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

Ran into this exact situation on the boat last summer. Had my Daly showing 73% while the BMV-712 was reading 58% — felt like I was going mad staring at two screens giving me completely different stories.

Short answer from my experience: trust the BMV-712, but only if you've got your settings dialled in properly. The Daly's SOC is largely voltage-based, which is notoriously unreliable mid-cycle, especially with LiFePO4's flat discharge curve. The BMV does proper coulomb counting — it's actually measuring current flowing in and out.

The catches though:

• You must set your charged voltage, tail current, and Peukert exponent correctly in the BMV or it'll drift badly over time
• The BMV needs a proper synchronisation point (full charge, hitting tail current) to reset its reference — if you're never fully charging, it loses accuracy gradually
• Your shunt placement matters enormously — anything drawing power that bypasses the shunt won't be counted

On my setup I've got the shunt handling everything except the engine starter, and I fully charge via solar or shore power every few days. Since sorting all that, the BMV barely drifts.

The Daly SOC figure is handy for a rough sanity check but I'd not rely on it for anything critical.

What does your charge cycle look like? Are you getting regular full charges? That's usually where the drift creeps in.

@VDH_Boats the Daly is guessing, the BMV-712 is measuring — trust the Victron every time, it's doing proper coulomb counting through a calibrated shunt whereas the Daly is just waving its finger in the air based on cell voltages.

That said, your BMV is only as good as its setup — wrong battery capacity entered, or you've never let it hit a proper 100% sync point (fully charged, absorption complete, tail current hit), and it'll drift too.

Classic fix: set your charged voltage and tail current correctly in the BMV settings, do one full charge cycle from near-empty, let it calibrate itself properly. Mine took a couple of cycles on the static caravan setup before it stopped being dramatic about it.

@Rob1963 is broadly right but it's worth understanding why the discrepancy happens. The Daly derives SOC primarily from cell voltage, which is notoriously non-linear across the middle of the charge curve — you can sit at 3.3V per cell for ages and it'll barely move. The BMV-712 is counting actual amp-hours in and out, which is far more reliable provided you've correctly configured the capacity, Peukert exponent, and charge efficiency factor.

The critical thing most people miss: the BMV-712 needs a proper full-charge synchronisation event to reset its reference point. If yours never reaches your configured tail current at absorption voltage, it'll drift over time. Check your charged voltage and tail current settings in VictronConnect — I had mine set too aggressively on my cabin system and the BMV was reading 6% low after a fortnight of partial-state cycling.

@VDH_Boats had this exact split-personality situation in my van build — BMV-712 and Daly disagreeing like a married couple on holiday.

The bit nobody mentions: your BMV-712 is only as good as its setup. If your charged voltage threshold, tail current, or Peukert exponent are off, it'll drift too. Spent an embarrassing afternoon convinced my Fogstar cells were dying before I realised I'd never properly configured the charge detection settings.

Key things to nail in the BMV:

• Correct battery capacity (actual, not rated)
• Tail current (~4% for LiFePO4)
• Charged voltage set just below your absorption endpoint

Once calibrated properly, the BMV wins every time — but a misconfigured BMV versus a Daly is basically a coin flip dressed in expensive Victron clothing.

@VDH_Boats the real problem is that your Daly has never actually seen a proper full charge to reset its reference point — sort that first and the gap usually closes dramatically.

In my motorhome I let the BMV-712 do the heavy lifting for day-to-day decisions, but I make sure it gets a genuine 100% sync at least once a week (absorption complete, tail current hit). Without that anchor point even Victron's coulomb counting drifts over time.

Key thing @Rob1963 and @TransitAdventure haven't mentioned: check your shunt is wired correctly — everything bar the battery negative should go through it, otherwise the BMV-712 is blind to certain loads and will read high, not low.