12 to 220 TL494 - drain waveform problem

[Vlado391]

Hello,
I'm trying to build a 12V to 220VDC 100W converter with TL494. For input source i'm using 150W 220VAC-12VDC smps. The load is a light bulb.

Works fine when the input voltage is low (maximum duty cycle) - efficiency around 92%.
After increasing it the waveform of the drains getting awful. The efficiency drop to 70 percent and the transistors starts heating up. The input current stay almost constant. Under 0.15 duty cycle its even disgusting for looking. For some reason the drain voltage of the first transistor dying when the other switching off. No matter what is the duty cycle, the drain voltage stay the same. Maybe thats why the input current stay constant. I already tried with different transistors, IC and transformer - same result.

Currently i'm using an EI-33 transformer from old ATX power supply. It's working in half-bridge so i'm expecting 155V, but it's around 65-70V. The middle point is connected to VCC and the two 12V to the drains.

I'm using optocouplers for driving the transistors, because I didn't have any complementary transistors. I tried even with dual power supply to shut down the transistors (+-12), but with no effect.

For explaining the problem in his best i added a potentiometer on DTC pin and removed all feedbacks.
YELLOW (gate)
BLUE (drain)

Questions:
1. What may be the reason for this? What i'm getting wrong?
2. Why the drain voltage is 2 times bigger than the supply voltage?
3. Why the output voltage of EI-33 is only 70V?

Thanks in advance!

 

[BradtheRad]

i'm using an EI-33 transformer from old ATX power supply.

1.
It looks as though spikes at high voltage are produced due to abruptly shutting-off current through the transformer.
The greater the current, and the more sudden the shut-off, the more severe is the spike.

2.
Operating a power transformer in reverse seems like it ought to work, however it appears we must not expect as much Watts as normally rated for. Notice we're trying to make a power source from a winding which is thin wires and many turns. It's designed to present high impedance to the electric company (low impedance):

I tried to homebrew the same topology of power inverter. When I applied greater voltage, I started getting severe spikes. Power throughput was meager. I gave up after I'd destroyed several mosfets.

 

[Easy peasy]

the big issue here is you know nothing about the transformer, are you driving the supposed 5V windings or the 12V pins somehow? in any event if you apply too many volt seconds you may run the Tx to saturation which causes the driving side currents to spike up and hence the heat etc.

You appear to not know the turns ratio - what you see is what you get, if the tx was in fact for a flyback you may never achieve your aims ...

--- Updated Sep 25, 2022 ---

for any push-pull the Vds is naturally 2x Vin due to the transfomer effect - i.e. if you put 12V on one winding you will get 12V on any other winding with the same turns - if they are in series they will add.