Processing Low V Cells

Announcement - Help us fight the BOTS! Please report all spam including stuff in your inbox!

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,610
I started this to show how to process Low V cells. At least how I do it and have had a reasonablegood success rate.

As you all know we run into those packs that just have a bunch of < 1V cells. What to do?
First off we know the cell is mechanically OK as there is some V so no CID trip or PTC issues.
At <1 V there is not much charge to worry about so I line them up in my 20P cell holder.
These are Sanyo UR18650FM all of them were <1V

image_tjcabi.jpg

I have a CC/CV BuckPower Supply hooked up to my modified Server Power Supply for 12V input andI set my CC/CV Buck to 4.2V and 1A.
That equates to 50mA per cell.
This picture shows some of the Sony "greens" being charged US18650GR G8 Theones that are inSony green cells suck thread

image_whujtt.jpg


Next I watch the Vbuild up and check the cell temperatures by hand.
(Thermal Imaging camera coming today If you want one that is a good unit and at a great price I got this one.https://perfectprime.com/collections/thermal-imaging-camera/products/ir0018)
After about 1 to 2 hrs you will notice some cells getting a little warmer than the others. I will have Thermal images in the future but for now I use a thermalgun and most cells will run around 22C to 25C. Then you get to the 27C and up cells. At this point you will notice on the DMM that the V has somewhat stalled at a certain V and doesn'tseem to rise anymore.
That means the high temp cells are absorbing all the A and are sucking the V down.
So I removed the "hot cells" and immediatelythe V started to rise.You can view the video here(Too big to post here)
Now you will see that the rise is inthe thousands of a volt so it would not show up on a regular DMM.
As the offending cells have been pulled out the remaining ones are now able to absorb a charge. What is to note though is that the IR of the "hot"cellswasn't bad at least not in the ionosphere.

I will make more detailed pictures and possibly more short videos of this process. as it continues.

Hope this explains some things.

Wolf
 

Generic

Member
Joined
Oct 23, 2018
Messages
217
image_grkrhx.jpg


^^ I do something similar - but instead of 20P, I do 4P with a TP4056. Ends up being about 25mA/cell until 2.93V.

I've had very little success with <1V cells, though. A lot of them will test very well initially, but most end up being self-dischargers and a lot will lose a lot of capacity on a second or third test. The 10% that were found at less than 1V, don't self discharge, and holda decent capacity must not have been at that voltage very long, but I have no way of knowing that. Also, brand seems to make a difference as the Sanyos are particularly bad at less than 1V.
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,610
Good Idea, In my case the 20p board serves as a discharger also.
I just have a rash of UR18650FM cells and a lot of them are below 1V so I decided to experiment.
We will see what happens.
So far the results aren't stellar

image_coxqmf.jpg

I will find out what their reading is after 30 days.
I may set some aside and do a series of C/D cycles on them to see what their longevity is.
On another note in the Sony Green cells suck threadthose cells seem to have come around.
Wolf
 

Generic

Member
Joined
Oct 23, 2018
Messages
217
I know people will ignore me as I ignored others' warning: Testing low voltage cells is only useful if you can't get your hands on cells very often. I'm bulk checking voltages of cells I tested 3 months ago, most are very good. I got to a pack that had low initial voltages, but tested with decent capacity. These are the Sanyo UR18650ZT rated at 2800mAh new. Well, 5 out of 6 have self-discharged significantly, while 1 is ok. I'm going to re-test the one non-self-discharger and see what capacity it has now. I won't be surprised if its lost a lot of capacity.


image_bjdiic.jpg


With my generics, I'm just putting my <1V cells in a box by themselves. Not going to bother testing <1V cells anymore.Easier to take to the recycler since I don't have to worry about discharging the cells!
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,610
These cells were somewhat low V cells all in the ?2Vbut<3V. Initial IR test showed good promise so all got charged at CC/CV 50mA, 4.2V.
All got stuffed into my three testers.
Results are pretty enlightening. All above 90% and pretty even across the board on all testers.
Kinda says that all my testers are really close to each other as far as their results in mAh is concerned.
Nice to see in living color. :D
Wolf

image_iauqss.jpg
 

Cherry67

Member
Joined
May 13, 2018
Messages
496
Dallski said:
I know people will ignore me as I ignored others' warning: Testing low voltage cells is only useful if you can't get your hands on cells very often. I'm bulk checking voltages of cells I tested 3 months ago, most are very good. I got to a pack that had low initial voltages, but tested with decent capacity. These are the Sanyo UR18650ZT rated at 2800mAh new. Well, 5 out of 6 have self-discharged significantly, while 1 is ok. I'm going to re-test the one non-self-discharger and see what capacity it has now. I won't be surprised if its lost a lot of capacity.


