Green Sony cells just suck

[BlueSwordM]

It takes a bit of time and more cycles for internal resistance to fall off a cliff.

 

[Overmind]

Interesting find. It kind of matches my own results. Let me tell you my experience with green Sony cells.

I have extracted hundreds of them from notebook batteries, capacity ranging from 2200 to 2600mAh. That was about 5 years ago. At that point, I did not had the possibility to test for capacity. during the years, I sold part of the cells and everyone no exception was happy with their performance in various applications.
Some even called after 2 years to tell me they want more because even after that time the cells work perfectly. Most sold cells were 2200-2400 with less 2600mAh ones.

Quite recently, I decided to give the remaining 94 cells a test. So I took 2 Opuses 3100 and started testing. To my surprise most of the 2600 cells were showing 100-200mAh, which is very very strange.

Here's the centralized result of the testing @1A discharge (after stored 5 years at full load):

Rate___Brand___Series__#___Measured
2600____SNY____8A____01____500
2600____SNY____8A____18____100
2400____SNY____7C____10____200
2400____SNY____G7____01____1500
2400____SNY____G7____02____1000
2400____SNY____G7____04____500+
2400____SNY____G7____06____300
2200____SNY____G5____02____1500+
2200____SNY____G5____22____1000+
2200____SNY____G5____04____400
2000____SNY____4E____02____1300

Basically, only the 2200mAh ones have some life left.

The following ones are a fresh rip-off some batteries:
Rate___Brand___Series__#___Measured
2100____SNY___6A/B___2____2000+
2100____SNY___6A/B___6____1500+
2100____SNY___6A/B___4____1000+
2100____SNY___6A/B___6____500+

By looking like this I can conclude the following:

1. Quality of green Sony cells decreased a lot. The 2200 ones are still working a lot better than newer like 2400 and 2600 ones. Larger capacities are just not that durable.

2. Discharge test is not 100% relevant because I could not set the 440mAh discharge rate in the cell's original specification. The measure results seems to conflict with the feedback from clients.

3. From various other brands stored on the same period only the Sonys 2400+, Panasonics (2250,2350 green, ugly purple) and LGs (blue)of 2200mAh presented the lowest capacities recorded while the pink 2600 Samsungs top the capacity table, followed by the 2200mAh ones.

If anyone is interested, I'll post a picture with the complete table of all brands, since I recently tested over 600 cells in total (power tools brands included).

 

[Wolf]

It takes a bit of time and more cycles for internal resistance to fall off a cliff.

Fall of a cliff? Sorry I'm confused
I am assuming you mean IR rise and capacity falls of the cliff.
I have these cells scheduled for a 8 cycle test and will check IR after that.
How many cycles should I do?

Wolf

---

Interesting find. It kind of matches my own results. Let me tell you my experience with green Sony cells.
If anyone is interested, I'll post a picture with the complete table of all brands, since I recently tested over 600 cells in total (power tools brands included).

That would be great if you could.

The question arises thought of the cells that were sitting for 5 years what V where they hibernating at?
What was the V when returning to be tested? And "I know I know" what was the IR of the cells?
Not with an OPUS quick test but a proper IR test.

1604 cells of all kinds tested and counting. (Actually a lot more than that but these are the ones I have recorded so far)

Wolf

 

[BlueSwordM]

I'd say 10-15 cycles at max.

Usually, if nothing changes, that usually means the cells have recovered well. or weren't damaged much.

 

[Wolf]

I'd say 10-15 cycles at max.

OK sounds great I got the sheet formatted for 8 right now and I can of course add another bunch of columns.

Results as of this evening.




Wolf

 

[Overmind]

@Wolf they were stored at full charge. Most of them had above 4.10 V when rested for capacity, but they were recharged at least once per year.

 

[Wolf]

Update on the "green" Sony Cells.

Here is the latest info.
The 0V cells where slow charged with CC/CV for 48 hrs and never recovered they got up to about 1V and as soon as they were removed their V dropped down to next to nothing. I determined these were dead.

The rest of the info is in the chart.
So far the initial test showed good capacity and then a slight fade. The >3V cells look like they are mixed but not bad as far as closeness in mAhs is concerned.
Anyway more results will becoming this week.
Wolf

 

[DCkiwi]

@Wolf they were stored at full charge. Most of them had above 4.10 V when rested for capacity, but they were recharged at least once per year.

*ouch*. thats what I used to do too, then what happened to you happened to me. i found a couple much by accident. then decided it would be prudent to retest them all again. (and again).

so now I store all at ~ 3.92V ... after triple testing for capacity. overkill? perhaps. but the results speak for themselves. all my tested cells are now stable and have reliable capacity gradings.

The 0V cells where slow charged with CC/CV for 48 hrs and never recovered they got up to about 1V and as soon as they were removed their V dropped down to next to nothing. I determined these were dead.

that was most likely a problem with unsupervised use of the 4p holder, or more specifically not removing the failed cell/s to let the rest continue. my method: connect CC/CV psu, and switch on. then one by one add cells, watch the precise voltage of the terminals. if it stops rising, then remove last cell. EG i do NOT just slam in (160) cells and press go. that wont work. not at all. they should be up above 2.0V within ... an hour really. if not up to 3.0V in a few hours then something is definitely wrong. it is key that vigilence is practiced on recovering 0V cells, especially the first bit, to make sure they are all playing ball. apologies for not being more detailed in my proposed testing method Wolf. dammit.

FYI also, I have found a thermal imager to be extremely useful in detecting 'heaters' super early. so early that I cant even feel them getting warm. shows up like a neon light on the imager. Thoroughly useful piece of kit for those -- like me! -- that insist on recovering 0V cells.

of course ... all four might be bad, but I would happily bet against that. loving your data Wolf... legend!

