(Maybe) a123 pack for recycling

[fabiograssi]

A friend gave me this package. Since he bought it never worked (more than two years ago).

He doens't have the knowledge to repair and just gave it to me. It was a duct tape style pack, with a external BMS in a small plastic house. I do not know the origin.

Original configuration: 16s14p

Incredibly, even after more than two years, 13 of the 16 paralleled blocks are with 3.3v.

What do you guys think as best to do? I was thinking in just took off 4 paralleled blocks and make a 12s14p pack.

Someone could indicate a BMS for this pack that could handle a 35A nominal, and where to acquire it? Maybe Paul?

Some pictures:

P6030065.JPG (49.62 KiB) Viewed 641 times

P6030066.JPG (63.82 KiB) Viewed 641 times

P6030067.JPG (132.61 KiB) Viewed 641 times

P6030068.JPG (98.25 KiB) Viewed 641 times

[cassschr1]

try charging those dead cells. they may come back seeing that there a123 cells.

[amberwolf]

If they're actually A123, you can probably revive the bad groups by recharging slowly. they might not have full capacity afterward, but some people have had good luck with A123 like that.

First, which groups are the bad ones? Are they all at the most negative end? If so, I'd guess the BMS being left connected probably drained them slowly while istting there, and none of hte cells are bad.

If the bad groups are random thru the pack, then ther eis probably a bad cell in each groups. Test each group to find the bad cell(s) in each one, by cutting a bad group's connection in half at one end (either positive or negative, both is not necessary), then monitor the voltage of each group as they sit there. One group will probably start going up, and the other will stay or go down (assuming it's not 0V already). The group that goes up is probalby fine, and hte gorup that goes down now needs to be divided in two again. Keep repeating this until you have just the bad cell(s) isolated, and you can reconnect the rest of the group together.

Leave the bad cell(s) disconnected (can be left physically in place for structural integrity), and charge up the group with a single-cell charger or lab power supply, at a low current (maybe 100-250mA) until it gets to the same voltage the rest of the groups are at. Then reconnect it to the rest of hte pack if you disconnected it, and recharge the whole pack with the regular charger. Monitor all the groups' voltages as they are charging so you can see if any spike up in voltage fast, as those probably have higher resistance in some cells and those cells might need to be replaced or removed.

The binary search can be used to find any of the bad cells in the pack.

[dogman]

12s pack might be the easiest. Unfortunately though, you still have the tiny cells with the hundreds of spot welds that may pop later.

Personally, I'd be inclined to make packs for my cordless stuff out of them. Drills, saws, weedwhacker, lawnmower. They are all 24v stuff. 5 ah 24v A123 packs on the saws in particular woud be nice.

[dnmun]

no, i bet the bottom 3 are dead since the BMS looks like it is still connected, after several years.

use the same BMS, and keep it 16S by cutting the 4 off the end of each, then use the 13 sets of 4P to make the 3 sets of 10P and you will have a buncha 2P sets for spares. that would be the simplest and you won't need to buy more parts.

did you get the charger too?

[fabiograssi]

First, thanks for the answers.

BMS discharging: don't appears to be the case. This is the lines with 0.2v:

P6030067.JPG (79.25 KiB) Viewed 525 times

I am a total dumbass in electronics. I try, but is dificult to me. So, I don't know how to test a BMS.

I Personally think that what ocurred was: the guy bought the pack and put it on the road. The (faulty) BMS don't cut when this 3 paralleled series becomes under LVC... The rest of the history is obvoius... He tried more a couple of times and nothing (maybe the bms have a circuit to measure the LVC of the entire pack, and this was cutting the power... Just don't know). Then he shipped the pack to the seller, who must be more dumbass then me, who keeped the pack for a whole year, untill the buyer lost her patience and request it back.