image_bjdiic.jpg


With my generics, I'm just putting my <1V cells in a box by themselves. Not going to bother testing <1V cells anymore.Easier to take to the recycler since I don't have to worry about discharging the cells!

I can report some strange positive things about Low-V cells, See my lucky thread, pls.
https://secondlifestorage.com/t-I-thought-i-was-lucky-when?pid=50123#pid50123


In this case Bike blocks, It may be a special case.
 

Generic

Member
Joined
Oct 23, 2018
Messages
217
Hi Cherry67, I just read your lucky thread. I've got my own research. And I've read a lot of Wolf's research on this topic, as well. I believe we can all agree that the proper way to recover cells that have less than 1 Volt at the time of discoveryis to charge them with as small a current as possible for the entire duration of their first recovery charge.That will not recover all cells, as some are permanently damaged from being in alow voltage state for too long. Just yesterday I was "trickle" charging fourLG cells from their original voltage up to just 3.5V at 25mA per cell,and I noticed the 0.8V cells held their voltage, but the ones that were at only 0.05V were already self discharging after just 24 hours. You cannot recover all cells, you can just have a better chance of recovering the ones that are not super damaged.
 

Cherry67

Member
Joined
May 13, 2018
Messages
496
Dallski, i agree to most what you say, but what i do is still slightly different.
I use much lower currents now for <3 Volt, and moderate charging to full then. 100 - 300 mA.
But, i dont believe that ist the last word in this issue.
What i have is 3 blocks 10s6p coming Up from 0.6, 0.8 and 1.1 V/cell, each Block 100 % successful and total in synch. That ist so totally different from the usual 70 % success rate that it triggers my interest.
1 Block would be Chance, 2 ist Strange, but 3 ist an indicator there ist Something in it.

I'll Just want to know WHAT.
There are so many possible Parameters...

Sorry i will Update later.
Patience typing mobile is Up.
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,610
Dallski said:
................... Just yesterday I was "trickle" charging fourLG cells from their original voltage up to just 3.5V at 25mA per cell,and I noticed the 0.8V cells held their voltage, but the ones that were at only 0.05V were already self discharging after just 24 hours. You cannot recover all cells, you can just have a better chance of recovering the ones that are not super damaged.

@Dallski

That is correct you cannot recover all cells BUT if you take an IR reading of the cell (no matter what the voltage) you will find that the ones that have a low IR reading your chances are improved dramatically.
I just charged a lot of LGABB41865 9of them had a V of <2V some of them below 1V using the 4.2V 50mA per cell CC/CV.
The ones with the V in yellow where those cells . It took 3 days for them to come up to 4.18V.
The other 6came out of the pack with 4.05V which went into the charger/testers right away.

The IR was perfect on the cells,no heaters, and the results were very good.
So just another feather in the IR cap.
Wolf

image_ogerjp.jpg
 

Generic

Member
Joined
Oct 23, 2018
Messages
217
Wolf, you are probably right. If I had more used cells, I probably would invest in a proper IR tester. It all comes back to IR, with a few exceptions as you have found.

Cherry, the reason I keep currents low for the whole first charge is because I think it prevents plating. I dont remember the specifics, but i believe high currents on low voltage cells produces plating on the cathode or anode of the cell. While it doesnt seem to affect the first test after charging, it does seem to cause the cell to self-discharge over time. I dont have the hard data on that, though, I have just had a lot of bad luck with <1V cells and I think it's because I was just throwing them in the TP4056 and charging them at 1A after they got to 2.92V. I lost a lot of cells that tested well initially. But then again, they could have been high IR cells to begin with, so who knows?
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,610
Dallski said:
Wolf, you are probably right. If I had more used cells, I probably would invest in a proper IR tester. It all comes back to IR, with a few exceptions as you have found.


