discharging caps

Sure, maybe Schottky diodes, depending on what is there, should protect just about anything.

How about a passive bleeder that's switched into the circuit when the power switch is turned off?

Seems obvious, but maybe I'm missing something?

I'd have to tell when the power switch is off, or actually when there's no AC power coming in. The switch is actually part of the AC inlet/filter assembly, so I can't add a pole to it or anything like that. So I'd have to sense 120-240 VAC and drive a discharge device. This box will be CE/UL tested, so any primary circuits have to be reviewed and hipot tested and all that. Hassle.

If anything went wrong with the logic, the resistors would smoke. Unless I used huge power resistors.

A constant-power bleeder might be interesting, something with bipolar transistors or depletion mosfets, but it would still need a heat sink.

Like an interlock switch on a Plexiglas cover that switches on high current bleeders? They worked well enough on high voltage, high current power supplies in broadcast transmitters. I've also seen them on PC board paneling routers, to keep idiots from losing their fingers.

Just a SMPS operating into a resistor would perform that function.

How about a bleeder to bring the supply out of tolerance and then switch in a load to kill it when it goes well out of tolerance (obviously it has to withstand a short transient at turn-on, and the supply must not current limit and fail to start- or it could be inhibited at power-on with a bit more complexity).

Eg. if supply >= 85V then draw 20mA = 2W (12 seconds to get to 85V) if supply < 85V then switch a 12 ohm resistor on (time constant 0.26 s), such as those Welwyn planar parts. Even the smallest can handle

Isn't that IRFR120 a bit risque here, with an absolute max of 100V VDS?

If you really want to short it out fast then an SCR + in-rush current limit= er combo would zero out the caps in 10s of milliseconds. 110 joules is on t= he low end for the inrush limiters, and almost any kind of SCR can take the= hit. It will cost more than a buck though.

Can you replace the switch with NC relay contacts driven by the switched side of the line?

That way, the dump starts automagically.

--Winston

On 2012-04-19, John Larkin wrote:

Detect the droop on the capacitors?

What's the thermal capacity of a TO220?

Something like this? (not these transistors though, ones rated for the voltage)

Version 4 SHEET 1 880 680 WIRE -784 32 -800 32 WIRE -672 32 -784 32 WIRE -576 32 -672 32 WIRE -528 32 -576 32 WIRE -432 32 -528 32 WIRE -224 32 -432 32 WIRE -576 48 -576 32 WIRE -528 48 -528 32 WIRE -432 48 -432 32 WIRE -224 64 -224 32 WIRE -784 96 -784 32 WIRE -288 112 -384 112 WIRE -864 128 -864 32 WIRE -672 128 -672 32 WIRE -576 128 -576 112 WIRE -528 128 -528 112 WIRE -528 128 -576 128 WIRE -480 128 -480 112 WIRE -480 128 -528 128 WIRE -384 128 -384 112 WIRE -480 144 -480 128 WIRE -784 192 -784 160 WIRE -224 192 -224 160 WIRE -384 208 -384 192 WIRE -864 240 -864 208 WIRE -832 240 -864 240 WIRE -832 272 -832 240 WIRE -864 304 -864 240 WIRE -784 304 -784 256 WIRE -784 304 -864 304 WIRE -672 304 -672 208 WIRE -672 304 -784 304 WIRE -480 304 -480 288 WIRE -480 304 -672 304 WIRE -384 304 -384 272 WIRE -384 304 -480 304 WIRE -224 304 -224 272 WIRE -224 304 -384 304 FLAG -832 272 0 SYMBOL voltage -864 112 R0 WINDOW 3 -59 232 VLeft 2 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value PWL(0.001 100 1 100 1.0001 0) SYMBOL cap -800 96 R0 SYMATTR InstName C1 SYMATTR Value 44000µf SYMBOL cap -800 192 R0 SYMATTR InstName C2 SYMATTR Value 44000µf SYMBOL diode -864 48 R270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D1 SYMBOL res -688 112 R0 SYMATTR InstName R1 SYMATTR Value 1k SYMBOL cap -464 144 M0 SYMATTR InstName C3 SYMATTR Value 1000µ SYMBOL npn -288 64 R0 SYMATTR InstName Q1 SYMATTR Value 2N3055 SYMBOL pnp -384 48 R90 WINDOW 3 -48 13 VRight 2 SYMATTR InstName Q2 SYMATTR Value 2N2907 SYMBOL diode -560 48 M0 SYMATTR InstName D2 SYMATTR Value MURS120 SYMBOL res -496 304 M180 WINDOW 0 36 76 Left 2 WINDOW 3 36 40 Left 2 SYMATTR InstName R2 SYMATTR Value 100 SYMBOL res -240 176 R0 SYMATTR InstName R3 SYMATTR Value .1 SYMBOL diode -400 208 R0 SYMATTR InstName D3 SYMATTR Value 1N4148 SYMBOL diode -400 128 R0 SYMATTR InstName D4 SYMATTR Value 1N4148 SYMBOL cap -544 48 R0 SYMATTR InstName C4 SYMATTR Value 100µ TEXT -938 350 Left 2 !.tran 10

I'd add a bleeder as well. I don't like high voltages sitting somewhere for a long time after a device is switched off. Is there no way to cut back on the capacitance?

Given your instrumentation capability i would try it and see. You may need a $1.00 mosfet though. (rated 2000 A pulse).

?-)