help with 4Mhz Class E SSTC

[iWei]

Hi all, back here trying to build a 4Mhz Class E SSTC, but im confused how to tune the values of the power section since I see all kinds of subharmonic oscillations and squegging no matter what I do (change rf choke inductance, add/remove DS capacitance, vary primary turn count, reduce secondary turns)

Granted all the above tuning was at very low voltages (0-20v limited to 1.5A) as Im down to my last 3 mosfets. I dont have a variac but I could use a buck converter to step down 60v but it is not current limited

Ive heavily referenced the EVR/uzzors2k circuit

Gate drive: CD4049 crystal oscillator with TC4422 mosfet driver, using antenna coax soldered directly below mosfet body, the signal looks ok to me.


Power section:
decoupling cap: 100nf
RF Choke: T106-26 toroid with 20Turns(gets pretty toasty maybe i should switch to type 2/red-black material?)
Mosfet:IXFH22N65X2
DS capacitance:660pf(Is it possible to calculate a good starting value since different mosfets all have different output capacitances?)
DC blocking cap: film 12nf 1.6kvdc
Primary turns: 4



My secondary is at approx 4Mhz self resonance frequency according to this
https://www.teslascientific.com/products/coil-resonant-frequency-calculator/
50mm height
62.9mm diameter
105 turns


Is there an algorithmic way to tune this coil while keeping mosfets from unaliving themselves?

[davekni]

Granted all the above tuning was at very low voltages (0-20v limited to 1.5A) as Im down to my last 3 mosfets.

Yes, I recommend low voltage initially.  FET Cds will be higher, so you might need to increase external DS capacitance a little for high-power use.
I've built two fixed-frequency class-E coils (6.78MHz and 13.56MHz).  For both, I found the best way to tune was by starting with analog simulation (LTSpice).  Simulation showed what waveforms will look like when tuned high and low.  My guess is that you are tuned too low, but it's been a while since I was staring at such waveforms.  I do recall subharmonic oscillation when detuned.

Gate drive: CD4049 crystal oscillator with TC4422 mosfet driver, using antenna coax soldered directly below mosfet body, the signal looks ok to me.

Looks fine presuming scope/probe is showing 5V/division.  20MHz bandwidth may miss some ringing.  Likely still good.  Tuning does appear to be your key issue.