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

Mine showed a 12% difference for weeks until I realised my Victron had the wrong charged voltage threshold set — basically it was never properly "syncing" because it didn't think the battery was ever reaching full. 🤦

Short version: the BMV-712 is essentially doing educated maths, whilst your BMS is (usually) reading actual cell voltages — so if your Victron's Peukert exponent or charged detection voltage is misconfigured, it'll drift further from reality every single cycle.

My money's on trusting the BMS for current state, but the BMV for trend over time, if that makes sense. A BMS that's reading cells directly knows where you are right now; the Victron knows how you got there.

Few things worth checking:

• Tail current setting on the BMV — is it set for your actual battery chemistry?
• Charge efficiency factor — lithium should be close to 99%, not the default lead-acid figure
• Which BMS are you running? Some cheaper ones (cough not Fogstar's own units cough) have wildly optimistic SOC algorithms

Had this exact saga with my emergency backup setup — turns out both were wrong in different directions, which is impressive really.

What's the voltage gap between what the BMS says is "full" and what the Victron thinks is "full"? That's usually where the mystery unravels. Others on here will know this rabbit hole far too well — what's everyone else's experience?

@SaltyMechanic that's a really common one — the sync threshold catches people out constantly.

Worth also checking your tail current setting in the BMV-712. If it's set too high, the BMV thinks the battery is full before it actually is, and the SOC resets prematurely. I had mine set at the default 4% for ages on my shepherd's hut setup and it was drifting badly over time.

For LiFePO4 specifically I'd suggest:

• Charged voltage: 0.1-0.2V below your absorption setpoint
• Tail current: 1-2% of battery capacity
• Charge detection time: 3 minutes minimum

Once those three are dialled in properly the BMV tends to track really accurately. My Fogstar cells now rarely differ more than 1-2% from the BMS reading.

The BMS is useful for protection, but I generally trust the BMV-712 for actual SOC — it's coulomb counting vs cell voltage estimation.

Good point from @Gill1982 — and once you've sorted the sync threshold, don't overlook the tail current setting either.

My BMV-712 was drifting badly on my shepherd's hut setup until I realised the tail current was set to the Victron default (4%), which was way too high for my Fogstar Drift LiFePO4 cells. Dropped it to around 1–2% and suddenly the BMV started agreeing with the BMS almost perfectly.

The BMV is essentially counting coulombs between resets, so if it resets too early it'll always read high. Small thing, but it makes a surprisingly big difference over several charge cycles.

Right, so I had this exact mess with my van build. Spent three weeks convinced my Fogstar Drift cells were drifting when really the BMV-712 had never once synced properly since install.

What nobody mentions is the Peukert exponent setting. Most people leave it at default, which is fine for lead-acid but wants adjusting for lithium. Mine was introducing cumulative errors that compounded every cycle — by the end of a week the BMV and BMS were living in completely separate realities.

Set it somewhere between 1.00 and 1.05 for LiFePO4 and watch how much closer the two readings track before you even touch the sync threshold.

Won't fix everything, but if you've already addressed what @SaltyMechanic and @Gill1982 covered and you're still seeing drift, that's likely your culprit.

Good thread. To add a bit of precision here — the BMV-712 and your BMS are measuring fundamentally different things. The BMS is typically tracking cell voltages and applying a lookup table, whereas the BMV is doing coulomb counting. Neither is "right" in an absolute sense; they just have different error modes.

The BMV's coulomb counting drifts over time if sync events are infrequent (hence the threshold issues @SaltyMechanic and @Gill1982 raised), but it's more accurate between syncs. The BMS voltage-based SOC is rubbish mid-discharge on LiFePO4 because the discharge curve is so flat.

Personally on my garden office setup I treat the BMV as authoritative day-to-day, but watch for the two figures converging at full charge as a sanity check. If they don't converge at 100%, your sync parameters need revisiting.

Really useful thread this. @VictronMaster makes a crucial point about them measuring different things, and it's worth adding that Peukert's exponent is another often-overlooked setting on the BMV side that can cause drift over time, especially if you're doing heavy discharge cycles. If your exponent is set for flooded lead-acid defaults rather than lithium, your coulomb counting will skew progressively worse under load.

Also worth checking your shunt is truly in-line with all loads and charging sources — I've seen installs where a secondary charging path (solar controller wired upstream of the shunt, for instance) was invisible to the BMV entirely, so it was always playing catch-up.

Between the two, I'd generally trust the BMS for protection decisions and use the BMV for trend monitoring once properly calibrated.