Anyone else had issues with JK BMS dropping cells randomly on cold mornings?

Had this exact drama last winter with my shepherd's hut setup on the fell. Woke up to the Victron Cerbo throwing fits at 6am, and sure enough the JK had decided one of my 280Ah EVE cells had simply... ceased to exist as far as it was concerned.

Turned out the cell sense wire had developed just enough resistance in the cold that the BMS was reading it as an open circuit. Temperatures had dropped to around -4°C overnight. Once things warmed up past 8 or 9°C it came back perfectly, which had me chasing ghosts for a good week before I twigged.

A few things worth checking if you're seeing the same:

• Sense wire connections — these crimp terminals on JK units can be surprisingly fussy. I resoldered mine and used a tiny dab of dielectric grease
• Cell compression — my Fogstar cells had loosened slightly in their bracket over summer, which apparently affects internal resistance readings in the cold
• BMS firmware — I was two versions behind. The current release handles low-temp cell resistance fluctuations noticeably better

The JK app does let you log the individual cell voltages over time, which is genuinely useful for spotting whether it's a real voltage drop or a sensing glitch.

Anyone running their packs in genuinely uninsulated spaces through winter? I've since wrapped the battery enclosure in 50mm PIR and it's made a significant difference to the overnight behaviour. Curious whether others have found the JK more temperamental than, say, a Daly or Overkill in the cold.

Been there @CumbrianWanderer! Mine did exactly this last February during that cold snap. Turned out my balance wires had developed a tiny bit of resistance at the connector - cold temperatures made it just enough to give the BMS dodgy readings on that one cell. Worth checking your JST connectors with a magnifying glass and reseating every one. I also nudged my low voltage cell differential alarm threshold up slightly in the JK app (not the cutoff, just the alarm) to give myself more warning before it trips. The EVE cells themselves are usually fine down to around 0°C for discharge, so nine times out of ten it's a connection issue rather than the cell genuinely dropping. What are your overnight temps up there on the fell getting down to?

@CumbrianWanderer this is almost certainly low-temperature lithium behaviour rather than a genuine BMS fault. EVE 280Ah cells have internal resistance that spikes sharply below 5°C — the JK sees voltage sag under load and interprets it as a weak/failing cell, so it drops it out of the pack.

What I did on my tiny house setup: added a cheap DS18B20 temperature probe directly to the cell body (not the BMS terminal) and configured the JK's low-temp charge cutoff properly. Also worth checking your cell-level voltage under load via the JK app during the event — if that dropped cell was genuinely lower IR than the others before winter, cold just exposed it.

Fogstar sell pre-matched EVE cells which helps. Also consider a small self-regulating heat mat under the battery tray if you're in an exposed fell location — passive insulation alone won't cut it at Cumbrian temperatures.

@SaltyRanger is on the right track but there's a bit more to it than pure cell chemistry.

The JK specifically has a known quirk where the NTC temperature sensor — that little black blob usually cable-tied somewhere optimistic — reads ambient rather than cell temperature. On a cold hull morning my readings were lagging actual cell temp by 20+ minutes. BMS sees -2°C, trips the low-temp cutoff, logs it as a cell undervoltage event because the protection cascade overlaps. Classic.

Check your BMS app logs carefully. If you see undervoltage and low-temp faults within the same second, that's your culprit — it's not actually a bad cell.

Fix: relocate the NTC so it's actually touching a cell body and insulated from ambient airflow. Took me forty minutes and a bit of self-amalgamating tape on the narrowboat last January. Zero issues since.

Really relevant thread for me — also running a shepherd's hut with EVE cells and a JK BMS, so watching this closely.

@SomersetNomad what's the "known issue" you were about to mention? Don't leave us hanging on that one.

One thing I haven't seen mentioned yet: has anyone checked whether the JK's temperature sensor is actually positioned on the cells rather than just floating loose inside the enclosure? Mine was barely touching anything when I first installed it, so it was reading ambient air temp rather than actual cell surface temp. Wondering if that could cause the BMS to make bad decisions about what it's seeing.

Currently wrapping my battery box in Thermawrap and considering a small self-regulating heat mat on a timer for the really bitter nights. Anyone gone down that route?