Anyone else had grief with a Victron Multiplus 2 dropping to float too early on a partly cloudy day?

Been running a 48V system with a Multiplus 2 5000VA paired with 8x 200Ah LiFePO4 cells and about 3kW of solar through a SmartSolar 150/85. On overcast but occasionally bright days the MPPT is hunting all over the place — which is fine, that's just the weather — but the Multiplus keeps deciding the batteries are full and dropping to float when they're sat at maybe 80% SoC according to the BMS. It's infuriating.

I've had a dig through VictronConnect and I'm fairly sure my absorption voltage (54.4V) and float (53.6V) are set sensibly for LFP, and the tail current is at 2% which should be right. My suspicion is that during a bright spell the solar is hammering in enough current to briefly push the voltage up, the Multiplus thinks "job done" and drops into float, then the cloud comes back and nothing ever properly tops up. The BMS is a Daly 200A with comms going to a Cerbo GX if that matters.

Has anyone cracked this? I've seen mentions of adjusting the absorption time or using ESS assistant settings but I don't want to go poking around blindly and end up cooking the cells. Genuinely not sure if this is a configuration issue, a Victron quirk, or something weird about how the Daly is reporting state to the Cerbo.

Seen this exact thing with my setup. The Multiplus sees a voltage spike when a cloud clears and briefly hits absorption target, then decides "job done" and boots it to float before the cells are actually full.

Worth checking your tail current setting in VictronConnect — if it's set too high it'll exit absorption the moment current drops, which happens constantly when your MPPT is chasing clouds.

Also have a look at the absorption time minimum. Setting a floor of say 30-60 mins stops it bailing out early regardless of what the voltage is doing.

Both sorted my shepherd's hut system after a frustrating summer of the same nonsense. 🌥️

Really common frustration with LiFePO4 and variable irradiance. The issue is the Multiplus is making charge state decisions based on voltage alone, which is notoriously unreliable with lithium - a cloud clearing can spike voltage without the cells actually being full.

Worth looking at your absorption time settings in VEConfig - if you're using a fixed absorption time rather than adaptive, try extending it considerably. Also check whether you've got the "BatteryLife" algorithm enabled, as that can interfere unexpectedly.

The proper fix though is getting a Victron BMV or SmartShunt talking to the Multiplus via VE.Bus or VE.Smart Networking so charge termination is driven by actual SoC and tail current rather than just voltage. Once @TonyRoss has the DVCC properly configured with a shunt providing SoC data, the system stops getting confused by those transient voltage spikes entirely.

Great thread, this catches a lot of people out. One thing worth checking in VictronConnect is your tail current setting - if it's set too high relative to your battery capacity, any momentary surge from cloud breaks can tick that box and trigger the float transition prematurely. For an 800Ah LiFePO4 bank I'd suggest dropping it to around 1-2% of capacity (so roughly 8-16A). Also worth enabling BatteryLife if you're not already, and making sure your Multiplus and SmartSolar are properly networked via VE.Smart so they're sharing the same charge state logic rather than making independent decisions. @EssexExplorer is right that LiFePO4 compounds this because the voltage curve is so flat - small fluctuations look disproportionately significant to the charger.

@GillGibson nailed the tail current bit — but also worth checking your absorption time isn't set stupidly short in the MPPT profile, because on a 48V LiFePO4 system mine was defaulting to something embarrassingly brief that even my static caravan's ancient AGM setup would've laughed at.

Also double-check your DVCC settings in Venus OS if you're running a GX device — letting the BMS actually control charge current via DVCC rather than letting the Multiplus guess based on voltage alone sorts a lot of this nonsense out. The Multiplus isn't psychic; it's just doing what the voltage tells it, and LiFePO4's flat discharge curve makes that a rubbish indicator.

Really good points from @GillGibson and @ForestDaz already. One thing nobody's mentioned yet — have a look at your BMS communication setup. If your LiFePO4 cells are talking to the Multiplus via a Venus device or a BMS that supports DVCC, the BMS itself could be signalling "I'm satisfied" and cutting absorption short before the cells are truly full. Worth checking in VenusOS under DVCC settings whether "Use charge current from BMS" is active. Sometimes the BMS is being overly cautious on a lumpy charge day and effectively pulling the rug out. Disabling DVCC control temporarily just to test the behaviour can be quite revealing. Obviously don't leave it disabled permanently if your BMS expects to be in control, but it's a useful diagnostic step to isolate where the decision is actually coming from.

Good shout from @WonkyWarden on the BMS side — worth adding that if you're using DVCC in VenusOS, make sure the CVL (Charge Voltage Limit) coming from your BMS isn't intermittently dropping below your absorption setpoint on those patchy days. If cell voltages spike briefly when a cloud clears and the sun hammers through, the BMS can momentarily signal a lower CVL, which the Multiplus obediently follows — and then it never quite claws back up to proper absorption. Check your Venus dashboard logs for CVL fluctuations. Catches people out more than you'd think.

Really useful thread, this. Running a similar setup in my motorhome — Multiplus 2 3000VA with a SmartSolar 100/30 and 200Ah Fogstar Drift LiFePO4 — and I've had exactly this on patchy days.

One thing nobody's mentioned yet: have you checked whether your battery BMS is momentarily tripping under cloud-edge spikes and confusing the Multiplus into thinking absorption is done? I noticed my system was logging brief overvoltage events when clouds parted suddenly, which seemed to kick the charge cycle into an odd state.

Also worth looking at the DVCC settings in Venus OS if you've got a GX device — forcing a minimum charge current can stop the Multiplus bailing out when the MPPT briefly dips. Made a noticeable difference for me on variable days.

Has anyone found a reliable set of DVCC values that works well for this sort of profile?