Need to wake this puppy up

[BigGun]

A123 cells are not your normal cells, + and - are reversed, the whole body of the cell is the +. Also if you're talking about the packs they have for sale, they're a bitch to get the nickel off and sometimes it makes holes in the cells when they come off. The ones I've fooled with, I just clipped the nickel as close as possible, then spot welded to the old nickel.

Good luck rebuilding a pack of that size with all the cells folded together . The possibility of shorting cells out is very high. It's time consuming just to rebuild a Ryobi 40v pack built in this manner.

I'm kind of gathering that you don't even have the understanding how these cells are put together in series to even test them. Going to be hard to rebuild a pack without that knowledge. Then there's the #1 Issue, does the BMS even function any more, Would be a shame to rebuild the battery to find out the unit isn't functional.

Usually when I rebuild ( lets say a 40v pack ) I charge the old cells up together monitoring closely. The ones that lag behind I charge separately. When I get the voltages high enough, if the pack / bms function , I rebuild it. Others I toss

 

[wpns]

I'm kind of gathering that you don't even have the understanding how these cells are put together in series to even test them. Going to be hard to rebuild a pack without that knowledge.

This. It’s not really a beginner’s project.

 

[STxFarmer]

This. It’s not really a beginner’s project.

Yes I realized that early on and this is why I am asking lots of questions before I do anything. I am still looking for someone in my area that has experience with building batteries as I am guessing there is but no luck yet in finding them yet. This is a brand new $3K UPS so I am assuming the BMS works as it was never taken out of the box before I got it. These have been sitting in a warehouse for several years and the recharge date on the box was Oct -2022 so I am guessing the BMS drained the battery pack over time.

Taking the battery pack over to an electronics repair shop that has done some work for me on other projects to see if I can get him to remove the solder on the top board so I can remove it and get to the 3 rows of cells. He is better setup for the solder removal than I am.

So,

Here is a pic (IMG_4508) of the top of the battery pack. I was assuming all groups were 6 cells but on the top row it starts and ends with 3 (little confused about this now). So would the guess be that the 2nd row all be groups of 6 or could there be a group of 3 on each end? The bottom row is all groups of 6 cells. Also in the 2nd pic it makes sense that B0 & B1 are the connections for the 1st group of 3. Then B2 & B3, B4 & B5 and B6 & B7 are the 3 groups of 6 cells. Then I have marked all common board connections with the same color circle in the 2nd pic (IMG_4512) B0 is the Negative connection for the BMS. B8 has a bussbar connection down to the bottom of the 2nd row of cells (blue boxes). B24 is the Positive connection for the UPS battery pack Anderson connector.

Starting with the top of the battery pack, the first location will always be my negative probe:
B0 - B1 0.727
B2 - B3 0.333
B4 - B5 0.275
B6 - B7 0.750

Bottom row of cells
B17 - B18 0.300
B19 - B20 1.328
B21 - B22 0.031
B23 - B24 .0634

So which pairs of connectors are the middle group of cells?

 

[MrSparkle]

So which pairs of connectors are the middle group of cells?

I would say, (if your measurements are correct), half of the pack's cell groups are waaay below recommended minimum voltage. Safe to say they have been in that extreme low state for a long period of time. The chance the other half are at similar level is practically 100%. If so, then cells have suffered severe internal and permanent degradation and should be considered as absolute past EndOfLife and even a potential hazard trying to charge them up.

Regardless if LFP (min 2.5V) or NMC (min 3V?), your cells are faaar below those values.

 

[BigGun]

Like this

B0-B1 B1-B2 B2-B3 and so forth . Do you have a adjustable power supply. If you do , set the voltage slightly higher than the pack voltage it's at now with a charge rate of about .25 amp. Check all the cells while it's charging. If all are well below 3.5v, raise the PS voltage around 3v. Then continue to monitor all cells. They'll eventually start to raise. This will take a long time, but it's the safest way to go about this.

 

[wpns]

 

[MrSparkle]

I'm all out of. Can you share some of yours?
On a serious note though. I hope OP can filter "safe practice" advice from "advanced experiment - danger zone".

 

[wpns]

I'm all out of. Can you share some of yours?
On a serious note though. I hope OP can filter "safe practice" advice from "advanced experiment - danger zone".

