Victron BMV-712 showing different SOC to my BMS — which one do I trust?

Had exactly this drama in my van build — BMV reckoned I had 73% left whilst the BMS was insisting on a cheerful 91%, and I'm sitting there like a referee at a very boring tennis match.

Short answer: trust the BMV-712 for day-to-day SOC, but don't completely ignore your BMS — they're measuring different things in different ways.

The BMV does coulomb counting (tracks every amp in and out), whereas most BMS units estimate SOC from cell voltage curves, which gets increasingly dodgy under load or when your cells are warm. Voltage-based SOC on lithium is notoriously flat across the middle range — could be 50% or 80% and the voltage barely blinks.

Key things that'll make your BMV drift over time though:

• Incorrect battery capacity setting — be honest about your usable Ah, not the sticker on the Fogstar
• Peukert exponent — tweak it properly for lithium (usually closer to 1.05)
• Charged voltage / tail current settings — if it's not resetting to 100% cleanly after a full charge, errors compound

The BMV will sync back to 100% on a proper full charge, which self-corrects the drift. If yours isn't doing that, that's your first fix.

Anyone else running dual monitoring on a lithium setup? Curious whether folk with a Victron Smart Shunt alongside a Daly or JK BMS are seeing similar gaps — or whether it's just me attracting electrical chaos.

@Chalky trust neither unconditionally — they're both just doing maths on dodgy inputs, like two sat-navs arguing whilst you drive into a lake.

The BMV-712 is coulomb-counting from a shunt, so if you haven't nailed your Peukert exponent and charge efficiency factor it'll drift over time. The BMS is typically just reading cell voltages and guessing, which is famously optimistic at high SOC — hence your suspiciously cheerful 91%.

My shepherd's hut setup: I treat the BMV as the long-term trend and the BMS as the low-voltage bouncer. Sync the BMV at 100% only on a genuine full absorption charge, and suddenly it behaves itself.

Fogstar cells in particular sit very flat on the voltage curve mid-range, so the BMS is basically useless between 30–80% anyway.

@Renogy_Nerd is onto something but I'd push back slightly — the BMV-712 with a properly sized shunt and accurate battery capacity entered is significantly more reliable for SOC than most budget BMS units.

The BMS SOC is usually just voltage-based estimation. Voltage lies, especially under load.

The BMV is coulomb counting — it's actually measuring current in/out. Much harder to fool.

That said, the BMV drifts if it never hits a proper full sync. When did yours last reach 100% absorption and hold it? If it's been weeks, that 73% reading could be pessimistic.

My narrowboat setup — Victron BMV-712 with Fogstar Drift cells — I trust the BMV almost unconditionally, but I make sure it syncs properly every few days.

Check your charged voltage and tail current settings in the BMV config first.

@ShedGenius makes a fair point about calibration, but here's what finally made sense of it for me — the BMV is counting coulombs in and out, so any small inaccuracy compounds over weeks like interest on a bad loan. My Fogstar Drift 100Ah cells taught me this the hard way last winter. The BMS meanwhile is typically doing voltage-based estimation, which is its own mess given how flat LiFePO4 curves are mid-range.

The real tell is what happens at the extremes. When I hit a genuine full charge — absorption complete, tail current dropped right off — I force a BMV sync. After that, trust the BMV for the next cycle. Think of the BMS figure as a rough sanity check rather than gospel.

@Jock is probably heading where I was going to go — the BMV tracks coulombs in and out, full stop. Your BMS is estimating SOC using voltage curves, cell balancing data, and whatever algorithm the manufacturer fancied that Tuesday.

On my narrowboat I run both, and they drift apart noticeably mid-cycle, especially with LiFePO4's famously flat discharge curve where voltage tells you almost nothing useful between 20–80%.

The BMV wins provided you've set your Peukert exponent correctly and — crucially — let it properly synchronise at 100% after a genuine absorption cycle. If it never sees a proper full charge, the drift compounds over weeks.

My workflow: trust the BMS for cell-level protection, trust the BMV for fuel-gauge purposes. They're answering different questions.