[SOLVED] Faulty Inverter Help

Trying to repair an old inverter of between 1Kw and 1.5Kw with a bank of 16 Mosfets working into 8 transformers to generate the HT on 3 x 47uF, I have got this section fixed though an IC marked CS2638 which drives them had me puzzled but seems to be a KA7500 type in disguise.

This section now works what seems OK as it can regulate then rest to maintain 300+ volts, a LM324 plays some parts in this inverter with slow start indicators and beeper.

As yet though I have not got its output end to work right, what seems to be a bank of H format Mosfets x 8, it did have an output of a modified sine wave but only seems to produce a narrow pulse in AC form (looks to be the peak pulse parts). What is lacking is the lower voltage wider pulses like rms section which I think should be from a second KA7500 but is only sending those narrow pulses.

The high voltage mosfets seem OK by basic measurement just need guidance and help to understand a bit better this section, found some help on your pages from a circuit included in a previous thread.

The only clue is some 1w resistors feeding rectifiers are cold and a 30v zener has only 20/25v across it so is lacking some drive force have measured parts in the output "front line" but nothing as yet, changed its three chips keep wondering if those mosfets are sick in some way that is not a short?

Thanks for any help.......

Deltatango :???:

Check the bootstrap capacitor on Boost pin and isolation diode to Boost.

Make sure you have BG drive waveform to bottom MOSFET. The bottom MOSFET charges the bootstrap cap (through isolation diode) when bottom MOSFET is ON. When bottom MOSFET releases the bootstrap cap provides the upper driver power supply for the top MOSFET. The bootstrap cap on Boost should have 15v for the high side driver.

Cracked it, the 0.22uF (electrolytic sort) allied to its 30v zener was to blame, though acting as a "capacitor" in checks turned itself into a resistor when voltage more than my meter (to see charge up action) was applied, changed it for a normal poly type. Also changed 2 x 100K feeding it plus a 16K with 1% metal film now output is 230v with a respectable wave form.

I have changed other "flaky" (technical term) looking capacitors with good brand low ESR types, going to improve power rail tracks with wire support bridges as this copper track is lousy thin. Thinking of lowering its 8 x 30 Amp fuses down to 20 Amp which at 12v gives a total 160A with 1,920W (if my calculations are correct) so a cushion of 400W above its 1500W rated output, at least they may blow next time rather than cleaning up its FET,s legs splattered all over.

Can anyone suggest any other tips to improve its durability please?