Victron SmartShunt showing different SOC to my BMS — which one do I trust?

Had exactly this on my narrowboat last winter — drove me mad for weeks.

Short answer: trust neither blindly, but lean on the SmartShunt once it's properly calibrated.

Here's what I found. My BMS (a Daly 100A unit) was reporting SOC based on a fairly crude voltage curve, which is notoriously unreliable with LiFePO4 because the discharge curve is so flat. The SmartShunt uses coulomb counting — it's actually measuring current in and out — so it's fundamentally more accurate if you've set it up correctly.

The critical bits people miss:

• Charged voltage threshold — set this to match when your cells actually hit full, not just a rough guess
• Tail current — I had mine too high, so it was declaring "full" before the battery truly was
• Peukert exponent — leave it at 1.05 for lithium, don't fiddle

Once I sorted those three settings in VictronConnect, my SmartShunt and Fogstar 200Ah battery were singing from the same hymn sheet.

The BMS SOC reading is really just there for cell-level protection, not accurate state tracking. Think of it as a bouncer rather than an accountant.

Anyone else running a split setup like this? Curious whether those with more sophisticated BMS units like a JK or Daly Smart are seeing tighter agreement — or whether the SmartShunt still wins out.

@PeteJames the calibration point is key — most BMS units are just doing voltage-based SOC estimation under the hood, which is notoriously inaccurate especially mid-range on LiFePO4's famously flat discharge curve.

The SmartShunt does proper coulomb counting, so once you've nailed the charged voltage and tail current settings it'll be far more reliable day-to-day.

One thing I'd add: make sure your shunt is the only path between your negative busbar and battery negative. Had a rogue earth somewhere on my setup that was bleeding current outside the shunt — both readings were wrong but for completely different reasons. Took me ages to find.

Also worth syncing them manually after a genuine full charge to at least get both starting from the same baseline.

Really interesting thread — this exact issue nearly had me pulling my hair out when I first set up my static caravan system.

One thing worth adding: the SmartShunt's accuracy improves dramatically once you nail the battery capacity setting and the Peukert exponent for your specific chemistry. I'm running Fogstar Drift LiFePO4 cells and spent a good fortnight logging data before the two readings converged properly.

Also worth checking your tail current threshold — if it's set too high, the SmartShunt never registers a genuine 100% synchronisation point, which causes drift over time.

@PeteJames your calibration point is spot on. @LindaJones69 is right about voltage-based estimation being unreliable under load — especially in colder months when internal resistance shifts noticeably.

If anyone's new to this topic, don't be put off by the complexity — it genuinely clicks once you understand what each parameter is actually measuring.

@PeteJames been through this on the narrowboat too — my Victron SmartShunt and Fogstar battery BMS were arguing like an old married couple until I nailed down the charged voltage and tail current settings properly.

The bit nobody mentions: your SmartShunt is only as good as its last full synchronisation. If you're doing lots of partial cycles (very narrowboat), it drifts. Set a sensible sync SOC threshold and let it hit 100% properly occasionally — even if that means running the engine a bit longer than your fuel wallet would like.

Once dialled in, SmartShunt wins every time. The BMS reading is basically decorative.

@PeteJames worth adding that synchronisation events are the real key to keeping the SmartShunt honest long-term. Mine drifts noticeably if I don't hit a proper full charge regularly — the shunt only resets its 100% reference point when it sees the tail current drop below your configured threshold at absorption voltage.

On my shepherd's hut setup (200Ah Fogstar Drift LiFePO4), I set the charged voltage to 0.1V below actual absorption and tail current to around 2% capacity. Took a few charge cycles to settle but the SmartShunt and BMS now read within 2-3% consistently.

Also check your Peukert exponent if you're seeing divergence under heavy EV charging loads — default 1.25 is conservative for lithium and causes the shunt to underestimate remaining capacity during high-current discharge.

@PaulCross makes a good point about sync events — and this is where most people fall down with their setup.

The SmartShunt needs a proper tail current threshold configured, not just the default values. I spent an embarrassing amount of time wondering why mine kept declaring full SOC prematurely in my garden office system until I realised Victron's defaults assume a different battery chemistry than my Fogstar lithiums.

Key settings to audit:

• Charged voltage — set ~0.1V below your actual absorption target
• Tail current — typically 1-2% of battery capacity for LiFePO4
• Charged detection time — I run 3 minutes minimum

The BMS is essentially just a protection circuit doing coulomb counting as a side-effect. The SmartShunt is purpose-built for the job. Once your synchronisation events are firing correctly, the two figures should converge to within a few percent.