Victron BMV says 100% but cells are drifting apart — BMS issue or just needs balancing?

Had exactly this with my shepherd's hut setup last summer. BMV was reading 100% and showing full charge, but when I dug into the VictronConnect app the cells were wandering — worst case I had about 120mV spread across a 280Ah LiFePO4 bank from Fogstar.

Turned out the passive balancing on the BMS only kicks in above a certain voltage threshold, and if your charge cycle never quite pushes each cell high enough for long enough, the balancer barely gets a look-in. The BMV calculates SOC from current in/out (coulomb counting) and voltage — it doesn't actually know what individual cells are doing. So 100% on the BMV just means it thinks the bank is full, not that it's balanced.

What fixed it for me:

• Bumped the absorption voltage slightly and extended the absorption time in the Victron MPPT settings
• Let it sit at absorption for a good few hours over several days
• The BMS eventually dragged the cells back into line — spread dropped to under 20mV

Worth checking whether your BMS has active or passive balancing too. Passive dissipates excess as heat and is slower. Active balancing is more effective but you're looking at a different class of BMS entirely.

Could also just be a weak cell starting to show its age, especially if the pack has some cycles on it. If the drift keeps getting worse even after proper top-balancing cycles, I'd be suspicious of one rogue cell.

What BMS are you running? And is the drift happening at top of charge or does it start appearing lower down the curve? That distinction matters a lot for diagnosing this properly.

PikeSeeker | 847 posts

@WD40Wizard11 That 120mV spread is definitely telling you something. Worth checking whether your BMS is actually doing anything at top of charge — some cheaper units will report fine but the passive balancing only kicks in above a threshold voltage that your cells might not be consistently hitting.

Before blaming the BMS outright, I'd look at your absorption stage. If you're running a float-heavy profile and not spending enough time in proper absorption, the cells never get the chance to drift back together. What's your charge profile set to in the Victron? That mismatch between BMV reading 100% and cells wandering often comes down to the BMV trusting voltage/current thresholds rather than actual cell-level data.

@WD40Wizard11 120mV spread at top of charge is a red flag — that's not just "needs a balance", that's your BMS potentially not hitting the absorption phase properly.

On my shepherd's hut build I had similar. Turned out the Daly BMS I was using had its cell overvoltage threshold set too low, so it was cutting charge before the cells ever had chance to balance properly. Swapped to an JK BMS and the difference was night and day — active balancer makes a huge difference vs passive.

Worth checking:

• What BMS are you running?
• Are you actually reaching absorb voltage on all cells, or just pack voltage?
• How old is the pack?

If it's Fogstar Drift cells they should balance pretty tightly once the BMS is set correctly. A 120mV drift on a healthy pack suggests something upstream is cutting out early.

@WD40Wizard11 120mV is significant but not necessarily a death sentence — depends on the cells.

Few things worth checking:

• Is your Victron absorption voltage actually high enough to trigger balancing? A lot of passive BMS units only balance above ~3.45V/cell, and if you're set conservatively you might never hit that threshold long enough
• What's the capacity of the drifting cell at rest? A cell that reads low at top-of-charge but recovers quickly is probably just unbalanced. One that stays low is likely degraded
• The BMV reads pack voltage, not cell-level — so 100% SOC is essentially meaningless when cells are drifting

Personally I'd run a deliberate full absorption cycle and watch the cells in the BMS app in real time. If they don't converge after an hour at absorption, you've likely got a weak cell rather than just a balancing issue.

Had this exact scenario on my boat build last year. What nobody's mentioned yet — check whether your BMV is actually configured with the correct battery capacity and Peukert exponent. A misconfigured BMV will hit 100% based on voltage alone (the "charged detection" threshold) while completely ignoring what the cells are actually doing internally.

Also worth running a proper absorption cycle at the correct voltage for your specific cells rather than a generic LiFePO4 setting. I found my Victron MPPT was cutting out of absorption too early, which meant the top balancer in my BMS never got a proper chance to work.

@MarineGaz is right that 120mV isn't automatically terminal — I've seen packs recover well after a few slow full cycles. But definitely pull the BMS data logs if your unit supports it before spending money on cells.