Locating Self Discharging Cell(s) in Packs

OhmGrown

Member
Joined
Sep 13, 2018
Messages
55
So I created a few "standard" sized packs of 100 paralleled cells in order to power a project of mine recently. I created four of them about two months ago and they've essentially been sitting, not connected to any loads or in parallel or series to each other since they were completed (I needed to create a tray, plexiglass cover and ensure they could fit in my enclosure in the time being hence the storage period).

They were all close to full at that time, hovering around 4.1V or so. Just tested them again and three of them are still above 4.0V, however one, of course, is now down to 3.2V. Obviously I have one or more bad cells in the pack (really disappointed in myself that I didn't weed them out during the cell harvesting and testing phase but I likely just mixed a handful up during one of those late nights testing or it possibly went bad on me since then).

So, my question is, now that I've got all these packs welded up with fuses and tabs soldered to busbars, essentially all completed with not so cheap terminals and materials, can anyone think of a way to locate the self discharging cell(s) in the pack of 100?

The only things I can think of, without having to undo the entire pack, is to maybe point a thermal camera at it and see if maybe one cell is hotter than all the others? But I'm also pretty dumb so I'm not sure that the cell would register a higher temp, even if I happened to have a nice FLIR camera (which I don't).

I suppose I could also undo the fuses on the positive end one by one, checking the current running to each cell with a multimeter then (I do have a DC clamp meter but no way in hell I'll get that clamp just around one fuse leg). I'd also rather not make a brand new pack of 100 but, honestly, that may be the quickest solution. I'm hoping against hope there's a better solution that's been mentioned on this board that I just can't seem to find, so if anyone knows of it, I would greatly appreciate it. I shall send you a 12 pack in the mail.

And sorry for a new thread for such a specific, selfish question, I tried searching for a like minded existing thread or post but came up with squat. Thanks for the help and time guys!

TLDR: Any ideas on how to find one cell in a pack of 100 that's self discharging/parasitic without taking apart the pack?
 
Could you upload pictures of the + and - sides of the pack so we can see how you have it wired?
 
What I would do is charge the pack back up to 4.2V, then disconnect 1/2 the pack. Watch what happens. Then the half that went low again, split it half, or there abouts. Depends on how you have it wired up.
 
Bubba said:
Could you upload pictures of the + and - sides of the pack so we can see how you have it wired?

But, of course. Here's the positive side of the pack in question. Actually, tried to get the entire thing in frame here so it's a weird angle:


image_iuivdt.jpg


Then here's the negative side. This is actually two packs just resting on top of each other but they're not electrically connected in any way. So just the first four top rows are of the pack in question again:


image_dxybqo.jpg



Korishan said:
What I would do is charge the pack back up to 4.2V, then disconnect 1/2 the pack. Watch what happens. Then the half that went low again, split it half, or there abouts. Depends on how you have it wired up.

That's not a bad idea at all. So just kinda whittle it down in halves basically, trying to find the correct part with the drainer in it instead of taking the whole thing apart? And I could add cells back in one at a time or so if it the part I left connected doesn't self discharge again. Better than ripping the whole thing apart definitely.
 
Yeah, problem is the way the design is. Hard to cut it down unless you cut the copper bus bar in sections.
 
I think what Korishan meant:
1. Charge whole pack to 4.2V (voltage drop will be more noticeable at high V)
2. Let it rest for an hour or so for all cells to settle
3. Disconnect 1/2 of the fuses
4. Wait for say 24h
5. Measure voltage of disconnected cells. Voltage drop -> leaker. There could be more than one.
6. Measure voltage of the still connected cells. Voltage drop -> return to step 1.
7. Replace leaking cells (match voltage first) and reconnect fuses.
 
Korishan said:
Yeah, problem is the way the design is. Hard to cut it down unless you cut the copper bus bar in sections.

Yeahhhhhhh... I haven't exactly made these modular or easy to disassemble by any means. I realized that while making them but just kept forging on. With those cheap 4x5 cell holders, you can break off the tabs that hold each cell in to slide a cell in & out but, other than that, not a lotta options to swap things. I wish I just had the foresight to make a 5th back up pack of 100 that I could substitute in. Oh well. Looks like there's no other option but to start tearing fuses off.


ajw22 said:
I think what Korishan meant:
1. Charge whole pack to 4.2V (voltage drop will be more noticeable at high V)
2. Let it rest for an hour or so for all cells to settle
3. Disconnect 1/2 of the fuses
4. Wait for say 24h
5. Measure voltage of disconnected cells. Voltage drop -> leaker. There could be more than one.
6. Measure voltage of the still connected cells. Voltage drop -> return to step 1.
7. Replace leaking cells (match voltage first) and reconnect fuses.

