Another Question Regarding Cell Voltages

[Dr. Dickie]

Okay, I am harvesting a couple of thousand 1865 LG M26 cells. While I wait on my Megacell Chargers, I am using two Opus BT - C3100 V2.2 testers.
To speed up the process I have two MiBOXER 8-Bay Smart Chargers that I use. One is to charge 8 cells up before putting them in the Opus to be discharge tested, and the other is to charge the tested cells coming out of the Opus back up after discharge testing--that way I can test 3 times as many cells.
Unfortunately, being a noob, I was not really paying close attention to the SOC of the cells when first charging them up (I have been doing this for the last few of hundred). They are mostly between 2.8 and 3.0 V at the beginning. I assumed (wrongfully) that the charger would let me know if a cell was too low to be charged (one of my chargers did this in the past--perhaps the Litto I used early on). I have now found out that if the battery is excessively discharged, this Miboxer will revive the cell and charge it up.
This morning I was putting in some cells that had been sitting for probably a month or two since being harvested and noticed that one had only 0.4V. Yikes. I pulled it out and set it aside.
Long intro, now finally to my questions:

1) Am I correct in thinking that an 1865 cell discharged to 0.4V is trash?--certainly not a cell I would want to put into a DIY powerwall, even if after being revived it charged and discharged showing a good capacity?

2) I am assuming this cell self-discharged to the 0.4V state (the packs they came out of are 10S2P), and so when I get ready to build my pack, I will be able to identify the bad cells because they will have self-discharged to a lower voltage than all the other cells after a couple of months?

3) If the above is not true, is there a way to tell these bad cells from good ones after they have gone through a charge-discharge-charge cycle? Or do I just have to put them into my packs and wait for them to fail

Thanks for the help!!

 

[OffGridInTheCity]

I've processed 1,000(s) of cells over the years. Some batches had 0.1-0.4'ish volts - and I've been able to revive some of them, maybe 40 out of 300. In my case they were NCR18650A from old medical packs. Ideally you want to start off with a gentle charge to bring them up - here's a great youtube on this by @Daromer -https://youtu.be/n_MKDta9M_I

So yes, you can fool with this but I found that many were 'heaters' as they got above 4.1v and wouldn't charge all the way - just got hot and stayed at 4.1v.(could be this particular brand). So it was extra work / worry BUT I did get a few that are in my powerwall and seem perfectly OK so far (about 800 cycles).

In general I think the advice is to skip them unless you're desperate or just really want to fool with them. Maybe save them for non-critical projects like a flashlight?

 

[floydR]

A IR meter is useful, The one I use is an Y1030+. https://secondlifestorage.com/showthread.php?tid=9714 https://secondlifestorage.com/showthread.php?tid=6575 for reference. Most of the last 800 cells tested for IR have been 36.9- 50mOhms, I started rejecting ones over 44mOhms about 10 cells rejected so far. Many use cells with IR up to 80mOhms.
The cells are LGGBM261865,LGEBM261865 from batteryclearinghouse.com .( I know the packing could be better)
The Y1030+ is also an accurate voltage meter.
later floyd

 

[jonyjoe505]

anything below 2 volts I would not use in a powerwall. Like floyd mentioned, if some of the low voltage cells got mixed in, I would definitely get an IR tester and check all your cells. Bad cells almost always have a high IR, you want to use every tool available when reusing cells.
The low voltage cells might eventually show themselves by self discharging when fully charged.

 

[Dr. Dickie]

It is interesting I do have an IR tester. And the first thing I did was test the IR of the one or two cells that I have found with very low volts. I found the IR to be zero!! While normal is about 15-20 mOhms. I suppose the zero is due to the fact that the voltage was about 0.4 V.
But such a small sample may be a fluke. I do plan to IR test ALL cells before putting them into my packs and reject any with high impedance. I think I will try to revive one of the bad cells and then test its impedance--also let it sit and see if it self discharges down.
Thanks for all the feedback.

 

[floydR]

18650's with 15-20 mOhms readings wow, what cells are they?
I just tested several hundred LG M26 cells and the lowest was 34.5 mOhms, new ones pass quality control if they are <60 mOhms

later floyd

 

[Dr. Dickie]

18650's with 15-20 mOhms readings wow, what cells are they?
I just tested several hundred LG M26 cells and the lowest was 34.5 mOhms, new ones pass quality control if they are <60 mOhms

later floyd

They are LG M26.
Before when I tested several, they were around 40 mOhms.
i just now tested about 20 and they were all 18-25 mOhmns.
Perhaps my IR tester is not so good.

---

 

[floydR]

Most of the ones I just tested were in the 38.3-40.5 mOhm range a few higher and a few lower. I have been testing with the sliding tester I got with my Y1030+ ,and the hand held probes give me the same readings. After capacity testing the ones i checked were 35 mohms or lower.

