weber wrote: Does anyone know what the conditions are that cause a PIP-4048MS to drop out of float mode...
I suspect that it is a fixed voltage lower than the float voltage parameter. In fact, 4.0 V less. There would be a time required to be at this level, say 10 seconds or so, to avoid temporary load surges from "prematurely" taking it out of float mode.
[ Edit: I now suspect that 52.02 firmware has a fixed return to boost voltage, where it needs to see less than 48.0 V for about 5 seconds. ]
This is in line with the "lead acid thinking" that seems to have shaped the design of this inverter. Unfortunately, this is a bad fit for LiFePO4 batteries. (It may well be fine for other chemistries, such as the NMC (?) used in Nissan Leaf cells.)
For LiFePO4, that means that if the float voltage is set to say 53.8 V (3.36 VPC), then the battery would have to go down to 49.8 V or 3.11 VPC. This is around 5% SOC, and the battery should almost never go that low. So the answer might be to go back to sending commands to change the float voltage, but instead of doing this every few seconds (and this constantly resetting the solar charger's MPPT algorithm), it would only be sent about once a day, when the battery falls to say 52.8 V (3.30 VPC).