Fusor Forums

I have a Spellman PTV50N200 (see this thread for reference: viewtopic.php?f=11&t=11477) which I hadn't used much except for scaring the living daylights out of myself w/ corona effects. I made some small modifications, specifically adding a switch to the line power and re-routing the high-impedance current and voltage monitor outputs through a pair of op-amps so I could drive a meter.

I had not yet attached it to my vacuum system due to lack of an adequate feedthrough and because to ensure safety (and to fit it on my bench) I needed some careful grounding, etc arrangement -- and added a resistive ballast of around 50 Kohm. Arcs that can obliterate any normal insulation and jump six inches need special precautions.

Anyway, I got it set up, I got it making a glow discharge at a very low pressure (fractional microns I think) that my small OBIT would not do anything at, and got some X-rays I could detect w/ my geiger counter...

And then I had some unpleasant flickery instability -- rapid flashing of glow and movement of voltage/current needle --

A bit later I smelled magic smoke, and the thing reads out full current and only a small voltage. And certainly does not make a discharge anymore.

I'm not really sure where to begin -- I can't see any visibly burned stuff, and I am leery of powering the thing on when testing it due to all the voltages (of varying levels of "high") inside the case. I'm concerned my amplifiers had some kind of short -- should I start by removing them and/or changing the socketed op-amps that are electrically near them?

how big was your ballast resistor?

Physically? About 30 inches long.

You may be able to smell the general area that let out the magic smoke.
Sometimes the FETs/ICs just have bubble or indent that may only be slightly discolored.
Electrolytic Caps will burst from the top so you may see some Caps with a bulged top.

Unfortunately I haven't got much closer smell-wise than the general area -- but I did find a burned capacitor -- the bypass / snubber on the PWM's outputs. Nothing else.

OK
Dead snubber cap
check or preferably replace the snubber diode there usually is one and check the snubber resistor.
Suspect one or both output FETs - replace
Check gate zener diodes preferably replace ditto any gate resistors and charge distribution series diodes.
The Gate drive transformer is usully bullet proof.
Test with series resistor placed in DC bus to limit smoke generation because of unknown downstream fault or fault in multiplier stack --usually dead cap shortly followed by nearby high voltage diode

Not sure if I can do much about a bad multiplier, since that's all cast into the Spellman Block.

Snubber diodes -- are these going to be schottky diodes between the snubber cap / PWM outputs and ground? TVSs are also present.

Snubber resistor = low-ish value resistor in series w/ snubber cap?

I unfortunately don't have a schematic for the actual driver portion, just the control board. But the MOSFETs are of course pretty hard to miss. I'm pretty fuzzy on the gate stuff you describe.

Ian
Snubbers are to take care off turn off transients so usually go from FET drain to DC bus
Cap to isolate DC Fast diode and resistor somewhere between a few hundred ohms to a few kilo ohms sometimes a transil around the other way as well

Gate diode isolates charge and discharge times ie two resistors one diode in series to gate series resistor diode and a resistor in parallel across those two

Zener from gate to ground or transil
transil from source to ground