Take a look at this Pylontech…
The charging current was set low, at about 8A, because the low the current is, the better the battery will behave when it has problems and you try to solve them.
Look how strange the min and max voltages have behaved, when the limiting charging algorithm kicked in at 23:00 in order to keep the max cell voltage below 3.52V.
At that moment the min cell voltage dropped and stayed even below 3.36V for a long period of time, with a voltage delta of 190mV !!!
Look at the charging current, when voltage was limited. Is only 100mA, the balancing current and some for the BMS.
But slowly the low voltage cell recovered and ultimately balanced, after 5 hours !!
Ignore the spike at 23:20, as it was a test from my part to see how the low voltage cell will behave.
So patience is gold in this case… In the end, all good things come to those who wait… 
