Help me correct my failed Inverter design

I have made an inverter but failed. The problem is IRF3205 MOSFETs always broken if I activate the inverter. My main inverter components consist of:
- step up transformer (11V CT 11V/20Amps to 220V 0V/2Amps)
- 2 IRF3205 N-MOSFETs
- MC34151P MOSFET Driver
- AT89C2051 as modified sine-wave generator.
- 6 12V/7,2Ah batteries
I would like to make a simple 500VA 12VDC-220VAC Inverter.
Can anyone help me to correct my schematic? Thank you very much.

irf3205 inverter

I recommend you to download and read a great book of Pressman:

what part of an inverter normally fails

The book that you recommend no longer available anymore because it was removed.

fet avalanche serial

example of inverter

why do mosfet fail in inverter

Hi aan_woodz, I think you need to place current sense to each source pin of the mos-fet, and feedback the voltage to the controller. When the controller sense the maximum current flow to the mos-fet, it should turn off the mos-fet. It will protect the mos-fet when it reach the maximum allowed current.

Re: Failed Inverter

Hello,

the transformer properties as well as the PWM frequency can't be seen from the schematic.

I also would add overcurrent protection, as suggested. Another point is that sinewave PWM implies normally an inductance (could be the transformer stray inductance) acting as a filter to achieve a smooth output current. The stray inductance sources a reactive current in the MOSFET, that possibly cause current flow in the FET reverse diodes. As these diodes are slow, excessive losses can be produced. On the other hand the stray inductance between both primary halve windings causes a positive overvoltage at switch off. Depending on the inductance, it may cause avalanche of the FET protection diodes and additional losses.

Antiparallel schottky diodes and snubber networks could reduce both effects.
However if the MOSFET are destroyed without output load, initial overcurrent is the most likely the reason. If the transformer has a tight dimensioning regarding winding count, it could get saturated when applying the first half cycle of output sine wave. To prevent from saturation, a transformer should be switched on at the peak of input voltage, this could be also considered with the start-up procedure.

Regards,
Frank

Failed Inverter

Frank's advice is good.

Also check your code - you need to make sure the mosfets on/off timing is correct and that they are in a known state during turn on and initialisation.

oddbudman

Re: Failed Inverter

I just now checked again. Pressman's book is still available in edaboard.

Failed Inverter

what is thebattery voltage
connected parreall of serial