diy solar

diy solar

bad cell in a 8s2p battery pack

FDM

New Member
Joined
Nov 19, 2021
Messages
6
I have built an 8s2p 24v pack with SAB 60280 cells. I built two packs one works perfectly charges and discharges great and all cells stay together within .010v while charging and discharging.

The second pack will become unbalanced both charging and discharging by over 300mv differences. while discharging the cells. The values cell #4 will drop to 2.6v. While cell #1 3.258, Cell#2 3.207 cell# 3 3.113 while the cell#'s 5, 3.282 cell6, 3.282 cell#7 3.283 , cell# 8 3.276. So cell #4 lowest with Cell#7 highest. Cells number 1, 5,6,7,8 are all within .020 thats fine. Cells number 2 and 3, (2) 3.207 and (3) 3.113 are about .080 difference. So cells 3 and 4 are the worse with #4- 2.6 and #3 - 3.113. Does this mean cells 4 and 3 are the bad cells? I have discharged and re-charged the pack with same results.

When I built the pack , I charged all cells to 3.5v then put them in parelle for a week and let them sit. Then placed them in their final series parallel combination and fully charged them. I'm only getting 87ah from the 110ah pack and all cells are new.

How do I find the bad cell to replace? what type of test to determine the bad cells causing the imbalance? SHould I break the pack apart into two 8series packs to find which row has the bad cells? I'm using a BMS from overkill BMS for all my packs and the BMS are all the same in the program structure so I'm confident it's not a BMS issue. Need help on how to ID the bad cell to replace.
When charging cell #4 is always behind in the charging if that helps.
 
Last edited:
I have built an 8s2p 24v pack with SAB 60280 cells. I built two packs one works perfectly charges and discharges great and all cells stay together within .010v while charging and discharging.

8S2P implies that you have two 8S batteries wired in parallel at the main terminals. It sounds like you have a 2P8S pack, i.e., 8S pack comprised of 2P "Cells"

The second pack will become unbalanced both charging and discharging by over 300mv differences. while discharging the cells. The values cell #4 will drop to 2.6v. While cell #1 3.258, Cell#2 3.207 cell# 3 3.113 while the cell#'s 5, 3.282 cell6, 3.282 cell#7 3.283 , cell# 8 3.276. So cell #4 lowest with Cell#7 highest. Cells number 1, 5,6,7,8 are all within .020 thats fine. Cells number 2 and 3, (2) 3.207 and (3) 3.113 are about .080 difference. So cells 3 and 4 are the worse with #4- 2.6 and #3 - 3.113. Does this mean cells 4 and 3 are the bad cells? I have discharged and re-charged the pack with same results.

Assuming "cells" are actually 2 cells in parallel, you've likely established that one or more "cells" are at least lower capacity than the other "cells," i.e., one of the cells forming "cells" 3 and 4 are lower capacity than "cells" 1-2, 5-8.

When I built the pack , I charged all cells to 3.5v then put them in parelle for a week and let them sit.

Unfortunately, this did absolutely nothing beneficial. It actually worked against you as it allowed strong cells to support weak cells. You would have been better served to let the cells sit individually.

Then placed them in their final series parallel combination and fully charged them. I'm only getting 87ah from the 110ah pack and all cells are new.



How do I find the bad cell to replace? what type of test to determine the bad cells causing the imbalance? Should I break the pack apart into two 8series packs to find which row has the bad cells?

Yes. This allows you to see how the individual cells behave vs. being masked by its "cellmate".
 
  • Like
Reactions: FDM
FIRST before getting all drastic. Double check all of your connections, busbar to cells, BMS Harness, including testing the suspect lead to cell to ensure that lead is good from ring terminal to connector. If you used busbars, also make sure there are no Burrs or Ridges around the holes, they really do have to be perfect, you would be amazed at how many troubles that causes.

Next also remember that oils and waxes can cause havoc as well so ensure everything is perfectly clean (wipe down with Isopropyl Alcohol (rubbing alcofool). You can add a Very Tiny coating of NoAlox/OxGuard and I mean "tiny" just a wipe on thin coat, NO GOBS and never ever on any threads, because it acts like grease and will mess up torque readings.

It is generally ill-advised to parallel cells within packs unless they are really genuinely Matched, Batched & Binned, a simple Voltage & IR test at static does not cut it.

