DRSSTC primary won't oscillate at correct frequency

[TMaxElectronics]

So after fixing the issue of the IGBTs blowing up I put the small coil together last week but for the life of me can't make it work... I've tuned it a couple of times to different frequencies but can never get it to oscillate correctly.

Coil parameters:
Primary cap 100nF
Primary inner diameter 140mm
Primary winding pitch 18mm
Primary winding count 5.3 (tried anything between 5 and 6)

Secondary length 350mm
Secondary wire diameter ~0.22mm
Secondary Winding count 1495

Toroid tube diameter 120mm
Toroid outer diameter 420mm

I'm also running a very sharp breakout on top of the toroid

Without the secondary in place it oscillates at 186kHz (which is a tiny bit under the 187kHz fres of the secondary). but as soon as I put the secondary with the topload into the primary the frequency shifts to 200khz. If I then tune it lower it starts oscillating at around 170khz. It almost seems like it actively avoids resonating at the secondary frequency...

Here's the plot of the UD3's tune sweep (fixed frequency, ~40V on the bus. Current is on the y axis with a peak of ~100A, frequency on the x in kHz) with the secondary in place:


And this is how it looks without and secondary resonator:


Seems extremely weird to me that the coil would pull so much energy out of the primary tank circuit that it wouldn't even oscillate at all at the right frequency.

Coupling is adjusted to ~11%, and I've tried lowering that even further without any luck. Trying to run it outside (with adjusted tap point, and better earth) also didn't make a difference. Earthing is done to the water main in the basement and directly into the ground next to the coil for the test outside.

Driver is a UD3, adjusted so the start oscillator runs for the first few cycles until feedback current exceeds 25A.

Has anybody seen such behaviour befor or has any idea what might be going on?

ADDITION:
I just tried two more things. One is to increase the cupling to perhaps 20%, still no change. Other is to try and fix frequency excite the primary, so I turned off the start cycle limiter and up the minimum feedback current with ~50V on the bus and the coil behaved extremely oddly:

Yellow is voltage induced into a loop of wire underneath the primary, sort of indicating current. Oddly enough it actually drops off and completely stops oscillating at all.

And another addition:
i tried tuning the coil much lower again, which at some point does result in some small streamers, but only around 30cm even at 800us ontime... definetely not the performance I'm expecting from a coil this size :/

[Mads Barnkob]

I assume that you checked phasing in all the places it could be reversed.

What about your MMC/resonant capacitor, does it check out good when under load? Could be that it behaves fine when just sitting, but does weird stuff when loaded with current or at high voltages.

Did you try a 10K power resistor across the inverter output, to drain any residual charge between pulses?

[coilerer]

I'm pretty new to this stuff, but could it be that the driver is jumping to the upper pole in case 1 and the lower pole in case two (when you lower f_res of the primary)? Not sure why you aren't getting streamers though. Maybe flip the primary leads?

Case 1 - "but as soon as I put the secondary with the topload into the primary the frequency shifts to 200khz."

Case2 -  "If I then tune it lower it starts oscillating at around 170khz. It almost seems like it actively avoids resonating at the secondary frequency..."