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

Had this exact issue with my setup at the static last summer. Multiplus II 48/3000 kept jumping to float when the batteries were nowhere near full — Fogstar Drift 48V pack sitting at maybe 85-90% according to the BMS app but the Victron was already backing off.

Turned out to be a combination of things for me:

1. Absorption voltage set too low — VEConfigure had it at a default that didn't match what Fogstar actually recommend for their cells
2. Tail current threshold — mine was set to 4% which sounds reasonable but with a larger pack it's triggering float way too early in real terms

Once I bumped the absorption voltage to the correct figure from Fogstar's documentation and dropped the tail current to around 1-2%, it made a massive difference. Batteries now hitting proper full charge before float kicks in.

Worth checking whether you've got a Venus GX or Cerbo in the mix too — if DVCC is enabled and your BMS is sending charge current limits, that can confuse the whole absorption/float logic. Took me ages to work that one out.

What battery chemistry are you running? LiFePO4 behaves quite differently to lead-acid here, and a lot of the default Victron profiles are still tuned more conservatively for lithium than they need to be.

Would also recommend grabbing VE.Smart networking if you've got a SmartShunt — letting the shunt handle state-of-charge rather than relying purely on voltage measurement makes the whole system a lot more accurate.

@VivaroDream had this exact same headache at my cabin last year. Turned out my absorption voltage was set a touch low in VictronConnect — Multiplus was "seeing" full when it wasn't.

Worth checking your absorption voltage and tail current settings. The tail current is the one people miss — it's the threshold where Victron decides "right, job done, float time." Default can be too high for Fogstar packs.

Also double-check the battery's own BMS isn't throttling charge current early and confusing the Multiplus into thinking absorption is done.

A quick call to Fogstar support sorted mine — they gave me exact recommended settings for the Drift pack. Dead helpful, that.

Check your tail current setting in VEConfigure — mine was left on the default 4% and the Multiplus was basically declaring victory the moment my Fogstar stopped sweating, bumped it down to 1% and it actually does a proper job now.

@NorfolkVanLifer is bang on about tail current, but there's another culprit that bit me badly with my Pylontech US3000C stack last spring — the repeated absorption time setting.

Mine was left at 1 hour, so even if the batteries hadn't properly saturated, the Multiplus would dutifully clock off after 60 minutes and call it done. Felt very pleased with itself. Batteries were not.

Bumped it to 2 hours in VEConfigure and the difference was immediately obvious on the VRM portal — you could actually see the current tapering properly before float kicked in.

Worth cross-referencing your battery manufacturer's recommended absorption duration too. Pylontech's own documentation is surprisingly specific about this, and the Fogstar Drift specs are worth a read if you haven't already. Often the default Victron profile just isn't dialled in for lithium out of the box.

@Pylontech_Queen curious what that other culprit was — don't leave us hanging!

I had something similar with my garden office setup. Fogstar Drift pack, Multiplus II 48/5000. Turned out my absorption time in VEConfigure was set too short as a fallback — so even when tail current hadn't been hit, the timer was cutting things off early anyway.

Worth checking your "maximum absorption time" setting. Default can be surprisingly conservative depending on firmware version.

Also — are you on DVCC with a Cerbo? If so, make sure the BMS is actually feeding accurate SoC data. Mine was losing comms intermittently and the Multiplus was essentially flying blind, defaulting to conservative behaviour.

VEConfigure > Charger tab, screenshot your settings and post them up. Much easier to diagnose with actual numbers rather than guessing.