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

Went through exactly this with my garden office setup last spring — maddening, isn't it?

The short answer I landed on: trust your BMS for protection, trust your SmartShunt for trends.

My Fogstar Drift 200Ah has its own BMS that watches individual cell voltages and will disconnect if anything goes sideways. That's non-negotiable safety kit. But the SOC percentage it reports is essentially calculated from voltage curves, which get wobbly in the middle ranges — you know, that annoyingly flat LiFePO4 plateau between roughly 20% and 80%.

The SmartShunt, meanwhile, is doing proper coulomb counting — it's watching every amp flowing in and out. Over time, if you've got the charged voltage, tail current, and Peukert settings dialled in correctly, it builds a much more accurate picture of what's actually in the battery. Mine took about a week of full cycles to really settle down.

Where it does drift is if you have small parasitic loads it's not monitoring, or if the shunt isn't positioned to see every single load. Worth double-checking your wiring.

For EV charging days when I'm hammering the system harder, I rely on the SmartShunt reading in VictronConnect almost entirely — the BMS SOC can sit at 40% while the shunt shows 28%, and the shunt has proven more accurate against actual capacity tests.

Has anyone found a smart way to reconcile the two automatically? I've been wondering whether Node-RED through a Cerbo could flag when they diverge by more than 10% as an early warning of something going wrong.

Reply by Gazza81:

@LindaLamb makes a solid point there. I'd add that the SmartShunt is only ever as good as its calibration — if it hasn't hit a proper full charge recently to resync, that SOC figure will drift noticeably over time.

Worth checking your charge parameters in the Victron app match your battery spec exactly. I had mine reading 8% low for weeks before I realised my absorption voltage was set slightly too conservative, so it never quite hit the sync point.

The BMS SOC (if your unit even displays one) is often just a rough voltage-based estimate anyway, so neither is gospel really. Between the two, I'd lean on the SmartShunt for day-to-day monitoring once it's properly dialled in. What BMS are you running with it?

Dead right on the calibration point @Gazza81 — and this catches people out constantly.

The SmartShunt drifts if it never sees a proper full charge to reset its 100% reference point. On my narrowboat I had a 12% discrepancy for weeks before I realised my absorption stage was cutting out too early. Fixed the charge profile and suddenly both readings were within 2-3% of each other.

Also worth checking:

• Peukert exponent set correctly for your battery chemistry
• Charged voltage threshold matches what your charger actually hits
• Tail current setting

LiFePO4 vs AGM settings are completely different and Victron's defaults aren't always right out the box.

Basically the two devices are measuring completely different things — BMS is counting cell voltages, SmartShunt is doing coulomb counting. Neither is "wrong", they just argue when your charge cycle isn't clean.

Has anyone actually mapped out how far apart the two readings can get over time without a full charge cycle to reset the shunt?

Mine sits in a static caravan so it doesn't always hit absorption properly — sometimes weeks go by with partial cycling and the SmartShunt just... wanders. Currently showing about 12% higher SOC than the Fogstar BMS app.

Is there a rule of thumb for how often you need a proper full charge to keep the SmartShunt honest? Or is the answer just "configure the tail current correctly in VictronConnect and stop worrying"?

Also — does the BMS SOC drift too over time, or is that more reliable long-term without resets? Genuinely not sure which number to panic at when things look low.

Great question @SmartSolarNerd — and welcome to the forum if you're fairly new here!

In my experience with the narrowboat setup, I've seen the SmartShunt drift 3–8% away from BMS figures over a week of partial cycling without hitting a full charge to trigger a sync. Longer periods of cloud in winter can stretch that further.

The key variable is your tail current threshold — if it's set too high in VictronConnect, the shunt thinks the battery is full before it actually is, and the drift compounds from there.

Worth logging both readings daily for a fortnight and noting when you last hit absorption/float properly. The pattern usually makes the culprit obvious pretty quickly.