use an exacto knife to delaminate the silicon steel plates one by one. I estimate it took about 45 minutes of focus.
Do not expect to get tight coils with 15 gauge wire. I calculated there was enough room to fit 40 turns of wire, but I ended up settling for half of that because my turns were so loose.
I planned to use LED's to check that the N-channel MOSFET (IRF530N) was actually conducting in both current directions when a 5V signal was applied to it, but I did not actually have any LED's on hand that would turn on with less than 1.5V, so I substituted a large coil as the load with a suspended magnet to detect the passage of AC through the MOSFET.
Unfortunately, there is leakage with 1.2 VAC applied to a MOSFET. The body diode begins to conduct at and below -0.7VDC, which I assume is similar to its action under AC cycling. The amount of current conducted was quite small, but did produce a detectable vibration in the suspended magnet, which is not acceptable for my planned application.
Fortunately, I can move forward with testing the concept of 3 phase electrical commutation at low voltage, high current with this small amount of leakage. I will eventually need to drop the voltage down to or just above 0.6 volts for the next version (which means using more MOSFETs paralleled to reduce the Rds(on) of the MOSFET at the lower Drain-Source voltage, which may become nonviable at such low voltages due to the cost).