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

Had this exact headache with my setup last year. Short answer: trust neither blindly, they're measuring different things.

The BMV-712 is coulomb counting from a shunt — it tracks charge in/out and calculates SOC from that. Dead accurate in theory, but drift accumulates over time if your synchronisation points aren't dialled in (peukert exponent, charged voltage threshold, tail current settings all matter).

Your BMS is typically reading cell voltages and inferring SOC from that. Problem is, LiFePO4's voltage curve is notoriously flat in the middle — so voltage-based SOC readings can be quite wobbly between roughly 20–80%.

What I'd suggest:

• Check your BMV-712 sync settings first. If it's not resetting to 100% properly at full charge, the drift compounds fast
• Compare both readings immediately after a full charge — they should agree at 100%. If they don't, the BMV needs calibrating
• In everyday use, I lean on the BMV for trend data and the BMS for protection thresholds

My Fogstar Drift 200Ah has its own BMS and I also run a BMV-712. They sit within 2-3% of each other most of the time, but after a week of partial cycling without a full charge the BMV can wander noticeably.

Curious what BMS you're running and how far out they are — makes a difference to the advice really. Also worth knowing if you're on LiFePO4 or a different chemistry.

@CotswoldNomad makes a solid point about them measuring different things.

Worth adding: the BMV-712's accuracy degrades if your shunt calibration is off or if you haven't set the Peukert exponent correctly for your battery chemistry. Lithium behaves very differently to lead-acid here — Victron's default settings are often tuned for AGM.

On my setup (Fogstar Drift 200Ah LiFePO4 with a Daly BMS), I saw a consistent 8-12% divergence until I:

• Set Peukert exponent to 1.05 (near-flat for LFP)
• Enabled the synchronisation at 100% SOC threshold properly
• Tightened the tail current setting to ~2% of capacity

The BMS SOC is also notoriously unreliable on cheaper units — they often just interpolate from resting voltage, which is nearly flat across most of LFP's discharge curve.

Use the BMV as your primary gauge, but let the BMS protect the cells.

Good points already made. What I'd add from running a Fogstar Drift 200Ah on the boat — the real problem is synchronisation drift over time.

The BMV-712 needs a proper "full charge" event to reset its 100% reference point. If your absorption/float settings don't match what the BMS considers full, they'll slowly diverge. Mine was showing 15% difference after a few weeks before I sorted the charge voltage parameters.

Check your charged voltage and charged current settings in the BMV match your actual charge profile. On LiFePO4 I've got mine set to:

• Charged voltage: 27.0V (24V system)
• Tail current: 4%

Once those align, the two readings track much more closely. The BMS is still your safety net — the BMV is your fuel gauge. Different jobs.

Great thread. One thing nobody's mentioned yet: the BMV-712 needs correct Peukert exponent and charge efficiency factor settings or its coulomb counting drifts badly over time. Most people leave them at defaults which are tuned for AGM, not lithium.

For LiFePO4 I run:

• Peukert: 1.05 (near-ideal)
• Charge efficiency: 99%
• Tail current: 2-4% of capacity

Also — your BMS SOC can jump wildly if it's purely voltage-based (many cheaper units are). LiFePO4's flat discharge curve makes voltage a terrible real-time SOC indicator; it's only reliable at full rest.

My practical workflow: trust the BMV-712 day-to-day, but force a full synchronisation cycle weekly by genuinely topping out your Victron MPPT until absorption completes properly. Both readings then converge and you've got a solid reference point again. Works a treat on my cabin array.

@DailySolar is right about Peukert but worth being precise — Peukert's exponent matters far less with LiFePO4 than lead-acid. It's essentially 1.0 for lithium, so if you're running LiFePO4 (which most people are these days) that particular setting is basically a non-issue.

The charge efficiency factor is the one that actually drifts and causes the BMV to slowly desync over time.

My Victron BMV on the boat — 280Ah EVE cells with a Daly BMS — I found the bigger culprit was an incorrect tail current setting for synchronisation. If it's set too low the BMV never registers a full charge and just accumulates error.

Get the tail current dialled in correctly and the two readings will converge pretty quickly after a proper full charge cycle.

Really useful thread this. Running a BMV-712 alongside a Daly BMS in my motorhome, and the thing that sorted my discrepancy was actually cable placement — my shunt wasn't truly catching all loads because of where I'd positioned it on the negative busbar. Even a small parasitic draw bypassing the shunt compounds badly over days.

Worth checking your wiring diagram carefully before diving into settings.

Also for emergency backup use specifically (my main reason for the setup), I now treat the BMS voltage reading as my hard floor — if that says danger, I stop regardless of what the BMV says. The BMV is better for day-to-day trend watching. Two instruments, two different jobs really.

@DailySolar makes a fair point about charge efficiency factors — what settings are you running for LiFePO4 specifically? Mine's at 99% but wondering if that's optimal.