Identifying a single bad 18650 in series

[thekubiaks]

I am building a 12s10p pack from 18650's salvaged from new laptop packs. The pack is constructed with simple pairs of ebay acquired spacers. I am charge testing every cell with an Opus BT-C3400 (excellent btw) and writing the capacity on the case. So far I have completed 3S10p of the build and am still testing cells.

My question is: Once I identify a series of cells in the completed 12s10p pack that is not close to the same voltage as the other series and know one of the 10 cells in that series is bad, how can I identify which cell it is?? All cells in the series are spot welded to each other so if I put a multimeter on each cell, wouldn't the other cells in the chain effect the reading??

Thanks.

ps. I searched the forum extensively for the answer, if it was there, I missed it...

 

[mlt34]

I think you might have series and parallel confused.

You said you're building a 3s10p pack, but that's only 10.8 volts.

Then you ask how to check a single cell in series since they are all spot welded together. Parallel cells are spot welded together so they share the same positive and negative terminals, not series.

Assuming you meant parallel, the answer is you can't. If they are wired in parallel (meaning they share the same positive and negative terminal) then all the cells will equalize their voltages. So if one is a weak cell, the others will prop it up but it will slowly drag the whole set of parallel cells down with it.

Assuming you meant series, then you can just measure the voltage at the positive and negative terminals of a set of 10s1p cells to determine the bad cell. However, if you have any cells wired in parallel, then as stated above, you can't determine which one is bad without unparalleling them.

I hope that made sense.

 

[amberwolf]

My question is: Once I identify a series of cells in the completed 12s10p pack that is not close to the same voltage as the other series and know one of the 10 cells in that series is bad, how can I identify which cell it is?? All cells in the series are spot welded to each other so if I put a multimeter on each cell, wouldn't the other cells in the chain effect the reading??

Not if they are in series.

If in series, every cell is only connected to one other cell at it's end, at most.

If in parallel, however, then yes, all cells will show the same reading.

At that point you have to divide the parallel group up to figure it out. If you look up my old Vpower / CammyCC pack repair thread, it shows a way to do that--but I don't recommend removing the tabs from cells like I did there. INstead just cut them between cells. If you have a dremel and cutting discs, that is probaby safer than a metal blade or knife or scissors.

 

[thekubiaks]

Thanks for the replies. I see where I goofed when I said All cells in the series are spot welded to each other, I meant to say All cells in parallel are.... When I said I have a 3S10p built, the way I said it was a little confusing I guess too. I am building 10 cells in parallel at a time (4.2V) and then connecting the 10 cell packs in series to eventually get to 12 rows of 10 cells, 48V 20Ah.

Thanks for the tip on Vpower/Cammy repair thread, good info there. I was afraid I'd have to cut out individual bad cells that are wired in parallel to find the bad one. I almost went with a no solder/weld design to make removing cells easier but I made a spot welder that is working nicely.

 

[arkmundi]

All cells in the series are spot welded to each other so if I put a multimeter on each cell, wouldn't the other cells in the chain effect the reading??

A multimeter will read across its poles. You can read one or more cells. No, adjacent cells do effect the reading.

 

[parabellum]

Do not know what is weak cell in your case, but if it is resistance related, discharge with reasonably high current and monitor temperature at negative side of the serie in question. Hotest cell is highest IR.

 

[thekubiaks]

Do not know what is weak cell in your case, but if it is resistance related, discharge with reasonably high current and monitor temperature at negative side of the serie in question. Hotest cell is highest IR.

I like that approach, I have an IR camera, and that should do it. Thanks again.

 

[dnmun]

this is the fundamental problem with the way everybody builds the batteries. i have tried and tried to get people to recognize it is impossible to isolate a bad cell if the battery is built up by connecting all the cans in parallel first.

the only way to solve the problem of repairing the pack later is to build it up as a serial pack first, then tie all of the serial packs together in parallel through a paralleling connecting wire soldered to the middle of the small copper links soldered between the edge of the bottom of the case and anode used to build the pack in series initially.

 

[tomjasz]

this is the fundamental problem with the way everybody builds the batteries. i have tried and tried to get people to recognize it is impossible to isolate a bad cell if the battery is built up by connecting all the cans in parallel first.

the only way to solve the problem of repairing the pack later is to build it up as a serial pack first, then tie all of the serial packs together in parallel through a paralleling connecting wire soldered to the middle of the small copper links soldered between the edge of the bottom of the case and anode used to build the pack in series initially.

Ducking and covering before asking....

Is there anyway you can with draw a picture or show a photo? That is as clear as mud, this early in my education. I just now figured out S and P. Building Serial first just doesn't register.

I'll ask my dumb C questions later.

 

[dnmun]

all the HK lipo packs are connected in series. they are all 1P. i did have some pictures of my battery builds but lost it when i had to reformat my disk during a crisis.