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

Been dealing with exactly this in my van build and it's doing my head in a bit.

My Fogstar Drift 100Ah LiFePO4 has the Daly BMS built in, and I've got a Victron BMV-712 doing its own coulomb counting alongside it. Most of the time they're within a few percent of each other, but after a big discharge and recharge cycle they can drift apart by 10-15% which feels significant when you're trying to manage loads carefully.

From what I've read, the BMV-712 is generally considered more accurate for day-to-day SOC tracking if you've set it up correctly — got your Peukert exponent right, charge efficiency factor dialled in, etc. The Daly tends to be a bit optimistic in my experience.

That said, the BMV needs a proper full synchronisation point (hitting absorption and holding it) to reset its reference. If you're doing a lot of partial cycles off solar like I am, it can drift without ever getting that full sync.

A few questions for anyone who's been through this:

• Do you trust one categorically over the other, or do you use both as a rough average?
• Has anyone found the ideal Peukert/charge efficiency settings for Fogstar or similar LiFePO4 cells specifically?
• Is there any way to get the Daly to talk to Victron kit so at least they're sharing data rather than operating blind to each other?

Would love to know what others are running. Feels like there should be a more elegant solution than just eyeballing two different readouts and splitting the difference.

@PaddyGibson classic "who do you trust" situation — and the answer is almost always the BMV-712, if you've set it up correctly.

The Daly BMS reports SOC based on cell voltage, which is notoriously flat across most of the LiFePO4 discharge curve — basically useless for precision. The BMV does proper coulomb counting from a known reference point.

Key things to check:

• Charged voltage — set it to ~13.6V (or wherever your charger absorbs)
• Tail current — I use 4% for LiFePO4
• Peukert exponent — set to 1.05, not the default

Once the BMV has seen a proper full charge to resync, it'll be far more accurate. Mine took about 3 full cycles to settle into something I trusted for my cabin setup.

The Daly figure is fine for a rough sanity check, nothing more.

Hey @PaddyGibson! I had almost identical confusion with my setup last year. One thing worth checking that nobody's mentioned yet — have you set the charged voltage, tail current, and Peukert exponent correctly in the BMV-712? Out of the box defaults are often wrong for LiFePO4 and that'll cause drift over time. Also, the Daly BMS SOC tends to be voltage-based estimation rather than true coulomb counting, which is why it can jump around a bit, especially mid-range where LiFePO4 voltage is so flat. The BMV-712 with properly calibrated settings will be far more reliable day-to-day. That said, both should broadly agree after a full charge cycle — if they're still miles apart after that, I'd look at whether your shunt is correctly wired as the last negative connection before the battery. Good luck with it! 🙂

Ran into this exact scenario with my narrowboat setup a couple of winters back. What finally made sense of it for me was thinking of the BMV-712 as a fuel gauge and the Daly BMS as the safety cutoff — they're doing fundamentally different jobs.

The BMV's coulomb counting drifts over time if your tail current and charged voltage thresholds aren't dialled in precisely for LiFePO4. That flat discharge curve catches people out constantly.

Worth doing a deliberate full charge cycle, letting it hit absorption properly, and watching whether the BMV resets to 100% cleanly — that synchronisation moment is everything. If it's not syncing, your settings are likely off.

@DailySolar is right that the BMV generally wins for day-to-day SOC, but only once it's properly calibrated. Raw out of the box it can be surprisingly wayward.

My Fogstar does exactly the same dance — the Daly BMS is basically the optimistic mate who says you're fine after three pints, while the BMV-712 is the sober one actually counting.