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

Had exactly this on Persephone last spring — BMV-712 was reading 74% whilst the BMS was showing 61%. Drove me absolutely mad for a fortnight.

Here's what I eventually worked out: they're measuring fundamentally different things, using different methods. The BMV does coulomb counting — it tracks every amp flowing in and out from a reference point. The BMS is typically reading cell voltages and inferring SOC from those, which on lithium is notoriously unreliable across the flat middle section of the discharge curve.

The real culprit in my case was an incorrect Peukert exponent and a charged voltage threshold that wasn't quite right, so the BMV never got a clean sync point to reset from. Once I dialled those in properly through VictronConnect, the two readings converged to within about 3-4% — close enough for narrowboat life.

A few things worth checking:

• Is your BMV getting a proper full charge sync? If it never hits your tail current threshold, it'll drift badly
• What chemistry are you running? LiFePO4 makes voltage-based SOC estimates particularly sketchy
• Has your BMS got a genuine coulomb counter built in, or is it purely voltage-based?

Personally, after sorting the calibration, I trust the BMV more for day-to-day monitoring — but I treat the BMS as the safety net that actually protects the cells. They're doing different jobs really.

What BMS are you running? That'd help narrow it down. Others on here will have opinions too — this comes up fairly regularly and there are some strong views!

@VanNicola — they're measuring fundamentally different things, which is the crux of it.

BMV-712 does coulomb counting from a reference point — it's integrating current in/out over time. Drift is inevitable, especially if your charge cycle never hits 100% absorption properly to trigger the sync.

Your BMS is likely working from cell voltage curves, which on LFP is notoriously flat across 20–80%, so tiny voltage measurement errors translate to massive SOC swings.

On my narrowboat setup with Fogstar Drift cells, I trust the BMV-712 long-term but only if I'm disciplined about letting it fully synchronise periodically. The BMS I treat as a protection device, not a fuel gauge.

Check your BMV synchronisation settings — specifically the Charged voltage threshold and Tail current percentage. Bet they're not matched to your actual charge profile.

@VanNicola @PanelEwan — ooh this is so relevant to my narrowboat setup right now! Quick question though — does it matter which one you use as your primary reference for day-to-day decisions, or do you basically need both readings together to get a proper picture?

I've got a Fogstar Drift 200Ah on Marigold and both my BMV-712 and the BMS app are constantly disagreeing with each other, usually by about 8-12%. I've been defaulting to whichever reads lower just to be safe, but I suspect that's not actually the right approach?

Also wondering whether a proper full charge cycle to sync everything up would help — I've read somewhere that the BMV-712 recalibrates its 100% reference point after a genuine full charge. Is that accurate?

@LindaClark90 the real answer is trust neither blindly — the BMV drifts if your shunt connections are even slightly dodgy, whilst most cheap BMS voltage-based SOC readings are basically astrology.

What I do on my motorhome: use the BMV-712 as primary during discharge (coulomb counting is decent when calibrated), but force a full sync reset periodically so it re-anchors to 100% — Victron calls this the "tail current" setting, worth dialling in properly for your chemistry.

Fogstar Drift cells in particular have a brutally flat discharge curve, so voltage-based BMS readings are useless until you're nearly empty.

If your two readings diverge by more than ~5% consistently, suspect a parasitic load the shunt isn't seeing — or a dodgy shunt connection.

@FormerCop @VanNicola so basically my battery has an existential crisis and two contradictory opinions about how alive it is 😅

Genuinely though — I had exactly this on my Fogstar Drift cells last winter. 13% gap between my Victron BMV and the BMS. Turned out my shunt had

@EcoFlowMaster closer than you think — the BMV-712 is a coulomb counter (integrating current over time), whilst your BMS is typically doing voltage-based SOC estimation, and they drift apart for fundamentally different reasons.

BMV errors accumulate from: shunt resistance tolerance, parasitic loads it can't see, and a charged/tail-current threshold that needs proper calibration in Victron's settings.

BMS errors come from: cell chemistry curves that are brutally flat in the middle on LFP — meaning 61% and 74% can look nearly identical voltage-wise.

Practical fix: trust the BMV for day-to-day relative tracking, but let a genuine full charge reset it to 100% — that's where the two usually reconcile. My Fogstar Drift 200Ah and BMV-712 agree within 3% once I nailed the Charged Voltage and Tail Current settings.