LiFePO4 pack rebuild - what are my options?

[oddjones]

Hi all,

Well I pulled my JimmyWu 36v 10Ah non-functioning pack to pieces and I got lots of blue Valence saphion IFR-18650EC cells.
They (were) grouped into 12 strings of 7 cells welded in parallel with thin (read sharp) tin strips.

The readings I was getting from the terminals of my BMS (prior to disassembly) were as follows

[pre]string reading value
1 3.1v 3.1v
2 5.7v 2.6v
3 7.1v 1.4v
4 10.2v 3.1v
5 13.4v 3.2v
6 16.5v 3.1v
7 19.6v 3.1v
8 22.7v 3.1v
9 23.9v 1.2v
10 25.0v 1.1v
11 26.1v 1.1v
12 27.3 1.2v[/pre]

SO: I split the pack up into good strings of cells and bad strings of cells. The bad strings I then cut into individual cells to see whether any were salvagable...

It turns out that from the bad strings of cells there's not a single cell which reads over 2v.

Question 1:
Is there any point whatsoever trying to "rescue" these bad cells? I have a Nokia 5.7v mobile phone charger hooked into a jury rigged single cell frame, but I've not connected it to a cell because I don't know if it's a dangerous thing to try.

Question 2:
I have 49 good cells and a BMS which is set up for 12 "strings" of cells - by my reckoning that'd give me 12 strings of 4 cells or about a 5-6ah battery. Is it that simple? - my daily commute uses about 2ah (each way) so this battery should just get me there and back.

Question3:
I'd be wiring the battery in serial with my other (working) 36v 10ah pack through an infineon controller fitted with 3 way current switch. I've been told that my Bafang motor won't handle the full power of a 72v system for long so intend using it on the "middle" setting of my current switch, which to my (limited) understanding offers the equivalent of a 48v battery. So I assume the actual load on each battery would be less than I'm currently using by running the "high" setting on a single 36v 10ah battery.

Does any of this make sense?

 

[dogman dan]

No, it's not going to work. The c rate you'll be drawing from the newly built, small pack will kill it immediately. Use the cells to make some nice flashlights, or whatever you can run on 12v packs or something like that. If you want to try to revive the low cells, they might get a use or two in a flashlight or something, but they are pretty much ruined.

You could however, build a 36v pack to paralell with another and use that OK.

 

[oddjones]

Ah - OK,

Well, my other thought was to make an 18V pack out of 6 strings of 8 cells to wire in serial with my 36 v - I reckoned that 54v would be just about perfect for my Bafang - Only question is where would I find a BMS with 6 channels (and an 18v charger) - is that viable?

Can't believe that the £300 battery I used about 20 times is completely toasted!

 

[oddjones]

I believe Fechter's "Even Newer 4-24 Channel BMS" would allow me to have 8 strings of 6 cells?

I've seen this earlier model described on V is for Voltage (http://www.tppacks.com/products.asp?cat=26)

Honestly - am I chasing rainbows?

 

[dogman dan]

Oh yeah, you want to salvage something out of this disaster for sure.

An 18v pack could still be built and have enough cells to be able to survive, and then you could buy one of the balancer/ voltage monitors from Hobby King to use as a bms for discharge, or one of TP packs(Goodrums) lvc's if you solder well. Then an inexpensive lipo/ lifepo4 6s charger would have you set for an 18v - 6s lifepo4 pack management setup. Using this stuff, would of course be great practice for you if you ever make the transition to lipo. You'd already own a charger and voltage monitor.

Our sympathies about the Vpower pack, unfortunately once cells in one paralell string die, they take the whole string with them. This effect gets worse if the cells have a low c rate to begin with. This is true of any brand, any chemistry. Once one paralelled cell has a problem, it puts extra strain on the rest.

 

[oddjones]

This is much better news! Thanks for bearing with me

Could you possibly point me at a couple of links for the exact balancer/monitors and chargers you're talking about?
(The hobby king website is fairly daunting at first visit, plus - of course it'd be great if I could find something similar in the UK to save the postage)

Also there's some (more) new terminology sneaking in there - what does 6s mean?

 

[robb]

In answer to Q1 I have a cammy battery which shorted out and had 2 strings of cells damaged. Though most of the cells in the damaged strings were 2v or less and some 0v, I did manage to save quite a few of them. Cammy sent me a replacement string and I made another good string out of the remaining cells with some left over. Still seems to be holding charge fine 6 months later. Is it worth contacting Jimmy Woo for replacement cells/strings?

 

[heathyoung]

As long as you keep an eye on the charge current and voltage on the bad cells - you could charge them back up and see if they - get hot (bin) hold a charge over time or fall back to 1v and get hot, hold a charge and have reduced capacity.