What I have in hands to try something: a meanwell 48v with Richard Fechter's limiter board (so I can input little current and I think I can reach sufficient voltage to 3.5 each cell). I have a turnigy watt meter. And I can monitor the paralleled lines voltage with a multimeter...

Another tool that I have is a Hyperion 1420...

In this first moment, I prefer don't cut the pack links yet...

What do you suggest?

BR,

- Fabio

[cal3thousand]

12s pack might be the easiest. Unfortunately though, you still have the tiny cells with the hundreds of spot welds that may pop later.

Personally, I'd be inclined to make packs for my cordless stuff out of them. Drills, saws, weedwhacker, lawnmower. They are all 24v stuff. 5 ah 24v A123 packs on the saws in particular woud be nice.

That's a great idea for the those cells. I would be doing that if those were in my possession. I have a drill that is basically crap until I frankenstein it.

[heathyoung]

Meh - turn it into a 14S14P then - cut out the two dead rows (or bypass (short) with heavy gauge cable) and reconfigure the BMS for 14S - this model allows this, have a close look at the solder pads just near the plugs. You just solder across the correct one (ie. 14s) for 14 series.

Hot glue would be a pain to remove those dead rows, so I would just bypass them with 10Ga cable personally, move the taps, and reconfigure BMS for 14S - easy.

[dnmun]

if it is a dead row, it would be down to zero. do you have any that are zero? you only said that they are all 3.3V except for three that are dead, but now you say they have .2V resting?

you need to put a single cell charger on that row and charge it up. do that to all three.

don't bulk charge the pack until the low rows can catch up, so first get a single cell charger and charge up those three rows and then watch how fast the voltage drops.

if they leak down rapidly, then you can follow AW's instructions and isolate the section of the row the bad cell is in by cutting the ribbon strap running across the ends of the cells.

i would cut out the four on each end first, doing only one at a time, and if you don't have a dead one in either end then the 6 in the middle can be cut with 3 on each side, test them to find the dead one, then cut 1 and 2, of course it will be the last group of 2 and you will have to cut it open to find the dead one. never fails.

once you find the bad ones then you can figure out what options you have. i do think it is best to just use it as a 48V pack and use the BMS as it is now so you will have to shrink the pack in width. i thought cutting 4 off each end would leave the most flexibility and spares.

be careful handling that pack. if you fold it back on itself without the insulating rubber layer it will short out the entire battery.

[999zip999]

Put a charge to the low strings to 3.3v and see if they stay youcan use your 1420 as a single cell charger lipo4 1s cutoff 75-80% at 1amp see if the string holds a charge ? Or could use a cell phone charger and watch it. If the charger says 5v it ok if you watch it. Just check polaity. And voltage.

[fabiograssi]

Well, charging now. Put in the 1420 hyperion: life, tcs 85%, charging at 0,5A...

Let's see....

F

[fabiograssi]

Well, 1 week charging the problematic groups. Incredible, 2 groups "appears" to revive... But the voltage drop much more than the other groups (in fact, the other groups are all in the 3.3v state...)

Facts: I have the original 16s charger and a 12s charger. So, I still think the best idea is to turn this pack a 12s, because I think I will can suck a lot of current from it, and I have the correct charger for 12s...

This is the BMS. Is there a way to test it?

Heathyoung: this small spots near balancing connectors is to convert this BMS to other configs? You told wrote something, hun? Is just bridge the 12s points and ok, turn it to 12s version? Then, I have to start from left to right the balancer wires?

Finally: is there a way to cut off the "low voltage protection" from hyperion 1420? To start the slow charge the bad groups, I need to stay present a lot of time in the begining, starting over and over again after five minutes, because the charger stop the process after 5 minutes, untill the groups reach at least 3.3v

Pics:

P6130101-001.JPG (157.47 KiB) Viewed 358 times

P6130094.JPG (41.85 KiB) Viewed 358 times

P6130098.JPG (70.9 KiB) Viewed 358 times

[FastDemise]

You just said you have a Hyperion 1420i charger. That's good for 14s and the 6v of extra juice will help your bike go that much faster.

