Batterieladung auf 100% klappt nicht immer?

Anyone else's LiFePO4 stubbornly refusing to hit 100% SOC via the Multiplus-II, or is my narrowboat just haunted?

Running a similar Victron stack — CerboGX, BMV-712, a pair of MPPT 100/50s — and my Fogstar Drift 200Ah cells occasionally decide 97% is close enough, thank you very much, and just... stop absorbing.

Suspect it's a tail current threshold argument between the BMV and the Multiplus that nobody told me about when I bought the kit.

Few things I've already ruled out:

• Not a BMS cutoff (checked comms, all happy)
• Absorption voltage is set correctly at 14.2V
• It's not just "winter laziness" — happens in summer too

The annoying part is it's inconsistent — sometimes charges to 100% no bother, other times it plateaus like it's given up on life, much like me when I'm doing a third engine-bay wiring job at midnight.

For context this is my emergency backup bank too, so I'd quite like to actually know it's genuinely full when the cruiser is moored and shore power's available.

Anyone running a similar Victron setup on a boat or van who's cracked this one — particularly around charge floor settings or whether the DVCC tail current config is the culprit?

Also curious whether Fogstar cells specifically have any quirks here vs the big-name alternatives.

@FormerMechanic43 classic one this — worth checking your absorption voltage and tail current settings in VEConfig.

A few things to look at:

• Tail current — if it's set too high (e.g. 4-5% of capacity), the Multiplus declares "full" before the cells actually are
• Absorption time — adaptive absorption can cut out early if the battery was only slightly discharged
• BMS comms — if your BMS is talking to the Cerbo via VE.Bus or VE.Can, it may be overriding the charger's voltage target prematurely

On my garden office setup I had similar grief with Fogstar cells until I dropped the tail current to around 1-2% and extended minimum absorption time.

Also double-check the BMV-712 charge efficiency factor — if it's wrong, your displayed SOC might be the problem rather than the actual charge state.

@FormerMechanic43 narrowboat here too, so this caught my eye. Had almost identical issues last winter — turned out my BMV-712 charge efficiency factor was set too low, so the SOC was perpetually pessimistic rather than the battery actually failing to charge.

Worth checking whether your actual cell voltages via the CerboGX match what the BMV is reporting as SOC. If the cells are balanced and sitting at proper resting voltage post-charge, the battery is probably fine and it's a calibration/synchronisation issue rather than a real charging problem.

Also — what BMS are you running? Some third-party BMS units don't communicate the "full" signal cleanly to the Multiplus-II, which can cause it to drop out of absorption early without ever triggering a proper SOC sync.

@FormerMechanic43 worth checking your battery's internal BMS communication if you're using CAN-bus. My Fogstar Drift cells were occasionally dropping the DVCC handshake mid-charge, so the Multiplus would back off before actually hitting full absorption.

Quick way to diagnose — pull up the VRM portal and look at the charge current graph. If it's cutting out abruptly rather than tapering smoothly down to your tail current threshold, that's your culprit.

Also double-check your BMV-712 charge efficiency factor and Peukert exponent — if those are slightly off, the SOC calculation drifts and you'll never see 100% even when the battery's genuinely full.

Really interesting thread this — reminds me of when I first wired up my static caravan setup and spent weeks convinced something was fundamentally broken.

One thing nobody's mentioned yet: have you checked your charge current relative to battery capacity? If you're pushing too low a charge current into a larger bank, the Multiplus-II can sometimes struggle to ever satisfy the tail current threshold and just... gives up before declaring 100%.

My BMV-712 was showing 97-98% perpetually until I realised my charged voltage and tail current settings in the BMV itself were slightly mismatched from what VEConfig had on the Multiplus side. Both need to agree, essentially.

@CE_Builds touched on the absorption side — but the BMV settings are a separate conversation that's easy to overlook entirely. Worth cross-referencing both simultaneously rather than one at a time.