I wouldn’t expect the user manual to get into the details of the charger’s absorption/float algorithm. Users generally would only care that the three stages happen at the appropriate times, not how the appropriate times are computed.
Another excerpt from Maine’s site - “In order to have an ideal recharge the charger really needs to know what is a house load and what is going to the battery. Seeing as battery chargers only know output current, or more accurately the percentage of its power supply being used, as well as voltage, they can’t have any idea what is flowing into the battery and what is flowing to house loads.”
So he’s saying that the charger does have a shunt to measure its output current. The limitation is that the charger doesn’t know how much of that output current is going to the batteries and how much is going to the house loads. Smart(er) chargers can make assumptions about those amounts to guess at the percentage going to the battery. From Maine’s description the Sterling does an acceptable job with those assumptions, but all chargers play it a little safe to not stay in absorption too long.
I have a Xantrex charger, generally regarded as smart but not as smart as the Sterling. It shows the percentage of current being output, and it definitely uses that as part of the algorithm for the absorption duration. When I go out for a long sail and my batteries are down lower the charger stays in absorption for several hours. When I plug in after only a short motor it falls to float after only a few minutes. Different depths of discharge fall in between those extremes. The process may not be as optimal as an ideal charger, but it’s better than an egg timer.
