I got the first assembled dual high-voltage pulse generator yesterday so I came to work today to fire it up.
It almost worked first try. Somehow the PCB footprint of the PV optos U3 and U20 got twisted, so there needs to be a small kluge. I hate it when that happens. We've used that part before with no problems.
Anyway, it's making 1200 volt pulses. The power supply was running really hot, especially T1 and the nearby C-W multiplier diodes, spotted with the FLIR. Pulsing at 10 KHz, it was pulling 320 mA from the 24 volt supply.
Seems like I authorized two sources for the SOT23 HV diodes, 1SS398 and MMBD5004S, mainly because we couldn't find the MMBDs in stock. Since the diodes were hot, I replaced all of the 1SS parts with some MMBD samples I had. The input current dropped to 110 mA, and the diodes are cool. T1 dropped from 80C to 44C!
The switcher is running at 200 KHz, and those 1SS diodes must have ghastly reverse recovery times.
The orange LED in the ground leg of the 78L15 regulator worked beautifully. I'm getting 16.7 volts and the LED looks great.