Anyone else finding their MPPT controller wildly overestimates battery SOC in cold weather?

Running a Victron SmartSolar 100/30 with a 200Ah 12V LiFePO4 (Fogstar Drift) in my van. During this recent cold snap we've had — temps dropping to around 3–4°C overnight in Cornwall — the controller has been showing 100% SOC by mid-morning even when I know the battery hasn't had a proper full charge in days. Been away from hookup for about a week and usage has been pretty heavy (diesel heater controller, 12V fridge, laptop).

The SmartSolar is set up with the correct LiFePO4 charge profile (absorption 14.2V, float 13.5V) and the Fogstar has its own BMS obviously. My gut says the cold is affecting the resting voltage and the controller is reading that as a fuller battery than it actually is. I've got the battery temp sensor connected, so it's not like Victron is completely in the dark about the conditions.

Has anyone else seen this with LiFePO4 in winter? I'm wondering whether tweaking the tail current setting might help it get a more honest picture of the SOC, or whether this is just a known limitation and I should be leaning on the Coulomb counting from a proper battery monitor instead. Got a BMV-712 sitting in a box I haven't fitted yet — maybe now's the time.

Same issue on my boat last winter. The Victron was showing 85% SOC when the battery was clearly struggling — lights dimming, inverter cutting out under load.

Turned out the main culprit was the tail current threshold. In cold temps, LiFePO4 acceptance current drops significantly, so the battery appears "full" to the MPPT before it actually is. Worth checking your absorption tail current setting in VictronConnect — I dropped mine from the default and it made a noticeable difference.

Also worth noting: Fogstar Drift has a built-in BMS that will throttle charge current below around 5°C anyway, so your MPPT and the BMS are essentially fighting each other for accurate readings.

The Victron app's battery temperature compensation helps but only really works properly if you've got a temp sensor wired in. Are you running one?

Yeah, this catches a lot of people out. The root cause is that LiFePO4 cells have a remarkably flat voltage curve at normal temps, but cold shifts that curve downward — so the controller's voltage-based SOC estimation goes haywire. At 3–4°C your battery's internal resistance also increases noticeably, meaning it'll sag under load more than the controller expects.

Worth checking if your Fogstar Drift has a low-temperature charge cutoff built in — most do around 0°C, which is sensible, but the SOC readings can still be misleading above that threshold.

My suggestion: connect via Bluetooth in the Victron app and set up a proper battery monitor rather than relying solely on the MPPT's estimates. A BMV-712 or the SmartShunt would give you coulomb counting rather than voltage guessing, which is far more reliable in cold conditions. Makes a proper difference.

Hey @CornishSolar — worth checking whether you've got the battery temperature sensor set up properly in VictronConnect. The SmartSolar does have temperature compensation built in, but it needs accurate readings to work correctly. Without the temp sensor, it's essentially guessing.

Also, LiFePO4 resting voltage curves get quite flat and compressed in the cold, so voltage-based SOC estimation becomes even less reliable than usual. The Fogstar Drift should handle 3–4°C fine operationally, but the BMS may be throttling things in ways the MPPT can't see.

If you're not already using a BMV-712 or SmartShunt alongside it, that'd be my main recommendation — coulomb counting gives you far more trustworthy SOC data than the MPPT alone can provide, especially in these conditions. Makes a massive difference over winter.

WonkyRanger | 847 posts

Worth checking your battery temperature compensation settings in the Victron app, @CornishSolar. LiFePO4 has a much flatter discharge curve than lead-acid, and at 3–4°C the internal resistance climbs noticeably — the MPPT can struggle to read voltage accurately under load, which throws the SOC calculation right off.

If your Fogstar Drift has a built-in BMS with temperature sensing, make sure VictronConnect is actually pulling that data rather than relying purely on voltage. Also worth enabling the "tail current" SOC sync option if you haven't already.

As @NeilJackson says, it's a known cold-weather quirk. A Victron BMV-712 or SmartShunt would give you proper coulomb counting and make the SOC readings far more trustworthy — honestly it's the missing piece in most Victron van setups.

