Struggling to get accurate SOC readings from my Victron BMV-712 — is my shunt wired wrong?

I've had my 200Ah lithium (LiFePO4) setup running for a few weeks now and the BMV-712 is driving me a bit mad. The state of charge seems to drift quite badly — after a full charge it'll show 100% fine, but by the next morning it's reading maybe 40% even though the battery voltage is sitting at 13.1V and everything seems perfectly normal. It doesn't add up.

The shunt is a 500A Victron unit wired in on the negative side between the battery negative terminal and the rest of the system. I've followed the diagram in the manual, but I'm wondering if I've got a rogue negative wire somewhere bypassing the shunt entirely — maybe from the inverter or the chassis earth. Has anyone else had this kind of drift with a similar setup?

For context it's a campervan build with a 300W solar array going through a Victron SmartSolar 100/30, a 1000W inverter, a compressor fridge, and a few USB/12V loads. Total draw overnight is probably around 40-50Ah. The BMV history is showing consumed Ah that seem way off compared to what I'd expect from those loads.

EdMason | 47 posts

@Turbo19 A common culprit with LiFePO4 is the synchronisation settings. The BMV needs to "see" a proper full charge to reset to 100% — make sure your charged voltage threshold matches what your charger actually hits, and that the tail current is set appropriately low (LiFePO4 has a very flat charge curve so default settings designed for AGM often don't trigger correctly).

Also worth checking: is everything running through the shunt? Any loads or chargers wired directly to the battery bypassing it will cause drift. The shunt must be the sole connection between battery negative and the rest of the system. Even a sneaky earth wire can throw it right off.

What are your current BMV parameters set to? Happy to help dial them in if you share the values.

OakSoul | 23 posts

Had almost the exact same issue when I first set mine up. Quick question — is every single load and charge source wired on the battery side of the shunt, or have you got anything connected directly to the battery terminals?

Even a small parasitic draw bypassing the shunt (a BMS balance lead, a Cerbo GX, anything) will cause the BMV to lose track over time. Mine was drifting badly because I'd connected my Victron SmartSolar directly to the battery rather than routing it through the shunt's load terminal.

Also worth checking — what have you set for Peukert exponent and charge efficiency factor? LiFePO4 wants very different values to lead-acid. I'm running 1.05 and 99% respectively on my Fogstar cells and it's been rock solid since.

What does your wiring diagram look like?

QuietTrekker | 184 posts

Worth double-checking your charged voltage threshold in the BMV settings. With LiFePO4 it wants to be around 13.4–13.6V (for 12V), not the default which is set more for AGM. If that's too high it never triggers a sync and the SOC just wanders off.

Also — tail current setting. Default is often 4% but lithium charges differently, some people drop it to 1–2%.

@EdMason's sync point is solid too, both things together usually sort it.

Mine was doing exactly this on my garden office setup before I adjusted those two values. Took maybe 10 minutes in VictronConnect and it's been rock solid since. Fogstar batteries, same 200Ah size as yours weirdly.

Check Settings > Battery in VictronConnect rather than fiddling with the buttons on the unit itself — much easier.

BoatGemma | 34 posts

On my boat setup I found the charged voltage threshold was the sneaky culprit — the default setting assumes lead-acid voltage curves. For LiFePO4 you want to set the "charged voltage" to around 13.4V (or 26.8V on a 24V system) rather than the higher default. That way the BMV actually triggers a SOC sync when the battery is genuinely full.

Worth also double-checking your tail current percentage — if it's set too high the unit thinks charging is complete before the battery's actually balanced, and your SOC drifts from the start of every cycle.

What voltage are you running — 12V or 24V? And do you know which firmware version the BMV is on? Some of the older versions had quirks with LiFePO4 profiles that a firmware update sorted.