Horse, water.
He’s not listening, he can’t follow instructions, he can’t parse 24S3P, he can’t solder, much less weld nickel strips, and did I mention he’s not listening?

 

[MrSparkle]

Horse, water.
He’s not listening, he can’t follow instructions, he can’t parse 24S3P, he can’t solder, much less weld nickel strips, and did I mention he’s not listening?

Sorry, what did you say? I wasn't paying attention...
Anyhow, thx m8! I do have a spare for you, as a thanks in return.

 

[fnnwizard]

Yes I realized that early on and this is why I am asking lots of questions before I do anything. I am still looking for someone in my area that has experience with building batteries as I am guessing there is but no luck yet in finding them yet. This is a brand new $3K UPS so I am assuming the BMS works as it was never taken out of the box before I got it. These have been sitting in a warehouse for several years and the recharge date on the box was Oct -2022 so I am guessing the BMS drained the battery pack over time.

Taking the battery pack over to an electronics repair shop that has done some work for me on other projects to see if I can get him to remove the solder on the top board so I can remove it and get to the 3 rows of cells. He is better setup for the solder removal than I am.

So,

Here is a pic (IMG_4508) of the top of the battery pack. I was assuming all groups were 6 cells but on the top row it starts and ends with 3 (little confused about this now). So would the guess be that the 2nd row all be groups of 6 or could there be a group of 3 on each end? The bottom row is all groups of 6 cells. Also in the 2nd pic it makes sense that B0 & B1 are the connections for the 1st group of 3. Then B2 & B3, B4 & B5 and B6 & B7 are the 3 groups of 6 cells. Then I have marked all common board connections with the same color circle in the 2nd pic (IMG_4512) B0 is the Negative connection for the BMS. B8 has a bussbar connection down to the bottom of the 2nd row of cells (blue boxes). B24 is the Positive connection for the UPS battery pack Anderson connector.

Starting with the top of the battery pack, the first location will always be my negative probe:
B0 - B1 0.727
B2 - B3 0.333
B4 - B5 0.275
B6 - B7 0.750

Bottom row of cells
B17 - B18 0.300
B19 - B20 1.328
B21 - B22 0.031
B23 - B24 .0634

So which pairs of connectors are the middle group of cells?

It is groups of 3. That is the 3P of the “24S3P”
It is 3P first, then 24 of those in series.

Ok, this side of the pack shows the 3P in an L shape config… which will make your build a tad harder, but offers slightly better current distribution between individual cells. That would make my previous posts detailing cells 17-24 incorrect. The other side would connect all 6 like the far right side cells 17+, but the groups are in a 3L shape instead of flat 3.

As others have stated. Your meter readings need to be for all 24 of each set of 3P cells…
B0 —> B1
B1 —> B2
B2 —> B3 …


Did they all self discharge back below 1V?

 

[STxFarmer]

It is groups of 3. That is the 3P of the “24S3P”
It is 3P first, then 24 of those in series.

Ok, this side of the pack shows the 3P in an L shape config… which will make your build a tad harder, but offers slightly better current distribution between individual cells. That would make my previous posts detailing cells 17-24 incorrect. The other side would connect all 6 like the far right side cells 17+, but the groups are in a 3L shape instead of flat 3.

As others have stated. Your meter readings need to be for all 24 of each set of 3P cells…
B0 —> B1
B1 —> B2
B2 —> B3 …


Did they all self discharge back below 1V?

B16-B17 was 2.904 and all others were below 1v.

Have the board off the top now. So what is the best guess on how to separate the 3 rows of cells? They seem to be glued together.

 

[fnnwizard]

B16-B17 was 2.904 and all others were below 1v.

Have the board off the top now. So what is the best guess on how to separate the 3 rows of cells? They seem to be glued together.

I rotated the pic so the image maintains consistency. With cell zero upper left.

The pack is built with submodules 8s3p.
Green line denotes connection points between modules. Sever that and see if modules pull off to the sides. Cant tell if the plastic holders are underneath the nickel strips or not. If not the cells should pull off easily.
If so, might have to do surgery to remove.

There's a piece of what looks like pcb (fr4) isolating between each modules. You want that intact.

