Anyone else had grief with JK BMS cutting out under high load? (48V LiFePO4 setup)

So I've been pulling my hair out this past week with my 48V 280Ah LiFePO4 bank (16 x EVE cells in series). Got a JK BMS — the 200A active balancer version — and it keeps tripping the overcurrent protection when I fire up my 3kW inverter under any serious load. Talking about running a kettle or a 2kW fan heater, nothing outrageous. The BMS is rated 200A continuous and I'm nowhere near that, so something feels off.

I've had a poke through the JK app and the overcurrent threshold is set to 150A with a 300ms delay, which should be plenty for a kettle startup surge. Wondering if the issue is actually the short-circuit detection being a bit hair-trigger, or whether my cell-level connections have enough resistance to cause a voltage sag that the BMS is misreading. Cells are all spot-on balanced sitting between 3.28V and 3.31V at rest.

Has anyone managed to tune the JK settings to get round this, or is there a wiring tweak I'm missing? I used 70mm² cable throughout and torqued all the busbars to spec, so I don't think it's a resistance issue — but I've been wrong before. Any experience with this specific BMS on a 48V bank would be dead useful.

RelayProject | 847 posts

@DriftMaster — classic JK headache! One thing worth checking that catches people out: the BMS samples current in short windows, so if you've got a capacitive load (inverter, motor controller etc.) the inrush spike looks like a sustained overcurrent even when your steady-state draw is fine.

Try bumping the overcurrent delay parameter up slightly in the JK app before touching the threshold itself — sometimes even 20-50ms makes all the difference without compromising actual protection.

Also worth confirming your cell-level connections are properly torqued. Loose busbars add resistance which the BMS can misread under load. What inverter are you running with it?

Has anyone checked whether the peak vs continuous current settings are configured correctly in the JK app? Mine was factory-set way too conservatively — the peak overcurrent threshold was basically the same as the continuous rating which meant any brief surge (inverter startup, pump motor, etc.) would trip it immediately.

Also worth asking — what's your actual load? Are you running an inverter directly off the BMS output, or going through a Victron SmartShunt or similar? I found having proper system monitoring helped me identify when exactly the trips were happening, which made tuning the BMS settings much easier.

I'm on a 24V setup on my narrowboat so slightly different, but the JK app behaviour seems consistent across voltage configs.

@DriftMaster had the exact same issue on my tiny house setup last winter. The JK app has a delay timer for the overcurrent protection — mine was set to basically zero so any brief inrush spike would trip it instantly.

Bump the overcurrent delay up to around 200-300ms. Genuine sustained overcurrent will still trigger it, but it'll ride through motor startup spikes and inverter inrush without nuisance tripping.

Also worth double-checking your cell-level wire gauge back to the BMS — undersized sense wires can cause false readings under load.

Something nobody's mentioned yet — check your cell wire gauge going into the BMS. I had mine tripping constantly on the shepherd's hut and turned out the issue wasn't the BMS settings at all, it was resistance building up at a dodgy crimp on one of the sense wires. JK was seeing a false voltage spike and shutting down as a precaution.

Also worth checking: are you running anything inductive? My inverter startup was hammering it until I added a soft start relay. Victron kit tends to behave better once that's sorted.

The 200A active balancer version is generally solid once dialled in — just needs patience with initial setup.

SolarNut | 312 posts

@DriftMaster my JK was doing the same theatrical shutdown routine until I realised the cell voltage differential was triggering protection before the overcurrent even had a chance — worth checking your weakest cell isn't sagging under load and spooking the BMS into thinking something's on fire. Stick a proper load on it and watch individual cell voltages in the JK app in real time; if one cell drops away from the pack, that's your culprit, not the current settings. Grabbed a cheap Victron SmartShunt alongside mine and the extra visibility was a game changer for diagnosing exactly this sort of drama. 16 EVE cells in series means one dodgy cell ruins everyone's day.

Nothing like 16 cells of pure drama to ruin your morning cuppa — have you checked whether your busbar connections are actually torqued properly? Loose connections inflate resistance, BMS sees a voltage sag under load and panics like it's read the energy bills.

On my narrowboat I sorted a similar phantom trip by hitting every terminal with a calibrated grunt on the torque wrench — night and day difference. EVE cells especially seem to hate finger-tight optimism.

Also worth double-checking your actual load profile — stick a clamp meter on the main cable and see what you're really pulling versus what you think you're pulling. Inverters starting fridges, pumps, and the kettle simultaneously are proper sneaky about it.

Hey @DriftMaster, one thing worth checking that nobody's touched on yet — have a look at your BMS settings in the JK app. The factory overcurrent threshold is often set quite conservatively, sometimes as low as 150A on a 200A unit. Also check the overcurrent delay time — if it's set to 1ms it'll trip on any brief inrush spike. I bumped mine up to 10-20ms and it made a massive difference for motor-type loads. Obviously don't go mad with the threshold itself, but the delay setting is often the quick win. What's your actual load drawing?

@DriftMaster my EVE cells needed a solid charge-discharge cycle or three before the JK stopped being a drama queen about it — fresh cells have tighter internal resistance tolerances than the BMS expects, so it panics like a Victron owner who's just discovered cloudy British skies exist.