Flyback charge pump for TC PSU

["LOCKWOOD, Chris J." <clockwood-at-orbeng-dot-com.au> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)]

Hi,

Does anyone have any experience with flyback converters as the main PSU for
small/medium power TC?

My initial impression is that this would be a fantastic way of delivering
adjustable power to a coil, without the dramas of big heavy xfmrs /
autoxfmrs... 

I built a coil using 3 car ignition coils, running off a 30V 15A PSU in
flyback configuration- through a simple cct using IGBTs. The secondaries
pumped up the tank cap through a series of 50 1kV 200ns diodes (resistored
etc). This was decoupled through a ~10mH inductor, and 1nf ceramic cap stack
to save the diodes (from excess dV/dT), and prevent power being pulled back
out of the tank.

The idea was to get a higher peak power (for longer sparks?) with a lower
rep rate by using a big cap.

The results aren't anywhere as impressive as I would have hoped. My tank cap
is 100nF (quite big)- an oil-filled type scored from a university physics
dept throw-away. I can only get 15-16" at around 20 zaps/sec, when the coil
is tuned to ~150kHz (static gap, ~30cm diam primary, tapped at 4 turns
(+-1/8 turn) for best results)...

I'm thinking that perhaps the cap is crap. After discharging from 10KV
(shorted for a second) the voltage rises again, to approx 1kV. Is this
normal for a HV oil cap???

Does the rep rate affect the spark length significantly?

Cheers

Chris
----------------------------------------------------------------