Cautionary tale of unintional pulse power and hearing loss

[davekni]

I purchased some bulk caps on EBay for DRSSTC use, so was testing them.  While connecting a resistor to discharge a cap after leakage-current testing, I bumped an alligator clip across the capacitor terminals.  The resulting bang hurt my ears, and it now appears I will have permanent hearing impairment in my right ear.  The capacitor is rated 20mF (18mF measured), and probably had between 350 and 400V charge at the time of shorting.  (Was charged to 450V, but it had discharged some before my accident.)

I regularly wear hearing protection when running my DRSSTC or Marx generator.  Going forward I'll wear protection in many more situations.

[Mads Barnkob]

I wear safety glasses and ear defenders in all cases of testing and high power/current/voltage. This is after several experiences with exploding silicon, flying scrapnels of plastic casings, copper pieces and arc flashes.

Worst was an underestimation of the remaining charge on my 4kJ maxwell capacitor, so with ear defenders around the neck after a shot, I just wanted to short the remaining out with the earthing crowbar circuit, it was however around 3kJ left on it and it was ear deafening for hours. No permanent damage that I can tell, short pulses are usually not damaging unless you have bleeding ears, compared to long exposure like going to a rock festival for 8 days straight.

[Hydron]

Ouch, sorry to hear that, hopefully it's something that recovers somewhat over time.

I have a rather similar capacitor (well a bank of 4 smaller ones) - 24mF @ 450V. When I did the maths on the energy stored it made me nervous, so I've recently build an over-current tripping device (using a massive low-Vce(sat) IGBT and DC current sensor) set to interrupt things in a few micro-seconds if it sees fault currents above what I'd expect to be drawn during use. This way I can use it to power my QCW coil (especially during development) without quite so much fear of loud bangs and flying silicon in the case of a mishap. I think I'll still don the safety glasses and ear defenders when using it at higher voltages though, at least during development.

[MRMILSTAR]

Sorry to hear about that Dave. A 2 KJ shot will definitely wake you up! Just one of those caps would make an excellent can crusher if it can withstand the voltage reversal. When you wrote "mf" I assumed that you meant 20 microfarads but then I looked at the link and saw that it was 20 millifarads!

Even though the solid-steel blast box on my coin shrinker greatly reduces the noise and contains the shrapnel, I still wear ear defenders and eye protection when operating it. Its still quite loud when firing at around 7 KJ, about like a small pistol. Before I improved the blast box seals I actually got injured by a piece of copper shrapnel escaping through some hair-width seams in the lid.

[klugesmith]

Sorry to hear about the accident.
Could be worse -- thread on laserpointer forum by an enthusiast who got a permanent vision defect from momentary accident, and on chemistry forum by a guy who lost most of his hands while preparing an explosive for the 10th time.
Steve, re. can crushing: I think there's a huge efficiency penalty when using an electrolytic capacitor, no matter how many joules.  Can't make a work coil with less than 1 turn to match the relatively low voltage, and I think the parasitic R and L don't support 10 microsecond discharges.   

[MRMILSTAR]

Rich, upon further thought, at a minimum the capacitance (20,000 uf) is much too high for can crushing. As you stated, you can't get few enough turns on the coil to get a resonant frequency high enough for a short pulse. As you know, my can crusher uses a 100 uf non-electrolytic cap with a 3.5 turn work coil which resonates at about 8.3 KHz which yields about a 50 us pulse to the first peak of the damped sinusoid.

All that being said, even with non-ideal ESL and ESR, electrolytics are commonly used (not what I would use) for can crushers and disk launchers even though they do experience voltage reversal which probably shortens their lifetime considerably. Their design details are usually not divulged but you can see electrolytic capacitors in their videos and images. There are several on Youtube.

[davekni]

The capacitor I accidentally shorted measures 18mF and 4mOhm, so a 72us time-constant without any added external impedance.  Would likely be possible to shrink coins with a few of these in series, but efficiency and life would be poor.

For coin-shrinking I have an old 14uF 20kV oil/paper pulse capacitor.  It sits in the corner of my garage.  I charge and trigger (discharge) it from inside my house.  No people are in the same room when it's charged.  Discharge lasts ~14us into a 17-turn work coil for US quarter-shrinking.  Between myself and different groups that come to see it, we've shrank ~1300 coins over the past 18 years with no safety issues.  I just wasn't in the habit of being as careful with electrolytic caps.