Microwave Oven Capacitors

Microwave oven capacitors are known for being somewhat un-usable for pulse applications, but I choose to use them anyway because they are cheap and can store a lot of energy if you push their ratings.  My destructive testing of a 1700V .85uF Aerovox capacitor (from a used MO) showed that these caps fail at about 10.5kV after repeated discharging (into a short circuit) of the capacitor.  The capacitor easily contains the breakdown, only a small snap can be heard.  The capacitor was storing 47J at this point.  Another test was to measure the ESL of a typical MOC (microwave oven capacitor).  This was performed by charging the cap to 500VDC and discharging it through a 3" length of wire (which had a current transformer on it).  The resulting waveform was captured and the resonant frequency was measured.  Working backwards I found the ESL to be about 180nH, not bad at all!  This is despite the fact that there are only 2 tab connections to the foil plates.  I assume that the close proximity of the foil plates works similarly to a coax conductor, so the overall inductance is low.

With these test figures in mind, I decided to see what MO caps are capable of.  The bank consists of 18 parallel caps, each rated 1uF at 2200VAC.  At 8kV, this amounts to 576J.  The switch is driven from a solenoid, and the contacts do not touch, but get *very* close.

Would this be enough to crush a can?  Certainly!  Not only that, but it tends to rip the can apart.  Below are some pictures of my setup and a can.

Ok, it can crush some flimsy aluminum, but how about a dime?  Yep!

This was with using 12 turns of 20awg magnet wire as the work coil.  Warning, wire shrapnel is very dangerous and must be properly contained!