LiMn2O4 48V pack low performance, strange, need idea

[urosv]

Hi guys...
I'm getting weird low performance from LiMn 48V pack. Pack was charged lets say 20 times. Approx after 21th and 22st charge the battery performs half or previous performance or even worse (5-7km). I disassembled the battery pack and without measurement I first changed the BMS, tried again and the problem was still there (2 discharge and charge). Then I disassembled the pack again and measure internal resistance of 3 packs by applying load (since I did not want to break apart cells wich are punked together not solded). The measurements of each pack (3 cells) is between 4.89V and 4.95V
The measurement of combine packs of 3 to get (12V) are this

delta T is ~ 10 secs
pack 1,2,3 (12.24V dropped to 11.95V  at 2.17A)
pack 2,3,4 (12.23V dropped to 11.94V at 2.2A)
pack 3,4,5 (12.23V dropped to 11.94V at 2.21A)
pack 4,5,6 (12.23V dropped to 11.94V at 2.25A)
pack 5,6,7 (12.25V dropped to 11.95V at 2.28A)
pack 6,7,8 (12.22V dropped to 11.93V at 2.27A)
pack 7,8,9 (12.23V dropped to 11.94V at 2.31A)
pack 8,9,10 (12.23V dropped to 11.93V at 2.33A)
pack 9,10,11 (12.22V dropped to 11.93V at 2.36A)

Increase of current is due temp of power LED which was used for load.

AS you can see there are not so significant difference to indicate that something is wrong. Does anyone have any idea how to test more or to get to the bottom of the problem.

[Lollandster]

If you have extremely cold temperatures you might be experiencing the same as me, that the capacity drops significantly in cold weather.

[urosv]

Ok, lets assume that the temperature is the case. If you look at the chemical point of view the problem is then irreversible and the battery will stay in this condition forever ? Explain your situation, what happened and what the indicators/meters are saying ?

[Lollandster]

http://goldenmotor.com/SMF/index.php?topic=4706

[urosv]

Ok, I see, but I don't think this is the case ? Why...
We did not get below 0 when this happened. I have at least 5 batteries, LiMN and LFP and this is the only one that has behaved this way. But thanks for the tip and information. Waiting for some other tip if someone can find a solution or answer.

[MAGICPIE3FOCUSPOWER]

Well I think you should charge every cell separate and measure how much capacity it will have.