Are your LGM26 cells the ones from the rubber covered packs? Have you found a simple way of extracting the cells from the holder? Take me a long time (1-2 hours) to remove the cells from one pack.

Most of the LG m26 cells I have are from ninebot batteries.
later floyd

 

[Wolf]

They are LG M26.
Before when I tested several, they were around 40 mOhms.
i just now tested about 20 and they were all 18-25 mOhmns.
Perhaps my IR tester is not so good.

---

What is the Y1030+?
A search for that does not return an IR or volt meter?

---

Try YR1035+ that will give you results

What IR meter are you using to get a 18-20m? reading out of a LGM26?

ICR chemistry batteries usually run in the 30-45m?category.
INR chemistry on the other hand will be in the 12-25m? catagory.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Edit:: Good ICR and INR batteries will run in that range.

Wolf

 

[Dr. Dickie]

Apparently a bad IR meter. A DIYFull. Gave me 40 or so mOhms the first time I used it, now it likes the 20mOhm range.
Going to buy a new one.

 

[floydR]

MY Bad, yes it is a YR1030+
later floyd

 

[resr]

YR1030+ uses an USB plug for connecting probes with device which is not great for measuring low resistance.

 

[Wolf]

YR1030+ uses an USB plug for connecting probes with device which is not great for measuring low resistance.

True the YR1030+ uses a USB connector and the YR1035+ uses a multi pin connector.
I own several versions of both the YR1030 and the YR1035 including several other brands with various different connectors.
The difference between the actual m? results is so small that it is of no concern. 0.1m?.
That being said I would of course get the YR1035+ or better yet the RC3563
Wolf

 

[Dr. Dickie]

Okay, I ordered a YR1030+ as well as a RC3563--of course they only come from China so a month for delivery.
Interestingly, I also have an XTAR Dragon tester, which can test for IR. I looked at about 10 cells using in, and got about 20-30 mOhms, pretty close to the same values I got from that other pos tester I have. I look forward to seeing what the new testers will give me.

The real question of all this was:
I am now tossing any cells lower than about 2.5V, BUT I do not know how many cells got "revived" by the charger I was using. My assumption was that testing the voltage after they have been setting for a month to 3 months after being capacity tested, and then measuring the IR before putting using any cells to build by powerwall should be able to detect any "bad" cells . Is that a valid assumption?
Thanks

 

[Wolf]

Okay, I ordered a YR1030+ as well as a RC3563--of course they only come from China so a month for delivery.
Interestingly, I also have an XTAR Dragon tester, which can test for IR. I looked at about 10 cells using in, and got about 20-30 mOhms, pretty close to the same values I got from that other pos tester I have. I look forward to seeing what the new testers will give me.

The real question of all this was:
I am now tossing any cells lower than about 2.5V, BUT I do not know how many cells got "revived" by the charger I was using. My assumption was that testing the voltage after they have been setting for a month to 3 months after being capacity tested, and then measuring the IR before putting using any cells to build by powerwall should be able to detect any "bad" cells . Is that a valid assumption?
Thanks

Excellent!
Especially if you ordered the RC3563 with the cell holder. Makes testing so much easier instead of struggling with the probes.


\

I also have the XTAR and it's IR testing seems to be OK although it is still basically a DC Voltage dropIR test. No that there is anything wrong with that. Just see how consistent the results are between each test.

Tossing cells <2.5V well these are my thoughts.
1. What is the cells IR? If it is within my cheat sheet parameters further investigation is required.
2. What is the condition of the other cells that came out of the pack? If they are all > 2.5 and IR is also good then maybe a bad BMS caused these cells to go low in voltage.
3. Date code. If the cell is less than 4 years old and again IR is acceptable I will mark it as LV and run a test on it. If it comes out at > 80% SOH I will give it a chance and see if it SD's over 30 days. If the cell passesit will go into the potential PW use pile. It will also be one of the first cells eliminated from my pack building as I sort through my spreadsheet filtering out the best of the best cells. I always test at least 20% more cells than the Battery requires so i can choose the best cells. The rejected ones although still very good will go into a noon critical battery of some sorts.

My first pack which was a 14s80p (1120 cells) took about 10,000 yes "ten thousand" cells to process. I have recorded over 6000 of those cells in this sheet.
https://drive.google.com/file/d/1NujY1eO6MKwGrpyEm185m6vpkMdb_Gp9/view?usp=sharing
As I started to learn more about IR, I would not even record the cells with an unacceptable IR anymore.

When it came time to build my pack I ended up filteringit down to 1200 cells to meet my criteria. ( I no longer scout for and or use laptop packs).
My cost analysis has shown it is better to buy scooter packs or medical packs with known 1st class cells. Failure rate is very low.
And yes before I commit a cell to a pack it gets a final V and IR check.