Perfect, that's kinda what I thought he was referring to but you thought it out much better than I could.

I suppose I could also start with the cells that I have an intuition might be the culprit based on the stats I got during testing. I was smart/dumb enough to record everything in a spreadsheet and some cells are undoubtedly "healthier" than others. Though none show any visible signs or issues, I might get lucky isolating and checking those. But most likely I'll have to do half and half to ensure I get em all. Goddamn you recycled cells, you've bested me again!
 
I recently did this to my first packs. and What I did was Charge up to 4.2 then remove the Positive Side fuses and Bus bar.

Let it sit for 2 weeks and then Tested the voltage of each cell. I found 7 cells out of 560 that were Really Bad. (3v) and another 30 that were kinda bad and I replaced those too. (got down to 4v where the rest of them were 4.18)
 
jdeadman said:
I recently did this to my first packs. and What I did was Charge up to 4.2 then remove the Positive Side fuses and Bus bar.

Let it sit for 2 weeks and then Tested the voltage of each cell. I found 7 cells out of 560 that were Really Bad. (3v) and another 30 that were kinda bad and I replaced those too. (got down to 4v where the rest of them were 4.18)

I have a feeling thats exactly what's going on. Which is weird cause I had waited at least two months before even choosing my cells to make these packs, taking the ones with the smallest voltage drops usually. Previously had you noticed that your packs were draining significantly without a load attached or were they just draining rather quick when connected to a load?
 
Not trying to be a pain in the butt here but where these cells checked for IR before the pack was assembled?
If they wereand there still was such a failure then it would be good to know which cell(s) causes this issue once you find them and recheck the IR to see how different it is now from when the cells wereassembled.
Certainly looks like something odd has happened.

Best of luck finding the bad cells and hope you get the pack up and together quickly.

Wolf
 
[quote pid='53554' dateline='1566348939']



Korishan said:
What I would do is charge the pack back up to 4.2V, then disconnect 1/2 the pack. Watch what happens. Then the half that went low again, split it half, or there abouts. Depends on how you have it wired up.

That's not a bad idea at all. So just kinda whittle it down in halves basically, trying to find the correct part with the drainer in it instead of taking the whole thing apart? And I could add cells back in one at a time or so if it the part I left connected doesn't self discharge again. Better than ripping the whole thing apart definitely.
[/quote]
Sounds like a good idea based on the pictures. You could sever the bus at the far end and always reconnect it when your done.
 
They were going out of balance with very little use. 2 cells were significantly higher after about 6 months of use and the Balancer I was using was too low of a current balance (80mA) that it could not keep up. After the Swap and re-balance of capacities of those cells and now they stay within 10mV during charge and discharge even without the balancer connected.
 
I would say the cells with self discharge will now be heaters. Good chance if you charge it up you will be able to spot it with a cheap IR thermometer, or even just by touch.

Hopefully it doesn't pop a CID before you find it. A cell that can self discharge a pack of that size, is probably self discharging rather quickly.

Also you could try the 'educated guess' given Sanyo cells are famous heaters, you could just disconnect those first - which is about half the pack. Also those green and black topped Sony cells have given me a bit of strife too.
 
Another Thing I have been doing is to keep the capacity of all my packs within 1Ah of each other when I build them (repackr to the rescue) so I can have a couple spare packs incase there is an issue or the Pack times out.

Times out = a pack has a limited time in the system before being tested. For my starter packs I set it at 6 months but now that it is soo stable I will be increasing that to 1yr. then that pack comes out and temp one goes in it's place.
 
Sorry to say, but i agree with others.
Charge up the pack to 4.2v disconnect one side(i would cut thestrips at the neg side)
Let it sit for 2-3 weeks, then check on v and ir.
Usually the ones with a high ir or sd or any other problem, will show itself in v drop.

For the late night testing, we all bin there i guess, i am no exception.
To tackle this kind of errors, i adopted a other test for my packs.
First the regular normal cell testing.
Assemble the packs, solder only wire to connect the whole pack, charge it up to 4.2v.
Let it settle for 24-48 hours, desolder one side, put it away for 4 weeks.
Test with a yr3035 meter on v and ir.
Bad cells/mistakes are easy to be found this way(for me, for sure).
This way you make sure that all the cells are top notch and healthy, before you put them in your wall.

