I think it must be someting linked to the q factor of converter transformer cause i found that even at higher output power of the converter, but different winding the sony power supply can have a cleaner output and no squeaking.Usually power supply  phase goes wrong when abused and the scope can't sinc properly for a clear signal picture.

apart from 6 X 470uf paralleled capacitors of the power supply , i have another 3 x 6800uf paralellled capacitors...

That amplifier topology was invented by Helge Granberg at Motorola in the early 1990s
They look like 9:1 transformers in this case if tyhey have 3 layers of coax each side 2 layers = 4:1 3 9:1 4 16:1 coax impedance has to be the arithmetic mean of the two impedances involved so different designs use 25 ohm 15 ohm and 10 ohm semirigid coax.
 Helge did reccommend that the input transformer was ferrite loaded due to the increased gain at low freqs to help stability

look up mid 90's motorola RF databook MRF141G MRF151G application notes the G suffix indicated Gemini package for push-pull applications

Thank you Sir!

I've never had repeatability issues with LTSpice.  It is a bit cryptic - many hidden commands that aren't officially documented.  However, there are great online resources - wiki and user groups.  It is the most powerful free simulator to my knowledge.

For optimum convergence, the 2016 release is better than later ones.  I've reverted to that version recently.  Don't know of any useful added functionality since 2016.

I am on the last spice version...
As for your converter design i can only say thank you! I finished the circuit and tried it today with 2sk941 as driver and 2 paralleled irfz46 on each side.Much better power transfer, much higher achievable frequency range, less heat dissipation.I tried it from 49khz to 190khz.At some frequencies it interfered with the power supply switching frequency which got squeacky at some frequencies.The power supply operates at 220khz and it produces less noise when the converter is operating at a quarter or half its frequency .At around  50khz , the converter got te best power transfer with transistor switching losses being the lowest so the transistors were  cold there. I may consider to modify the power supply frequency and hopefully it will work.At higer frequencies the irfz46(with no heatsink) would get warmer but not by much...
I got +-11.5v with one winding and  a doubler. 420v on 20 kohm and 6v on 3 ohms which is all i need for my phono preamp.I will use this converter to check also an an electrostatic headpones amplifier that needs 600 vdc for anode and 12v for fillaments which was halted in development a year ago due to the lack of a suitable supply.

The 100v cap , the bigger one at the right is the low loss one.It might be polypropylene foil or polystirene foil...i ave no idea but usually the low loss capacitors are the bigger ones.They are from a very high quality resonant Hitachi  power supply for high speed trains converters.

Cool, now i know the name so i can download some tech papers to maybe understand that better.
The history of RF pumped lasers starts with UK based AEA technology bought probably by  Rofin Sinair and implemented in their  transverse gas  flow lasers too  which used  parallel anodes placed at about 2mm one from another working as a capacitor which was in series with some paralleled inductors resonating at a fixed frequency delivered by this amplifier in short PWM bursts up to 100khz.With the help of a UV ionizing lamp or a secondary RF amplifier the laser discharge was initiated.This system provided a very high quality laser compared with other pumping methods. You can check their early patents:
https://patents.google.com/patent/US4288756
https://patents.google.com/patent/US4077018
Not really sure about the whole history, but transversal flow was better to be used with an RF pumping system.
COHERENT , based in the US used same technology later , got better bussiness and a few years ago aquired its mother company, Rofin which had better but more expensive technology.
The second patent is the origin of later circular flow Trumpf lasers in Germany which integrated very precise CNC in their machines.
You can check any patent linked to rf  pumped transverse flow lasers online like this going to Raytheon:
https://patents.google.com/patent/US4651325
I'm probably  wrong about who's the true inventor of this tecnology, but that's not really important today.

I'd be interested in the schematic for any of  their 40w+ CO2 laser transformerless(no ferrite core)  amps like this one:
/>I used to fix their lasers 12 years ago and i remember that kind of amplifier being their last embodiment which blew me away at the time.Looks like their RF manufacturer is alpha telecom apparently doing the amps for yaeshu...I'm not knowledgeable in rf field although i used to have a HAM radio operator licence, but i'd like to learn a bit more on that.

With these types of symmetrical circuits the transient sim can sometimes get stuck in perfect balance - since all the parts are mathematically identical. It helps to deliberately unbalance the circuit (like the real world) to get things started. I find adding a few pF to one of the power fet gates is usually enough.
Oh, and V1 should be a single step from 0 to 12V, not a fixed 12V.

And always include the series resistance of any inductors, even just a token milliohm - that helps with convergence.

here's what i "love" about LtSpice: restarting it next day all its results are screwed:

And always include the series resistance of any inductors, even just a token milliohm - that helps with convergence.

Indeed , i couldn't make even one transformer work untill i added some 0.2 ohm to the input inductor, then naturally added simillar series resistance to other inductors too, also added some 5...10pf winding capacitance too as i had some convergence rules already that didn't help enough.I still need to find out how to properly sim two transformers in parallel and actually learn to sim true transformers.Just saw some CUK material on TDK PC40, but its a bit too theoretical for me at the moment:
https://core.ac.uk/download/pdf/4876634.pdf

loading a single transformer there's no waveform anymore...
until i add some parasitics to input inductor and the output looks as it should.
also the driver power dissipation looks pretty good to me at half the load althoug at full load some oscillations show up but i am really far from a true simulation of te real ting anyway.Now i'm more confident.