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

Had exactly this on my narrowboat setup last winter. MultiPlus-II 3000 was calling float at around 80% SOC according to my BMV-712 — drove me mad for weeks.

Turned out to be a combination of two things:

1. Absorption voltage set too low — I'd followed some generic advice online and had it at 14.2V for my Fogstar Drift lithium cells. Bumped it to 14.4V and it helped considerably.
2. Absorption time — I had it on the adaptive setting, which on a cold boat (baseplate temperatures dropping to near zero) was timing out way too early because the initial charge acceptance was sluggish.

The fix that actually sorted it was switching from adaptive to fixed absorption time — 2 hours for my 200Ah bank — and letting the Victron do its thing properly. You can set this in VictronConnect under the charger tab.

Worth also checking your battery cabling for voltage drop. If the MultiPlus is seeing a slightly inflated voltage at the terminals due to resistance elsewhere, it'll think the battery is fuller than it is and drop to float prematurely.

What's your battery chemistry? LiFePO4 and AGM behave quite differently here and the troubleshooting path is fairly different for each. Also worth posting your current charge profile settings — absorption voltage, float voltage, and whether you're using BMS assistant or not.

If you've got a Cerbo GX or similar, pull the VRM logs and look at the charge current curve when it drops to float. That'll tell you a lot about whether it's a voltage issue or a timing issue.

@PennineNomad the "two thin" cliffhanger is killing me, finish your post mate 😄

But yes, seen this exact behaviour on my boat. Nine times out of ten it's either dodgy battery temp sensor readings throwing the charge algorithm off, or the absorption time is set too short for your bank size.

What's your battery chemistry? If you're on LiFePO4 and haven't set the charge profile correctly in VEConfigure, the MultiPlus will basically give up early because the voltage curve is so flat it thinks it's full.

Also worth checking your DC cable sizing — voltage drop under load can confuse the SOC calculation even with a properly shunted BMV. I had 3m of slightly undersized cable causing grief before I bothered properly sorting it.

Run VE.Configure and post your absorption/float voltage settings — that'd narrow it down quickly.

Great thread timing — I've got a MultiPlus-II 3000 on my widebeam and wrestled with this for ages last autumn.

@PennineNomad please do finish that post, genuinely curious whether your root cause matches mine! @ExBrickie I feel your pain on that cliffhanger 😄

In my case the culprit was the absorption time being set too short in VEConfigure — it was defaulting to a fixed 1-hour absorption rather than adaptive. Once I switched to adaptive absorption and bumped the repeat absorption interval down, the BMV-712 and MultiPlus started agreeing with each other much more sensibly.

Worth checking your battery temperature compensation settings too if you've got the temp sensor fitted — mine was reading slightly high near the engine room and effectively telling the charger the batteries were warmer than they were, which shortens absorption automatically.

@PennineNomad the cliffhanger aside, I've had almost identical behaviour on my shepherd's hut build — MultiPlus-II 2x120 calling float embarrassingly early on a 200Ah Fogstar Drift lithium bank.

Root cause in my case: absorption voltage set too conservatively combined with a tail current threshold that was far too high. The unit was seeing what it thought was a sufficiently reduced charge current and declaring job done, when really the battery was nowhere near full.

Key settings to audit in VEConfigure:

• Absorption voltage — needs to match your battery's actual spec, not the default
• Repeated absorption time — extend this
• Tail current as a percentage — drop it significantly for lithium

Also worth checking your DC cable sizing. Resistive losses were masking my actual charge current at the battery terminals, which skewed the whole algorithm.

@PennineNomad the "two thin" is genuinely the most suspenseful thing I've read since the last series of Line of Duty. I'm assuming it's either two thin cables or two thin excuses for why you didn't check it sooner 😄

Had something similar on my narrowboat last spring — BMV-712 and Multi