You could reconfigure the cells as a 'boost' pack if you get stuck with not being able to get new cell strings.

 

[dogman dan]

Since I'm a dork that can't seem to get it figured out how to post a link here, uh, no. It was explained to me once in words that were gibberish to me, I don't DOS or know keyboard shortcuts. If I can't drag it and drop it with a mouse, I'm clueless. Maybe if you could explain it to me in idiot language I could start posting links.

Look in the hobby king site Battery chargers section, and for starters, ignore all items not in stock. The battery medic is a fave, but only if you can buy it. But you would be looking for devices that monitor voltage on 6 cells, and have adjustable warning settings so you can set it for lipo or lifepo4, or nicad, etc.
Same thing on the chargers, pick a price range, and look for at least a 6 cell one to do 18v lifepo4. The chargers typically need a power supply too, such as a computer 12v supply, a 12v lead battery, or whatever. Expect to spend about $100 for a charger and a voltage monitor.

 

[dak664]

Dogman, you post a link with

[url=http://www.google.com]click here[/url]

click here
From what I've read LiFePO4 doesn't immediately die at 1 volt, but every day at low voltage degrades the capacity a bit more. I don't think there is a fire danger trying to recharge low LiFePO4 cells, but I have no experience with that. Other lithium chemistries that are over discharged can explode during recharge via thermal runaway, for sure.

 

[dogman dan]

ebikes.ca

ebikes.ca

www.ebikes.ca.com

 

[dogman dan]

Ok, I seem to be getting warm here. Now explain the rest of the obvious stuff any 10 year old would know and I may get it. Do I type the entire url into there? I won't be doing that a lot for sure.

 

[oddjones]

but every day at low voltage degrades the capacity a bit more.

Ah, well - this battery had been left for 18 months(!)

As long as you keep an eye on the charge current and voltage on the bad cells - you could charge them back up and see if they - get hot

So will my 5.7 volt Nokia charger be OK for this job?

Re contacting Jimmy Wu - nah, I don't think that's going to happen

Oh and links - copy the link you're linking to out of the address bar of the browser, paste it into your text then highlight it and click the URL button (in between "Img" and "normal")- that'll make it into a link.

 

[oddjones]

Ok - so I've charged a couple of cells with my nokia charger and they seem to be holding steady at 3.3v

thing is I've got no idea how long to leave 'em on charge for! So I've dug out a camera charger used for 3.7v li-ion batteries - this has a charging/charged led on it.

Will this give me an idea of when my lifepo cells are fully charged?

I'm encouraged by the other cells holding 3.3v - obviously I don't know at this stage how long they'll hold it for.

Can anybody suggest a good discharge test?

 

[dogman dan]

Some kind of light bulb like a halogen headlight, then look for sag when the load is applied. Compare the voltage of your known good cells under load with the recharged ones under load. The bad ones will sag compared to the good ones. If you get some revived, keep em in the same paralell string so they don't drag down the good ones if they have less capacity now.http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=10952&Product_Name=Cell-Log_8M_Cell_Voltage_Monitor_2-8S_Lipo

Thanks for teaching an old dog new tricks. Amberwolfs PM finally did the trick. The link above is to one of the HC voltage monitors that can keep track of an 18v pack, and warn you of low voltage.

 

[999zip999]

I was looking for a single cell I checked out10 chargers cell phone ect. They were all over the place the 3.8v checked out 5v. the 4v. was 7v.ect. I just didnt let the cell get over 3.8v just by wacthing it closely. If they get over 3.9v they very well be damaged. if over charged..I replaced a set 4 cells in parelle charged the cell to 3.8v for ballance. It been working great for the last three mos.GOOD LUCK!!!

 

[oddjones]

Just tested my first recharged cell.
I had it wired to a 12v 5w car brake light. It drew 0.9A for 1 hour 15 minutes before the current started to drop slightly. Voltage remained at 3.3v throughout.

Now then, please attempt to follow my flaky maths and point me out if I'm wrong here.

My "good" cells are wired in strings of 7 to give 3.3v and 10ah. If I divide that 10ah by 7 I get about 1.4ah per cell.
Now if I just got an hour and a quarter at 0.9a does that mean my "bad cells" are charging back up to at least 1.125 ah? (.9 /60 * 75)

If so, then that's pretty damned good for cells I was told were probably toasted.
If that's the case and they all turn out to he that good, what do people think about my joining them up with the "good" cells again?

Or am I best keeping them as a reserve pack

 

[Battboy]

What is the final Spec. you want to get from a LiFeP04 Pack? Voltage / Ah please.

