Seplos 200 BMS stuck

[Logic28]

I can almost guarantee you that :
1. Busbars aren't torques to 4-5 nm
2. Because of that heat that was created causing an even worse connection for sense wires.
3. Conducting paste is bad

4 possibly the connector in pcb to and from BMS aren't seated right/pushed all the way in

Bus bars torqued correctly to 4-5nm
All BMS sensor's ring terminals are well seated and firmly tightened (unless there is a bad crimp - see end note)
Why is conductive paste bad?
The issue has repeated in reverse this time causing an under-voltage protection alarm as opposed to the over-voltage of last time, curiously a similar cell voltage mismatch occurred between cells 1 & 2 instead of 15 & 16.

Notice that it is always the first two or the last two cells affected, those closer to the battery terminal that leads me to think that a voltage ripple is affecting the closest cells. Seconds later the voltages are perfectly matching again but unfortunately, the alarm has already been set halting the inverters so the battery needs to be recharged for a little while to bring it away from the erroneous 00 SOC, obviously wrong as the voltages are all over 3.38v per cell.

This happened at night when there was a minimum demand never exceeding 200watts but never in the day with loads up to 5/6kw
Also, I suspect the MBS sensors connector/PCB soldering, that being dodgy could also lead to a voltage misreading.

 

[Logic28]

3. Conducting paste is bad