Del58 | 1,203 posts

Good shout from @PennineDweller and @WonkyRanger on the temperature sensor, but I'd also add — have a look at your tail current settings in VictronConnect. At low temps, LiFePO4 acceptance rate drops significantly, so the battery might be cutting off absorption early and the controller's calling it "full" when it genuinely isn't.

I had almost identical symptoms with my Fogstar setup last winter. Bumping my absorption time minimum and reducing the tail current threshold helped enormously. Also worth double-checking your charge voltage hasn't crept down due to temperature compensation actually working too aggressively — sometimes it overcorrects and you end up undercharging in the cold rather than overcharging.

Cornwall at 3–4°C overnight isn't extreme, but it's enough to throw things off noticeably with LiFePO4. What's your current tail current set to?

RussGreen | 312 posts

Something nobody's mentioned yet — LiFePO4 voltage curves flatten out massively in the cold. The SmartSolar is estimating SOC from voltage, and at 3–4°C that voltage-to-SOC relationship shifts considerably. Even with perfect temp compensation dialled in, you'll get drift.

Ran into exactly this in my Transit conversion last January parked up in the Peak District. Controller was reading 85% when I was actually closer to 60%.

Practical fix that helped me: set a manual SOC reset each morning once the battery's had chance to warm slightly from the habitation heater. Also worth enabling the Fogstar's Bluetooth monitoring alongside VictronConnect — cross-referencing both gives you a much more honest picture than relying solely on the MPPT's estimate.

KangooSolar | 156 posts

@RussGreen makes a really important point. To add to that — the flat voltage curve issue gets worse because Victron's SOC algorithm relies heavily on voltage when it loses coulomb counting accuracy. So you end up with a sort of compounding error. What I'd suggest @CornishSolar is logging your actual consumed Ah via a BMV-712 or SmartShunt if you haven't already — gives you a proper reference point independent of voltage readings. The Fogstar Drift is a solid battery but any LiFePO4 will fool a voltage-based system in those conditions.

MuddyNomad | 634 posts

@RussGreen nailed it — that flat discharge curve is the real villain here. In my shepherd's hut last February, I watched my Victron app show 87% SOC right up until the system suddenly started pulling from a nearly exhausted bank. Genuinely alarming at 6am.

What actually helped me was switching away from voltage-based SOC estimation entirely. Enable the Tail current setting in VictronConnect and let it judge full charge by absorption tail rather than voltage thresholds. Far more reliable once temps drop below 5°C. The SmartSolar 100/30 supports it natively — took about ten minutes to configure properly.

FrostyTinker | 847 posts

Worth adding a practical workaround nobody's touched on yet — if you're using the Victron app, have a look at your tail current settings. In cold weather, LiFePO4 cells absorb charge more slowly, so the battery can appear "full" to the controller well before it actually is. Bumping your absorption time up slightly during winter months helps compensate. Also, @CornishSolar — are you using a battery temperature sensor with your SmartSolar? Without one, the controller's making educated guesses at best. They're cheap and make a genuine difference. Mine transformed the accuracy of SOC readings overnight once fitted.

JulieGraham | 23 posts

Really useful thread — I've been seeing something similar with my garden office setup (Victron SmartSolar 75/15, Fogstar 100Ah 12V LiFePO4). Temps have been brutal lately and my SOC readings have been all over the place.

Quick question for the more experienced folks: does the same cold weather SOC confusion affect the charge algorithm itself, or is it purely a display/calculation issue? I'm wondering whether the controller might be ending a charge cycle too early as well, thinking the battery is fuller than it actually is?

MuddyMaker | 89 posts

Has anyone tried pairing a dedicated battery monitor like the Victron BMV-712 alongside the SmartSolar to get a more accurate SOC reading? I'm wondering whether the MPPT's built-in estimation is just fundamentally less reliable than a proper coulomb-counting monitor — especially in cold when the voltage/SOC relationship shifts. On my motorhome I've been considering this setup but haven't pulled the trigger yet. Does the BMV actually track SOC better through temperature swings, or does it still drift without a temperature sensor connected directly to the battery?