Do avalanching diodes "latch up"? Answered
While working and revising my flyback driver design, I had one of the 1N4148 diodes that I was using as protection get cooked! (it seemed to fail short circuit) I have not tested to see if the MOSFET survived yet. Obviously the diode needs to react fast and handle avalanching energy well. Any suggestions? are some beefy schottky diodes suitable? Also, does making any ordinary diode go into that avalanche or zener reverse breakdown region cause it to potentially lock up and conduct for a period longer than the EMF spikes I'm dealing with? Or with any sort of hysteresis (like when using neon lamps)?
Originally my wimpy glass 1N4148 diodes was wired in parallel directly across the MOSFET drain and source reverse biased. However, carefully examining this tesla coil schematic,it appears that there is both a forward biased diode in series with the MOSFETs, and another reverse biased diode(s) in parallel across the MOSFET & diode. It does not look like a bad idea given one the reverse biased diodes can shunt EMF spikes that are too large while the other isolates them from the MOSFET. My only concern is how lossy that would be at 12V operation.