resonant generator - colpitts osc.

[H2OPWR]

I'm glad you mentioned the the oscillation going back into the pwm. I believe a good diode can solve that. Just connect it in series on the pwm's output. You say you have diodes, Hmm, that should still be the one way valve to protect the pwm's.
Russ

I am not positive that it was the ocillation that destroyed them. I have not investigated the electronics further. I also think the frequency of my PWM's may have been too high. They were running at 400hz. The fet's may have overheated due to the high inrush each time the PWM cycled. The faster you run the fet's the more they heat. I probably should change the resistor that controls timing and slow one down to around 15 to 20 hz and try again. I just haven't had time and do not want to burn another one out. I will change out the fet's to bigger ones on one of the bad PWM's, see it it still works, slow it down and try again.

[fisher]

H20, It could be inrush current, you may have to use a power FET or a DIAC and let the PWM turn on the device and then let the device feed current to the generator. Or it could be the negative spike. But a diode will block a positive spike, but not a negative spike, maybe using the FET or DIAC would solve either problem.

Hmmmm, gotta think this one over a bit. Reply above is just my first guess. Eager to hear results either way.