Seplos BMS "Output Short Circuit Protection"

[edcoopered]

My Seplos 150A BMS is hard wired to my Sofar ME3000SP Charger/Inverter, everything runs normally most of the time, however once a week the BMS disconnects with a "Output Short Circuit Protection" condition, it reconnects shortly after but the error shuts-down the inverter/charger until I can restart it. Has anyone else experienced this? It's only happened at fairly low power stead-state conditions at high SOCs on the pack. The inverters maximum load is 65A.

If I understand correctly from the manual this feature of the BMS is implemented in hardware (perhaps some fixed firmware too) and is not user adjustable. Im wondering if it's triggered by some sort of transient event.

The only thing I've done that's perhaps not ideal is use diagonally opposite terminals for the P-/B- connections, as I was only running 70A through it peak I felt this was probably sufficient - perhaps I should fit some sort of bus bar across the pair of three sets of terminals.

 

[mciaio]

Any solution to this? I am running into the same problem. Seplos told me to increase the ms for the alarm and it didn't work.

 

[edcoopered]

As I understand it you can't adjust threshold for this protection as its implemented in hardware, but if you look in the GUI theres a bunch of (poorly formatted) toggle switches - I believe I disabled it there - I've not had any more incidents yet anyhow.

 

[mciaio]

I saw that and I just couldn't bring myself to turn it off. This is the first time hooking it up and I am afraid if there really is a short circuit. I have tested ohms between the inverter leads and the BMS leads. There are megaohms while disconnected but only about 450 ohms when connected.

 

[edcoopered]

Yeah I'd rather of left it on, but I doubt those little FETs could do much under an actual short-circuit with the huge cells I've got connected, that's what my appropriately rated HRC fuse is there for.

 

[mciaio]

Turned it off. Big spark and now it's cooked. Hopefully they will replace under warranty.

 

[edcoopered]

oh no, do you think there's something up with the inverter?

 

[Roswell Bob]

Turned it off. Big spark and now it's cooked. Hopefully they will replace under warranty.

ouch!

 

[Roswell Bob]

A mosfet can survive a direct short across a very low impedance source - even a LiFePO4 array. In most cases it is possible to get the fet off in plenty of time. I would be surprised if a setting made in software could disable a hardware based overcurrent protection circuit.

 

[mciaio]

I check the ohms across the inverter leads and it was megaohms. There was no short.

 

[jsolar]

What is the final result? You get new BMS, just connect and all work normally?

 

[testtenor]

I'm also getting this (amongst other) errors. Any idea on its cause, or a fix?

 

[edcoopered]

I'm also getting this (amongst other) errors. Any idea on its cause, or a fix?

Mine is still working perfectly with the short circuit protection disabled, if you have an established, static, otherwise working system, with an appropriately rated fuse then I see no harm in turning it off. If you're struggling with this error during commissioning of a. system then I think you need to be really sure its not currently saving your arse.

 

[Lt.Dan]

If you had a dead short, I would not be relying on the BMS to stop that. I would be relying on a quality Class T Fuse.

 

[edcoopered]

A mosfet can survive a direct short across a very low impedance source - even a LiFePO4 array. In most cases it is possible to get the fet off in plenty of time. I would be surprised if a setting made in software could disable a hardware based overcurrent protection circuit.

There's probably a comparator doing the short circuit protection, would be trivial to have the micro hold its signal low, or use an internal DAC to generate a bunch of different thresholds for different revisions of the board.

 

[testtenor]

Mine is still working perfectly with the short circuit protection disabled, if you have an established, static, otherwise working system, with an appropriately rated fuse then I see no harm in turning it off. If you're struggling with this error during commissioning of a. system then I think you need to be really sure its not currently saving your arse.

The thing is, it's not all the time. It's happening randomly, so I'm confused. I've checked the wiring on the battery and there isn't a short anywhere. But it keeps cutting out. Seplos think it might be a BMS issue

 

[edcoopered]

The thing is, it's not all the time. It's happening randomly, so I'm confused. I've checked the wiring on the battery and there isn't a short anywhere. But it keeps cutting out. Seplos think it might be a BMS issue

If you read my original post, I was having the same problem, my caution is regarding mciaio's experience though there could have been a defect with their BMS anyway, and disabling the protection, just turned it into a more catastrophic fault.

 

[testtenor]

If you read my original post, I was having the same problem, my caution is regarding mciaio's experience though there could have been a defect with their BMS anyway, and disabling the protection, just turned it into a more catastrophic fault.

Yeah turning off short circuit protection doesn't seem wise. Seplos think it might be a BMS issue, hopefully they'll send me a new one

 

[jsolar]

I had a similar problem. I turned off shortcut protection in the Solar monitor app and then bms exploded.
One FET burned out.. I replaced the FET and now it works fine. I would not recommend repeating my procedure because the PCB on the BMS could have burned. Maybe, you can try to replace the MOSFET before.

 

[edcoopered]

I had a similar problem. I turned off shortcut protection in the Solar monitor app and then bms exploded.
One FET burned out.. I replaced the FET and now it works fine. I would not recommend repeating my procedure because the PCB on the BMS could have burned. Maybe, you can try to replace the MOSFET before.

Was your system otherwise working, with a transient short circuit shutdown issue that happened once a week/month, or was it behaving like this out of the box? I suspect there's two very different issues getting mixed up.