STM32 DAC controlled AC Power Supply / Push Pull Amp --- HELP

[raum]

I came across some videos about electrets and thought about making some.
The process require a HV DC source but as the Electret is curing,the voltage changes and needs to be adjusted.

I have designed a voltage multiplier with 2Gohm sense resister and will be using an STM32F373 as a controller.
The voltage multiplier simply uses a 240v to 6v transformer in reverse to supply 1KV.

Now I need a power source to drive the transformer.
Originally I wanted to use an H-Bridge from a single 30v power rail but found that doing so from an analog signal is impractical.
From everything I have looked at, a push pull amplifier is the best way to get an AC signal.

Why am I using Analog, 48v, Amp Circuit, instead of a flyback transformer circuit and effectively over engineering it all.
1. I want to learn ...
2. Its fun.
3. I have 6 250VA transformers I picked up from Jaycar for 10 Bucks.
4. I can use the power supply for other things.
5. Too dumb to know any better (may be)
6. Maybe someone else may want one too.

Here is the circuit I have so far and the questions.



Firstly will it work?
I feel like R10 is needed so that if no load is connected the MOSFETS will still switch on otherwise there is no Vgs.
What key paramaters do I need for the Mosfets like VGS Max 20v, I max 10A, Vmax +-60V.
Are the diodes simple rectifier diodes or should the be low voltage shotky.


your help would be greatly appreciated
Note: there is already too much negative feedback in the circuit, R5 should be 310K.

[ace1903]

Embedded developer here so no expert knowledge in analog electronics.
Few remarks:
(48v+48v)/10k =9,6mA seems too much for TL082.
Output stage as drawn is highly questionable.
Info about transformer is conflicting(250VA transformer is huge maybe 5kg).
You can not overdrive 6v transformer with 96 volts.
Coil of the transformer is designed for ~250V. At ~1kV dielectric will break down for sure.
I would approach this problem a bit differently.
Instead of DAC I would use PWM and will drive the transformer with standard 12v.
On 220v side small cascade diode capacitor doubler will multiply voltage up to 1kV DC.
Transformer with 2x6v secondary and two transistors driven by complementary outputs from the timer in F373 is simple, nice and clean solution.
 

[Kleinstein]

The circuit does not look right. The power stage would not give more than some +-5 V (the amplitude from the TL081 minus the gate threshold). A 250 VA transformer would be way to large. I would expect a low power requirement.

The more resonable size would be more like a 5 VA transformer and than using a capacitive multiplier. This still has a problem as with a relatively low frequency (e.g. 1 kHz for a normal iron based transformer) the capacitors would be relatively large. The larger the caps the nastier the effect touching the supply.

One could use smaller caps with a higher frequency (e.g. 20-50 kHz range), but this often does not work that well anymore with a mains transformer. It would need more like a ferrite one.

One may find suitable plans for something like Geiger counters or PMT supplies.

[raum]

Thanks for looking at this, I agree the Power transformer is overkill, I just have 6 of these collecting dust.

I tried using SPICE to simulate this circuit and can see where the problems are.
The FETs are arranged to follow the input of the gates and as such the op amp needs to have the same voltage range. I had thought the gates would have been triggered at +-10v difference but as the FETs switch on the required voltage increases as well. the output from the opamp leads the final output.

Also if the output draws 5 Amps and is only at 1 volt, then the remaining 30+ volts accross the FET would dissipate 150+ Watts, the heatsink would be bigger than the circuit board.
to compensate for this, I was considering to use a buck step down and have the MCU adjust as required.

ace1903, Thanks for pointing out the 1KV limit of the transformer, it had not occurred to me, the voltage multiplier has 1KV caps, I was only ever intending to use 24vAC accross the 6V input, the power the big transformer provides was in case i wanted to do other things later on.

I was considering this could be a usefull signal generator / power supply but its crept well out of scope for what I need. I'll try to look into simple low power solutions, I'd much rather use the lab power supply of 24v instead of this.