Separate packs also add Fault tolerance & Fail-Over should one fail or have issues, as well they share & split Load & Charge which is less stressful on them in both directions and allows you more leeway for charging.

Hope it helps, Good Luck
 
Thank you, for your replies gentlemen, I guess I used the 8s2p backwards as pointed out :) . Yes t have a 24v pack with 8 cells 55amp each in series for a total of 24v. The second battery is made the same way as the first. Then I parallel them into a single 24v pack for 110amps. consisting of a total of 16 cells. So I should refer to them as 2p8s thank you for letting me know this as well. I did it this way so I could use one BMS per 2p8s cell packs as I have a total of 3 sets of 2p8s and it's the one pack thats driving me nuts. I kid you not the other packs charge up and discharge down within .020 of each other and most often as close as .005v they behave nicely. If I split them all into individual 24v batteries then I'd need 6 BMS one for each 8 cell battery. I was trying to keep the cost down was thinking a 2p8s battery with one BMS for a total of 3 bms for my total setup.

Ok, so the connections are all good the plates were made on a milling machine at my buddies machine shop. They were all sanded (actually to a smooth finish) from aluminum 1/4 plates and the bms leads are all crimped and affixed to the plates through tapped threaded holes in the plates. I'll buy some of the NoAlox/OxGuard to treat plates for good measure.

I took that pack apart today.
So should I charge each cell individually to 3.65v then do a capacity test on each cell one at a time to find the bad cell?
or do I use the information from the BMS voltage readings ? Is it likely the low voltage cell is the bad ones?
You can see 5 of the 8 cells are very close in value with 3 cells with greater differences specifically cells 3 and 4. Or are cells 3 and 4 being drained by another cell causing then to deplete faster?

If I have to replace a cell or two I want to get the right ones out of the pack of course and thats where I'm not sure how to ID the exact bad cell or cells. I plan to use this as an off grid supply for lights and a few small items in a small cabin with a 4000w 8000w surge inverter and six 385w solar panels when I get the batteries right. I may add another 16 cells for a total of 600 amps of batteries. It want be for daily use as I'll visit the cabin every so often.


thank you
Frank
 
Last edited:
Thank you, for your replies gentlemen, I guess I used the 8s2p backwards as pointed out :) . Yes t have a 24v pack with 8 cells 55amp each in series for a total of 24v. The second battery is made the same way as the first. Then I parallel them into a single 24v pack for 110amps. consisting of a total of 16 cells. So I should refer to them as 2p8s thank you for letting me know this as well. I did it this way so I could use one BMS per 2p8s cell packs as I have a total of 3 sets of 2p8s and it's the one pack thats driving me nuts. I kid you not the other packs charge up and discharge down within .020 of each other and most often as close as .005v they behave nicely. If I split them all into individual 24v batteries then I'd need 6 BMS one for each 8 cell battery. I was trying to keep the cost down was thinking a 2p8s battery with one BMS for a total of 3 bms for my total setup.

Ok, so the connections are all good the plates were made on a milling machine at my buddies machine shop. They were all sanded (actually to a smooth finish) from aluminum 1/4 plates and the bms leads are all crimped and affixed to the plates through tapped threaded holes in the plates. I'll buy some of the NoAlox/OxGuard to treat plates for good measure.

I took that pack apart today.
So should I charge each cell individually to 3.65v then do a capacity test on each cell one at a time to find the bad cell?

Only if you can't charge them to 3.55+ as a battery.

or do I use the information from the BMS voltage readings ? Is it likely the low voltage cell is the bad ones?

Likely. Once you identify the outliers, you can test them individually if you choose.

You can see 5 of the 8 cells are very close in value with 3 cells with greater differences specifically cells 3 and 4. Or are cells 3 and 4 being drained by another cell causing then to deplete faster?

Likely.
 
I'm charging Cell#3 that was 3.113v been 22hrs power supply set to 3.62v amps set to 10 but cell only drawing at first 7amps dropping to 3.5amp over the total charging time of 22hrs,,
Cell voltage is currently at 3.335v while charging . Voltage likely would be lower if I took it off the power supply. I'm trying to get it to at least 3.55v.
Should 3.55 be the resting voltage?
Do I put the cell on and off the power supply until 3.55 becomes the rested voltage of the cell even off the power supply?
The batteries always read less after a rest period. Is that normal ?
Should it hold steady at 3.55?, with no decrease in voltage a few hours later?