I have just had a lot of bad luck with <1V cells and I think it's because I was just throwing them in the TP4056 and charging them at 1A after they got to 2.92V. I lost a lot of cells that tested well initially. But then again, they could have been high IR cells to begin with, so who knows?

@Dallski

What do you mean probably right? You now I'm right, probably. :p

Anyway I did the same thing as you initially. I just threw the cells into the TP4056 board that I built and called it good. Then the sadness came when the cells had no mAhs in them. The bad bucket was growing faster than the good bucket. It was just such a waste of time to charge these cells even with good voltage and have such a low return. There had to be a better way. There is of course and we all know what it is.

So my first giveaway was ~1000 cells that I tested and and found to be lacking.

image_zntonm.jpg


My second giveaway consists of ~3000 cells maybe more. Another ~1000 or so cells cells that I tested and found lacking but the rest of the ~2000 cells never made it to the chargers the IR was to high. So I saved myself countless hours of testing cells just to be dissapointed.

image_bpzxgg.jpg

image_zaavhy.jpgimage_gsyhqx.jpg

image_yqvrsl.jpg


Now that I have my 2800 + cells I can concentrate a little more on experimenting with cell charging methods and so forth.

Amazing how fast you can accumulate good cells when you place 44 cells into your chargers and pull out at least 42 to 43 good cells if not 44 at a time. :D

Wolf
 

Cherry67

Member
Joined
May 13, 2018
Messages
496
Wolf said:
- snip -
Amazing how fast you can accumulate good cells when you place 44 cells into your chargers and pull out at least 42 to 43 good cells if not 44 at a time. :D

Wolf

In effect you have made my pipedream come true. Exactly true. Not only true, you PROVED it.
Chapeau !

And now i do want to know, why do i have a 100 % successrate on reviving 180 low-U cells?

And others have not ?

I will sacrifice the next low-U block i get to a binary parameter check. Half of the cells with a minimal current, the other half with a binary increase.
Any other ideas or wishes ?
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,610
Cherry67 said:
In effect you have made my pipedream come true. Exactly true. Not only true, you PROVED it.
Chapeau !
You are welcome and thank you
Ah yes it only took 1500+ cells to slowly come to the conclusion how relevant IR is to asuccessful harvesting strategy.
Not only does it virtually eliminate heaters, SDs and low performing cells, you will be confident that your powerwall cells will have a reasonable balance with IR andacceptable SOH cells.
As of now anybody that does not believe in the results is just a denier of the facts and just wants to spend more time testing cells that wont do anything but frustrate.

And now i do want to know, why do i have a 100 % success rate on reviving 180 low-U cells?
And others have not ?

IR my friend IR. If the cell has an acceptable IR it is a great candidate for recovery.
One thing you cannot do though is take a ~1V cell and stuff it into a TP4056 and expect it to survive good IR or not.
Now that I have my allotment of cells I will join you in experimenting with very low V cells. Just because I am just as curious as you.
I have recovered cells as low as .0852Vthat after a CC/CV 4.18 @50mA charge was at a 100% SOH as itwent through 3 normal 500mAh CDC cycles and had absolutely no degradation. I still have to give them the "30"day sit test to see if they will SD but all indications are good that they will not.
At present I have some cells again <2V and some <1V charging at 4.2 and this time 25mA each it will probably take 4 to 5 days for them to reach 4.18V then we will see. But again the IR is acceptable on these cells so I expect them to be fine.

I will sacrifice the next low-U block i get to a binary parameter check. Half of the cells with a minimal current, the other half with a binary increase.
Any other ideas or wishes ?

Poor cells to take such an abuse but its all in the name of the harvest :D

Wolf
 

Overmind

Member
Joined
Jan 16, 2019
Messages
451
Wolf, you have a lot of experience with IR.

From your observations, is it possible for IR to become better after you do recovery attempts on the cells ?
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,610
Overmind said:
From your observations, is it possible for IR to become better after you do recovery attempts on the cells ?

