It really looks like we reach an asymptote at about 3.5V. I believe it is the batteries that can take no more charge that limits the process. But we also saw one datum that says that the solar panels are at less than half power from what they were at the beginning of operations. Again, although we got the satellite to switch to outreach at least twice it got back to basic beacon because of a reset in eclipse.
I analyzed the 18 passes of AS-1 to see the dependency of the battery voltage from the time of the satellite in sunlight. Due to bandwidth of the values I assume it’s really a question of the orientation of the satellite.
The maximum voltage dropped under 3.6 volts (more than 4.5 volts at the beginning).