 

[STxFarmer]

I rotated the pic so the image maintains consistency. With cell zero upper left.

The pack is built with submodules 8s3p.
Green line denotes connection points between modules. Sever that and see if modules pull off to the sides. Cant tell if the plastic holders are underneath the nickel strips or not. If not the cells should pull off easily.
If so, might have to do surgery to remove.

There's a piece of what looks like pcb (fr4) isolating between each modules. You want that intact.

It looks like the nickel strip might be glued to the pcb isolating board (both sides) as I can see a strip that seems to be spot welded to the battery. I can separate it a little as shown in the pic. I was guessing/hoping the the pcb isolation board was just a divider but it seems to be a bit more. They were glued and have attached pics of the top row with B1 being the top left corner. 4517 shows the top of the row of cells with B0 being the top left. 4518 shows the bottom with B1 being the top left

 

[BigGun]

The good thing is, the cells are all dead. Other wise prying it all apart you'd likely see Fireworks. Looks like some thick nickel also.

 

[STxFarmer]

The good thing is, the cells are all dead. Other wise prying it all apart you'd likely see Fireworks. Looks like some thick nickel also.

I could see where the gap between groups was so I was able to start with a sharp thin blade to start each group. Then I could get a small putty knife in to finish each group.

 

[BigGun]

There you go, my point exactly. You have no idea what you're doing, if those cells had energy in them, you would of shorted them out, maybe even hurt yourself " BAD " You do realize , you're not fooling with AA batteries. In between those rows you can direct short series cells.

 

[STxFarmer]

There you go, my point exactly. You have no idea what you're doing, if those cells had energy in them, you would of shorted them out, maybe even hurt yourself " BAD " You do realize , you're not fooling with AA batteries. In between those rows you can direct short series cells.

If they had energy in them I wouldn’t be doing this. If I’m such an idiot then why waste ur time?

 

[fnnwizard]

I could see where the gap between groups was so I was able to start with a sharp thin blade to start each group. Then I could get a small putty knife in to finish each group.

Sorry to rely so late. I had written a reply at the bottom of this post, but I guess never hit the post reply.
You've done a good job so far and I am confident you will learn quite a bit from this so it's a positve in my book.

Sometimes going onto a forum to ask for advice can be daunting when you get more responses that are negatives vs condusive and members not looking or understanding the bigger picture. We should all remember that we all start somewhere.

Anyway, see if you can find prebuilt 8S3P packs like what you have. This is the easiest way, worst case = you can solder nickel tabs to form the BMS and inter module connections.

I really don't see much danger with how you have proceeded. Kudos on a decently clean dissection.


"I would use a plastic scaper to see if you can cut into the glue. Maybe an old credit card/ hotel key sanded sharp along an edge.
Extra precaution can be had by balck taping the tabs that stick up.
The pcb helps to prevent shorting between columns. Which can be if ou are using a metal tool and it touches a set of cells along with the tabs over an across the pcb."

 

[STxFarmer]

Sorry to rely so late. I had written a reply at the bottom of this post, but I guess never hit the post reply.
You've done a good job so far and I am confident you will learn quite a bit from this so it's a positve in my book.

Sometimes going onto a forum to ask for advice can be daunting when you get more responses that are negatives vs condusive and members not looking or understanding the bigger picture. We should all remember that we all start somewhere.

Anyway, see if you can find prebuilt 8S3P packs like what you have. This is the easiest way, worst case = you can solder nickel tabs to form the BMS and inter module connections.

I really don't see much danger with how you have proceeded. Kudos on a decently clean dissection.


"I would use a plastic scaper to see if you can cut into the glue. Maybe an old credit card/ hotel key sanded sharp along an edge.
Extra precaution can be had by balck taping the tabs that stick up.
The pcb helps to prevent shorting between columns. Which can be if ou are using a metal tool and it touches a set of cells along with the tabs over an across the pcb."

I never worry about people that don't offer constructive criticism and I understand the concern over shorting cells out and the dangers. I bet most on this board has done something that wasn't correct but learned a valuable lesson. So no biggie on the criticism. In my teenage years I was working at a grain elevator that had 440v 3 phase so I learned quickly to take precautions when dealing with power as that was truly deadly. I'm 69 so I have made it this far without killing myself and doing some electrical work of some sort over my lifetime. Will be wiring my nephew's 26kw solar system with 60kWh of battery storage this weekend so that isn't for the faint of heart.

