There are multiple device Voltages "negotiating" to produce a "compromise" result on the circuit.
When the bank is truly 100% Full as per Lifeline specs, say
14.4V held until 0.005C endAmps
that is the **charging** voltage.
Isolate the bank completely, and 24+hrs later you will see a much lower **resting** voltage.
Given a newish healthy battery
, some of that might still be a "surface charge", so drawing say half an Ah will be a better reference point "X".
That bank is still at 100% SoC
So, an input voltage **any higher** than that X will be enough to Float the bank, keeping it at 100%.
As long as the sources' current
is sufficient to cover any loads fed, no current
is drawn from the bank, stays at 100% SoC and the circuit voltage is maintained at that "negotiated" point Y.
But, when current drawn is **greater** than sources' input, then current must come from the bank, and both SoC and Voltages start to drop.
Depending on the load current rate, then circuit voltage Y will drop much faster than bank voltage X, which can be seen by isolating the bank, and watching bank V "bounce back".
R, try dropping your bank SoC% down lower than say 70%, and see how that changes the numbers you are reporting.