Victron MultiPlus keeps switching to float too early — anyone else had this?

Running into something similar with my MultiPlus-II 48/3000 on the garden office build. It's been cutting to float at around 80% SoC according to the Cerbo GX, which clearly isn't right.

A few things I've been working through methodically:

Battery settings first — are you running lithium or lead-acid? I'm on Fogstar Drift 48V with the Victron DVCC enabled, so I wanted to rule out whether the BMS was signalling prematurely. Took a while to establish it wasn't the culprit.

Absorption voltage threshold — worth double-checking your absorption voltage is actually set correctly in VEConfigure. Mine had somehow defaulted to a lower value than the Fogstar spec sheet recommends. Could be worth comparing against your battery manufacturer's figures.

Tail current setting — this is the one I'm currently investigating. If the tail current percentage is set too high, the MultiPlus interprets the charge as "done" well before it actually is. What's yours set to?

A couple of questions for anyone who's solved this:

• Did you find the repeated absorption timer approach (setting a fixed absorption time rather than adaptive) more reliable?
• Is there a known firmware version that handles this better? I'm on v502 currently.

Has anyone gone down the route of letting the BMS handle charge control entirely via the VE.Can integration rather than relying on the MultiPlus's own charge algorithm? Wondering if that sidesteps the issue altogether or just introduces different headaches.

@MuddySkipper had exactly this on my narrowboat setup last winter. Nine times out of ten it's the absorption voltage set too low, so the MultiPlus thinks it's "done" before the batteries are actually full.

Check your charge profile in VE.Configure — what are your absorption and float voltages set to? For 48V LiFePO4 you want absorption around 55.2V (3.45V/cell), float at 53.6V or so. If someone's left it on a lead-acid profile that'll cause this instantly.

Also worth checking:

• Tail current setting — if it's too high it'll terminate absorption early
• Whether your BMS is sending a charge current limit via DVCC that's confusing things
• Battery temperature compensation (should be off for lithium)

What battery chemistry are you running? Makes a big difference to where the fault likely sits.

@MuddySkipper worth checking whether your battery capacity setting in the Cerbo is actually correct — if it thinks the bank is smaller than it really is, the tail current threshold (the "end of absorption" trigger) gets hit way too early. Classic gotcha.

Also, what chemistry are you on? If you're running Fogstar Drift cells or similar LiFePO4, make sure you've got a proper BMS assistant installed in VE.Configure rather than just relying on the default lead-acid charge algorithm. The MultiPlus genuinely has no idea what to do with lithium without it.

Run a VE.Configure session and screenshot your absorption voltage and your tail current percentage — those two together usually tell the whole story. Mine was doing identical nonsense until I dropped the tail current from 4% down to 1% on my 280Ah bank.

@MuddySkipper oh this took me ages to figure out on my shepherd's hut build — turned out my battery temperature sensor wasn't properly seated, so the MultiPlus was compensating voltage downward and calling it "done" way too early.

Worth checking your adaptive charging history in VictronConnect too. If the absorption time has been consistently short over several cycles, it starts learning that as the new normal, which just compounds the problem.

Also — what battery are you running? My Fogstar Drift cells needed the charge profile tweaking quite specifically before things settled down properly. The default Victron profiles are a bit conservative out of the box for certain LiFePO4 chemistries.

Had mine sorted by resetting the adaptive algorithm and starting fresh with manually-set absorption voltage. Took about three full cycles before the Cerbo was reporting sensible numbers again.

@MuddySkipper had something similar crop up during my van build — worth checking your absorption time setting in VictronConnect. If it's set too short, the MultiPlus can hit its time limit before the batteries are actually full and just drops to float regardless of SoC.

Also, are you running LiFePO4? If so, double-check your charge profile is actually set up for lithium — I accidentally had mine on the AGM preset for weeks before I noticed. The absorption voltage ceiling was completely wrong for my Fogstar cells.

What does the charge current look like just before it switches? If it's still pulling significant amps when it drops to float, that's a strong clue it's a time/voltage threshold issue rather than the batteries genuinely being full.

@MuddySkipper one thing nobody's mentioned yet — check your tail current setting in VE.Configure. If it's set too high, the MultiPlus interprets the naturally tapering absorption current as "battery full" and exits to float prematurely, well before actual capacity is reached.

On my LiFePO4 bank (Fogstar Drift 200Ah cells, 16S 48V), I had this exact behaviour until I dropped tail current from the default 4% down to 1% of battery capacity. Made an enormous difference — absorption now runs its full duration properly.

Also worth cross-referencing your Cerbo's SoC reading against a proper coulomb counter if you have one, since the Cerbo's SoC estimate can drift significantly, especially if the synchronisation point (typically 100% at end of absorption) never actually triggers correctly because you're dropping to float too early. It becomes a self-reinforcing problem.