DH

 

[dogman dan]

well, that's progress, you've proven that those batteries are still good for a 5 watt ebike. Remember my comment about a flashlight? You will need to test them with a bit higher wattage load now to really tell if they are good for use on the bike. A headlight, or a large halogen bulb of some kind is what you need to try next. Other ways to test under load use heating elements of some kind. So put more load on em somehow, and if they perform as good as the others, similar voltage sag, and similar run time, then they survived it.

 

[oddjones]

5w eBike gonna try me one of those

There doesn't appear to be any voltage sag at all (from the 5w bulb any ways - gonna find myself something more meaty)

should I load the cells individually or string them together to get something close to the load that the test device is expecting (ie 4 cells = roughly 12v for a halogen bulb)

 

[oddjones]

What is the final Spec. you want to get from a LiFeP04 Pack? Voltage / Ah please.

DH

Well, in an ideal world I'd like the 36v 10ah pack I originally paid for!

If I can get a valid 36v 10ah pack then I can just wire it back up using my existing Ping BMS and charge using my existing 36v charger - plus the current switch wired in to my controller would then work properly (it was set up for 2 36v 10ah batteries)

Second Favourite is an 18V 10ah pack - which I can build with my remaining "good" cells (ie the ones reporting 3.3v without needing recharging) - indeed from comments I've heard on other threads this battery may even be preferable in respect of preserving my motor (72V through my little bafang isn't advised)

The issue is that I'd rather not have to shill out the additional $100 for a new charger and monitoring/balancing equipment if I can avoid it.

Plus, if my "bad" cells turn out to be actually quite good cells then I'm simply wasting them - I was thinking about building another 18V pack and paralleling that off the back of the "good" pack but that would give me 20ah of 18v - which isn't much use to me as I only have 10ah of 36v from my other battery - so I'd run out of usable power before I even needed to call on the additional pack

 

[dnmun]

you should not use a lipo charger to charge up the cells unless you actually can watch the voltage or it will get pushed over 4V.

lifepo4 cells will self discharge if left unattended for years and will pop back to life from lower voltages than this instantly. like starting at .5V and charging with a single cell charger, the voltage immediately jumps to 2 then 2.5 then 2.7V and will charge up and hold a charge as normal so just because the cell has self discharged does not mean it is 'dead' as you say.

there may be nothing wrong except that the pack has not been charged up for a year, but if it has a BMS then some of the cells could have overdischarged during that time since the first 4 cells provide the circuit current for the BMS.

 

[oddjones]

you should not use a lipo charger to charge up the cells unless you actually can watch the voltage or it will get pushed over 4V.

lifepo4 cells will self discharge if left unattended for years and will pop back to life from lower voltages than this instantly. like starting at .5V and charging with a single cell charger, the voltage immediately jumps to 2 then 2.5 then 2.7V and will charge up and hold a charge as normal so just because the cell has self discharged does not mean it is 'dead' as you say.

there may be nothing wrong except that the pack has not been charged up for a year, but if it has a BMS then some of the cells could have overdischarged during that time since the first 4 cells provide the circuit current for the BMS.

Err wow - that's not what other people said - I'm no expert here, I was assuming the were dead coz other people told me they were! {http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=18445}

I have checked the voltage going in from the charger periodically, and it's not gone over 4v - I think it's a fairly clever charger - it charges multiple voltages so presume it must check the cell to see what voltage it should be giving out somehow.

This is more in line with what I was expecting to hear originally (although of course I've since pulled the pack to pieces!) - It's no issue as I presume I'd have had to take the pack to pieces to individually charge the cells anyway.

The "first four cells" bit certainly figures as these cells are the worst of the lot - down to 1.1 - 1.2 V

It's certainly great news if it's the case - does this fit with your experience dogman, heathyoung, dak664? (I'm equally ignorant of all your superior knowledge!)

 

[dogman dan]

You should be able to put a 12v halogen light on a 3.7v cell, or string of paralell cells to test it.

I thought the cells had been drawn down in another way, that would have been more likely to ruin them. One cell dies, then the c rate increases on the other paralelled ones, killing them, and so on. So that would have been the too high c rate treatment, vs the slow discharged by the bms treatment.

Still. I thought below 2v was going to mean at least low capacity, lots of voltage sag in use, and very few cycles if revivied. That's why I say don't mix the good cells paralell with the revived ones. Worth the trouble? Depends on what your time is worth. Depends on if you have a use for a lower voltage battery. Depends on if you have money.

 

[oddjones]

I have very little money!

This bike was working in its present form 2 years ago, it was only when I went chasing speed that it all started to go wrong(!) since then I've bought a cycle analyst, another 36v 10ah battery, 2 new bms's, 3 chargers, disc brakes, 4 battery housings ... I've spent too much chasing the extra velocity and so far I've seen nothing for it. If I can rescue the battery and not spend an extra $100 then I'm going to try that - the purpose of this whole exercise in the first place was to save money (how naieve can a person be!)