Sprite Small DRSSTC

After Finding some cheap small IGBTs on ebay I decided to build a small DRSSTC

Secondary: 900T of 0.125mm enamelled copper wire on a 40mm diameter PVC former 127mm Tall.
Primary: 6T 1mm copper.
Topload: 6"x1.5" Torroid
Primary Cap: 22nF @3KV  (4x47nF @1.5Kv in series parallel)
Transistors:  2 x 4PH30KD Half bridge.
Fres: ~500Khz

The coil is mounted on to a 5" x 8" box that contains the half bridge and the driver. The corners of the box are surrounded buy 12mm copper tube, to act as field control as the coil will arc up to twice its secondary height.

As I was trrying to make a small DRSSTC I wanted to keep the electronics as simple and small as possible.

Primary feedback would have involved a couple of large CT's and I'm not sure how well they would have performed with such a small coil, so I settled on Secondary feedback.

A simple CT driving a logic gate is used in many circuits, but It always gives an un-even drive signal so I made a symmetric circuit with two CT's hopefully to reduce this issue.  

As the IXDD414 has an enable input all I would need is to drive this enable from an interrupter. I have seen many designs that interrupt the GDT drive directly from the interrupter, but I didnt like the idea. As the IGBT's were fairly small and I was driving them rather hard (5x rated max) I thought  a syncronous turn off was a better option.

The Driver circuit can be found  here 

The coil runs very well indeed, the only problem I have is that I have made an interrupter to give 200uS pulses as used in bigger coils. Unfortunatly at 500Khz this equates to 100 cycles.
If I wind up the on time to 200uS I get two loud bangs from the IGBTS.

I need to change the max timing to around 40uS to give around 20 cycles. This should limit the primary current to around 150A which appears to be ok for these IGBTS.

When running the heatsink on the half bridge dosen't even get warm to the touch.