On Sat, 18 Aug 2012 01:23:15 -0700 (PDT), wrote:

I need to be around 60 - 100 Khz. There will be no active feeback loop, no regulation just brute force. Primary impedance will be so low that any spikes in the load will be immediately reflected to the source. It will run at resonance.

Square or sine, or something within reason at a constant frequency, able to deliver say 200 V P-P at eight amps or so.

Actually 75 Khz would probably be optimum.

J

In order to run resonant, you'll need some kind of control to sync the
driver to the power transfer frequency. Self-oscillators tend to be
supply volage dependent, and the resonant frequency of the power
components will drift, all by themselves, with temperature.

Resonant converters tend to run intentionally above or below
resonance, FMing into the resonance to increase/regulate power.
Ungapped ferrite is unsuited to this application. Pick a ferrite core
shape that allows for introduction of a gap, or use a powdered core
material of suitable mix permeability and loss density for the
application.

You have to determine what you need, before you can go looking for it.

RL