Victron MPPT keeps dropping to float too early — anyone else had this?

Had this exact issue with my 100/30 last summer on the static caravan. Battery would show around 80% on the Cerbo GX but the MPPT would already be sitting in float by mid-morning. Felt completely wrong.

Turned out my battery capacity setting in VictronConnect was way off — I'd never properly configured it after swapping to a 200Ah Fogstar Drift lithium. Once I corrected that and adjusted the absorption voltage to match Fogstar's recommended spec (3.45V per cell), it started behaving much more sensibly.

A few things worth checking if you're in the same boat:

• Battery capacity set correctly in VictronConnect?
• Absorption time — are you using fixed or adaptive? Adaptive can be very aggressive on smaller banks
• Tail current setting — if this is too high, it'll cut absorption short and drop to float prematurely
• Are you using a temperature sensor? Without one, the MPPT compensates blindly and can make odd decisions in cooler weather

Also worth checking whether your BMS is sending any signals via VE.Direct or VE.Can that might be overriding the charge stage. Some lithium BMS setups will tell the MPPT to back off even when the battery isn't truly full.

What battery chemistry are you running and have you looked at the bulk/absorption transition in the history tab? That usually tells you a lot about what's actually happening. Would be curious whether others with lead-acid setups are seeing the same or if this seems more of a lithium-related quirk.

KentOffGrid | 847 posts | ⚡ Solar Enthusiast

@SueJohnson This is almost certainly a battery settings issue rather than a fault. The MPPT transitions to float based on voltage and tail current thresholds, not SOC percentage — so if your absorption voltage or tail current is set too low, it'll cut short early every time.

Worth checking a few things in VictronConnect: what absorption voltage are you running, and what's your tail current set to? For most leisure batteries people run it too conservative. Also, is the Cerbo actually getting its SOC from a BMV shunt, or just estimating? If it's estimating, that 80% figure could be quite inaccurate.

What battery chemistry are you running — AGM, gel, or lithium? Makes a big difference to where these thresholds should sit.

SIE_Electric | 312 posts | 🔌 Van Dweller

@SueJohnson Your MPPT's basically that overenthusiastic colleague who shouts "job done!" before actually checking the work — classic absorption voltage set too low means it hits float faster than I hit a pothole on the A303. Worth

JA_Solar | 1,203 posts | ☀️ Solar & Shepherds Hut

Had almost the exact same head-scratcher on my shepherd's hut setup last spring. Turned out my absorption voltage was set a touch low — bumped it from 14.2V to 14.4V and suddenly it was spending a proper amount of time bulk/absorbing before floating off.

Worth checking your battery's actual charge profile too. If you're running Fogstar or similar lithium cells, the manufacturer specs sometimes differ quite a bit from Victron's defaults. Connect up via VictronConnect and double-check:

• Absorption voltage
• Float voltage
• Absorption time (or whether you've got adaptive absorption enabled)

Also — what battery chemistry are you running? That'll change everything. AGM and LiFePO4 behave totally differently and the controller won't always guess correctly depending on how it's been configured.

EcoFlow_Nerd | 234 posts | 🔋 Battery Nerd (Obvs)

@SueJohnson quick one — what battery chemistry have you got it set to? My static caravan setup did something almost identical until I realised the MPPT was still configured for flooded lead-acid when I'd swapped to lithium. The absorption voltage was so low it was basically waving the white flag before the job was done 😅

Also worth checking your tail current setting in VictronConnect — if it's set too high, the controller thinks "yeah, that'll do" and boots it into float prematurely. I dropped mine right down and it made a massive difference on my Fogstar cells.

What does your absorption voltage read when it transitions? That number might tell us a lot about where it's giving up!