Anyone else finding MPPT controllers lie about battery state of charge?

I've got a Victron SmartSolar 100/30 feeding a 200Ah lithium (LiFePO4) bank and I'm increasingly convinced the SOC readout is basically fiction. Yesterday it was showing 87% all afternoon even though the panels had been in cloud for hours and I'd been running a 12V compressor fridge plus a laptop off it. Pulled out my Kaiweets clamp meter and the resting voltage was sitting at 13.1V — more like 70% if you go by the discharge curve.

From what I can gather, MPPT controllers estimate SOC based on voltage and charge/discharge history rather than doing proper coulomb counting. Without a dedicated battery monitor like a Victron BMV-712 or a Smartshunt in the circuit, the controller is just guessing. Bit annoying when you've spent £££ on "smart" kit and it still can't tell you accurately how full your batteries are.

Has anyone else seen this? And for those running lithium banks specifically — do you think a Smartshunt is worth the £90-odd, or is there a decent cheaper alternative that plays nicely with Victron's app? Seen the Renogy one mentioned a few times but no idea if it integrates with VictronConnect.

@Neil1962 mate, MPPT controllers don't actually measure SOC — they're guessing based on voltage, which on LiFePO4 is famously flat as a Norfolk road for 80% of its range, so yes, your 87% is essentially creative writing. Get a proper Victron SmartShunt doing coulomb counting and suddenly your numbers will actually mean something. Mine lived in blissful ignorance for two years before I sorted it — van thought it was fine, I thought it was fine, battery was absolutely not fine.

@Neil1962 this is a fundamental misunderstanding of what MPPT controllers are designed to do — they're charge source managers, not battery monitors. The SmartSolar has no idea what's going into or out of your loads; it can only see its own output.

What you actually need is a dedicated coulomb counter on the battery negative — a Victron BMV-712 or SmartShunt is the obvious choice and integrates natively into VictorConnect/VRM. It tracks every amp-in and amp-out continuously.

On my narrowboat I had exactly this confusion before fitting a SmartShunt. The MPPT voltage-based estimate is particularly useless on LiFePO4 precisely because the discharge curve is so flat between roughly 20–90% — voltage tells you almost nothing meaningful in that range.

The BMV/SmartShunt also lets you dial in your specific battery's Peukert exponent and charge efficiency factor, which tightens SOC accuracy considerably over time.

@Neil1962 yeah this caught me out with my shepherd's hut setup. LiFePO4 voltage curve is basically flat across 20–80%, so voltage-based SOC is pretty much useless in that range.

What actually helped me was adding a BMV-712 — proper coulomb counting rather than voltage guessing. Night and day difference. The SmartSolar talks to it over VE.Direct and suddenly everything made sense.

Your MPPT is doing its job fine, it just was never designed to track SOC accurately on lithium. Think of it as a charge controller, not a battery monitor — those are two separate jobs really.

@Neil1962 this is the thing that caught me out badly when I first switched to lithium. Spent a whole winter trusting my SmartSolar's SOC figure, then got caught short during a three-day storm with far less capacity than I thought.

The flat discharge curve on LiFePO4 is the villain here — voltage barely moves between 20% and 90%, so voltage-based estimation is essentially meaningless across that middle range.

What actually sorted it for me was adding a Victron SmartShunt. It does proper coulomb counting — measuring actual current in and out — rather than guessing from voltage. Night and day difference in accuracy.

The SmartSolar is brilliant at what it does, but SOC tracking genuinely isn't it. Treat that figure as decorative until you've got a dedicated battery monitor in the circuit.

@Neil1962 the fix for this is a dedicated battery monitor — I've got a Victron SmartShunt sitting between my battery bank and everything else in my garden office setup, and it transformed how accurately I can track state of charge. It uses coulomb counting (measuring actual current in and out) rather than trying to infer SOC from voltage, which as @WillWebb says is almost useless with LiFePO4.

The SmartShunt talks to the SmartSolar over VE.Direct so everything shows correctly in the Victron Connect app. Worth every penny. Renogy do a cheaper alternative if budget's tight, but I'd personally stick with Victron given you're already in that ecosystem — everything plays nicely together.

@Neil1962 the Victron SmartSolar is doing voltage-based SOC estimation, which is basically useless for LiFePO4 given how flat the discharge curve is. What you actually want is a dedicated battery monitor with coulomb counting — a BMV-712 or similar will track actual amp-hours in and out rather than guessing from voltage.

In my tiny house setup I've got the BMV feeding into VictronConnect alongside the SmartSolar and the difference in reported SOC between the two can be laughable sometimes. The MPPT figure I now completely ignore.

Worth noting — even coulomb counting drifts over time without a proper sync point, so make sure your BMS is set up to trigger a full charge periodically so the monitor can reset to 100% accurately.

@Neil1962 to add to what @Thommo is getting at — LiFePO4 has that notoriously flat discharge curve, sitting around 3.2-3.3V per cell across most of its usable capacity. Your MPPT simply cannot distinguish between 90% and 40% from voltage alone. The controller is designed primarily to manage charging, not accurately track SOC. It's essentially guessing. A coulomb-counting monitor like @HamishMitchell suggests is the proper solution — it tracks actual ampere-hours in and out rather than inferring from voltage. Made a massive difference to my confidence in the system once I fitted one.