Wolf

 

[Dr. Dickie]

Thanks Wolf.
Yes, I think I got the cell holder (have to go to Aliexpress to check).
I also found out the MiBoxer I have gives internal resistance (the cells are coming in at around 40 mOhms). Of course I don't trust it, I will check when I get the better testers.
Unfortunately, I started just taking packs apart and getting the cells out--collecting them in a big box. So, for the 16 hundred or so I have I can't compare to other cells in the packs. I will for the 500 or so I have left to go through.
When I get the two Megacell charges that should be coming, I am going to go through all the cells I have capacity tested and first test the voltage (they have been anywhere from a day or two to three months). If they are down to 3 V or so, then out they go. If the voltage is good, I will test IR. If that is good, I will finally top off the voltage, then let them sit for a week, then test them two weeks later to see the drop. I figure the best way to test them is let them drop a bit for a week, then give them two weeks to see if they are self-discharging too much (toss if they are below about 4.1 V).

These cells are all from medical packs, all LG M26, for my first powerwall, I wanted to eliminate as many variables as I could (but did not know have of them?

I will start looking into the date codes as well. Haven't checked them all, but a random sampling all had an R number, so 2018.
IF, almost all my cells are withing 100 mAh of each other, would it be better to pack the packs by IR? Seems to me that would make sense.
Thanks for all the help--this place is fantastic!

 

[daromer]

Ask the question if the IR is of that importance? Are you going tha close to its limits in terms of current?
If you have that many cells i promise that randomizing it will get you close enough and if you happen to end up with all highs in 1 pack and then lows in another you should buy a lotto-ticket...


The IR is of larger significance when you push the limits.

 

[Wolf]

I will start looking into the date codes as well.
IF, almost all my cells are withing 100 mAh of each other, would it be better to pack the packs by IR? Seems to me that would make sense.
Thanks for all the help--this place is fantastic!

LG M26 date code is quite easy.
Look at the secondline there you will find a letter and 3 numbers



In this case R333 which translates to 29-Nov-2018. Q302 would be 29-Oct-2017 and S077 is18-Mar-2019 and so on
I have a spreadsheet with all the formula and results of 1593 LG M26 cells tested including the cells I chose out of the bunch which are marked in tan in the Cell Number Field.



If nothing else you should look at it and try it out.https://drive.google.com/file/d/1-N4GD-BVR1IXVzVZnD3uoVz6GCafA2zN/view?usp=sharing
There are a bunch of sheets there including a database of cells.
If your IR is close on all cells the results will be close on the mAh. I for one filtered my cells by a bunch of criteria in excel. Pretest V, mAh results, IR range > and <.
Then I ran all these cells through a repacker which randomizes all the cells into packs. So no worry about matching IR. If your cells test well at 1A you will be fine.
https://drive.google.com/file/d/1NRgUXvnwGH06Bb2oC-pgaHvQYhkJHjLR/view?usp=sharing
Have fun
Wolf

 

[Dr. Dickie]

I will start looking into the date codes as well.
IF, almost all my cells are withing 100 mAh of each other, would it be better to pack the packs by IR? Seems to me that would make sense.
Thanks for all the help--this place is fantastic!

LG M26 date code is quite easy.
Look at the secondline there you will find a letter and 3 numbers



In this case R333 which translates to 29-Nov-2018. Q302 would be 29-Oct-2017 and S077 is18-Mar-2019 and so on
I have a spreadsheet with all the formula and results of 1593 LG M26 cells tested including the cells I chose out of the bunch which are marked in tan in the Cell Number Field.



If nothing else you should look at it and try it out.https://drive.google.com/file/d/1-N4GD-BVR1IXVzVZnD3uoVz6GCafA2zN/view?usp=sharing
There are a bunch of sheets there including a database of cells.
If your IR is close on all cells the results will be close on the mAh. I for one filtered my cells by a bunch of criteria in excel. Pretest V, mAh results, IR range > and <.
Then I ran all these cells through a repacker which randomizes all the cells into packs. So no worry about matching IR. If your cells test well at 1A you will be fine.
https://drive.google.com/file/d/1NRgUXvnwGH06Bb2oC-pgaHvQYhkJHjLR/view?usp=sharing
Have fun
Wolf

Thanks. Seems these cells are all R, so 2018.
1A? I have been doing all of my capacity testing at 0.5 A. Is it that much better to do the test at 1A?
If so, looks like when I get the megacell charger I will be retesting everything

 

[daromer]

Depends on what you are after? If you dont plan to run the cells at higher than 500mA MAX per cell then you can possibly test at that. It will yield a tiny bit more capacity.
1A weeds out more cells and it also goes 2x faster

Most manufactures test capacity at rather low C rating and that often is around 0.5-0.7A. But i prefer to test at as high current as possible and the Opus do 1A. thats alot more than my cells ever will see.