Battery percentage (again)

[MoffKalast]

So a client has done some battery characterization and found out that the percentage values seem to not be quite as linear they should be. Instead the voltage seems to be linear, but the variance is super high.

The robot was supposedly left stationary with a more or less constant discharge current for some time, resulting in the following graphs. I would assume the empty spaces on the vertical axis would be the resolution of our ADC.

So theoretically, if the discharge is constant then so should the decrease in percentage, while the voltage wouldn't be quite linear but close to it.

This probably means the source values I used for that percentage PR don't quite align with our usual batteries I guess. I suppose we could use the voltage graph to make new values that would fit better, but it's hard to tell what the actual average voltage is with that massive vertical spread. Like apparently 100% charge varies from 4.15 V to 4.28 V per cell or something.

[mjstn]

We should NOT try to show 'percentage' for lead acid gel cells. Instead we (sadly) need to show things like 'good', 'moderate' or 'poor' charge state. We cannot do 'science' without monitor hardware that can also read current consumption at any given time. This is not 'LiPo' battery. The load greatly impact lead acid performance and we don't know the load at the time of the evaluation of the battery. We also don't know the differential battery voltage drop as a function of loading which is the ONLY way to judge percent or even close to that for lead acid. LiPo you CAN do that, Lead Acid you cannot.