exploding wirewound resistors

Why are you conflating CURRENT LIMITERS with screwdrivers?

You wrote this, didn't you?

Tim

Those green "vitreous enamelled" vishay wirewounds were pretty impressive. I remember putting them across car batteries until the coils glowed yellow through the molten glass. Never seemed to do them much harm :)

Most resistors that are joule rated, including wirewounds, are spec'd handle X joules for 1 second, and drop off hard for shorter durations. Some level off at some short time, and some keep tapering off forever as the times get shorter.

We found a Vishay 3-watt part, WSN6927 at 0.33 ohms, that specs 75 j from 1 to 10 seconds, and 10j from 0 to 100 ms, with a slope in the gap. That makes our SSR circuit work. I'll get some and blow them up to see how much margin we'll have. Maybe pulse them for a few months to see if they fatigue.

We'll put eight (!) ADCs on the board to measure channel ON voltage drops fast, to shut things down quick if a customer puts a hard voltage across one of our SSRs. The series resistor buys us enough time to do that.

Eventually the glass melts and drips off.

Ceramic has a high specific heat but very poor thermal diffusivity compared to metal. At short times all you have to work with is the metal itself.

Cheers

Phil Hobbs

There are some impressive bulk-ceramic resistors that will absorb hundreds of joules in any time domain, but they are gigantic leaded parts. Wouldn't fit on a VME board and bad for vibration.

Den mandag den 18. september 2017 kl. 21.30.52 UTC+2 skrev John Larkin:

how about a IR2127 to do the over current shutdown, seems like it would be cheaper than that fancy resistor

A resistor of uniform cross-section will dissipate heat uniformly, if it's reasonably small. I have data showing this happening in ~10 ps time scales. It's only thermal conduction that's slow.

Cheers

Phil Hobbs

Carbon composite? or is that too noisy?

George H.

Carbon has a negative temperature coefficient of resistance and you get hot channels. A former boss had a party piece where he would put roughly an amp through a 0.6W 10k carbon film resistor for half an hour at a time.

The resistor still measured 10k afterwards, but it had a dark line in the paint over the top of where the hot channel had been.

You aren't worth reading.

The Vishay 3 watt resistor is cheap.

We want floating switches, so we'd have to power up that chip somehow.

If we put a 0.33 ohm, 10 joule or so resistor in series with the fets, the resistor takes most of the voltage drop and dissipation if the customer gives us a hard voltage source up to maybe 75 volts. We'll digitize the voltage drop across each switch (resistor included) and then we know the current and can do the shutdowns in FPGA code. We needed the voltmeter function already for BIST.

Really, in the last couple of weeks, we've explored a zillion options, including fuses, self-protecting fets, current limiting the fets, all sorts of crazy schemes... each with at least one fatal flaw.

Some loads, like AC relays and solenoids, and bi-level DC relays and solenoids, need high current for a short time to pull in. A fast current limit gives us problems there. We're talking 2 amps continuous, 20 amps for 100 ms, and shutting down ASAP at 100 amps or more. The FPGA can tune all that.

Who would have thought that a simple SSR would be so complex? It has been fun, though, brainstorming all those ideas.

I'd consider a transformer, myself. I've got a dozen transformers with hundred-amps-capable windings, not sure if any of my capacitors can do that.

Faster than millisecond, you need ferrites, but at a millisecond... well, that's about what an auto spark coil is doing. Higher voltage, lower current, but you could rewind one, or just retask a soldering pistol.

OK, I'm ready to blow up some resistors, as soon as UPS arrives.

That should do numbers like 300 amps, 80 joules, 50 volts.

Where'd you get that information?

Vishay's WSR series is an interesting set of 2W and 3W metal-strip current-sensing parts.

Vishay sent us a graph. I guess it's public info.

The WSN is non-inductive, Ayrton Perry winding, which we understand has more metal inside than the usual resistors. That helps absorb energy on short time scales.

On a sunny day (Tue, 19 Sep 2017 11:02:38 -0700) it happened John Larkin wrote in :

Why blow anything up? I made this to test 30C lipo batteries, 4 x 1 Ohm 100 W:

John's got to be kidding ?

You both contrbute worthy information here

boB

On a sunny day (Tue, 19 Sep 2017 12:03:13 -0700) it happened boB wrote in :

Actually in the flash tube designs I did I used a small inductor in series to limit the current. Else BAD THINGS can happen.