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

Had exactly this debate with myself last year when I set up my garden office system.

Short answer — trust your BMS for protection, trust the BMV for trends.

The BMS is watching cell voltages directly and will cut things off if needed, so it's your safety net. But it's rubbish at actual SOC percentage because most cheaper BMS units just estimate from resting voltage, which is notoriously unreliable under load.

The BMV-712 does proper coulomb counting — it's literally measuring every amp going in and out. Once you've got it calibrated correctly (nailed down your Peukert exponent and charge efficiency factor), it's far more accurate day-to-day.

Main things to check if they're drifting apart:

• Has the BMV synced recently? It resets to 100% on a full charge — if you're never hitting absorption properly it'll drift over time
• What's your charged voltage threshold set to in the BMV settings?
• Is the shunt wired correctly with everything running through it? Missed a direct connection once myself, took ages to figure out why numbers were off

What batteries are you running? Fogstar Drift cells tend to have a very flat discharge curve which makes voltage-based SOC even less reliable, so the BMV becomes more important there.

Curious what the actual gap is between the two readings — 5% or 20%? Makes a difference to where the problem likely is.

@BerlingoSolar makes a solid point, but there's a bit more nuance worth adding.

The BMV-712's coulomb counting drifts over time unless it gets a proper synchronisation event — typically hitting absorption voltage with low current. If you're rarely fully charging (common in winter with solar), that drift compounds significantly.

My narrowboat setup runs both, and I've found the BMV's percentage becomes unreliable below about 20% SOC precisely because accumulated counting errors stack up. The BMS cell voltages are absolute truth at the extremes.

What I'd actually recommend:

• Calibrate your BMV's Peukert exponent and charge efficiency factor carefully — defaults are rarely right for LiFePO4
• Set a realistic charged voltage synchronisation point
• Use the BMS low-voltage alarm as your hard floor, not the BMV percentage

Both instruments together tell you more than either alone.

Been through this exact thing on the boat. My Fogstar lithium pack's BMS was reading 87% while the BMV-712 had 71% — drove me mad for weeks.

Turned out my BMV charge efficiency factor was set wrong. Default is 95% but lithium is closer to 99%. Made a massive difference once I corrected it.

Also worth checking your tail current and charged voltage settings — if the BMV never sees a proper sync point it just drifts further and further out. Mine syncs reliably now whenever I hit absorption.

@DevonDweller is right about drift, but half the time it's just misconfiguration rather than a fundamental flaw with coulomb counting. Worth going through the BMV setup properly before deciding which to trust.

@Jason1970 that 16% gap screams your BMV hasn't had a proper sync in a while — mine did the same on the narrowboat until I nailed the charged voltage and tail current settings (I run 13.5V / 2% for LiFePO4).

Key settings to dial in:

• Charged voltage — just below float
• Tail current — 1-4% of capacity
• Peukert exponent — set to 1.05 for lithium, not the lead-acid default

Once calibrated, my Victron BMV-712 and Fogstar BMS sit within 3% of each other all day long. The BMV just needs a reliable full-charge reset point to recalibrate its coulomb counting — without that it's basically guessing in the dark like me in the galley at midnight.

@Jason1970 that kind of gap usually sorts itself once the BMV gets a clean full-charge sync. Mine was all over the place on the cabin setup until I tightened the charged voltage and tail current settings properly — Victron's defaults aren't really tuned for lithium out of the box.

Worth double-checking:

• Charged voltage set just below your BMS high-voltage cutoff
• Tail current around 1–2% of your battery capacity
• Peukert exponent at 1.05 for lithium (default is higher, assumes lead-acid)

Once those three are dialled in, my BMV-712 barely drifts between syncs now.