Anyone else had grief with JK BMS dropping cells during cold nights?

Been running a 280Ah LiFePO4 pack in my van for about eight months now — four EVE 280Ah cells in series with a JK BMS (the 2A active balancer, 100A model). Setup's been brilliant through summer but now we're heading into winter I'm getting random low-cell alarms triggering around 3–4am when temps are dropping to 4–6°C inside the van. Cell 3 seems to be the culprit every time, reading about 50–80mV lower than the others at the point of alarm, then it recovers once the heating kicks on in the morning.

I've already checked my bus bars and torqued everything to spec (4Nm on the EVE cells), so I don't think it's a connection issue. My suspicion is that Cell 3 is just sitting in a slightly colder spot — it's nearest the external wall — and LiFePO4 internal resistance does climb noticeably below around 10°C. I've got the JK low-cell cutoff set at 2.8V which might be a touch aggressive for cold-weather use.

Has anyone dealt with something similar? I'm wondering whether to bump the cutoff down to 2.75V or even 2.7V to stop the nuisance trips, or whether that's masking a genuine weak-cell problem I should be investigating properly. Also curious if anyone's added cell-level insulation or a low-wattage heat mat under the pack — seems like that might be a more sensible fix than fiddling with protection thresholds.

@Turbo19 classic cold-temperature behaviour from LiFePO4 — the cells' internal resistance rises sharply below about 5°C, causing the BMS to see voltage sag under load and trigger low-cell-voltage protection prematurely.

A few things worth checking:

• JK app → Cell Voltage Undervoltage Protection — drop it to 2.80V if you haven't already
• Temperature charging cutoff — JK defaults are quite conservative; ensure your discharge low-temp cutoff isn't set too high (some ship with 5°C discharge cutoff, which is overly cautious)
• Critically, are your cells actually cold? Insulating the battery box makes a huge difference on a van floor

I fitted a small self-regulating heating pad (the Fogstar ones work well, or a generic 12V silicone pad) beneath my cells on my narrowboat — transformed winter performance completely.

The 2A active balancer won't cause this, red herring there.

@Turbo19 worth checking whether your JK is configured with a low-temperature charge cutoff — mine was factory-set at 5°C and would disconnect the entire pack rather than just blocking charge. Critically, if you're also drawing load when it trips, the BMS can latch into a protection state that looks identical to a dead pack.

In my motorhome setup I added a Victron Smart Battery Sense paired with a dedicated 10W self-regulating heating pad (wrapped in closed-cell foam around the lower cells) — triggered via a Victron temperature relay at 3°C. Keeps the pack above the threshold overnight for pennies.

Also pull the JK app logs — the event history will tell you exactly which protection triggered and at what cell voltage/temperature. That rules out balancing issues versus genuine cold cutoff.

Ran into exactly this on the boat two winters back — woke up at 6am to a dead inverter, a very cold cabin, and a BMS that had basically decided it was having a lie-in.

What nobody mentioned yet: check your cell voltage differential in the JK app during those cold mornings. Mine looked fine at rest but the moment I kicked

Good timing on this thread — just dealt with something similar in my shed setup last month. One thing nobody's mentioned yet: even if your low-temp cutoff is configured correctly, check that your temperature sensor is actually making proper contact with a cell body rather than just floating loose inside the case. Mine had vibrated away from the cell surface and was reading ambient air temp instead, so the BMS thought everything was warmer than it actually was. Stuffed a bit of foam behind it to hold it firmly against the cell and the readings immediately became far more accurate. Also worth grabbing the JK app and logging overnight temps — you can see exactly what the BMS was seeing when it tripped, which takes a lot of the guesswork out. @Turbo19 what temperatures are you actually seeing overnight where you're parked?