ENERGY AMPLIFICATION

[Dave45]

  for every problem there is a solution, we just have to find it.
This may or may not be it.

[MarkE]

For the coil currents to get back to where they started greater than zero each cycle the V*T product across the coils during the off time has to match the V*T of the coils during the on time, CCM.  If the V*T during the off time exceeds that of the ON time then the coils stop conducting by the end of the off time: DCM.

The capacitors accept charge from the coils during the MOSFET off times.  The amount of energy that can go into a capacitor in any half cycle is determined by a second order differential equation.  But in basic terms the capacitors each have to charge up enough so that the charge added each cycle matches the charge drained out.  Usually the safest thing to do if you are just experimenting is go bigger than needed on capacitor value.  A better solution is to calculate what you need.  If you let me know more or less the parameters that you know then I can suggest some values.

I would not go with the arrangement that you propose above as it uses a flying capacitor.  What you might do is walk before you run and build-up just the top half or bottom half of your circuit at first jsut to keep it manageable.  See what kind of results you get from that, what you need to fix and get yourself on a solid footing before hooking up the second half.

[wattsup]

@Dave45

What about this way to cut the cap off while the inductor is being charged?

wattsup

[Dave45]

I see what your saying, in my last diagram the cap would probably never fully deplete.
Your diodes may be the only way.
Will ponder the situation awhile.

[Bob Smith]

Hey everybody
Here's something interesting posted by our friend Mags:

Tito once told me in pm that, yes, he uses a zener as a spark gap.  So possibly putting it all together here, we would need a rectifier diode in series with a reverse biased zener(probably high voltage), in which the rectifier diode forces the one way street and the zener stops conduction from the cap at a particular voltage level. One more item would be needed. A timed/voltage controlled  switch in series with the diode and zener. The trigger level to turn on the transistor would be what ever voltage that is higher than the zener voltage that you desire. This would effectively produce a disrupted discharge. Say the trigger turns on the transistor at say 300v, and the zener cuts it off at 250v. Just an example. The transistor should stay on after trigger voltage is met, possibly also  triggered off when the zener cuts out. Would have to be worked out.

Source: http://www.overunity.com/13460/teslas-coil-for-electro-magnets/msg416153/#msg416153
Bob