Author Topic: Recovering lithium batteries that have been overdischarged for a long time.  (Read 2583 times)

0 Members and 1 Guest are viewing this topic.

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
How many cycles does it take for a lithium battery, that has been recovered from a severely overdischarged state, in which it had been for a very long time, to recover to its full remaining capacity?
The place where i interned let me have a dead battery pack from one of their segway thingies. The pack has 64 SONY US18650V3 cells, of which 16 were immediately either shorted or open.
The specs for the cells are dated but beggars can't be choosers, right?
Anyways, after recovering the cells with the help of a 5V PSU and a 150 Ohm resistor i plug them into my Lii-500 battery charger to test them and the first 4 came out at a pretty healthy 2030-2098 mAh, which is pretty good for a cell that sat at some 1V for what might be over a year. I hadn't used the charger for a while so before i figured it out i put two cycles on the cells.
The next two, however, came out at a much lower 1900 and 1847 mAh and that makes me wonder, will the capacity go up if i cycle the cells since these are only on their first cycle after recovery?
It's important to know because if i want to assemble a pack out of these i need to balance the capacity of each set of parallel cells and if they're not yet fully recovered that will become a problem once the pack is cycled a few times.
What do you guys think?

Also a tip for your Lii-500: You can make the normal test as fast as the fast test if the cells are precharged. I use my spare TP4056 modules to precharge the cells before they go into the tester.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RoV

  • Regular Contributor
  • *
  • Posts: 175
  • Country: it
How many cycles does it take for a lithium battery, that has been recovered from a severely overdischarged state, in which it had been for a very long time, to recover to its full remaining capacity?
In my experience, good and relatively new cells do recover after an overdischarge (but I haven't experience with long times), but they permanently loose part of their capacity. According to what I have read, this is due to metallic lithium forming over one electrode (called lithium plating), that reduces the active area.
The same happens when the battery is charged at very low temperatures and high current, in fact generally they can be fully charged only over 0 °C, charged at reduced current up to -10 °C and not charged at all below.
 
The following users thanked this post: AmnevaR

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
How many cycles does it take for a lithium battery, that has been recovered from a severely overdischarged state, in which it had been for a very long time, to recover to its full remaining capacity?
In my experience, good and relatively new cells do recover after an overdischarge (but I haven't experience with long times), but they permanently loose part of their capacity. According to what I have read, this is due to metallic lithium forming over one electrode (called lithium plating), that reduces the active area.
The same happens when the battery is charged at very low temperatures and high current, in fact generally they can be fully charged only over 0 °C, charged at reduced current up to -10 °C and not charged at all below.
Funny how some tutorial on the internet said to freeze and thaw the cells to "hopefully" help them.
So far most of the cells i have tested do confirm what you say about lithium forming and reducing the surface area in a sense that the internal resistance of the cells has become 3-4x larger than the rated spec.
This could also be error in the charger and also due the fact that they still have their nickel strips (just folded up) on the ends.
The pack was made in 2016 so these cells are quite old and from what i gather these knockoff segways were not used very much so the cells probably have less than 50 cycles on them.
What i noticed is that the charger for the particular device would overcharge the battery pack and over time an imbalance in the cells would form, this imbalance would then overload the balancing circuitry, which would eventually overheat and burn up.
After that all hell would break loose, which is what happened to the pack i have.
I might have to cycle the weaker cells to see if their capacity improves any.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 9410
  • Country: gb
I am very wary of internet tutorials that say freeze things. We had a guy the other day asking if freezing would recover his NiCd batteries. Thankfully I haven't seen any advising sticking them in the oven (Dont!).
« Last Edit: April 19, 2021, 08:53:34 am by Gyro »
Best Regards, Chris
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
We had a guy the other day asking if freezing would recover his NiCd batteries. Thankfully I haven't seen any advising sticking them in the oven (Dont!).
Freezing nickel batteries has been a thing for a while, perhaps it carried over to lithium cells.
Should have told him to stick them in the microwave, sure helped the iPhone people.  :-DD

As i test more cells seems like more and more fall in the ~1900 mAh range. There are some in low 1800's so i might cycle them a few times to see how the capacity changes. If it does at all.

