2 cells lagging behind

[jay_jay]

Hey there,

I noticed a strange behavior in one of my 8s battery bank. When the bank reaches the upper knee, two cells (#5 & #6, red and yellow in the attached graph) shoot up faster than the others. When the system reaches absorbtion voltage (15min at 27.6v) they fall behind the others. The delta goes up to 100mv.
I redid top balance, checked the connections, left absorption voltage up to 2 hours to give the cells time to balance (which resulted in even higher delta). I also tried helping #5 & #6 with a bench power supply when absorbing. Nothing improved the situation. However, once the charging stops the cells stay within 10mv or less during the whole discharge cycle.
Batteries came from docan.
What do you think? Is it just that the two cells have less capacity or is it pending doom?

 

[sunshine_eggo]

Cell voltage issues only arise in the upper and lower legs of the voltage curve UNLESS there is a significant resistance issue in either the cells themselves or the sense leads.

Did you properly torque each connection, typically 4-6Nm?

10mV is normal in the operating range even with notably imbalanced cells.

Have you confirmed the voltages you see with a voltmeter?

Worth noting that you can lose a top balance without regular use and charging into the balancing range. Your BMS should be set to balance at or above 3.40V with a 20mV difference. Normally, during charge only, but leaving it on without charging is fine when trying to recover a top balance.

 

[jay_jay]

Yes, i redid all the connections and checked the torque. Also, the voltages reported by the BMS are on point, triple checked with voltmeter.
The batteries had plenty time to balance the last weeks. Roughly every second day the batteries are full and absorb. That led to battery bank #2 staying within <20mv delta.

 

[jay_jay]

Situation is resolved. If anyone experiences similar issues, redo the top balance. It seems like the faulty BMS introduced the issues. It restarted up to 300 times per day. When I installed the replacement BMS I thought the battery was still top balanced. After top balancing again the two misbehaving cells are back in order.

 

[Ucatchmydrift]

Situation is resolved. If anyone experiences similar issues, redo the top balance. It seems like the faulty BMS introduced the issues. It restarted up to 300 times per day. When I installed the replacement BMS I thought the battery was still top balanced. After top balancing again the two misbehaving cells are back in order.

Hi, for how long did your bms stay switched off for when switching on and off? Just half a second, or for longer periods of time? As ive got one cell each on two banks unbalanced. The software add on for the bmses in home assistant has a software button to power cycle them (not actually sure what that means though) and ive had this switched on, meaning they could have been switching on and off many times a day? Maybe causing my issue, just abguess though

 

[jay_jay]

Hi, for how long did your bms stay switched off for when switching on and off? Just half a second, or for longer periods of time? As ive got one cell each on two banks unbalanced. The software add on for the bmses in home assistant has a software button to power cycle them (not actually sure what that means though) and ive had this switched on, meaning they could have been switching on and off many times a day? Maybe causing my issue, just abguess though

Hi there,

It seemed like it was constantly rebooting. So a reboot loop. Do you have similar issues with yours?

 

[Ucatchmydrift]

Hi there,

It seemed like it was constantly rebooting. So a reboot loop. Do you have similar issues with yours?

Hi, i think mine is a different situation, its not in a boot loop, just think the HA software add on could have been rebooting it, or just rebooting the bluetooth many times a day, not sure yet. But i have the software turned off now, so not doing it anymore, was just contemplating things when i read your post, as one of the cells in each bank has became unbalanced.

 

[jay_jay]

I see. I am using home assistant as well, my jks are read from an esp using sysi's code. As I was also suspecting the ha integration being at fault I removed the switches to turn the bms's charging on and off. Didn't help. I also suspected the Bluetooth connection could lead to reboots but my BMS had the same symptoms with no Bluetooth connection active.

Edit: after replacing the BMS everything works fine with esp and sysi's ha integration. Very stable (as long as the wifi reception is good for the esp)

 

[Ucatchmydrift]

I see. I am using home assistant as well, my jks are read from an esp using sysi's code. As I was also suspecting the ha integration being at fault I removed the switches to turn the bms's charging on and off. Didn't help. I also suspected the Bluetooth connection could lead to reboots but my BMS had the same symptoms with no Bluetooth connection active.

Edit: after replacing the BMS everything works fine with esp and sysi's ha integration. Very stable (as long as the wifi reception is good for the esp)

Im very confident that both my jk bmses are not faulty, i did look at the esp route using sysi's code, but used the batmon add on instead as it looked more simple. (i may change this now though! - i have an esp32 at home to try) It works fine without having the 'power cycle' setting enabled for one battery bank, but when i add another, it will only connect to both with the power cycle setting enabled, and then it seems to have to power cycle both bmses many times a day as they then have connection issues. Both bmses never stay connected for more than about 30 seconds. I just wondered if this was causing the inbalance in two cells. See this thread i started yesterday:

2 runaway cells on 2 banks

Hi, i have two 24v lifepo4 banks in parallel (via busbars). They were top balanced when new to 3.65v. I have had them running for 6 months now. Coincidentally? the first cell on both banks seems to run away when all other cells are at about 3.33v (according to the jk bmses). I have...

diysolarforum.com