Re: build a flyback?

> Build a flyback with a silicon steel core? I doubt it can be done....  
> ....flybacks use a very different core material, called ferrite. Flybacks
> are designed as very high frequency transfortmers, while those mains
> trannie cores are probably for only 50/60Hz
Thanks, to all others who answered my question! 
I nevertheless couldn't keep from trying it anyway :), and with the E-I
you are right, it doesn't work at high freqs. 
But, I managed to run an unchanged pocketsized O core transmormer
(220V->5V or so), which gives about 2kV at 1.5kHz, but with very little
current output... maybe that's the way to go?

>  and while the gap breaks down 3 times/sec, 
> have you tried fiddling around with the capacitance? 

Not yet, but I intend to. 
I've a [crappy] theory on increasing streamer length: have a larger cap +
a more powerful PSU, adjust primary inductance accordingly so the f(res)
remains  unchanged. The spark gap size and break rate kept about the same
as before. And the size of the secondary should remain exactly the same.

Then, a very small current limiting resistor (no idea, maybe limit to
50A or so?) between cap and primary inductance so that the cap discharge
time will be longer, meaning that the streamers from the topload
(i.e. candy box) will have power pumped into for a longer time and thus
hopefully will grow larger, before power is cut off and they break down

Could this work, or is it complete nonsense?

The reason is that currently, I found that after the 1.7" to air streamers
there's some additional, very pale, white 3" long streamers (only noticed
them after bringing a fluorescent tube a bit too close... ouch... ;o)
So, I hoped a longer resonant time would let them get more visible?