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

Had this exact headache last year with my setup. Short answer: neither is 100% right, but for day-to-day decisions I lean on the BMV-712.

Here's my reasoning — the BMV is doing proper coulomb counting with a quality shunt, and once you've nailed the Peukert exponent and charge efficiency factor for your specific cells, it's genuinely accurate over time. Most budget BMS units (looking at you, anything bundled with a cheap Fogstar knockoff) are estimating SOC from voltage curves, which is notoriously unreliable especially mid-charge and under load.

That said, the BMV does drift if you don't let it sync properly. It needs to hit your set charge voltage and hold there long enough to trigger a full sync — if your system rarely does a proper absorption cycle, the displayed SOC creeps away from reality over weeks.

A few things worth checking:

• Is your BMV shunt correctly sized and wired as the only path for current?
• Have you set the charged voltage threshold to match what your BMS actually considers full?
• What's the discrepancy — 2-3% or more like 15-20%?

A small gap is probably just calibration. A large consistent gap suggests either your BMS is being optimistic (common) or your BMV settings need revisiting in VictronConnect.

What BMS are you running? Some of the JK units actually have decent SOC tracking once you've run a few full cycles through them. Would be useful to know the specifics before anyone gives you concrete advice.

@JubileeClipHero95 solid reasoning, but I'd push back slightly on dismissing the BMS entirely.

The bit people miss is that the two are measuring fundamentally different things. The BMV-712 is counting coulombs in/out (with all the accumulated drift that entails), whereas your BMS is responding to actual cell voltages in real time. Neither is ground truth.

What I do on my narrowboat setup — Victron BMV-712 paired with Fogstar Drift cells and a JK BMS — is use the BMS as a sanity check for the extremes. If the BMS is screaming about cell imbalance or hitting protection thresholds, I trust that immediately over whatever the BMV says.

The BMV wins for mid-range SOC decisions though. Just make sure your:

• Peukert exponent is properly configured
• Charged voltage threshold is correctly set
• Tail current % matches your actual battery spec

Those three settings cause most of the BMV drift issues I've seen reported here.

Yeah both valid points tbh. The thing I'd add — calibration matters massively with the BMV-712. If your battery capacity figure is slightly off (and it usually is with Fogstar Drift cells that haven't been properly cycled in), the coulomb counting drifts over time and compounds the error.

I had mine reading ~8% high after 3 months on my garden office setup before I did a proper full charge reset.

So practically speaking:

• Use BMS for protection decisions (it's seeing cell-level data)
• Use BMV for daily planning (how much have I actually used)
• But sync the BMV regularly via a full absorption charge so it can self-correct

The two disagreeing is often a sign one needs recalibrating rather than evidence either is broken.

There's a pattern I've noticed on the narrowboat — the gap between the two readings tends to widen after a partial cycle rather than a full one. The BMV-712 drifts if it never gets a proper synchronisation point, because it's never had a clean 100% reset to anchor itself to.

What sorted it for me was being deliberate about getting a full absorption cycle through regularly, letting the BMV sync properly. After that, the two figures came much closer together.

@WattHamish is right that capacity matters, but I'd also check your tail current setting. If it's too high, the BMV calls 100% too early, and everything downstream is skewed from that point onwards. Worth reviewing before you start suspecting the hardware itself.

Great point from @CraigCross — that drift on partial cycles is exactly what I see on the boat too.

The bit I'd add: synchronisation triggers are your friend here. The BMV-712 resets its SOC to 100% when it detects a full charge, but if your absorption/float thresholds aren't dialled in correctly for your specific cells (especially important with LiFePO4 vs AGM), it'll sync at the wrong moment and you're immediately chasing errors.

On my setup I've got the tail current set to around 1-2% of capacity and it made a noticeable difference to how well the two readings track each other over time. Worth checking in VictronConnect what yours is set to — the factory default is often way off for lithium.

Fogstar Drift cells in particular seem to behave differently enough from the BMV's assumptions that a bit of manual tuning pays dividends.

Really useful thread this. One thing nobody's mentioned yet — what happens when you've had a power cut to the BMV-712 itself?

I've found on my shepherd's hut setup that if the BMV loses power even briefly (loose connection, swapping fuses etc.), it loses its SOC reference point entirely and starts counting from whatever's in the battery at that moment. Suddenly it's reading 94% when the BMS is showing 67%.

In that scenario I'd actually trust the BMS reading short-term, then let the BMV resync on a full charge before relying on it again.

Have you checked your Peukert exponent and charge efficiency factor settings @JubileeClipHero95? Those two settings wrecked my accuracy for months before I dialled them in properly for my Fogstar Drift cells.