cell voltage irregularity in DIY lfp 16s battery bank, mirroring

[Atomize]

hello,
I have a DIY lfp battery bank for my offgrid greenhouse, and I've been getting irregularities in the voltages. When this happens, it's always two cells that are next to each other, and they seem to perfectly mirror each other. Does anyone know what is going on? I'm suspicious that the bms is starting to fail and give faulty readings, but I have pretty low confidence on that answer.

Attached are the pics for both the cell-level voltages given by the jdbms (16s configuration, 100amp) and current from my shunt.

Scanning back through recent history, it's mostly cells 14/15 doing this voltage mirroring, but it happens a little between cells 6/7 too.

Any ideas?

 

[RCinFLA]

First suspect is poor bus bar/cell terminal connection.

Check if cell voltage irregularity correlates to inverter load current, indicating voltage drop across bad connection. BMS sense wire is shared between two adjacent cells with one side having series bus bar in path and other adjacent cell not. The flipped mirror voltage rise and slump between two adjacent cells points to the bus bar connection between two cells.

Also look for a bus bar/cell terminal getting warm under moderate load current.

 

[Atomize]

First suspect is poor bus bar/cell terminal connection.

Check if cell voltage irregularity correlates to inverter load current, indicating voltage drop across bad connection. BMS sense wire is shared between two adjacent cells with one side having series bus bar in path and other adjacent cell not. The flipped mirror voltage rise and slump between two adjacent cells points to the bus bar connection between two cells.

Also look for a bus bar/cell terminal getting warm under moderate load current.

Thanks, I'll check the torque on the terminals when I go down there in a couple days. Maybe they are loosening over time.

Check if cell voltage irregularity correlates to inverter load current

Inverter load was almost nothing(1.25 amps, as seen by the shunt) during the irregularity. Night time, so no solar.

 

[RCinFLA]

Hard to tell the colors but looks like adjacent cells 14 and 15 and looks like maybe a bad connection on sense wire to BMS or issue with BMS voltage sampling snapshot.

BMS's usually shuts down any balancing being done for a moment and MUX samples to storage capacitors across all the cells. Then the sampled voltage on storage capacitors is MUX'd down to microcontroller's analog to digital converter, 0-3.3v level, to read the sampled voltage. Microcontroller ADC round-robins all the sampled cell voltages on the sample and hold capacitors.

Being the load current through battery is low points the problem to the BMS voltage sensing measurement.
- sense wire (cell 14,15) bad connection.
- BMS MUX transistors intermittent.
- Sample and hold capacitor intermittent with bad capacitor or intermittent solder joint to capacitor.

 

[Atomize]

Hard to tell the colors but looks like adjacent cells 14 and 15 and looks like maybe a bad connection on sense wire to BMS or issue with BMS voltage sampling snapshot.

BMS's usually shuts down any balancing being done for a moment and MUX samples to storage capacitors across all the cells. Then the sampled voltage on storage capacitors is MUX'd down to microcontroller's analog to digital converter, 0-3.3v level, to read the sampled voltage. Microcontroller ADC round-robins all the sampled cell voltages on the sample and hold capacitors.

Being the load current through battery is low points the problem to the BMS voltage sensing measurement.
- sense wire (cell 14,15) bad connection.
- BMS MUX transistors intermittent.
- Sample and hold capacitor intermittent with bad capacitor or intermittent solder joint to capacitor.

Thanks for the info. I was wondering what could cause the mirroring, and this makes total sense.

 

[400bird]

Hard to tell the colors but looks like adjacent cells 14 and 15 and looks like maybe a bad connection on sense wire to BMS or issue with BMS voltage sampling snapshot.

100% agree
I've seen this plenty of times in the BMSs I work with. It's always the cell tap wiring. Anything from soft seated pins, poor crimps, to failed wire (broken copper inside the insulation)
The cell tap between the high and low cells is the wire to inspect. But because you have 2 issues, I'd recommend Atomize checks them all.