Have no idea what type of glue they used but it is a booger. I would be using a plastic scraper but the stiffness needed to get between the nickel strip and pcb board requires something very stiff and thin. That is why they first separation is done with a exacto knife just to get a little gap, that is the only thing thin enough to get between the two. Then getting the group separated can be done and on to the next group. But the gap between nickel strips gives me enough room to put a spacer between the cell pack and the pcb board to start on the next group. As they say you learn from experience and separating the next 2 rows will be much easier. So the help I have received so far has gotten me to the place I can start working on the cells themselves. Starting the search for 6s3p packs as you have suggested today.

Any suggestions on a good spot welder as I am sure I am going to need one before this is over.

 

[STxFarmer]

Does this look correct? Can't see all of the symbols on the batteries but from what I can see it looks like this layout.

Is there any good method for prying the spot welds off of the batteries? Looks like 6 spots on each battery on the thinner center section.

Or any idea what would be a good replacement nickel strip that I can make the same configuration out of for replacement batteries? Are these fused nickel tabs?

 

[fnnwizard]

Does this look correct? Can't see all of the symbols on the batteries but from what I can see it looks like this layout.

Is there any good method for prying the spot welds off of the batteries? Looks like 6 spots on each battery on the thinner center section.

Or any idea what would be a good replacement nickel strip that I can make the same configuration out of for replacement batteries? Are these fused nickel tabs?

That looks correct. It’s not worth the effort to take the strips off. It’ll tear at the weld points. But if you really want to try, needle nose at one end of strip and rolling motion.

Those nickel strips are heavy duty. If you want same ratings as oem, you’ll need to source same gauge/size. Search for 2P nickel strips for 26650.

That will also mean a hd spot welder… easily over $200 for capable one. If you don’t care about full power then can go smaller gauge

The plastic housing is cheap enough. They have 2p that snaps together to get proper length for 8S (not 6S).
A123 26650 have aluminum positive casing, but for > a decade now, they have a thin nickel “coin” you can easily spot weld/ solder to.

I’m sure you can find a builder on Alibaba.com to build you 9 modules. Might be best way unless you intend to do many packs over the years.

 

[max2k]

Will be wiring my nephew's 26kw solar system with 60kWh of battery storage this weekend so that isn't for the faint of heart.

are you sure you want to do that? One thing is to play with bunch of dead cells and completely different- work on live system, I am worried after reading this thread.

 

[fnnwizard]

Does this look correct? Can't see all of the symbols on the batteries but from what I can see it looks like this layout.

Is there any good method for prying the spot welds off of the batteries? Looks like 6 spots on each battery on the thinner center section.

Or any idea what would be a good replacement nickel strip that I can make the same configuration out of for replacement batteries? Are these fused nickel tabs?

Oh, btw.. those are double strips. probably .15-.25mm on cells and then thicker one over that.

And to answer your fused cell ???. the strip under looks to be fused, but the overlaid ones seem to connect to the unfused portions so overall it doesnt look like individually fused cells.

Maybe something like this underneath.

 

[STxFarmer]

That looks correct. It’s not worth the effort to take the strips off. It’ll tear at the weld points. But if you really want to try, needle nose at one end of strip and rolling motion.

Those nickel strips are heavy duty. If you want same ratings as oem, you’ll need to source same gauge/size. Search for 2P nickel strips for 26650.

That will also mean a hd spot welder… easily over $200 for capable one. If you don’t care about full power then can go smaller gauge

The plastic housing is cheap enough. They have 2p that snaps together to get proper length for 8S (not 6S).
A123 26650 have aluminum positive casing, but for > a decade now, they have a thin nickel “coin” you can easily spot weld/ solder to.

I’m sure you can find a builder on Alibaba.com to build you 9 modules. Might be best way unless you intend to do many packs over the

are you sure you want to do that? One thing is to play with bunch of dead cells and completely different- work on live system, I am worried after reading this thread.

Not my first rodeo with solar so no biggie.