With so many cells around it is easy to make a mistake.
I also think, but a second opinion would be nice:
Some cells are in "storage/sleeping mode" for years, we wake them up.
So the first test rounds are good, then they had it and sink back to "not for work anymore".

Good luck with tracking down the bad ones, best.
 
Wolf said:
Not trying to be a pain in the butt here but where these cells checked for IR before the pack was assembled?
If they wereand there still was such a failure then it would be good to know which cell(s) causes this issue once you find them and recheck the IR to see how different it is now from when the cells wereassembled.
Certainly looks like something odd has happened.

Best of luck finding the bad cells and hope you get the pack up and together quickly.

Wolf

So, yes, I did record the IR for the cells though did not factor that in nearly as much as mAh & other stats when assembling the packs with Repackr. Also, halfway thru the testing phase I realized that the testers do not give a very accurate IR reading and only then switched over to a more manual method. So, admittedly, that is one area of info that I didn't very good on at all. I have a number for each cell but I'm not willing to bet much on its accuracy. I'll recheck em once I unassemble the cells and try to see if there's any corresponding factor to any of the bad ones. Thanks for the tip!


Yep I knew you guys would all have good info for me. I'm gonna take a closer look at the usual suspects (read Sanyo heaters) first but either way I think there's no way around disassembling this piece o' crap. I thought I was doing pretty good selecting cells, using Repackr, but I may have just gotten lucky with the other 3 packs. It is very disconcerting that this guy is draining so quickly and so far down though none of my fuses have popped so I think that could point to multiple cells self discharging instead of just one bad apple dropping REALLY fast?

In any case, I think I'll take the advice here and make a back up pack or two to swap in and out as I'm sure another will undoubtedly crap the bed after I fix this one. Ya know, I thought that would be the fun part for me, putting packs together and welding/soldering everything up while listening to a podcast and clearing my mind with just rote muscle memory but, so far, it has been the most annoying and monotonous portion of this whole process. Dah well, time to clear off all the hand tools on the work desk.
 
This morning i was checking a pack that i left alone for 2 weeks.
Tested cell per cell, and i found a sd, 0.78v, the rest was still above 4.1....oke 2 x 4.05
To bad i have to replace him, but at the other hand, what if i did not find this sd?

The ir and everything was good when i was assembling the pack.
 
100kwh-hunter said:
This morning i was checking a pack that i left alone for 2 weeks.
Tested cell per cell, and i found a sd, 0.78v, the rest was still above 4.1....oke 2 x 4.05
To bad i have to replace him, but at the other hand, what if i did not find this sd?

The ir and everything was good when i was assembling the pack.
100kwh-hunter,
How long did you let the cells sit beforetesting them again for IRand for SD?
I let mine sit for at least 30 days before the second IR and V test.
When I checked IR and V for the second timeI found somethat had risen in IR,somethat had below 4.12 V and some that were below 1V.

I am still 96 cells shy of my 2856 cell goal. As I am waiting the 30 days to retest the missing cells.So most of the cells that I will put into my powerwall will be checked again for IR and Va third time and a lot of them will have sat at least5 months. Better to find them now than when the whole pack is assembled.
Does that mean we will find every one ?Probably not but I for one am going to give them a good chance to show their true colors before I assemble them in a pack.

Wolf
 
I'm on the fence over IR. If I had a really good tester I might do it more often but I Look more at the Capacity measured vs Rated. But until we use brand new cells we always be in the test, check, Check, CHECK method.
 
jdeadman said:
I'm on the fence over IR. If I had a really good tester I might do it more often but I Look more at the Capacity measured vs Rated. But until we use brand new cells we always be in the test, check, Check, CHECK method.

Jdeadman,


Hm, Slightly off topic but still relevant to the OPs post.
I agree on the capacity VS rated criteria as it is a good indicator of SOH of the cell.
On the other hand if you have 500 cells you are about to test and just by measuring IR you can eliminate 300 of those as you know by there IR that their capacity VS rated will be insufficient then first of it has saved you a lot of time and you have eliminated some verypoor cells.
Second the cells that do pass the IR test will still have to be tested and some of those will not be great either but you have eliminated the majority of non performers. For a $50.00 investment in a very good tester that is not much of money to spend to speed things up considerably and have at least some peace of mind knowing that you have eliminated 95% of potential heaters and SDs.
Yes as we are using reclaimed cells there will always be a check ,check and check again but to have IR in your toolkit is as far as I am concerned an imperative measurement to have.
Wolf
 
Back
Top