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

Had this exact issue with my 100/20 last summer. Turned out the absorption time was set too short in the custom settings — Victron's default adaptive algorithm can be a bit aggressive in deciding the battery is "full" if your panels are oversized relative to the bank.

Worth checking a few things:

• Tail current threshold — if this is set too high, it'll drop to float prematurely. I knocked mine down to 2% of battery capacity and it sorted it immediately
• Absorption voltage — are you using AGM or lithium presets? The factory profiles aren't always ideal for third-party cells (I run Fogstar Drift LiFePO4 and had to dial in custom parameters)
• Temperature compensation — if you've got a battery temp sensor connected, a warm battery will lower the absorption target, which can look like early float

The other thing to check is whether you're actually seeing the drop in VictronConnect or if you're reading it off a BMV/SmartShunt. Sometimes there's a lag in reported state that's just a display artefact rather than the controller genuinely switching modes.

What does your absorption tail current actually look like when it drops? If it's still pulling 10-15A when it switches, that's definitely a config issue. If it's genuinely tail-ending at low current, the algorithm might be working correctly and your bank is simply smaller than you think relative to the charge input.

Post your battery capacity, panel wattage, and current MPPT settings — someone on here will almost certainly have seen the same combination before.

@BarryWood interesting one — did you also check your battery temperature compensation settings? I had something similar on my motorhome build with a 100/30 and Fogstar Drift 100Ah lithium. The BMS was occasionally throwing a signal that confused the MPPT into thinking absorption was complete when it wasn't.

Worth checking:

• Is your battery sense cable connected properly (or using the built-in VE.Smart networking)?
• What battery type are you set to — user-defined or one of the presets?
• Are you monitoring via VictronConnect? The history graphs will show exactly when and why it's dropping to float

Also, what's your tail current set to? If it's too high relative to your bank size, the algorithm will cut absorption short thinking the battery is already full.

@BarryWood worth checking your battery voltage sense wiring too — if the MPPT is only measuring at the controller terminals rather than at the battery itself, voltage drop across the cables fools it into thinking the battery is fuller than it is. Ended up with exactly this on my shepherd's hut setup. Ran a separate sense wire directly to the battery terminals and it sorted the premature float completely. Victron's own documentation buries this but it makes a meaningful difference, especially if your cable run is more than a metre or two.

@BarryWood had this exact problem on my narrowboat setup running a 100/30 into 200Ah of Fogstar Drift LiFePO4.

Worth checking your absorption voltage threshold — if it's set even slightly low, the adaptive algorithm interprets the battery as "full" prematurely because current tapers off quickly at the wrong voltage point.

Also: what battery type is selected in VictronConnect? If you're on LiFePO4 but the controller's set to AGM or gel, the voltage curves are completely wrong and you'll see this behaviour constantly.

The tail current setting is the one most people miss — it's the percentage of rated capacity at which the controller decides absorption is complete. Default is often 1A or 2%, which is far too aggressive for lithium. Drop into the expert settings in VictronConnect and set it properly for your specific bank.

@SaltyHiker's point on voltage sense wiring is solid too — compound problem if both are off.

Had this with my garden office setup last year — drove me mad for weeks.

One thing nobody's mentioned yet: check your battery profile is actually set to LiFePO4 if you're running lithium. Victron defaults can sometimes get reset after a firmware update and silently switch back to a flooded profile. Absorption voltage ends up way too high so the algorithm thinks it's done earlier than it should.

Also worth connecting via VictronConnect and watching the absorption current in real time. If it's dropping below the tail current threshold too quickly, bump that threshold down slightly.

@LH_Marine's point about Fogstar Drift is relevant too — those cells are very efficient at accepting charge so the current tail-off genuinely can be fast. Might not actually be a fault, just the battery behaving well.