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

Had this exact headache last year when I commissioned my garden office setup. Short answer: trust neither blindly, but lean on the SmartShunt once it's properly calibrated.

Here's why the discrepancy happens:

• BMS SOC is typically voltage-based or uses a basic coulomb counter with factory assumptions about your specific cells. It rarely accounts for temperature, ageing, or your actual charge/discharge patterns.
• SmartShunt uses a proper calibrated coulomb counter, but it's only as good as its synchronisation events — it needs to reach 100% (tail current threshold met) regularly to re-anchor itself.

The key settings people overlook on the SmartShunt:

If your SmartShunt isn't seeing a proper full charge cycle frequently enough, it'll drift. My Fogstar Drift 200Ah LiFePO4 and the SmartShunt disagreed by nearly 12% until I sorted the tail current threshold.

Also worth checking — are you monitoring the shunt on the correct side of the battery? Loads and charge sources all need to go through it, otherwise your readings are nonsense from day one.

What BMS are you running? Some (JK, Daly, JKBMS) have notoriously optimistic SOC reporting. Would help narrow this down further for everyone here.

@EcoFlow_Queen makes a solid point about calibration, but here's the thing that tripped me up on my narrowboat build —

The SmartShunt is only as good as its shunt placement. I had mine upstream of the BMS initially, meaning parasitic loads from the Victron kit itself were skewing readings by 4-5%. Moved it downstream, recalibrated with a known full charge, and suddenly both devices were agreeing within 2%.

Also worth checking your charged voltage threshold and tail current settings in VictronConnect. Mine shipped with defaults clearly set for a lead-acid world — completely wrong for the Fogstar Drift 100Ah cells I'm running.

The BMS SOC is essentially a rough estimate based on cell voltages. Useful for catching drift, but the SmartShunt doing proper coulomb counting will always win for accuracy once dialled in properly.

@EcoFlow_Queen @LochWalker — one thing neither of you have mentioned yet: the SmartShunt's accuracy lives and dies by your shunt wiring. Every single current path must pass through it — I made the schoolboy error of running my Renogy DC-DC charger direct to the busbar, bypassing the shunt entirely. Victron Connect was reporting wildly optimistic SOC because it couldn't see that charge current at all.

Also worth setting your Peukert exponent correctly for your specific chemistry. LiFePO4 wants something close to 1.05, whereas the default assumes lead-acid behaviour. Most BMS units don't model Peukert at all, which is partly why they drift.

My cabin setup now uses the SmartShunt as primary reference, with the BMS purely as a protection layer. Once the SmartShunt hits a genuine 100% absorption cycle, the calibration holds remarkably well for weeks.

@EcoFlow_Queen @LochWalker @TransitAdventure — my Fogstar Drift cells spent three weeks arguing with the SmartShunt before I just let the BMS handle low-voltage cutoff and the SmartShunt handle everything else, bit like having a bouncer who checks IDs but lets the landlord run the till.

Key thing nobody's said: sync events are your friend — configure the SmartShunt to reset to 100% on a proper float tail current, otherwise it'll drift further apart from your BMS every cycle like two clocks that've never been set to the same time.

Mine eventually settled within 2% of each other on my garden office setup after I sorted the tail current threshold in VictronConnect. Took longer than it should've, admittedly.

@RelayBuild three weeks is nothing — my static van's Fogstar pack and SmartShunt were having a full-blown diplomatic incident for nearly two months until I finally nailed the charged voltage threshold and tail current settings together.

The bit everyone ignores: your BMS SOC is essentially counting from an assumed starting point after each protection event, whereas the SmartShunt is doing proper coulomb counting from a known full-charge reference. Neither is gospel, but at least the SmartShunt's errors are measurable and correctable via VictronConnect.

Set your tail current to roughly 1-2% of your battery's Ah capacity and your charged voltage maybe 50mV below the BMS absorption endpoint — stops false "full" triggers resetting your SOC mid-cycle like an idiot.