SDU Overcurrent Always On

[coleasterling]

Hey All,

I've had my SDU spinning in manual mode for a bit now. Today, I wired a full bench setup with contactors, throttle, precharge, etc...
I had it spinning well for a bit with a 56V battery and 3A circuit breaker, then pulled a little too much current and tripped the breaker. I tried flipping the breaker while the motor was spinning down and got a hard stop on the motor, then the OC alarm mentioned in the title.

My assumption is that I fried a set of IGBT's. What do y'all think?

Edit: From some more reading, sounds like that's what happened. Any options for replacing these guys or am I just down an inverter now? Can't really find the inverters by themselves, sadly.

Edit 2: New news. Pulled the inverter and there are no shorted phases like I've experienced with IGBT failures in the past and nothing obviously wrong on the driver board. The inverter behaves the same way as before when disconnected from the motor, same OC error.

I then disconnected the openinverter board from the inverter and tried again, same issue. Unless I'm reading something incorrectly, the board is seeing the same OC/desat error when completely removed from the inverter.

Thanks!

[arber333]

Hey All,

I've had my SDU spinning in manual mode for a bit now. Today, I wired a full bench setup with contactors, throttle, precharge, etc...
I had it spinning well for a bit with a 56V battery and 3A circuit breaker, then pulled a little too much current and tripped the breaker. I tried flipping the breaker while the motor was spinning down and got a hard stop on the motor, then the OC alarm mentioned in the title.

My assumption is that I fried a set of IGBT's. What do y'all think?

Edit: From some more reading, sounds like that's what happened. Any options for replacing these guys or am I just down an inverter now? Can't really find the inverters by themselves, sadly.

Thanks!

What, on 60V? Can you try starting your setup again? You could fry one of the driver chips, not necessarily IGBTs. If you can check how the chips look, maybe you can spot the one that is toasted. Usually they use NPN transistor with a pullup as a failsafe fault output...

Just yesterday i made a dumb mistake and connected 12V to my Volt board 5V circuit. Of course i fried something on the driver section that was connected to 5V. My guess its the driver chips. I will try to replace them...

[coleasterling]

What, on 60V? Can you try starting your setup again? You could fry one of the driver chips, not necessarily IGBTs. If you can check how the chips look, maybe you can spot the one that is toasted. Usually they use NPN transistor with a pullup as a failsafe fault output...

Just yesterday i made a dumb mistake and connected 12V to my Volt board 5V circuit. Of course i fried something on the driver section that was connected to 5V. My guess its the driver chips. I will try to replace them...

Ya, I tried a couple times and went back to manual mode. It trips my breaker immediately on powerup now. I'll pull it apart and we'll see if anything is obvious.

[coleasterling]

Reassembled the board onto the inverter (still disconnected from motor) and was able to boot it without throwing the OC alarm and thus opening the contactor. However, the second I apply throttle, it trips. That's the start of PWM and would point to an IGBT/driver issue, eh?

[johu]

Yes, no PWM until you press throttle

[coleasterling]

Thanks, any idea why it would behave the same way with the board completely removed from the inverter?

[arber333]

Thanks, any idea why it would behave the same way with the board completely removed from the inverter?

Do you have Wifi module connected? Try taking it off board when you next start it. I had some problems with olimex Wifi modules. I think they drew too much for the poor 3V3 regulator and inverter reference dropped too low. I think that is why i got so much OC events.
I will try with D1 mini board because it can be fed from 5V directly.

[johu]

Thanks, any idea why it would behave the same way with the board completely removed from the inverter?

When you remove the board the fault outputs are no longer connected so all fault lines will read as asserted, i.e. not 5V. To turn them off you need to find the 3 driver fault inputs and pull them high and also the HVIL.

[coleasterling]

Makes sense, thanks. I'll continue to dive into it and start looking for another inverter in the mean time.

[johu]

Yes, and don't use that breaker next time. We've seen multiple blowups in the past with BMS operated main relays that randomly opened and turned the inverter into a scrap pile in milliseconds. Not quite sure why it happens, probably a small amount of regenerative braking with nowhere to go for the energy.