Electronics help required with Sofar ME3000SP

[Tinbum]

The voltage to the top and bottom diodes look the same now.

Gate to source voltages measure 6.15v to 6.6v

[Marcus]

Sorry- hectic weekend.

I seem to be getting more confused: my last post was when looking at the gate-gate scope traces, but i hadn't seen the gate-source traces. Possibly I'm mistaken on the scope sensitivity but based on previous scope traces i was interpreting the scope settings: 50v and x10 as 5v/div on the trace. Looking at the last two lots of traces the volts are too low, but the gate pulse shapes look ok, and look much the same for all 4. Not sure about the shapes of the gate-gate waveforms either - some seemed a bit off square, but I'd be looking at the output waveforms.

Looking at your voltmeter diagram: 6 and a bit volts on the gate would be right if your voltmeter is averaging a 13v square wave. Is the voltage on the diodes still 13v?

[Tinbum]

No problem. Thank you for helping.

I'm really baffled. I reinstalled the capacitor banks but left the transformer on the output disconnected. I powered it up with the light bulbs and all was ok.

Powered down, connected the transformer and changed the fuse for a 5 amp. (i figured as it ramps up that may be enough). I've set it to discharge at 20w.
As soon as it did its self checks and powered up the fuse blew.

Replaced the fuse with a 20 amp. Fuse blew and bottom high fet went bang!! I've not looked at the bottom low fet yet.

I'm just wondering. Could a problem with the inverter sync with the mains cause this?

I'll test voltages again

[Marcus]

I don't know anything about this inverter:

I took the 50khz square 50/50 square wave on the h-bridge to be a drive for a 2 stage boost converter- i.e. it's not directly connecting to 50hz a.c.

[Tinbum]

I don't know anything about this inverter:

I took the 50khz square 50/50 square wave on the h-bridge to be a drive for a 2 stage boost converter- i.e. it's not directly connecting to 50hz a.c.

The outputs are connected to a transformer to step up the voltage and the output of the transformer then go the ac circuitry.

I'll try replacing the fets and swap the transformer for another and try again. I measured the coil resistance of the transformer and compared it to a know working on and both were 0 ohms.

[Marcus]

Yes, substitution is an easy way to rule out the transformer if you have a spare: even if you have a low ohms meter, it wouldn't show a shorted turn.

Did you take a scope trace of the waveform across the lamps?

[Tinbum]

Yes, substitution is an easy way to rule out the transformer if you have a spare: even if you have a low ohms meter, it wouldn't show a shorted turn.

Did you take a scope trace of the waveform across the lamps?

Yes I did.
It looked ok. I cant find it so will repeat once i have replaced the fets.

I'm just taking a rest- this heat has done me in. The joys of long covid.
Heart rate went above 150 just swapping 2 hoses on the tractor. Had to sit under the Aircon for an hour to bring everything down to normal!!

[Tinbum]

Replaced the fets and took a trace with the bulbs. Looks good.



Replaced the transformer with a known good one, 10A fuse- fets went bang!!

[Marcus]

Is that the trace across the outputs with the bulbs between them? The voltage should be 50v(?) Peak to peak and flat across the top, which doesn't tally with my interpretation of the voltage sensitivity of the scope (20v x10 = 2V/div). Of course if the sensitivity is actually 20v/div then that would suggest the battery is almost 60v (initial peak) but then the waveform falls away to about 52ish (could be due to high impedance 60v source(?) But if there are capacitors across b+ b- they ought to flatten the wave tops.

With the bulbs lit i would expect the Vds to fall to zero from b+ for each of the fets.

Do you know what happens to the transformers ouput? Is it rectified into a capacitor bank? Are there any shorts there?

[Tinbum]

Is that the trace across the outputs with the bulbs between them? The voltage should be 50v(?) Peak to peak and flat across the top, which doesn't tally with my interpretation of the voltage sensitivity of the scope (20v x10 = 2V/div). Of course if the sensitivity is actually 20v/div then that would suggest the battery is almost 60v (initial peak) but then the waveform falls away to about 52ish (could be due to high impedance 60v source(?) But if there are capacitors across b+ b- they ought to flatten the wave tops.

With the bulbs lit i would expect the Vds to fall to zero from b+ for each of the fets.

To be honest I'm not exactly sure on the scope. The batteries are Pylontech so voltage max is 53.2v when charging. I can't recall exactly what they were at the time but probably in the region of 50 - 51v. Yes the bulbs were on and pretty bright but not over bright.. (only 2 x 21w 24v bulbs in series). Output of the inverter was set at 10W but as the transformer wasn't connect I dont know if it can limit it. Inverters capacitor bank was connected across the battery input.

Do you know what happens to the transformers ouput? Is it rectified into a capacitor bank? Are there any shorts there?

I'm not certain, that was the next thing to look at. Their are 3 x 680 micro farad capacitors. I was going to leave the output of the transformer disconnected initially but I forgot.