Yes I have noticed that on occasion. Mostly with Sony and some models of LG cells.
I have had IR changes as high as -70% after charging and testing.
Now mind you with my harvesting strategy now these cells would never make it into the testing stage. The IR is too high.
I would not have these cells in my packs as they are sure to cause trouble even though they show a good result.

image_srvgrr.jpg

I would absolutely not use theses cells.

Now these cells are in the -15% to -11%.

image_zvnlxr.jpg

Some of these cells I would use as long as the Voltage is OK after the 30+day test and the IR is ?75m?
The Sony GR G* series will not be used at all no matter how good they recover. I find that their IR swings wildly between charge and discharge.
One of my projects is to build a single cell arduino based charger/tester with 1kHz AC IR monitoring during all the stages of recovery to find out what is actually going on there.

The cells that I have documented with a yellow font on the Voltage pretest are the ones that have been recovered from <2V but they are just now coming up on the 30 day retest and I am a little bit behind on that as I am doing April cells right now.

image_iqbmpx.jpg

Also the IR was already acceptable and if the IR is on the margin it isindicated by the red font on the IR pretest.

image_hmkpap.jpg


OK enough screen shots you can always check the Excel file out on your own and sort by whatever criteria you so desire. With a cell count of 5855 and constantly growing there is a lot of information there.

Wolf
 

krobertson

New member
Joined
Jun 28, 2019
Messages
9
What is your process for testing IR? Ive read that the Opus tends to be quite off with its IR numbers.

Since Im just at the beginning of my cell harvesting, I was thinking it be good to set a start out trying to collect it rather than going back and doing it later.
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,610
krobertson said:
What is your process for testing IR? Ive read that the Opus tends to be quite off with its IR numbers.

I use aYR1030. It is a 4wire kelvin 1kHz AC IR tester.
You can get it herehttps://www.aliexpress.com/item/32817594188.html?spm=a2g0s.9042311.0.0.5ca84c4dqsxf2D
This is the best one I found. But you can get any one of different brands as long as they say YR1030.

All comercial Charger/Discharger/Testers are very poor on IR as they only do DC IR instead of AC IR. (AC IR is what the manufactures spec sheets show when available.)Also the contacts cause a varying measurement on a consistent basis so they cannot be trusted.

Wolf
 

Bubba

Member
Joined
May 9, 2018
Messages
300
Wolf said:
krobertson said:
What is your process for testing IR? Ive read that the Opus tends to be quite off with its IR numbers.

I use aYR1030. It is a 4wire kelvin 1kHz AC IR tester.
You can get it herehttps://www.aliexpress.com/item/32817594188.html?spm=a2g0s.9042311.0.0.5ca84c4dqsxf2D
This is the best one I found. But you can get any one of different brands as long as they say YR1030.

All comercial Charger/Discharger/Testers are very poor on IR as they only do DC IR instead of AC IR. (AC IR is what the manufactures spec sheets show when available.)Also the contacts cause a varying measurement on a consistent basis so they cannot be trusted.

Wolf

Are there any versions of these testers that can connect to a network either through WiFi or Ethernet or have USB PC connection?
Then you could barcode and read the cell for a more automated entry.
I have looked and not found the information.
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,610
Bubba said:
Are there any versions of these testers that can connect to a network either through WiFi or Ethernet or have USB PC connection?
Then you could barcode and read the cell for a more automated entry.
I have looked and not found the information.

I bought this Unithttps://www.aliexpress.com/snapshot...orderId=100049778505788&productId=32982590554
and it has a serial communication protocol setting screen

image_nstwau.jpg

so I questioned the seller about it and here was the response.

image_yztsnm.jpg

I guess they are working on it but when it will function.....No Idea.

Wolf
 

Bubba

Member
Joined
May 9, 2018
Messages
300
I was looking into the AI features Microsoft is building into Office.
Here are my plans.

1. Have a unique number on the charger/tester read.
2. Have a it read the number on the slot. (if slot is indicating complete... if not don't get values)
3. Have it read the value directly from the screen
4. Read the Cell number
5. Directly enter it into my database... with minimal interaction.

Currently Microsoft is only released for Cellphones. Waiting for PC use.

Of coarse I could be dreaming..... What do you think?
 
Top