Original poster: "resonance" <resonance@xxxxxxxxxxxx>You need to use Erms x 3, so 6 kV x 3 = 18 kV DC pulse rating for the capacitor. 10.5 kV is not enough. I also suspect you may have used a polyesther dielectric cap and they die rapidly as you discovered. Check with Sprague to determine the caps dielectric and construction features.
You need to use polypropylene or polyethylene diectric caps with metallized foil to handle the current. Otherwise they heat up rapidly wih the RF currents and the dielectric begins to melt. I trust you were running your coil. If you didn't have a sec coil in place the rise in Q factor will blow caps or transformer (sort of like holding you gas pedal to the floor in a car in neutral --- the energy has no where to go, ie, no load). In RF circuits without sufficient load the potentials and currents can rise to very high levels, typically the Q factor times the applied potential.
Most experimenters use Cornell-Dublier capacitors, typically 0.15 uF 2 kV, with typical of 16 pcs in series at 12 kV Erms. With your smaller xmfr potential you could use 12 or 14 in series.
Dr. Resonance
Are you guys using any special type capacitors?I built my first MMC capacitor today it lasted a whole 2 seconds then went up in smoke. My power supply is a 6K 20ma transformer. The capacitors are .1uf 1500VDC Sprague 430P 227084-43F. I soldered 7 capacitors in series to get 10.5K volt rating. My TC is a tiny little thing I built 5 years ago it had an oil filled poly cap.Here is a picture on the TC. http://i43.photobucket.com/albums/e358/gary350/tc4.jpg Notice the small TC in the bottom left with a 1.5" diameter 5" tall secondary with a door knob top load.Gary