[999zip999]

Maybe you can leave it whole and wire around the 2 dead strings. Turn up your 12s charger for 14s ? I have a 1420 too. Need to spend some time with it. The lvc might not work on the bms just the balance. Or if it stays balance go bareback with balance wires. Or ?

[dnmun]

you should keep it 16S and use the BMS as it is. AW gave good advice about finding the bad cells in the bad row, you can charge it up and cut the bad rows into sections and then watch the sections to see which side drops faster, and if the other side remains stable, keep doing that until you find the bad cells, just like he did with the Vpower pack.

once you know which ones are dead, come up with a scheme to cut and paste so you can keep all 16 rows but cut 2-3 off the end and use them for the spares to paste in. it sould be neat if you could cut just one off the end of each row and that would be enuff to solder up the replacements.

seems complicated, but if you can keep the 48V it seems worth trying, jmho. i think it will be easier in the long run because less risk of BMS mods going bad. jmho.

[999zip999]

I agree 16s is best. If it is one bad cell you could wire around it, you could add it on the end. Just looking at cutting a cell out of the middle.

[fabiograssi]

Well... At least for me, this is unbelievable...

Using this:

viewtopic.php?f=11&t=32967

seted to 3.3v and 0,30 Ah (If the cells are really a123, they are 1,1Ah each, so x14=15,2 Ah each row, so 0,02C of charge), the bad rows APPEARS to revive.

Tomorrow I will do the first test with the bms: I will buy a string with 4 car ligths in series, put as load, and start to check voltage of all the rows... and if they still in the same voltage of another rows with this small load, i will check if the bms can cut the rows and in wich voltage it does...

Lets see, I will keep you informed.

F

ps: this inexpensive dc-dc converter and charger (and led driver too) is a very nice tool to have at home... The price is ridiculous...

[fabiograssi]

Finally, a little bit of courage to start this work!

[skeetab5780]

to bad you had to pull the pack apart, i bet that was a real good battery when it was in operation! those cells pack quite a punch

[dnmun]

did you pull the straps off? all the cells in a row are now torn apart for all rows?

you can isolate the cells that are dead by using a binary search like AW mentioned.

if you do not have a 3.65V single cell charger to charge up the row, but have a 12V auto battery charger you can charge up the rows using that if you have a voltmeter to monitor.

you will have to solder some leads onto each row, or use jumpers with alligator clips if you have them.

connect three rows in a series, put them on the charger and charge them up to the 3.65V or fully charged while watching them closely with the voltmeter, then pull them off the charger. or you can 4 rows in series and see if you can get them to 3.4V/row.

when you have them charged up, find the weak rows by allowing them to discharge so you can see which are the worst. some rows will not lose charge, the voltages will remain constant, save them to recharge individual rows that drain down.

when you identify the rows with the bad cells, cut them into two sections, the binary search AW mentioned, and watch to see which sections drains down first after charging the sections up with some of the good rows that remained charged. that allows you to charge just one section at a time.

then cut that weak section apart, recharge again with a good row of cells, find the bad subsection, cut it again, then recharge, repeat until you can isolate a bad cell.

you only have to cut the strap on one end of the cell at a time to isolate the bad section, or subsection, so you have less repairs to make later when you reassemble. but looks like a lot are free already so you gonna need a spot welder to reassemble. BOL

[fabiograssi]

I have a individual charger... Take a look in this device http://endless-sphere.com/forums/viewtopic.php?f=11&t=32967

Bad cells: easier than that to detect: three entire rows... Took one row, aparted in two blocks... exactly the same voltage in each half... So I made a bet and separated all the cells... all with exactly the same voltage!

So, 2 options: rearrange cells to stay with 16s pack, but with less Ah (harder), or took off 4 rows and make a 12s pack (easier)...