Matching 70 cells in 10S7P

dopravopat

New member
Joined
Feb 25, 2019
Messages
4
Hello!

I would like to build a 10S7P battery pack for my e-bike. I harvested many cells from old notebook batteries. I picked identical Samsung ICR 18650-28A cells for the battery, I do not want to mix different brands and types.

I measured each cells capacity,charged themto 4,2Vand let sit forabout 8 weeks.Today Imeasured the voltage. The results are in the attached file asmAh;voltage.

My question is how to best match these cells to a 10S7P configuration.

Should Ibuild each 7P module with batteries that have capacities as close as possible and as a result have ten modules with bigger difference in overall capacity, or mix cells so that all modules have similar total capacities and each module will consist of mixed capacity cells? Matching voltage is more of a theoretical question, as the differences are very low, basically between 4,17V and 4,19V, with only two cells at 4,16V (as you might have noted, I posted the results for 73 cells, I need only 70 in my battery, so 3 cells are extra).

How would you match the cells for the 7P modules and why?

The finished battery is intended for a load of 7A constantly (which is 1A per cell ~ 0,4C) and 20A peak (2,8A~1C).

EDIT: I already have a BMS with ballacing fucntion, designed for 10A constant load and 20A peak load.
 

Attachments

  • Samsung ICR 16850-28A.txt
    801 bytes · Views: 211
Same capacity in each pack since you have same brand. There is not much more you need to care about. The self discharge you have there is the error marginal of the meter you got
 
daromer said:
Same capacity in each pack since you have same brand. There is not much more you need to care about. The self discharge you have there is the error marginal of the meter you got

Thanks. I also have 20 used and tested Samsung ICR 18650-26C cells, which all have a capcacity over 2400 mAh.

Icould, theoretically, build two modules out of these, to get higher overall capacity of the battery. Basically replace the most worn 18650-28A cells which are 2317 to 2387 mAh.

But I am not sure about mixing the cells in the battery. Yes, they are same brand and similar model, but I have my doubts. I did not find the discharge curves for the 26C model, but since each cells has been already used, the real discharge curves might be more different between cells of the same type than between individual28A and 26C cells. The word is "might". I am not sure and do not have the experience yet. Maybe someone else has experience with this. I certainly do not want to mix cells in a module, but I can imagine building two out of ten modules with similar cells.
 
daromer said:
Same capacity in each pack since you have same brand. There is not much more you need to care about. The self discharge you have there is the error marginal of the meter you got
Sorry to ask, but I am not sure I understand correctly.

Do you mean that seven cells in one pack should be as close as possible in capacity, or the ten packs should have a capacity as close as possible to each other?
 
The 10 packs should be as close as possible. Ideally identical if possible
 
OK, so I plan to build the modules like in the attached txt file.

The differences are pretty small. Each module will have about 17Ah maximal capacity in theory.I am counting the battery as 15Ah, as I do not want to place high currents on cells which still hold some charge after the weaker ones in the paralell group stop delivering.

So the final battery will be 36V, 15Ah, 540Wh. Theoretically it should be capable of running for 2 hours with a continous load of 7,5A (270W @ 36V nominal voltage) if my calculations are correct.
 

Attachments

  • Samsung ICR 16850-28A modules.txt
    638 bytes · Views: 209
Back
Top