Using a two-winding choke can be a good idea (I did not check).
For correct simulation, it is necessary to substitute parasitic parameters (resistances) for transformers and chokes.
Also a disadvantage is the galvanic connection. On the sonic earth are large currents (dozens of amperes), which causes interference.
It should also be taken into account that in a real transformer the coupling factor is less than 1 (there is inductance of the winding dispersion).
Also in this microcircuit the control pulses are not symmetrical (the pulse lengths are different), and this will lead to the bias of the transformer.
At first, I built a transformer without scattering inductors. Then I applied the windings taking into account the parasitic inductance.
This sharply reduced the load capacity of the converter. Substitute the coupling factor K = 0.998.
For proper modeling, it is necessary to use a nonlinear core. While this dramatically slows down the calculation. First, I used a nonlinear core.
And then to speed up the account, I switched to linear. The non-linear core has hysteresis loop parameters and dimensions.

 

Using a two-winding choke can be a good idea (I did not check).
For correct simulation, it is necessary to substitute parasitic parameters (resistances) for transformers and chokes.
Also a disadvantage is the galvanic connection. On the sonic earth are large currents (dozens of amperes), which causes interference.
It should also be taken into account that in a real transformer the coupling factor is less than 1 (there is inductance of the winding dispersion).
Also in this microcircuit the control pulses are not symmetrical (the pulse lengths are different), and this will lead to the bias of the transformer.
At first, I built a transformer without scattering inductors. Then I applied the windings taking into account the parasitic inductance.
This sharply reduced the load capacity of the converter. Substitute the coupling factor K = 0.998.
For proper modeling, it is necessary to use a nonlinear core. While this dramatically slows down the calculation. First, I used a nonlinear core.
And then to speed up the account, I switched to linear. The non-linear core has hysteresis loop parameters and dimensions.

ok yeah are you able to edit the file you made with the UCC38084 so that the output voltage is constant with a dynamic load? I really just need this with the currents I think I can find FETs and adjust I just need the dynamic load part working now (varying the load from 1ohm (pretty much no load) to full load of 375W) , if you could please help with this that would be greatly appreciated.

 

I made such a load, imitating the amplifier.
The result disappointed me. Negative stress went into razzing (becoming larger in size) I tried (in two ways) to eliminate it. But it got worse. I came to the conclusion that we need to make two autonomous sources. A little later I will give the results of modeling and schema files. By the way, in this case the power dissipated by the transistor will decrease fourfold.
I came to the conclusion that the negative result was due to leakage of inductances (dispersion)!

 

See

 

See
View attachment 149240 View attachment 149241

Brilliant thank you two quick questions:
1) The diodes used at the drain of the FET: SM15T24A and RBR30T40A, I assume they are there to clamp the voltage there, are both needed or can only one diode be used here?
2) Is the third coil used for feedback critical (I assume so that a similar transformer can be used for both the Positive an negative)? Can I just use isolated DCDCs practically (smaller power that I can buy online) if so what voltage do they need to produce?
3) I also noticed when measuring the current at the drain of the FET it goes negative and some points is this something that will appear on the simulation only and practically won't happen?

 

See
View attachment 149240 View attachment 149241

Also would you mind listing the properties of the transformer so I can get transformer built up please.
Kind Regards
Art

 

A=100u Lm=150m Lg=0 mU=2000 A=Area=100u=100mm^2, Lm=150m=150 mm Effective length of the magnetic line,
mU - Effective relative permeability of the core.
Lser=110u Rser=100m Cpar=10p - Lser= Lleak =110uH=AL*N^2*(1-k)
AL~1.5uH*190^2*(1-0.998)

 

When the scheme is turned on, extra currents and extra voltages develop. In a stationary mode, this is not the case.

A=100u Lm=150m Lg=0 mU=2000 ==> A=area
=100mm^2, Lm=length=150mm,
mU=2000 effective relative permeability of the core

Lser=110u Rser=100m Cpar=10p Lser=Lleak=AL*N^2*(1-k)
AL=mU0*mU*area/length

 

When the scheme is turned on, extra currents and extra voltages develop. In a stationary mode, this is not the case.

A=100u Lm=150m Lg=0 mU=2000 ==> A=area
=100mm^2, Lm=length=150mm,
mU=2000 effective relative permeability of the core

Lser=110u Rser=100m Cpar=10p Lser=Lleak=AL*N^2*(1-k)
AL=mU0*mU*area/length
View attachment 149255

ok thank you very much will D2 need to be replaced with something else for both circuit the power seems really large?

 

Try to change the diodes to others. But you should apply the calculation from zero voltages on the capacitors for the study (V0=0).

 

Try to change the diodes to others. But you should apply the calculation from zero voltages on the capacitors for the study (V0=0).

Ok It's a little tricky I will try to do that have been trying it at the moment, it's a lot trickier than I thought, not sure if you could also help with this as this is the only thing I can see that needs to be fixed with this design. I also changed the FETs in the attached schematic, I also put in a constant load (max power) just to compare the load will in reality be dynamic.

 

Try to change the diodes to others. But you should apply the calculation from zero voltages on the capacitors for the study (V0=0).

Will the FETs require a diode in parallel and snubbers?

 

Try to change the diodes to others. But you should apply the calculation from zero voltages on the capacitors for the study (V0=0).

Also is there a way to add a feature say a FET that allows the power converter to be turned off via external control?

 

 

View attachment 149324

Great thank you so much quick questions about the diodes used on the drain of the FET the TVS diode seems to be fine in terms of ratings can you recommend a different diode for the schottky diode I think the ratings are exceeded? Is it also worth adding a snubber circuit across the FETs or will the be unnecessary?

 

Snubber chains always dissipate parasitic power. The limiters I apply work only when the device is started, or in the operating mode, but very briefly. It is energetically more profitable than using a snubber.
Attention. You wanted to control the converter. In the last figure, I showed how to disconnect (turn on) electronically a weak signal. Also it will allow using a low-power switch to control this powerful device (instead of a switch for tens of amperes)