VPower / CammyCC 48V20Ah LiFePO4 Repair by Amberwolf

[amberwolf]

Fixed up DayGlo Avenger finally, and thought I'd take this pack off the storage shelf and take it for a spin, but it didn't work. Yet another cell group has a cell dragging it down, to 0.12V this time! It's been a while since I used it, maybe a couple of months or so, (well, whenever I was working on RTLSHIP's pack, I think) but it was fully charged when I put it away, intending to put it on DGA once DGA was fixed.

I'll have to do some troubleshooting on it later, ATM I've got some flu variation and have enough trouble concentrating on stuff that I am afraid to work with teh battery right now. Too easy for KFF or worse.

 

[amberwolf]

Couple days ago (I think, can't remember for sure) I sat down and divided stuff up to find the bad cell, and ended up having to divide all the cells on half of the group to find it, probably from my own current physical/mental state of confusion. Anyway, I got these voltages from that half of the group:
1.86
1.82
1.88
1.58
1.58
1.41
1.42
0.86
1.82

So the 8th cell is the toasted one, and some others don't bounce back as well as most.

Charged up each cell till it hit 3V and let it off at that moment, going to the next one in turn (including the bad one), then checking them all again right after that:
2.85
2.86
2.83
2.81
2.78
2.65
2.85
0.51
2.84

Letting them sit for a day after that gets these readings:
2.80
2.72
2.67
2.65
2.74
2.61
2.81
0.00
2.72

I haven't checked them today...I'm trying not to get out of bed if I can help it, see if I can get enough rest today to get over this flu.

Thursday (thanksgiving day here in USA) I will probably work on the pack again if I feel well enough.

 

[amberwolf]

so almost 2 months later i got back to pack. has just sat as-is till now. cells =
2.64
2.28
1.53
1.56
2.68
2.51
2.75
0
2.25
so a couple have worse self dischage but not horrible compred to cell 8.

other half of group is still 2.79v.


rest of pack is 3.35v/cell +/- .02v so no other duds hiding yet.


blocked cell 8 off w/plastic since i cant get it out w/out both hands working, and soldered rest back together. now charging on little sorenson at 3.35v.


after it charges to 0 current flow i will put pack on bms and charger to let it fill/balance for a while.

 

[amberwolf]

finally finished above, reconnected pack halves and bms, hooked up to charger and it took only about 15 minutes to go green. but it is turning back on evry couple of minutes, for about 6-7 seconds at a time. will let it run overnite till i have to leave for work tomorrow.

 

[amberwolf]

i've been sticking this back on the charger whenever i am home, and taking it off when i leave, so it typically has several hours at a time just sitting there with teh bms balancing or idle. then when i plug it back in, a few minutes of charger running then sitting for a couple-three hours, then charging kicking on for a minute or two, then off again for hours.

after a few days of that, the charger didn't bother kicking back in again after the initial plug in, unless i pull it off the pack and let the pack sit for a day.

after letting it sit for 7 hours off charger, checking the balance plugs without them plugged into anything, i find all groups at 3.48v except the one i just finished fixing, whihc is at 3.42v. so it does still have at least one slightly leaky cell in it, just not as bad as the dead one i pulled. it is probably the one with the lowest voltage on it out of the group-cell checks i did before reassembly.


for now i won't do anything about it (mostly because I don't have any cells to replace the duds with, and capacity is still capacity).

i'm doing a discharge test of letting it just sit on dayglo avenger with the lights on (headlight, taillight, brake light, and one turn signal non-blinking), which run off a dc-dc from pack voltage down to 12v, and it takes about 1.27a constant. we'll see how many ah i get out of it before bms hits lvc on a cell group, and then i can check the group voltages. since i am only testing at a very low c-rate, i don't expect to get the same ah out of it while riding.

 

[amberwolf]

discharge test completed down to 13.199Ah, 687.1Wh. Dunno the cutoff voltage becuase I had the TWM powered by external source, so it showed monitored voltage of 1.09V, probably the leftover voltage in controller caps plus any leakage thru teh BMS FETs that shut off at LVC.

I powered the bike off and on to reset the BMS, and it came back up to 49.8V, and I let it run again for a few minutes, down to 48.95V, while I pulled both blance cables and measured them under the small load.

most negative first:
3.08
----
3.06
3.15
3.12
2.20*
2.68
3.15
----
3.05
----
3.12
2.96
2.91*
3.10
3.10
3.10
2.78
----
3.07

So it is the first problematic group i'd fixed that is cutting out first; the one next ot it isn't a lot better. But the most recently-fixed one doesn't appear to be the the problem, as it's still up at 2.91v.

Anyway, it seems to be working again, and still gives enough useful power to be worth carrying around, without major pack surgery to cut out the worst cells (whcih i'll have to do eventually).

I still kinda wish i could get enough still-good cells to replace the dead ones, and make it worth going thru the pack to find the worst once and replace them, but so far the offers made have not come thru, so I'll just keep using it as it is whenever i need it.



I started it charging up again after the test, and it is now in it's balancing stages, which I will leave running for as long as possible until I need it again.

 

[amberwolf]

I forgot to note this, but during charging at the beginning, with higher current flow, one of the low groups (2.78V in previous post's list, 1st inward from most positive end) read higher voltage by almost a millivolt than any other group, including ones that were at a higher SOC/voltage to begin with. So it has at least one higher-IR cell in it that I should dig into the pack and locate and remove (or at least test separately to see if it's also low-capacity or leaky).

The others were proportionally about the same voltages as they started out as, until it got to nearer full, although the groups with removed cells were a bit higher proportionally. That's to be expected given that they have less cells and thus a slightly higher internal resistance than the groups with more cells.


I've left it on the charger since the above, letting it do it's thing, whenever I've been home. Only taken off when I had to leave the house for work and such, until four days ago when i decided to let it sit and see what stayed up and what went down.

I tested all the voltages of the pack and each group, with balancing connectors still plugged into the BMS. I was going to also measure with the connectors disconnected, but I got distracted by the dogs barking at something outside and forgot until I was typing this up.

Right now It's at 55.4V pack voltage, and every group is within 0.03V of each other, most at 3.46V, a couple at 3.47V, and some at 3.45V, with one at 3.44V. The voltages don't appear to have anything to do with the in-use or discharged voltages, and are probably a result of either meter-to-cell connections thru the back of the balance connectors or slight variations in the balancing circuitry.

most negative first, * is repaired groups:
3.46
----
3.46
3.46
3.47
3.46*
3.47
3.45
----
3.46
----
3.44
3.46
3.46*
3.46
3.45
3.46
3.46
----
3.45


I'm thinking about making up some 6s "test lead" sets to solder to the pack so I can plug a battery medic into them one at a time and watch them with the pack under load. Maybe while riding, or maybe just on the bench with lightbulbs or stove elements or something.