Electronics help required with Sofar ME3000SP

[Tinbum]

Now with one FET fitted in - low side

Right of working diode




Right of diode D58




left of diode working



Left of diode D58



Across outputs




Diode voltages

[Marcus]

Ok, i think those scope traces seem reasonable. If the left of d58 is showing a steady 13ish volts for the gate drive, then we don't have a short after all - could try reconnecting pin5 of the opto and see if there's still 13v and if there's a gate drive square wave.

Although I'm not really making sense of the voltages diagram - im assuming the outs are the output terminals which are common with the sources of the high side fets. And the voltages are to the high side diodes. Suggests there's now 19v on the high side gate drive supply.

[Tinbum]

Ok, i think those scope traces seem reasonable. If the left of d58 is showing a steady 13ish volts for the gate drive, then we don't have a short after all - could try reconnecting pin5 of the opto and see if there's still 13v and if there's a gate drive square wave.

Although I'm not really making sense of the voltages diagram - im assuming the outs are the output terminals which are common with the sources of the high side fets. And the voltages are to the high side diodes. Suggests there's now 19v on the high side gate drive supply.

Should have marked it up better.



The 19v reduces as I measure it- should that happen? (As does the voltage across D58)

I will try connecting pin 5 as you suggest, I suspect it won't change anything (from past results) which I guess means it could be the opto faulty. Would that result in those FETS being on? They would have to be to cause them to create a short with the low side fets and all 4 blow. I think i did put the scope on the gate in the past and it didn't show anything.

[Marcus]

Yes i would expect the 19v to drop under measurement down to a more normal 13 14v ish. Because there's no load on the gate drive supply the caps are charging up to the nominal 13v ish plus any transients, or ripple. It does beg the question: is there no zenner or tvs diode there to limit the voltage. It would seem there are no shorts in the gate drive components that are connected.

So as far as i can tell there's no faults outside of the gate drive circuit and fets. It's a little hard to be sure about the low side gate drive but if the gate drive is a good clean 13v ish square it's probably ok*.

We still don't have the full circuit for the gate drive - the totem pole outputs might not be common to each other but drive the gate each via it's own resistor near the fet gate; and there may be another semiconductor there possibly a zenner diode to limit the gate voltage.

Hopefully we will find something faulty as there must be a reason for the fets blowing.

*Without a fet, or without the 50v drain voltage on the fet it's difficult to be sure, but maybe test its shape once the high side is working, or by using a pullup resistor on the output so you can see the gate waveform with the fet switching a voltage.

[Tinbum]

Done a few more tests. This is all with no fets in Bottom high.

The traces dont look bad (i think) which has me baffled. I may put a fet in and try (i've put a 1a fuse in the DC supply)

Top high to top low doesn't look like bottom high to bottom low, which looks ok, which I think it should do. Is that because the fet is missing on the bottom high?

Top high and top low shouldn't be on at the same time. Bottom high and bottom low shouldn't be on at the same time.
Top high and bottom low should be. Top low and bottom high should also be on at the same time.


Gate to source (out) pin 5 isolated



Gate to source (out) pin 5 connected




Gate to gate (50v setting) (Bottom high to bottom low).





Gate to gate

Top high to top low




Top high to bottom low




top high bottom high




bottom high to bottom low




bottom high to top low




bottom low to top low

[Marcus]

2st one looks plausible. 2nd and 3rd are what I'd expect.

Gate to gate:-

no.4, top high to top low (the working 1/2 bridge): would have expected 63v ish (50+13) when top output is high, and -13v when low. Or +13 / -63, depending on which way the probes connect.

Top high to bottom low: would expect 50 / 0.

Top high bottom high: if there's no bottom high fet, then 63 / ? (Could be 0 or -ve depending on if the bottom output floats up when bottom low is off). Looks plausible.

Bottom high bottom low (again): looks ok but with more ringing than in #3

Bottom high top low: would expect approx 0/0

Bottom low top low (same as no 3): would expect no.3 waveform

[Tinbum]

I've just put the fet in on the bottom high and all was ok. Very strange

I then put 2 24v 21w light bulbs in series accros the output and turn on. Lights come on (ramped up slowly)

I've then done the gate to gate scopes- one is very noticeably different.

Top high top low




Top high bottom low




Top high bottom high




Bottom high top low




Bottom high bottom low




Top low bottom low

[Tinbum]

(Top high to Bottom low) and (Bottom high to top low) look virtually the same

(Bottom high to bottom low) and (Top high to top low) look virtually the same

Its the (top low to bottom low) that looks a bit strange to me. (But then those two gates should never be on at the same time nor should Top high and bottom high).

Next is to scope gates to source.


EDIT
All look good, so not sure why it blew the FETS initially and then the replacements. Perhaps time to reconnect the output transformer but keep a low value fuse in?

[Marcus]

Hmm... those all seem to be much lower amplitude than I expected. Is the 50v still 50v? Is the 13 supply still 13v? Be interesting to see gate to source volts.

[Tinbum]

Hmm... those all seem to be much lower amplitude than I expected. Is the 50v still 50v? Is the 13 supply still 13v? Be interesting to see gate to source volts.

I'll check with a voltmeter.