Anyone else had issues with a JK BMS dropping cells offline mid-charge on a 280Ah LiFePO4 bank?

Been running a 4S 280Ah LiFePO4 setup in my van for about eight months now — four EVE cells wired up with a JK BMS (the 200A active balancer version). Generally it's been brilliant, but lately I'm getting a fault where one or two cells show as "offline" in the JK app mid-charge, usually somewhere around 3.45V per cell. The BMS cuts the charge completely, throws an error, and I have to manually reset it via the app to get things going again. Bit of a pain when you're parked up on a cloudy week and trying to eke every amp out of a 400W solar array.

I've checked the cell sense wires and they all seem seated properly in the JST connector. Resistance across each sense wire reads fine on the multimeter. The cells themselves balance to within about 5mV at rest, so I don't think it's a genuine cell issue. My suspicion is it might be a dodgy contact in the JST connector housing itself rather than the wire — possibly vibration related given the van use.

Has anyone else seen this on a JK BMS, particularly the active balancer variant? I'm wondering whether it's worth reseating and heat-shrinking all the sense wire connections, or if there's a firmware update that addresses this — I'm on v11.25 at the moment. Also open to the idea that I've got my overvoltage protection set a bit tight (currently at 3.65V cutoff) if that could somehow cause this behaviour.

@RogerHobbs mine did exactly this in the garden office last winter — turned out the balance wires had developed a dodgy crimp that only showed up when the cells were warming up and expanding slightly, classic "works fine until it doesn't" energy.

@RogerHobbs had almost identical grief with my EVE 280Ah bank in the motorhome last spring. Before you go too deep down the rabbit hole — check your cell-level voltage readings in the JK app during charge. If one cell is reading suspiciously low compared to the others, nine times out of ten it's a bad connection at the cell terminal rather than the cell itself.

Torque those M6 bolts properly too. Mine were finger-tight from the factory and I was losing millivolts across every joint under load.

Also worth checking — what firmware version are you on? JK pushed an update a while back that changed the overvoltage protection thresholds. Mine started cutting out until I reflashed it.

Doubt it's the cells themselves if they're genuine EVE — those things are pretty bulletproof in my experience.

@RogerHobbs following this with interest as I've got a very similar setup on my boat — JK BMS with EVE cells, though mine's a 12V 100Ah bank rather than 280Ah.

One thing I haven't seen mentioned yet: have you checked the actual cell voltage readings in the JK app against a multimeter directly on the cell terminals? I had what looked like a cell dropping offline but it turned out the BMS was seeing phantom voltage spikes due to a loose main busbars connection rather than anything wrong with the balance wires themselves. The resistance was causing weird readings under load.

Worth ruling that out before assuming it's the BMS itself. What firmware version are you running on the JK? There were some known issues with earlier versions handling active balancing during bulk charge.

@RogerHobbs worth pulling the JK app and checking your overvoltage protection threshold on the individual cells. JK defaults are sometimes a bit tight out of the box — if one cell is hitting the OVP limit before the others have caught up, it'll drop that cell and kill the charge cycle.

Also check your balance start voltage. If it's set too low it's trying to balance during bulk charge rather than waiting for absorption — wastes balancer capacity and can confuse the BMS logic.

My EVE 280Ah bank on the boat had similar weirdness until I set:

• Cell OVP: 3.65V
• Balance start: 3.40V
• Balance stop: 10mV differential

Sorted it completely. Takes about 10 mins in the app once you know what you're looking at.

@RogerHobbs one thing worth checking that nobody's mentioned yet — have a look at your cell wire connections at the BMS sense leads. I had my JK doing something very similar and it turned out one of the balance wire terminals had a dodgy crimp causing intermittent high resistance readings. The BMS was seeing a false voltage spike on that cell and pulling it offline as a precaution. Completely logical behaviour from the BMS, just not an obvious cause.

Reseat all the sense wire connectors and check the crimps are solid — if you've got a multimeter, measure resistance along each balance lead while gently wiggling the connectors. Mine was fine until I flexed the wire slightly, then jumped to several ohms. Replaced the crimp and never had the problem again.

@IslandDweller

@RogerHobbs one thing worth checking that nobody's mentioned yet — have a look at your cell internal resistance readings in the JK app. If one cell is showing noticeably higher IR than the others, it'll hit the overvoltage threshold earlier under charge current, which can cause the BMS to trip that cell out. Seen it happen with EVE cells that've had a few shallow cycles without ever being properly top-balanced.

If the IR looks uneven, a proper top-balance session with the cells in parallel overnight at 3.65V often sorts it. Also double-check your bus bar connections are properly torqued — a slightly loose connection can mimic a dodgy cell in the app readings.

What charge current are you running through it? If you're pushing hard above 50A that can exaggerate the problem considerably.

Good shout from @Hamish1980 on the connections — I'd add that it's worth checking the sense wire terminals on the cells themselves, not just at the BMS end. I had almost identical symptoms on my EVE 280Ah bank and it turned out one sense wire had just enough resistance from a dodgy crimp to fool the BMS into thinking that cell was spiking. Recrimped the lot with proper ferrules and haven't had a dropout since. Takes about 20 minutes and costs nothing if you've already got the tools.

@RogerHobbs had almost identical on my narrowboat build last winter — turned out to be the overvoltage protection threshold set slightly too low on one cell input. JK's default settings can be quite aggressive. Worth connecting via the Bluetooth app and logging individual cell voltages in real time during a charge cycle; you'll quickly see if one cell is spiking ahead and triggering the protection cutoff rather than an actual fault.

Also check your balance start voltage — if it's set higher than your charge voltage, the active balancer never kicks in and you'll see cell divergence grow over time.

@RogerHobbs had this exact thing on my garden office bank — JK BMS, EVE 280Ah cells, same symptoms. For me it was the cell overvoltage recovery threshold set too close to the protection threshold. BMS was tripping a cell at 3.65V then not recovering until it dropped back far enough, causing this weird oscillation mid-charge.

Worth pulling the JK app and checking the gap between your OVP and OVP recovery values — I ended up setting recovery at 3.50V with protection at 3.65V and it's been solid since. The defaults on some firmware versions are surprisingly tight.