Had this exact issue with my MultiPlus-II last spring when I was commissioning the shepherd's hut setup. Batteries would sit at maybe 80% and the unit would cheerfully declare them full and drop to float. Maddening.
Turned out my battery voltage sense wires weren't connected properly — the MultiPlus was reading voltage at the unit rather than at the battery terminals, so it was seeing a slightly inflated figure due to cable resistance. Once I ran dedicated sense wires directly to the Fogstar Drift 200Ah cells, the absorption phase started behaving itself properly and actually finishing before dropping down.
Worth checking a few things:
- Voltage sense wiring — is it reading at the battery or the inverter terminals?
- Absorption time settings in VictronConnect — the adaptive algorithm can be quite aggressive about shortening absorption if it thinks previous cycles were shallow
- Battery temperature — if you're running a temp sensor, oddly high readings can cause it to back off early
Also had a forum member suggest checking whether "BatteryLife" mode was enabled somewhere in ESS settings — apparently that can interfere in unexpected ways if you're not running a full ESS configuration.
Mine's been solid since sorting the sense wires. Still gets me how such a small wiring oversight caused so much head-scratching.
Anyone else found other causes? Curious whether this is more common on the 48V units — mine's a 24V system and I wonder if the voltage thresholds leave less margin for error.