Well, here's the deal. I have an off-the-grid neighbor that has a 12v solar/hydro/generator system for his electric. We built a little board from a magazine circuit that regulates the output of a 12v ambulance alternator by varying the field in proportion to the battery charge....to accomplish battery charging for the dreary weeks. The alternator is driven by a 5 hp Honda engine with an appropriate step-up pulley arrangement. Nominal current with low batteries is
30-50A.
It's a little PWM circuit implemented with an LM723 and a 555 timer to make a quick-and-dirty PWM output. The output of the 555 switches a
30A rated series-pass transistor to supply about 2A max to the alternator field. Surprisingly, the circuit works quite well....most of the time.
With the engine comes vibration, and with Tennessee weather, 70F one day, 7F the next things get loose. Like the connections to the alternator field, which defines our failure mode. When the field connections get loose, the field will dump ~300V onto the dc 12V bus. Despite a hefty flyback diode (6A4) across the field and a fat 24V MOV at the output of the board, the board is destroyed when this happens.
The 6A4 will fail shorted, the Output transistor (like a 2N3055, but PNP) will be open E-B, the MOV is a cinder and even the 50v rated .1 bypass caps around the board will short. Of course, the 723 and 555 are toast.
We have tried fuses (make a handy indicator to tell the board has blown), a couple ohms to the field circuit as a dump (no joy there), bigger MOV's, and assorted blue language. Just for s**ts-'n-grins, we put an NE-2 across the field and wire-wiggled. Now I know how to make black Neon lamps.
I HAVE changed the 6A4 diode to a stud-mounted 200A fast-recovery type commonly used in induction heaters, but I have to wait for some parts from Digikey to try again. Short of soldering the wires everywhere external (not practical), I dunno.
Spike and surge experts come out come out where ever you are....thanks.