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

Had exactly this head-scratcher last winter when I was finishing the electrics in my Transit. The Daly was telling me I had 74% whilst the BMV-712 was reading 58%. Two very different stories, and one of them was about to ruin my plans for a wild camp in the Peaks.

Short version: trust the BMV-712 for day-to-day SOC, but don't completely ignore what the Daly is doing.

Here's my thinking. The Daly BMS calculates SOC based on cell voltages, which is notoriously unreliable in the flat middle section of a LiFePO4 discharge curve — the voltage barely moves between roughly 20% and 80%, so it's essentially guessing. The BMV-712 uses coulomb counting, meaning it's actually measuring every amp flowing in and out. Far more accurate in practice.

That said, the BMV-712 needs to be properly calibrated — Peukert exponent, charge efficiency factor, tail current settings all matter. Out of the box defaults are built around lead-acid assumptions. Get those dialled in for lithium and it transforms.

My Fogstar 200Ah cells behaved brilliantly once I adjusted the charge efficiency to around 99% and set a sensible tail current.

The Daly reading is still worth watching as a cross-reference for catching any cell imbalance weirdness, but I wouldn't park my van somewhere remote based solely on what it's telling me.

Anyone else found their Daly wildly optimistic compared to Victron? Curious whether it's consistent across different battery brands or just something in how Daly handles the algorithm.

@OddJobBob trust the BMV-712, full stop. The Daly BMS is just doing basic cell protection — its SOC estimate is pretty rough and resets oddly after resting.

The BMV-712 is doing proper coulomb counting, so as long as you've got the shunt wired correctly (all loads and charge sources passing through it) and you've set the charged voltage/tail current right, it'll be miles more accurate.

Main thing to check — has the BMV ever hit a proper 100% sync point? If it's never seen a full charge cycle since setup, it'll drift badly.

Set your charged voltage about 0.2V below absorption and tail current around 4% of bank capacity. Let it sync a couple of times and it'll settle down nicely.

@NZ_Marine is broadly right but it's worth understanding why the Daly is so optimistic. These cheap BMS units typically estimate SOC from resting voltage alone — which is notoriously flat across 20–80% on LiFePO4. Hopeless.

The BMV-712 does proper coulomb counting, meaning it's actually tracking electrons in and out. That said, it's only as accurate as its setup — if you haven't nailed your Peukert exponent, charged voltage threshold, and tail current settings, it'll drift over time too.

Classic gotcha: if the BMV hasn't seen a proper full charge cycle to reset its 100% reference point, it'll read low permanently.

Run the bank to genuinely full (absorption + tail current satisfied), let the BMV sync, and I'd wager your 58% figure climbs considerably. Had exactly this nonsense on my narrowboat with a Fogstar 200Ah — took three full cycles before the BMV settled.