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

Had this exact headache last year when I set up my garden office system. Short answer: it depends on what's causing the discrepancy.

The BMV-712 works by coulomb counting — it tracks every amp in and out from a reference point. Problem is, if your synchronisation settings aren't dialled in properly (tail current, charged voltage threshold etc.), it drifts over time. Mine was reading 8% higher than reality after about 3 weeks.

Your BMS, on the other hand, is working from cell voltages directly. That's more "honest" in real-time but voltage-based SOC on lithium is notoriously flat in the middle range — tiny voltage differences = massive SOC swings depending on load.

What I'd actually do:

• Short term — trust the BMS for protecting the cells (that's its job)
• Long term — trust the BMV-712 once you've calibrated it properly

The Victron documentation on sync settings is decent, but the key thing people miss is setting the correct battery capacity. Fogstar Drift cells, for example, can drift from their rated capacity slightly depending on temperature, so factor that in.

Also worth checking whether your BMS is reporting usable capacity or total capacity — some budget units report the full chemical capacity which skews everything.

What BMS are you running? And how far apart are the readings — a few percent or wildly different? That'll narrow it down considerably.

Anyone else had persistent drift issues with the 712 on larger LiFePO4 banks? Curious whether increasing the tail current percentage helped or made things worse.

Great summary @BreezySparky. One thing worth adding — the BMV-712 is only as good as its calibration. If your charged voltage threshold or tail current settings are off, it'll never get a proper sync point and the drift compounds over time. I'd check those first before assuming either device is faulty.

Also worth knowing: most BMS units report SOC based purely on cell voltage, which can be wildly optimistic mid-discharge on LiFePO4 given how flat the discharge curve is. So paradoxically, the coulomb counter often gives you a more reliable picture during actual use, whilst the BMS tends to be more accurate at the extremes (full and empty).

If they're more than 5-10% apart after a full charge cycle, I'd trust the BMV and re-examine your BMS settings.

My Fogstar Drift cells laugh at my BMV-712 every single morning until I sorted the tail current setting — had mine at 4% when it should've been 1% for lithium, so it was calling "full" about 20 minutes too early and the whole SOC just drifted off into fantasy land over a fortnight. 🧱

Basically: BMS knows cell voltages, BMV knows coulombs — neither one's the full picture on its own. Cross-reference both and if they're more than ~5% apart after a proper full charge, something's calibrated wrong rather than broken.

Good shout from @ExBrickie94 on tail current — that fixed half my issues too.

The other thing I'd add: peukert exponent. BMV default is set for lead acid. Running LiFePO4? Set it to 1.00 or your coulomb counting will drift over time, especially under heavy loads.

My setup's a 280Ah Fogstar pack and once I sorted:

• Peukert to 1.00
• Tail current to ~1% of capacity
• Charge efficiency to 99%

...the BMV and BMS were within 2-3% of each other consistently.

BMS SOC can also be rubbish if it's a cheap unit that just reads cell voltage — those jump around all over the place under load. Victron's coulomb counting wins there once properly calibrated.

Really useful thread this, bookmarking it for future reference.

Can I throw a question into the mix though? I've got a Victron BMV-712 paired with a Fogstar 200Ah LiFePO4 in my motorhome emergency backup setup, and I'm seeing maybe a 3-5% drift over about a week of light use.

Following on from what @ExBrickie94 and @PennineSolar have said about tail current and Peukert — for LiFePO4 specifically, should I essentially be setting Peukert exponent to 1.05 (near-linear) and letting the synchronisation at full charge do the heavy lifting for accuracy?

Or is the weekly drift more likely a shunt calibration issue — zero drift offset creeping in? I've read the BMV can accumulate small errors if there are parasitic loads the shunt isn't seeing properly.

Anyone had similar with a motorhome install where you've got multiple 12V circuits running off the same bank?