I recently built a 12s3p Molicel P42a pack using the NESE system. Before I charged it, the p-groups were around 3.47v, and the maximum voltage difference between p groups was less than 0.01v. I charged the pack for the first time using a bench power supply set at 3A, and the voltage of the 6th p-group gradually got ahead of the rest of the pack. Eventually this group hit 4.15v and triggered the BMS’s over-voltage protection, which disables charging of the entire pack.
I don’t think there’s anything physically different about the 6th p-group, although I haven’t opened the pack to check. I probed the voltages of each p-group with my multimeter and they were all within 0.02v of what the BMS is reporting.
Here’s a screenshot of the BMS data after I had stopped charging for while and enabled static balancing. (While charging, there was a bit more spread in the other p-group voltages, but the 6th group was significantly ahead of the rest)
I usually use a cell charger and charge all my cells to 4.1V before starting to weld. Its a good practice that havn’t failed me so far.
Anyway, if you’re using NESE it should be fairly easy to dissassemble the P group. Measure the cells and see which one is the culprit. Sounds like you might have som bad connections somewhere like previously stated. (also check the lead to the BMS, dsn’t have to be a loose cell)
If you’re still unsure of that state of your pack after charging it full. I’d recommend doing one discharge cycle to empty. check voltages when BMS cuts off (if discharge bms) and then charge it to full to see that you have equillibrium. If it passes this test it should be all good.
That’s one of the benefits of the NESE system… you can easily break it back down to the cell level. Bad cell or connection issue feel like the most likely causes.
The behavior of P-group 6 increasing voltage at a faster rate than the others – may point to a connection issue. Imagine if a single cell in that group was unconnected (as if you forgot to install it). Then the charging current for that P-group would be allocated to only 2 cells rather than 3. Charging them up faster.
Disassemble P-group 6 and measure the voltage of each of those 3 cells (individually, separated).
I’ve had dud genuine P42A cells before on a few different battery builds that was unrelated to the connections (welded) or construction. Just lower capacity than the rest. But they are heaps better than 30Q cells where I used to experience this problem quite often
