How should MAX_VOLTAGE_TIME_SEC work?

[cflenker]

I often observe the following behaviour:
When reaching MAX_CELL_VOLTAGE and the current is dropping towards zero the pack of high voltage cells come down and cross the pack of low voltage cells come up and end up below their voltage level.
At the short point of time when they are crossing the battery pack seems to be balanced an the timer of MAX_VOLTAGE_TIME_SEC starts.
So after MAX_VOLTAGE_TIME_SEC the control_voltage is lowered to FLOAT-Voltage.
In reality the battery is not completely balanced and MAX_VOLTAGE is stopped to early:

I would prefer that voltagediff has to be lower than CELL_VOLTAGE_DIFF_KEEP_MAX_VOLTAGE_UNTIL for the whole time of MAX_VOLTAGE_TIME_SEC. If the voltagediff gets bigger than CELL_VOLTAGE_DIFF_KEEP_MAX_VOLTAGE_UNTIL the MAX_VOLTAGE_TIME_SEC should restart.

What Do you think about that?

I think the reason for the crossing is, that the cells with the bigger resistance come down as the current drops. When the current is close to zero you can see the real inbalance (but MAX_VOLTAGE_TIME_SEC already has started).

[mr-manuel]

A lot of batteries would never change to float in this way.

[cflenker]

Hello @mr-manuel,
I agree that there is a high risk that some batteries would never change to float this way.
So maybe it is a possibility to offer a third threshold for celldiff to reset the MAX_VOLTAGE_TIME_SEC-counter if celldiff is above the threshold. Maybe you can have a look at #895.
Thank you!