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

Been wrestling with exactly this on my shepherd's hut build. My Fogstar 200Ah LiFePO4 has a Daly BMS built in, and the BMV-712 is reading about 8-12% higher SOC pretty much constantly. It's genuinely confusing when you're trying to manage loads sensibly.

From what I can gather, the Daly is working from cell-level voltage readings, whereas the BMV-712 is doing coulomb counting — tracking actual current in and out over time. In theory the coulomb counting should be more accurate for LiFePO4 because the voltage curve is so flat it's nearly useless as a reliable indicator mid-charge.

The issue I think I've got is that my BMV-712 isn't properly syncing. It only resets its 100% reference point when it hits the "charged" conditions you set (tail current, voltage threshold, time). If those aren't dialled in correctly for LiFePO4, drift creeps in over days and weeks.

A few questions for anyone who's sorted this:

• What charged voltage and tail current are you using to trigger a full sync on LiFePO4?
• Does the Daly ever communicate actual SOC via anything the Victron kit can read, or is it always going to be a separate estimate?
• Is there a rough rule of thumb for which one to trust when they diverge significantly?

Wondering if anyone's running the Victron SmartShunt instead and finding it more consistent — or whether it's essentially the same algorithm and I'd hit the same drift issues.

@Bazza the BMV-712 is almost certainly more trustworthy here, but only if you've configured it correctly for LiFePO4.

Critical checks:

• Charged voltage — set this to around 14.2V (not the default 13.2V which is lead-acid territory)
• Tail current — 1-2% of capacity, so 2-4A for your 200Ah bank
• Peukert exponent — set to 1.05 for lithium

The Daly's SOC algorithm is notoriously basic — it largely guesses from cell voltages, which are nearly meaningless across most of LiFePO4's flat discharge curve. Mine was wildly optimistic until about 20% actual capacity remaining.

The BMV uses coulomb counting from a properly synchronised full-charge point, which is far more accurate in practice.

When did you last hit a genuine full charge that triggered a proper BMV sync? That's usually where the discrepancy originates.

@Bazza had exactly this on my static caravan setup. The Daly SOC is basically just a voltage-based guess, which is pretty meaningless on LiFePO4 given how flat the discharge curve is.

The BMV-712 does coulomb counting so it's tracking actual amps in/out — far more reliable. Key thing though is making sure your charged voltage, tail current and Peukert exponent are dialled in properly for LiFePO4 in the Victron settings. Default values are set for lead acid and will throw it off.

Also worth doing a full sync cycle — charge to 100%, let the tail current drop right off, and let the BMV reset its reference point. Mine was miles out until I did that.

@Bazza the key thing nobody's mentioned yet is charge efficiency factor. In the BMV-712 settings, if you've left it at the default 95% rather than dropping it to around 99% for LiFePO4, it'll accumulate overcounting errors over time and drift high — which matches your 8-12% discrepancy pretty closely.

Also worth checking your tail current setting. If the BMV isn't seeing a proper full-charge synchronisation event regularly (LiFePO4 flat voltage curve makes this harder to trigger), the coulomb counting just compounds any initial error.

My setup — Fogstar 230Ah cells with a Victron SmartShunt — drifted similarly until I sorted those two parameters. Now it stays within 2% of actual capacity tested against a known load.

The Daly SOC readout is essentially decorative in my experience.

@Bazza worth adding to what @CarlBaker is getting at — once you've sorted your charge efficiency factor, the other critical step is making sure your BMV-712 is actually syncing correctly. It only resets to 100% SOC when it hits your configured tail current threshold and absorption voltage simultaneously. If either of those parameters is slightly off for LiFePO4, it'll accumulate drift over time and the gap between your two readings will wander rather than stay consistent.

Check what you've set as your charged voltage and tail current in the BMV settings. For a 12V LiFePO4 I'd typically suggest around 13.8V charged voltage and tail current somewhere around 2-4% of capacity — so roughly 4-8A on a 200Ah bank.

The consistent 8-12% gap does suggest a calibration issue rather than a fault though, which is the good news.