Currently the TP4056 jig is all taken up, but once it frees up i'll modify one slot to have a discharge function.
« Last Edit: April 19, 2021, 12:18:03 pm by Refrigerator »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
So far it doesn't seem like there's any correlation between internal resistance and capacity.
Some high ISR cells pump out more mAh, some less. Some of the lowest ISR cells produce the lowest capacity values.  :-//
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Online tunk

  • Frequent Contributor
  • **
  • Posts: 969
  • Country: no
If you're going to use them in a battery pack, then you
may also want to check their self discharge as well.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
If you're going to use them in a battery pack, then you
may also want to check their self discharge as well.
True
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 8110
  • Country: fi
You should understand what you are doing is dangerous and the risk of fire is, while not the most probable thing since sliced bread, still very real.

To make it somewhat safer, apply very low charging current (like C/100) for a few hours max, and if the cell voltage doesn't get into normal range (> 3V) during that time, or falls back after sitting for a few hours after such treatment, just give up with the cell.

The problem is copper dissolution into electolyte, and the possibility of copper dendrite formation, which can penetrate the separator and short the cell causing localized hotspot. High quality cells are unlikely to go into full thermal runaway, but it's still risky.

In addition to reduced capacity, you are likely to see increased self-discharge because of the dissolved copper in electrolyte. Increased self-discharge manifests itself as difficulty maintaining balance in series packs.
« Last Edit: April 19, 2021, 06:36:32 pm by Siwastaja »
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
Seems like some cells have huge internal resistance.
One cell cranked out only 647mAh but right after the discharge ended it bounced from 2.9V up to 3.64V, which is alot more than you'd expect.
Might be interesting to cycle these to see what they do.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RoV

  • Regular Contributor
  • *
  • Posts: 175
  • Country: it
So far it doesn't seem like there's any correlation between internal resistance and capacity.
I also have the Lii-500 and my impression is that it's not reliable in the internal resistance estimate. If I repeatedly try on the same battery I get variable results and it doesn't seem to be able to measure values lower that 31 mohm. Besides, as also you noticed, I often get low values from batteries performing poorly.

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
A thing i noticed with the Lii-500 is that the normal test is also the fastest way to check cell capacity if they're already fully charged.
The normal test charges the cell and then measures the capacity as it discharges, as a final step it charges the cell fully again.
But if all you care about is the capacity then you can pop in a fully charged cell and just let it discharge.

I had a cell that only managed to crank out a measly 369mAh, so i let it go for a second test, which resulted in 937mAh.
Now the cell is going for a third test to see how much it will produce this time around.
Cells like these wouldn't be used in any pack but they're still good for other stuff where i need a high current battery but don't want to risk ruining a good cell and where a laptop 18650 cell doesn't have the current capability.
The other weak cell i had is also on its second rodeo and so far it has pumped out double the capacity in mAh and still going. oops got my cells mixed up.
I might have to build a cycling jig for all cells to just to make sure they're stable.
« Last Edit: April 19, 2021, 10:45:49 pm by Refrigerator »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
The TP4056 jig finally freed up so i used a spare SPDT switch to add a discharge function to it.
To make it work you connect the IN- of the TP4056 board to common on the switch and then NC and NO to ground of the 5V suply and a load, i used a 3,9 Ohm 10W resistor.
The other end of the load goes to B+.
I also added an LED with a 150 Ohm current limiting resistor to show when the load is applied.
This way you can rapidly switch between charge and discharge and the DW01 IC on the TP4056 board cuts off when the cell discharges.

So far i've also noticed that the Lii-500 is quite sensitive to bad connection. I had a cell that only squeezed out 70mAh but on a second test it managed 1940mAh and could have gone longer had i not bumped it.
So i might have to tear off the nickel strips because they might be getting in the way. The reason why i left the nigkel strips in the first place is because i have neither nicklel strips nor a tab welder so i left the strips as something that i could solder to without heating the cells too much.

