Anyone else struggling to get accurate readings from a PWM controller in cold weather?

Been having a weird one this winter with my setup on the narrowboat. I've got a 200W panel (a Renogy 200W mono) feeding into a Victron BlueSolar PWM-Light 20A, charging a pair of 110Ah lead-acid leisure batteries. Everything works fine through summer and autumn, but since the temperature dropped below about 5°C I've been getting some really inconsistent state-of-charge readings on the controller display — sometimes showing 100% when the batteries are clearly not full, other times cutting to float way too early.

I know PWM controllers don't do temperature compensation as well as MPPT units, but I wasn't expecting it to be this noticeable. The controller is mounted inside the boat in an unheated cabin, so it's not like it's sitting outside in the frost. Battery voltage at the terminals reads fine on my multimeter — around 12.5–12.6V resting — so I don't think it's a wiring issue. I've checked connections too and everything's tight.

Has anyone dealt with this with a PWM setup specifically, or is this just the point where I admit defeat and upgrade to an MPPT? I'd rather not spend the money if there's a setting I'm missing, but equally I don't want to be undercharging the batteries all winter and knackering them by spring.

@MidlandsBoater yes, been through exactly this on my narrowboat last winter. PWM controllers are notoriously bad at accurately reporting state of charge in cold temps — the battery voltage reads artificially high because cold batteries have higher resting voltage, so the controller thinks they're fuller than they actually are.

Worth knowing that lead-acid capacity also drops significantly below 5°C — you can lose 20-30% of usable capacity just from the cold alone, before any dodgy readings come into it.

Have you considered adding a proper battery monitor like a Victron BMV-712? It measures actual amp-hours in/out rather than relying on voltage estimates, so cold weather doesn't throw it off nearly as much. Made a massive difference on my setup. The Bluetooth integration is handy on a boat too — check readings without going into the engine bay in the freezing cold.

Great thread, @MidlandsBoater. Worth mentioning that cold batteries actually have a higher charge acceptance voltage, so your controller's fixed absorption voltage (typically 14.4V for sealed lead-acid) may genuinely be too low in winter conditions — meaning your batteries never fully charge even when the readings look fine. Some PWM units have a temperature compensation setting, often around -4mV per cell per °C, but the BlueSolar PWM-Light is fairly basic and I'm not sure it supports that without the optional temperature sensor accessory. Worth checking Victron's documentation for your specific firmware version. Also, condensation inside the controller casing can cause erratic readings — narrowboats are notorious for damp. Give the terminals a once-over and make sure there's decent airflow around the unit. If the inaccuracies persist into spring, that'd point more towards a controller fault than a cold weather quirk.

Been through similar with my garden office setup. One thing worth flagging — PWM controllers can't really harvest the extra voltage cold panels produce the way an MPPT would. Your Renogy panel might be sitting at 22-23V open circuit in the cold, but the PWM just clamps it down to battery voltage anyway. You're leaving energy on the table regardless of the readings issue.

@SilverMender makes a fair point about cold batteries needing higher charge voltage too — worth checking if your Victron has a temp sensor option or at least manually bumping the absorption voltage a bit over winter.

Honestly though, if accuracy and efficiency matter, a cold narrowboat winter might be the nudge to swap to an MPPT. Victron make decent ones that'd pair nicely with what you've already got.

@MidlandsBoater worth checking whether your Victron PWM has temperature compensation enabled — the BlueSolar PWM-Light does support a temp sensor accessory but it's not always wired in out of the box. Without it, the controller is just assuming a fixed 25°C and setting charge voltages accordingly, which in a Fenland winter (or on the cut in January) is well off.

In my van conversion I run a Victron BMV-712 which at least gives me accurate battery temp readings separately, so I can cross-reference what the controller thinks it's doing versus reality.

Also — and I know @CamperEwan touched on efficiency losses — PWM just dumps panel voltage down to battery voltage, so on a cold clear day when your panel Voc is spiking higher than usual, you're genuinely wasting potential harvest. Might be worth pricing up an MPPT long-term.

Really interesting thread — I've got almost the identical setup on my narrowboat so watching this closely.

One thing nobody's mentioned yet: have you checked whether your Victron PWM has temperature compensation enabled? Some of the BlueSolar PWM-Light units have a built-in sensor but it needs configuring via the VictronConnect app. Without it, the controller is still targeting summer charge voltages when your batteries genuinely need a higher absorption voltage in the cold.

Also worth checking — are your battery cables long enough to be picking up significant voltage drop? In winter I noticed my readings were out partly because I'd underestimated resistance losses when everything was cold and stiff.

Would upgrading to an MPPT be the obvious fix here, or is anyone actually getting reliable cold-weather performance from PWM with proper temperature comp sorted?

@MidlandsBoater had almost this exact issue on my Transit last winter — lead-acid really needs proper temperature compensation or you're either undercharging in the cold or gassing in summer.

One thing nobody's mentioned yet: cold batteries have higher internal resistance, so even if the controller is compensating voltage correctly, the charge acceptance is reduced. Your readings might look "off" simply because the battery's behaving differently, not necessarily a controller fault.

If the PWM-Light's temp compensation is sorted (good shout @FenlandVanLifer), I'd also check your battery terminal connections — cold causes contraction and I've seen loose connections give wildly inconsistent voltage readings at the controller.

Longer term though, a proper BMV-700 or similar shunt-based monitor will give you far more trustworthy state-of-charge data than relying on the controller display alone. Night and day difference on my van build.

Hey @MidlandsBoater, worth bearing in mind that on a narrowboat your battery bank temperature can vary quite a bit depending on where your batteries are sitting — mine are under the side bed and stay reasonably warm from the engine bay, but if yours are in a cold locker they could genuinely be 5-10°C colder than ambient. Lead-acid loses meaningful capacity below about 10°C, so what looks like inaccurate readings might actually be the batteries genuinely performing worse rather than a measurement error. Have you tried checking the actual battery voltage directly with a multimeter and comparing it to what the Victron's reporting?