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

Had this exact situation with my shepherd's hut build last summer. Short answer: trust the BMV-712, but understand why they disagree.

The Daly BMS calculates SOC primarily from voltage, which is notoriously unreliable mid-charge-cycle — especially with LiFePO4's flat discharge curve. It also has no concept of charge efficiency losses or self-discharge over time.

The BMV-712 uses coulomb counting through a proper shunt, so it's tracking actual electrons in and out. It's far more accurate provided you've got it configured correctly:

• Charged voltage — set this to whatever your charger's absorption voltage is (typically 14.2–14.4V for 12V LiFePO4)
• Tail current — I run mine at 2% of battery capacity
• Peukert exponent — set to 1.05 for lithium, not the default lead-acid value
• Charge efficiency factor — 99% for LiFePO4

The BMV will also drift over time if it rarely sees a full synchronisation event. If your system never properly tops out and triggers that "charged" threshold, the SOC reading slowly becomes less reliable.

One thing worth noting — the two units should broadly agree when the battery is full or nearly empty, because those are the points where voltage actually means something on a LiFePO4 cell. If they're miles apart at 100% or at low SOC, that's telling you something's misconfigured somewhere.

What battery are you running, and what charge source? Worth knowing whether the Daly is getting accurate cell voltage data too — a dodgy balance lead will skew its readings completely.

@LiamFrost70 nails it on the voltage-based limitation, but there's another layer worth knowing.

I ran both on my tiny house system for months before I understood what was happening — the Daly resets its SOC reference point every time it hits a full charge voltage threshold, but it doesn't account for how slowly you got there. A float arrival isn't the same as a proper absorption cycle.

The BMV-712 tracks coulombs in and out continuously, so its history is richer. The key is getting your charged voltage, tail current, and Peukert settings dialled in properly — Victron's defaults are conservative.

Mine drifted until I set the tail current to match my Fogstar cells' actual charge acceptance curve. Once synced properly, both readings converged within about 3%.

The BMV wins, but only if you've done the setup work.

Really useful thread this — exactly the kind of thing I've been puzzling over with my motorhome setup!

Quick question for the group though: does the BMV-712's accuracy depend heavily on how well you've configured the Peukert exponent and charge efficiency factor? I set mine up fairly quickly following a YouTube guide and I'm wondering if I've left accuracy on the table by not dialling those in properly for my Fogstar Drift lithium cells specifically.

@LiamFrost70 @Chippy — did either of you spend time tweaking those advanced settings, or did you find the defaults were close enough for LiFePO4? I've seen some folk recommend setting charge efficiency to 99% for lithium but I'm not 100% sure that's right for every cell chemistry. Would love to know what values you're actually running!

@LiamFrost70 good summary but I'd add one practical gotcha — the BMV-712 is only as accurate as its synchronisation points. If you're not hitting 100% SOC regularly (full absorption + tail current dropping below your set threshold), it'll drift over time regardless.

On my static van setup I had the tail current set too high initially so it was syncing prematurely. Took weeks to notice the creep.

Worth double-checking your charged voltage and tail current settings in VictronConnect match your actual battery spec. Fogstar Drift cells in particular have a fairly low recommended tail current — easy to get wrong out of the box.

The Daly will always look optimistic around the middle SOC range because LiFePO4 voltage curves are so flat there. That's not a bug, it's just voltage-based SOC being fundamentally limited.

@RelayDream raises the sync point issue which is genuinely the crux of it all.

My van's BMV-712 was reading 15% high for weeks last winter. Turned out I'd set the "charged voltage" threshold slightly too low — the unit never thought it had reached 100%, so it never reset its baseline. Spent a miserable evening in a layby in November chasing phantom capacity before I twigged.

Worth checking your Tail Current and Charged Voltage settings in the Victron Connect app. Both need to be dialled in for your specific cells. LiFePO4 wants different values to AGM — obvious in hindsight, less obvious when you're copying settings from some random YouTube tutorial aimed at a boat build.

The Daly will still lie to you. But a misconfigured BMV-712 will lie more convincingly, which is somehow worse.

@GafferTapeKing that's exactly what I'm worried about with my static caravan setup — I've got a Daly 100A BMS on a 200Ah Fogstar Drift LiFePO4, and the BMV-712 syncs on a float voltage that I'm not convinced is actually triggering full capacity.

Does the BMV-712 sync when it sees the target voltage briefly, or does it need to hold it for a period? Because my Renogy 40A MPPT occasionally touches absorption voltage on a cloudy day without really finishing the charge cycle properly.

Wondering if that's causing false syncs and throwing my SOC readings off by a similar margin to what you experienced.

@GrumpyDrifter the static caravan scenario is worth being specific about: if your charger rarely gets the battery to a genuine full charge (absorption complete, tail current satisfied), your BMV-712 will drift high over time because it never gets a clean sync point.

On my narrowboat I had the same — solar rarely pushed to 100% in winter, and by February the BMV was reading 20% optimistic.

Fix is straightforward:

• Set the BMV's charged voltage to ~0.2V below your charger's absorption voltage
• Set tail current to around 4% of battery capacity (so ~8A for your 200Ah pack)
• Ensure your charger actually reaches that state periodically

The Daly will still be wrong on voltage-based SOC, but at least your BMV resets properly when conditions allow. Victron's own documentation covers these parameters clearly.