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

Had this exact issue with my 100/30 last summer. Turns out the default absorption voltage was set slightly low for my Fogstar Drift LiFePO4 cells — the BMS was reporting full before the cells had properly balanced, so the MPPT saw the voltage tick up and bailed out into float prematurely.

Worth checking a few things:

• Absorption voltage — are you using the correct setting for your battery chemistry? LiFePO4 is not the same as AGM. Victron's defaults are often conservative.
• Absorption time — are you on fixed duration or adaptive? Adaptive can cut short on partially cloudy days because it's calculating off the tail current, which gets noisy with intermittent irradiance.
• Tail current threshold — in VictronConnect, check what percentage is set. If it's too high (say 4-5%), it'll exit absorption early even if the battery isn't genuinely full.

My setup is a 48V system running into a shepherds hut with a separate EV charging circuit, so getting proper charge completion matters — I can't have a half-full bank when the car needs topped up later in the day.

The fix for me was dropping the tail current threshold to 1% and setting a fixed absorption time of 2 hours. Stopped the premature float drops immediately.

Also — what does your battery comms look like? If you've got a Victron-compatible BMS talking over VE.Can or VE.Direct, it may be the BMS telling the MPPT to go to float, not the MPPT itself deciding. That's a completely different problem and the MPPT logs in VRM will tell you whether the transition was internally triggered or externally commanded.

What battery

@SolarNotSure — worth checking whether your 100/30 is actually hitting the absorption voltage or just timing out. Victron defaults to a fixed absorption time (or adaptive, depending on firmware), and if the tail current threshold is too high it'll bail to float prematurely even if the cells aren't genuinely full.

On my shepherd's hut setup I had to drop the tail current from the default down to around 0.02C before it stopped cutting short — the Fogstar Drift's BMS was masking what was actually happening at cell level.

Also double-check: are you using the VE.Direct temperature sensor? Without it, the controller's voltage compensation can skew things depending on ambient. Pull up the MPPT history in VictronConnect and check whether it's logging a proper absorption plateau or just a blip.

@SolarNotSure this is a classic one. On my 100/30 I found the absorption duration was the culprit rather than the voltage threshold — Victron's adaptive absorption algorithm calculates hold time based on how long bulk took, so on a partially-discharged battery on a bright morning it was cutting to float in under 20 minutes.

Worth checking in VictronConnect under Battery > Adaptive Absorption — if you're on LiFePO4 you arguably want this off and a fixed absorption time of around 1–2 hours set manually. My Fogstar Drift settled perfectly once I set:

• Absorption: 14.2V
• Float: 13.5V
• Fixed absorption time: 1hr

Also confirm your battery sense wiring is clean — a poor connection here throws the algorithm completely.

Had this exact same saga with my boat setup last spring. Spent three days convinced the 100/30 was dying, turned out I'd fat-fingered the absorption time in VictronConnect to something embarrassingly short. Classic "it's always the user" situation.

@BordersExplorer is right about absorption duration being the usual villain. Worth also checking your **

@BordersExplorer and @SomersetVanLifer have covered the main suspects, so I'll add what nobody's mentioned yet — tail current.

On LiFePO4 specifically, the absorption phase should end when current drops to a set percentage of battery capacity (typically 1-2%), not purely on time. In VictronConnect, check your "End of absorption" setting under the LiFePO4 preset. If it's not configured properly, the controller will exit absorption on the timer alone regardless of actual state of charge.

Ran into this myself last winter in the Cairngorms — my Fogstar Drift cells were consistently sitting at 96% according to the BMS whilst the MPPT cheerfully declared float. Sorted it by tightening the tail current threshold down to 1%.

Also worth enabling the Battery Life algorithm off if you've got it on — it artificially limits absorption on smaller arrays and causes exactly this symptom.

@LiFePO4Nerd raises the tail current point which is the one most people overlook. On my setup I've got the tail current set to 2% of battery capacity — any higher and the Victron calls "full" when there's still meaningful capacity headroom, particularly at lower ambient temperatures where LiFePO4 acceptance naturally slows.

Worth noting: if you're using VE.Direct or Bluetooth to monitor, check the actual absorption time logged versus what you've configured. On my 100/50 I found the logged absorption duration was consistently shorter than the set value, which confirmed the tail current threshold was being hit prematurely rather than a voltage misconfiguration.

Also ensure your battery temperature compensation is appropriate — LiFePO4 generally wants zero temperature coefficient, and if the controller's applying a lead-acid curve it'll skew the target voltage downward in cooler conditions.