I have a 105Ah car battery that is practically new but it froze and developed a short on one of the cells but i can still use it to make a simple tab welder.
« Last Edit: April 20, 2021, 02:27:44 pm by Refrigerator »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 28142
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
FWIW a 14AH 12V jump start pack with a dud mainboard for the blinky lights and 5V USB plus other supplies it can do was down to just a few volts after some years sitting on the shelf but the battery pack with dunno how many flat cells probably 4 with an integral BMS charged up just fine and has held 12.4V for some weeks.
Not tried any capacity tests yet but will in the next few weeks and compare it against an identical jump start pack that's still in working order.
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
Had a cell with 1820mAh and cycled it three or four times, after that it now tests at 1878.
So there are some gains but not huge, but IMO still relevant for balancing.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
Looked in my scrap board pile and found an old 775 motherboard with a total of 17 2SK3919 MOSFETs, each has a max current of 64A so i think 10 in parallel should suffice for a quick and dirty spot welder.

Edit: Also found some inexpensive nickel strips so for my next test i'll be pulling all the residual nickel strips off the cells for more reliable test results.
« Last Edit: April 21, 2021, 10:06:31 am by Refrigerator »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
Tore the nickel strips off most of the cells, including the dead cells, which will be crucial during the maiden voyage of my homemade spot welder.  :-BROKE
The welder might not come any time soon since i'm busy with university stuff right now but i like having several ongoing projects in case one stalls i can use the momentum on another.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
Also do you guys have any tips on removing the metal nibs?
Dremel, file, chisel?
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
Ok so to keep track of cells i've numbered about half of the cells and plugged the data into excel and calculated the distribution, which is better than i expected.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
All cells have now been tested and numbered so that their capacity can be put into a table.
Included picture is the capacity distribution of all of the cells, seems pretty ok.
One cell measured 1462mAh on first go and 1396 on second go. My brother told me that one of the slots in the Lii-500 reads about 100mAh high but i thought that it was only in the fast test mode, i'll have to investigate.
It it turns out to be true then that might have jeopardized my measurements.
Another cell that had previously measured 1433mAh after a about three cycles now measures 1729, which is above the ~100-ish mAh error that might have been on one of the ports so i think i'll have to cycle all cells at least two or three times before i re-test them all and then i'll know their true remaining capacity.
Re-testing might take about three days, but that's not including cycling the cells.
But since cycling the cells takes much longer than testing the total time of cycling and testing will be almost the same as just cycling.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
Made a spot welder but seems the one shorted cell in my lead acid battery is too high in resistance to make it work.
Whenever the welder fires there's audible bubbling coming from the battery but not enough current comes through to make the weld.
So i think i'll just stick a screw driver inside the dead cell and mash it up to short it out completely.
Other than the one bad cell the battery is brand new.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
Decided to try my other lead acid battery that was only old but no shorted cells and the welder worked fine.
That was until i plugged the charger in to top the battery off because it had been sitting for a while and all 10 mosfets popped on the first go.  :-BROKE
Made a nice weld tho  :-DD :-DD

Makes me wonder why it let go. The battery is only rated for 470A and it's very old and each of the MOSFETs were rated for 64A.
Anyways i've pulled 9 NTD70N03R MOSFETs from my parts bin, i think these came from a DELL motherboard so i might go look for more.
« Last Edit: April 23, 2021, 10:21:41 am by Refrigerator »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
Datasheet says 70 amps continuous but then you look further in and find this:
72.0A - Continuous @ TC = 25°C, Chip
62.8A - Continuous @ TC = 25°C, Limited by Package
32A    - Continuous @ TA = 25°C, Limited by Wires

Now that i've looked into it more it's a miracle these worked at all.  ;D
« Last Edit: April 23, 2021, 10:40:54 am by Refrigerator »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline f4eru

  • Super Contributor
  • ***
  • Posts: 1086
  • Country: 00
    • Chargehanger
It could be a failure related to alcaps gone bad.
Check those for ESR before replacing the MOS...

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: lt
It could be a failure related to alcaps gone bad.
Check those for ESR before replacing the MOS...
What "alcaps" ?  ???
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf