Gonna start testing soon, but first you can do this a few ways. Firstly you can build a simple circuit like this with two bar magnets, a pulse width modulator and a primary and a secondary on a ferrite core with a resistive load.
[attachment=4919]
I've already successfully done that.
Then you can copy Stan's schematic like this:-
[attachment=4920]
Now if you do this you will need a variable resistive load marked R1. A and B are the electromagnets. I would start off small and work up with the windings. Follow Stan's schematic for coil winding direction to begin with and include the diode so the poles are correct on the magnets. Same with the primary marked 'p' and the secondary marked 's' don't be going heavy. Primary and secondary don't want to be too big so I suggest we go for a step up of 3-1 with fine wire. We don't want to saturate the core with too powerful electromagnets. PWM needs to be monitored so connect oscilloscope to monitor frequency and pulse width. If we get too much saturation we can increase the time in between the pulse widths. The resistive load will decompress the time intervals but the electromagnets cannot so don't worry about power going into the load. If the electromagnets are too powerful then you can increase the load at R1 too. Not bothering with fuel cells at this point because we need R1 to be as static as we can get it for testing purposes and a fuel cell will screw that up.
We need to connect an amp meter in between the power supply and the PWM to monitor current draw. Try to use ferrite not iron or any efficient core material that will not saturate easily.
Then we play.
It will take time and experimentation. Good luck.
[attachment=4919]
I've already successfully done that.
Then you can copy Stan's schematic like this:-
[attachment=4920]
Now if you do this you will need a variable resistive load marked R1. A and B are the electromagnets. I would start off small and work up with the windings. Follow Stan's schematic for coil winding direction to begin with and include the diode so the poles are correct on the magnets. Same with the primary marked 'p' and the secondary marked 's' don't be going heavy. Primary and secondary don't want to be too big so I suggest we go for a step up of 3-1 with fine wire. We don't want to saturate the core with too powerful electromagnets. PWM needs to be monitored so connect oscilloscope to monitor frequency and pulse width. If we get too much saturation we can increase the time in between the pulse widths. The resistive load will decompress the time intervals but the electromagnets cannot so don't worry about power going into the load. If the electromagnets are too powerful then you can increase the load at R1 too. Not bothering with fuel cells at this point because we need R1 to be as static as we can get it for testing purposes and a fuel cell will screw that up.
We need to connect an amp meter in between the power supply and the PWM to monitor current draw. Try to use ferrite not iron or any efficient core material that will not saturate easily.
Then we play.
It will take time and experimentation. Good luck.