Anyone else struggled to get accurate readings from a cheap battery monitor in cold weather?

Just wanted to share something that's been driving me a bit mad this winter. I've got a Renogy 500A shunt monitor wired up to my 200Ah LiFePO4 bank (two 100Ah Epoch batteries in parallel), and I've noticed the state of charge readings going completely haywire once the temperature drops below about 5°C in my shed. It'll show 74% and then suddenly jump to 61% within a few minutes with barely any load on it.

I know LiFePO4 has a notoriously flat voltage curve anyway, which makes SOC estimation tricky, but I'm wondering if the shunt itself or the monitor unit is getting affected by the cold. The shunt is mounted on the negative terminal inside the shed, which isn't insulated, so it's basically sitting in near-freezing air most nights. I've read that some shunts have temperature compensation built in but I'm not sure if the Renogy one actually does anything useful with it in practice.

Has anyone found a reliable fix — whether that's a better monitor, insulating the shunt somehow, or just accepting the readings are rubbish below a certain temp? Would love to know if the Victron BMV-712 is actually worth the extra £80 or so compared to what I've got now.

@EmmaPowell74 oh I feel your pain. Spent most of last January convinced my narrowboat battery bank was dying because the shunt monitor kept showing wildly pessimistic SOC readings. Turns out the monitor itself was sat next to the engine bay bulkhead where temps were dropping to about 4°C overnight — the shunt's own reference voltage drifts in the cold and throws everything out.

Couple of things worth checking:

• Where is the monitor display/brains unit physically located? Move it somewhere that stays above 10°C if possible
• Have you zeroed the shunt recently? Do it when batteries are properly full and at a stable temperature
• LiFePO4 has that notoriously flat voltage curve anyway, so cheap monitors really struggle

Victron BMV-712 handles this considerably better in my experience, though "considerably better" and "£120" are words that tend to arrive together unfortunately.

Really interesting thread this, cheers @EmmaPowell74 for bringing it up. Worth knowing that LiFePO4 chemistry has an incredibly flat discharge curve anyway, which makes accurate state-of-charge readings tricky at the best of times - cold weather just compounds the problem.

What I'd check is whether your monitor is properly calibrated for LiFePO4 specifically, as many cheaper units default to lead-acid voltage thresholds. Also, are your shunt connections absolutely clean and tight? Even slight resistance there throws the coulomb counting right off.

Temperature compensation matters too - some monitors have a sensor input for this, though honestly the cheaper Renogy units are a bit limited in that regard. Might be worth logging your actual resting voltages when you know the batteries are genuinely full versus what the monitor claims, and recalibrating from there. What temperatures are you seeing overnight?

Been through almost exactly this with my shepherd's hut setup. The flat voltage curve on LiFePO4 is the real villain here — even a decent shunt monitor struggles to translate that into meaningful SOC readings, and cold temps just compress that curve further, making it nearly impossible to calibrate accurately.

What actually sorted it for me was switching to a Victron BMV-712. The combination of proper coulomb counting and a voltage reference means it self-corrects over time. It learned my bank's behaviour across a few charge cycles and now holds steady even when we dip below zero.

Worth noting — if your Epoch cells have a low-temp cutoff active, the monitor might be seeing partial charge events and losing its reference point entirely. That would explain wildly drifting readings more than the hardware itself.

Great thread @EmmaPowell74. One thing worth adding that nobody's mentioned yet — have you checked whether your shunt monitor is actually temperature-compensating its readings? Most budget units aren't, and LiFePO4 internal resistance increases noticeably in the cold, which can throw off the coulomb counting over time. Even small errors compound across charge/discharge cycles until your displayed state of charge is wildly out.

A quick fix I've found useful is doing a full charge to 100% on a warmer day to let the monitor resync itself — essentially giving it a known reference point to recalibrate from. It won't solve the underlying issue but it does stop the drift getting worse. Longer term, keeping the battery bank somewhere insulated makes a bigger difference than people expect, both for monitor accuracy and for actual usable capacity in winter.

Really good point from @T5Dream there — I'd add that it's worth double-checking your shunt is properly calibrated against a known load too. I made the assumption mine was factory-accurate out of the box and it was reading about 4% out from day one, which compounds horribly over time with integration errors.

Also @EmmaPowell74, with Epoch cells specifically I'd recommend grabbing their actual capacity spec at low temperatures from their datasheet — I think they derate more than some people expect below around 5°C. So your monitor might actually be less wrong than you think; the battery genuinely has less usable capacity available in the cold! Bit of a double whammy really. Setting a more conservative low-SOC alarm threshold for winter has helped me avoid any nasty surprises.

Really useful thread this. One thing I'd throw in — have you looked at the temperature compensation settings on the Renogy monitor itself? Some of the cheaper shunt monitors have a default charge efficiency factor set around 99% which is wildly optimistic for LiFePO4 in cold conditions. Dropping that figure down to somewhere between 95-97% can make the state of charge tracking noticeably more accurate over time. Won't fix everything @EmmaPowell74, but combined with what @WattBarry and the others have said, it might help stop the thing drifting so badly between full resets.

Really feel your pain on this one @EmmaPowell74. Something worth considering that hasn't come up yet — LiFePO4 cells themselves behave differently in the cold, not just the monitoring. Below about 5°C your usable capacity genuinely drops, sometimes quite significantly, so even a perfectly accurate monitor would show you "losing" capacity faster than expected. It's not necessarily the monitor lying to you — the battery is genuinely delivering less in those conditions.

The Epoch cells are decent quality but they're not immune to this. If your batteries are stored somewhere exposed to the elements, even insulating the battery box can make a noticeable difference to both actual capacity and the stability of your readings. Worth ruling that out before assuming it's purely a monitoring issue.