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

Mine was doing exactly this — 94% on the Victron, 87% on the BMS — and I nearly lost my mind trying to figure out which one was lying to me.

Short answer: neither is lying, they're just measuring differently. The BMV-712 works on Coulomb counting (tracks charge in/out) and drifts if your charge tail settings are off, whereas your BMS is typically doing voltage-based estimation which has its own bag of problems depending on cell chemistry.

For my Fogstar Drift 200Ah in the shepherd's hut, I ended up trusting the BMS for protection decisions (it's the one that'll actually disconnect the pack) but using the BMV-712 as my day-to-day gauge once I'd properly set the:

• Charged voltage threshold
• Tail current %
• Peukert exponent

The BMV-712 will sync itself to 100% once those conditions are met consistently, and after a few full cycles mine was reading spot-on.

Key question though — what's your charged voltage threshold set to in the BMV? That's almost always the culprit. If it's set too high and your charger never quite hits it, the BMV never resets and the drift compounds over time.

Also, are you on LiFePO4? If yes, make sure Peukert is set to 1.05 or thereabouts — it's not a lead acid.

What BMS are you running? Someone here will have had the same combo, I guarantee it. 🔋

@ValleyChild this is something I went through methodically when I first set up my static caravan system.

The key question I'd ask is: has your BMV-712 been properly synchronised? The Victron uses coulomb counting from a reference point — if that sync point was off, the error compounds over time.

Two things worth checking:

1. Charged voltage threshold — is it set correctly for your chemistry? LiFePO4 needs a different value to AGM
2. Tail current setting — if this is too high, it syncs prematurely

My experience was that the BMS cell-level voltage reading is more trustworthy for current state, but the BMV-712 is better for tracking trends over a cycle.

What BMS are you running? Some cheaper units report SOC based purely on voltage, which is notoriously inaccurate under load.

The BMV-712 is coulomb-counting from a known full-charge reference point, while your BMS is estimating SOC from cell voltages — and LiFePO4's flat discharge curve makes voltage-based estimation about as reliable as a chocolate teapot between 20–80% SOC.

Trust the BMV-712 for day-to-day usage tracking, provided you've set the charge efficiency factor, Peukert exponent, and — critically — the "charged voltage" and "tail current" thresholds correctly in VictronConnect; get those wrong and it'll drift badly within a fortnight.

The BMS SOC reading is really just there to trigger protection cutoffs, not for accurate metering — Fogstar and Fogstar Drift BMS firmware both explicitly state this in their docs, which nobody reads.

Sync your BMV-712 whenever you hit a proper absorption-complete cycle and the two figures will eventually stop arguing like a pair of wet cats.

@OldSailor nailed it but worth adding — the BMV accuracy lives and dies by your Peukert exponent and charge efficiency settings. Mine was way off until I dialled those in properly for my Fogstar Drift cells.

Also check your tail current setting. If it's not syncing to 100% at the right point, that drift compounds over weeks and you end up with exactly what @ValleyChild is describing.

Practically speaking I trust the BMV for day-to-day usage tracking, and the BMS for actual cell-level protection. Different jobs really.

@BatterySam raises a crucial point about the charge efficiency factor — I had a maddening 8–9% discrepancy on my static caravan setup for weeks before I realised my CEF was set to 99% (Victron default) when my Fogstar Drift cells were actually closer to 97% under real-world conditions.

The synchronisation trigger is worth examining too. If your BMV isn't hitting the tail current threshold consistently — perhaps because a small parasitic load is always present — it'll drift further from reality over every cycle without ever resetting its reference point.

Check your tail current setting in VictronConnect. I run mine at 2% of battery capacity with a 3-minute timer. Once I tightened that up alongside correcting the CEF, both readings converged within 2–3%.

The BMV becomes your long-term truth-teller once properly calibrated; treat the BMS figure as a useful cross-check rather than gospel.