I'll do cells #4 2.8v after this Cell#3 is tested.
 
Last edited:
I'm charging Cell#3 that was 3.113v been 22hrs power supply set to 3.62v amps set to 10 but cell only drawing at first 7amps dropping to 3.5amp over the total charging time of 22hrs,,

3.5A * 22 = 77Ah input. If these are 55Ah cells, and your numbers are accurate, this is a bad sign for this cell.

Cell voltage is currently at 3.335v while charging.

If a 55Ah cell is at 3.335V while charging after taking 3.5A for 22hr, it's bad. Stop charging.

Voltage likely would be lower if I took it off the power supply.

This is normal.

I'm trying to get it to at least 3.55v.

If you've input 77Ah per your numbers, stop. A completely empty 55Ah cell will be full once you put in 55Ah of charge.

Should 3.55 be the resting voltage?

Depends on the cells. The big 280Ah cells don't settle much, but smaller ones can settle very rapidly. If you charged to 3.65, most cells will rapidly drop into the 3.5X range.

Do I put the cell on and off the power supply until 3.55 becomes the rested voltage of the cell even off the power supply?

No.

The batteries always read less after a rest period. Is that normal ?

Very.

Should it hold steady at 3.55?, with no decrease in voltage a few hours later?

No.
 
Thanks Sunshine Eggo ... I was thinking the same for the charge and number of hours the current and the voltage weren't adding up. This cell should have been charged withing 8-10 hours from 3.113v to max Voltage 3.62v set on the supply.... well it looks like I've found the first bad cell so now onto the others to be sure there are no more. Cells 5,6,7,8 are all holding there original voltage and I believe they are fine. All my low voltage cells were in groups 2,3,4 with 3 and 4 the lowest by far. thats why I started with cell 3it being next to number 4 maybe dragging #4 down while #5 was normal with the other cell voltages of 5,7,and 8 even cell 1 wasn't bad 2 a little worse 3 even more so then 4 was in the dumps. 3 being in the center of the two made me start to look there. I'll try charging #4 cell 2.8v to 3.62v if it gets there I'll do a cap check on it and same for cell #2. Hoping its just #3 thats bad.
 
Strange.... I took battery number three that I had a hard time charging and decided to discharge it to 2.5v and then tried again to charge it . This time it charged up to 3.62v and I did a full voltage discharge again to 2.5v and obtained a 60ah discharge from a 55aH battery. This would indicate the battery is good? Correct?
I also did the companion Number 3 cell (the parallel cell) that also took a full charge and discharge reading of 57ah exceeding the rating by 2ah's.
Cell #4 I can not get it to charge all the way up to 3.55v in fact 3.46 but as soon as I disconnect it from the charger it drops back to the low 3.30v but so far it will not charge to 3.55v.
I have been charging each cell to 3.630v and doing a capacity discharge check to obtain the discharge rating reading and then recharging them back to 3.63v So far all exceot Battery #4 has been able to be fully recharged. c

So question is , if a battery does recharge to 3.63v is this evidence of a good cell? Do i need to do a discharge capacity check?
 
Strange.... I took battery number three that I had a hard time charging and decided to discharge it to 2.5v and then tried again to charge it . This time it charged up to 3.62v and I did a full voltage discharge again to 2.5v and obtained a 60ah discharge from a 55aH battery. This would indicate the battery is good? Correct?
I also did the companion Number 3 cell (the parallel cell) that also took a full charge and discharge reading of 57ah exceeding the rating by 2ah's.
Cell #4 I can not get it to charge all the way up to 3.55v in fact 3.46 but as soon as I disconnect it from the charger it drops back to the low 3.30v but so far it will not charge to 3.55v.
I have been charging each cell to 3.630v and doing a capacity discharge check to obtain the discharge rating reading and then recharging them back to 3.63v So far all exceot Battery #4 has been able to be fully recharged. c

So question is , if a battery does recharge to 3.63v is this evidence of a good cell? Do i need to do a discharge capacity check?

I'm suspecting faulty charging hardware or you've mixed up the cells. That's very odd behavior.

What are you using to charge and discharge?
 
Back
Top