Anyone else had issues with JK BMS dropping cells offline during high discharge?

Picked up a 200A JK BMS a few months back to run with my 280Ah Eve LiFePO4 cells (4S configuration). Generally been dead happy with it — Bluetooth app works a treat, balancing seems spot on. But lately I've been noticing one or two cells dropping out of the active count in the app when I pull heavy loads through my inverter (2000W Victron Multiplus). Voltage shoots back up the moment the load drops off, so I don't think it's a genuine cell problem.

I've checked my busbars and the main cable runs — everything's torqued down properly, 95mm² cable throughout from battery to inverter. Wondering if it's maybe a dodgy sense wire connection on the BMS board itself, or possibly a firmware thing. I'm on firmware v11.25 if that means anything to anyone.

Has anyone seen similar behaviour? Specifically whether cell 3 going "missing" briefly under load is a known quirk with these units or if I should be looking harder at the sense wires. Happy to pull the whole thing apart if that's what it takes, just want to rule out the obvious stuff first.

@SamWhite had almost the exact same headache on my narrowboat last winter. Turned out my busbars weren't torqued properly — looked fine visually but under a big inverter surge the resistance spiked just enough to spook the BMS into thinking a cell had gone rogue.

Grabbed a proper torque wrench, went back over every connection, applied Denso tape over the terminals afterwards to keep moisture out (boats, y'know), and the phantom dropouts vanished completely.

Worth checking your cell interconnect resistance in the JK app — there's a screen that shows individual cell internal resistance. If one reads noticeably higher than the others, that's your culprit nine times out of ten, and it's the connection rather than the cell itself causing grief.

@SamWhite worth checking your cell-level internal resistance readings in the JK app — if one cell is even slightly higher than the others, it'll sag harder under load and trip the undervoltage protection before the pack as a whole looks low.

Had a similar scare in my tiny house setup with a 4S 280Ah pack. Turned out one cell had crept up to around 0.35mΩ versus the others sitting at 0.18mΩ. Under a 150A+ draw it was dropping out of tolerance.

A few top-balance cycles sorted mine. Also — what's your undervoltage cutoff set to? The JK defaults can be a bit aggressive out of the box. I dropped mine to 2.90V per cell and the nuisance trips stopped entirely.

What inverter are you pulling that load through? Surge currents from certain units can spike well beyond the rated draw.

@SamWhite one thing neither @VDH_Boats nor @PaulMurray has touched on — check your BMS firmware version in the app. There was a known issue with earlier JK firmware where the over-voltage protection threshold would drift under sustained high-current discharge, causing phantom cell trips even when cell voltages looked reasonable. Updating to the latest via the JK app sorted it for several people on the EVE 280Ah cells specifically.

Also worth confirming your discharge overcurrent setting is actually configured for 200A continuous and not defaulting to a lower figure post-update — mine reset to 120A after a firmware flash and I spent two days chasing a ghost fault before I spotted it.

Had this exact issue in my static caravan setup last summer. Ended up being a dodgy cell interconnect — looked fine visually but had just enough resistance to cause a voltage drop under load that the BMS read as a low cell.

Worth grabbing a cheap clamp meter and checking voltage directly at each cell terminal during a high discharge event rather than relying solely on the BMS readings. The JK app is decent but it's only as accurate as what it's measuring from.

@PaulMurray's point on internal resistance is solid — if one cell is climbing noticeably higher than the others that's your culprit straight away.

Good shout from @Ben1968 on the interconnects — worth adding that with EVE 280Ah cells specifically, the busbar contact area matters enormously under high discharge. Even a marginally undersized or slightly proud busbar can introduce enough resistance to skew the voltage reading at that cell terminal, making the BMS think a cell is dropping when it's actually a connection artefact.

I'd measure the voltage directly at the cell terminals (not the BMS sense wires) under load and compare. If your sense wire pickup points are on the busbars rather than the cell terminals themselves, you'll potentially be reading a false low.

In my shepherd's hut setup I had near-identical symptoms — turned out the sense wire crimp on cell 3 had marginal contact. Recrimped with proper ferrules, problem gone entirely.