Anyone else running a Victron SmartSolar in a shepherd's hut through winter? Seeing weird float behaviour

Noticed my 100/30 SmartSolar has been dropping out of float and cycling back to bulk most mornings even when the batteries are clearly full. Running 2x 200Ah Fogstar lithiums wired in parallel, 24V system.

Temperatures here have been dropping to around 3-4°C overnight and I'm wondering if that's messing with the voltage readings. SOC sits at 98-99% on the BMS app but the MPPT seems convinced it needs to do a full charge cycle again by 8am.

Checked the absorption voltage settings — I've got it at 27.4V with a 2hr absorption timer. Wondering if I need to bump that slightly or if this is more of a temp compensation issue. The hut isn't insulated brilliantly so the batteries do get cold.

Anyone dealt with this on a lithium setup? Tempted to just disable temp compensation entirely since Fogstar recommend fixed voltages, but not sure if that'll make it worse.

Moorey79 | 📍 Yorkshire Dales | ⚡ Off-grid since 2019

@Lefty I had almost identical behaviour with my setup last winter. Worth checking your absorption voltage and time settings in VictronConnect - lithiums need a much tighter absorption window than the defaults are often set to. If absorption time is too long, the controller overshoots slightly, the BMS pushes back, and you end up with that bulk/float cycling loop.

Also check whether Fogstar have published specific charge parameters - I'd be running absorption at around 14.2V with a tail current setting to exit early rather than relying on a fixed time. The SmartSolar handles this really well once configured properly.

What firmware version are you on? There were some quirks in older versions with lithium profiles worth ruling out.

@Lefty this is almost certainly your absorption timeout being too short for winter solar profiles. When irradiance is low, the controller hits absorption voltage but can't sustain the current long enough to satisfy the algorithm, so it falls back to bulk at dawn.

In VictronConnect, check your absorption time setting — if you're on adaptive absorption, it scales with bulk duration, which in December can be trivially short.

For Fogstar lithiums specifically, I'd also verify your tail current threshold. Mine was set far too high (default lead-acid value carried over), causing premature float entry followed by the exact cycling you're describing once resting voltage sagged overnight.

Worth grabbing a VRM portal log if you have a GX device — the charge state graph makes the pattern immediately obvious rather than guessing from panel behaviour alone.

@Lefty worth checking whether it's actually a temperature compensation issue rather than the timeout. On my narrowboat the Victron was pulling down the float voltage on cold mornings based on the battery temp sensor readings, which then looked like the batteries weren't holding charge properly and triggered a fresh bulk cycle.

If you're not running a Smart Battery Sense or the temperature sensor dongle, the MPPT will be making assumptions about battery temperature that are almost certainly wrong for a shepherd's hut in winter. Mine was essentially phantom-cycling for three weeks before I realised.

Pull up the history graph in VictronConnect and look at your float voltage over a 24-hour cold period — if it's dipping below what you've set, that's your culprit right there.

@Lefty one thing nobody's mentioned yet — check your tail current setting in VictronConnect. With lithiums, if your tail current threshold is set too high relative to your actual panel output on a grey winter day, the MPPT will never see the battery reach that endpoint current and will time out of absorption prematurely, kicking it back to bulk.

On my garden office setup (similar 100/30 with Fogstar 200Ah cells), I had to drop the tail current down to around 1–2% of capacity (roughly 4A on 200Ah) before the float behaviour stabilised through December/January. Default Victron presets tend to assume higher irradiance.

Worth cross-referencing with the VRM portal if you've got a Cerbo or even the dongle — the charge cycle graphs will show you exactly where it's dropping out.

@KeithMurray makes a good point on tail current — that was exactly what caught me out first winter at my cabin with a similar setup.

One thing worth adding: check your re-bulk voltage threshold in VictronConnect. On lithiums it should sit quite a bit lower than lead-acid defaults — I had mine triggering re-bulk at 13.1V which on Fogstar cells is nowhere near empty, just normal resting voltage after a settled float.

Bumped it down to 12.8V and the phantom re-bulk cycles stopped almost entirely. Winter mornings with cold batteries and low voltage sag can easily nudge you over that threshold otherwise.

@Lefty also double-check your absorption timeout — mine on the narrowboat was set to the default 2 hours which sounds plenty until you realise a gloomy British December morning means you're basically restarting the whole charge cycle before the kettle's even boiled 🫖