disabling cap draining resistors

Sure. Just pull the gate negative somehow when AC is on. You could do the equivalent with an enhancement mosfet circuit, except that it would have to leave a few volts in the caps, ultimately.

There are some normally-on SSR's, which encapsulate a depletion-mode fet and a photo-isolated driver. That might be handy.

Or use a real relay!

John

The best way is to just leave it alone. At 400 volts, the current is

Just use a larger value resistor. It may take longer to discharge, but it will eventually do so and burn less power. Check if there's a safety spec on how much time is allowed for discharge. You might have some leeway.

If the caps are fed by a string of series diodes, that happen to have "equalizing" resistors across them, you can cheat and use those to discharge the filter caps. However, that's doubtful at 400V which doesn't need series diodes. It's also safer to put the bleeder directly onto the cazapitor.

Incidentally, I kinda like to put neon lamps across high voltage cazapitors so that I know that the cap really is mostly discharged. The bleeder resistors do tend to burn and open after many years of operation. I hate getting zapped when that happens.

They are more than bleeders. They also serve to equalize the voltage drops.

Especially if this is for a product I'd be careful because the TUEV/UL/CSA or whatever blessing will be gone afterwards.

I've done active cap dischargers, but they need fuses or big power resistors or whatever to handle the case when something goes wrong.

John

LEDs in series with the bleeders is fun, and serves as a warning too.

Some of my bigger amps have a manually-operated bleeder switch, and a biggish resistor+led thing always connected. Modern led's are visible at very low currents.

John

Talking about bigger amps, why is it that EMC labs in your neck of the woods don't always have a huge RF amp on the shelf? After resigning to the fact that we can't get a tractor-trailer rig cert'ed in the Bay Area on account of its weight and size at least we wanted to have a mock chassis blasted. Whenever I asked about >>20V/m the guys at the other side began to choke a bit. That was a lot easier in Europe. Most places I used there were geared for mil-level.

The EMC labs are usually out in the boonies, not around here. The background fields here are outrageous. The AM, FM, TV, and certain mysterious bands are pretty much saturated. I get -20 dBm on a couple of FM frequencies with a clip lead hanging out the front of my spectrum analyzer.

Besides, FCC only checks for radiated emissions, not susceptibility. So nobody here needs amps unless they're doing CE too.

John

That's why I was surprised that there are any labs in the Bay Area. It used to be that we had to schlepp everything to Mariposa.

Well, most stuff that I design or help to design is eventually marketed world-wide so we have to. Plus automotive where you can't get anything on the road without some serious susceptibility testing. You don't want the electronic brake assist to come on hard just because the trucker next to you keyed his CB radio mike.