Victron Multiplus keeps dropping to float too early — anyone else had this?

Mine did exactly this for about three months before I finally admitted it wasn't "just a quirk" and actually investigated — turns out my battery temp sensor had fallen off the negative terminal and was just dangling there reading ambient air temperature like a useless little liar.

Worth checking a few things before you go full conspiracy mode on the Multiplus itself:

• Absorption time — if it's set too short in VictronConnect it'll bail out early regardless
• Tail current setting — if your battery bank is small relative to your charger, it hits the tail current threshold fast and thinks it's done
• Battery temp sensor placement (see above, ask me how I know)
• Battery voltage sense wires — loose connection here will give the Multiplus a wildly optimistic picture of what's happening

What batteries are you running? I've got Fogstar Drift 200Ah lithium in the motorhome and the default Victron lithium profile needed tweaking before mine stopped doing the exact same dance. The absorption voltage was set slightly low so it was touching float conditions almost immediately after hitting target voltage.

Also — are you monitoring via Cerbo or just the app? Sometimes what looks like float is actually a brief recovery absorption, the display labelling can be a bit ambiguous depending on firmware version.

Could genuinely be nothing dramatic, but could also be masking an underperforming cell somewhere. What does the SOC look like when it drops?

@FormerCop classic case of the temp sensor gremlins — I spent weeks chasing phantom absorption cutoffs on my Multiplus-II before realising the same thing. Worth adding for anyone else debugging this: even if the sensor appears seated, corrosion on the negative terminal can give wildly incorrect resistance readings that the unit interprets as elevated battery temperature, causing it to derate the charge curve aggressively.

Check your VRM portal data — if you're seeing the reported battery temp spike erratically rather than tracking ambient gradually, that's your culprit rather than any DVCC or absorption time misconfiguration in VE.Configure.

Also worth confirming your BatteryLife state in VictronConnect isn't set to Optimised if you're running a proper LiFePO4 bank — that algorithm is really designed for lead-acid behaviour and will absolutely cut absorption short in ways that look like faults.

@FormerCop @RustyRoamer — worth also checking your absorption voltage threshold in VictronConnect if the temp sensor checks out fine. On my Multiplus-II (running a 280Ah Fogstar Drift LiFePO4 bank), I found the default charge parameters were set for flooded lead-acid from a previous owner. The absorption end-current setting was ludicrously high, so the charger was declaring "done" after about 20 minutes regardless of actual SOC.

Specifically, look at Battery > Charge voltages and cross-reference against your BMS documentation. For LiFePO4 the tail-current cutoff is typically around 2-5% of capacity — if yours is set higher, absorption will terminate prematurely every time.

You can log the charge cycle data via VRM portal and overlay current against voltage to pinpoint exactly where it's jumping to float. That graph usually tells the whole story without any guesswork.

@KangooNomad makes a solid point on the absorption voltage, but I'd add one more culprit that catches people out: the absorption time setting itself, particularly if you're running an adaptive absorption algorithm.

On my Multiplus-II 48/3000, I found that with Fogstar lithium cells the adaptive timer was cutting absorption absurdly short during partial-state-of-charge cycles — the algorithm inferred the battery was "full enough" based on previous cycle history. Worth pulling your VRM logs and looking at the actual absorption duration trend over several days rather than just diagnosing a single event.

Also check your tail current setting in VEConfigure — if that's set too high relative to your bank's actual resting acceptance current, it'll trigger float prematurely regardless of voltage or temperature compensation. 2–5% of C10 capacity is typically the right ballpark for LiFePO4.

Had this exact issue in my shepherd's hut setup last winter. Turned out to be the absorption time setting — mine was on adaptive and the algorithm was cutting short because it thought the battery was fuller than it was.

Worth checking Battery Safe Mode in VE.Configure too, not just VictronConnect. Some settings only show up there.

Also, if you're running Fogstar lithium cells with a third-party BMS, double-check whether the BMS is sending a premature "battery full" signal via the VE.Bus. Mine was doing exactly that — Multiplus trusted the BMS over its own voltage readings and bailed out of absorption early every time.