 

[Wolf]

that was most likely a problem with unsupervised ..........or more specifically not removing the failed cell/s to let the rest continue............

FYI also, I have found a thermal imager to be extremely...................

of course ... all four might be bad, but I would happily bet against that. loving your data Wolf... legend!

So they where not quite unsupervised. I have experienced that before and have pulled cells out of my 20Pcell charger/discharger board hooked up to my 4P
so 24P in total and have found the one or two that drain to whole pack. I can seperate the 2 and discharge 20 cells at once for safety and storage.
Anyway for these I was just using it in 4P mode Ok I did it kind of different but similar and from now on I will insert 1 cell at a time and see what happens with V.
After about 5 hrs took all the cells out and tested for V a couple had some V .005so those I put back in and let them sit for the remainder but no difference even lowered the mA for 50mA per cell.

Thermal Imager will beone of my next acquisitions. I do have access to a very high resolutionSnap-on Diagnostic Thermal Imager EETH300 $650.00
That I can use anytime and I will have it on hand for the next experiment.
I am doing these experiment for the fun of it and for educational purposes and possibly help some of us here out.
I have a great supply of cells so I really don't need to mess around with 0 V cells but hey what else am I going to do Just charge and test >3V cells?
Naw I want to see how far we can stretch the limits of reclaimed cells and when to give up and when to try to recover.
So keep the challenges coming and I will continue to rise to it. As I charge and test my "good" cells on the remaining testers.
904 cells >2200 so far another 2100 cells to go.


Wolf

 

[Wolf]

Latest update as I continue to document the "green sony phenomenon"

Wolf

 

[Chablism]

Those clusters of 4 reds are strange

When 1 cell goes red so do the other three in the tester

Any idea why?

 

[Wolf]

Those clusters of 4 reds are strange

When 1 cell goes red so do the other three in the tester

Any idea why?

Not really sure. Could be temp related but house temps don't really change much.The cat could be sleeping on them or the dog panting waiting for the results.
No just kidding but no I'm not sure but here are the latest results in this ongoing investigation.
It may be that the cells are slowly recovering due to the charge and discharge cycles. Like it's being exercised.
I also have changed from immediately retesting the same cells in the same charger to letting them rest a day or so between testing. Always putting them back in the same charger of course. Got to keep things scientific.
That could be it. They may be finding their equilibrium and finally settling down to their normal levels.
I should have all of them at Test 8 this weekend and will continue for at least 10 tests if not 15 tests. Don't want to go too much more as they are all within my 2200mAh range and I want to use them not abuse them.

Wolf

 

[Wolf]

Latest data in the "Green Sony Cell Debate"

I don't know but they look relatively consistent too me. Variances of ~200mAhs but not per say a steady decline.
So I am giving the cells a rest for a couple of days. I need the 12 slots for a couple of days to catch up on my cells waiting to be tested.
Took a V and IR reading after the 8th test just to check.
Did a Min/Max on the results so far so you be the judge. I will go to test 11 unless you guys want to go a full 15.
I still have 37 cells I havent tested yet so I want to get started on those.

Here's the sheet.




Wolf

PS:Built a Cell Testing launch platform this weekend.0V to 1V on the left 1V to 3.3V in the middle and 3.4V to 4.2 on the right.

 

[Wolf]

UPDATE:
Here is the latest result from the cycling test of the "GREEN" Sony cells.
Conditional formatting is for highest mAh is green and lowest is red. Just because its red does not mean it is "bad" just the lowest mAh.
I would have no problem using them in my powerwall.
Tell me what you think. I just don't see a loss of 100mAh per cycle consistently.
Should I do more cycles or is 11 enough?
Got another 34 cells to go we will see.





Wolf

 

[Wolf]

Ok so here are the results of 24 Sony US18650GR G8 (except for 10)all of these cells where below 2.5V andeach tested 11 times.
For you math wizards thats 264 individual tests.
I have 32 more cells each indifferent stages of the test. As soon as they are done I will post the results.
Procedure,
Initially I tested the same 4 cells in rapid succession on the same tester giving the cell no time to rest.
I did notice that temperature was not decreasing and the cells where no longer being tested within the manufacturer ~23C recommendation.("room temperature")

I changed that and allowed the cell to rest one day after charging and then continued the test cycling between sets of cells but always to the same charger and the same slot. The temp issue seems to show up as the sharp dip in the chart.
Initial mAh test shows a good value and subsequent tests seem to indicate some trending signs to a lowering of capacity on some cells but not all.
At least not the ~100mAh or so per cycle as discussed in this thread.
Also I have changed to charging all the cells initially with a CC/CV system at 4.2V and 50mA per cell.
After their initial test they are charged with any charger that is available at 500mA per cell.
Please keep in mind that this is a rescue a low V cell experiment and not on Cells that were pulled out of a pack with 3.8V+ in them.
So without further adieu here are the test results so far.
Your comments are welcome and asked for. Please let me know as to anything you would like to see changed or kept the same for the remaining 32 Cells






Wolf

 

[Wolf]

Update:
Here are the results so far got a few more to go and then we will have the complete results and I can go back to my life!
Oh wait this is my life what am I saying.
Questions or comments? Would you use them in your powerwall?
As far as rescued cell with less than 3V and some at below 2 I don't think they are that bad.
I have attached the Excel sheet just in case its too difficult to see the picture as I had to cut and paste to get it all in one.
Wolf




Edit:
Ah picture sucks here are the 2 that were combined.







Much better

Sony Green.xlsx

Sony Green Cell Number,Number by Voltage,Part Number,Manufacturer,Format,Wrap Color,Insulator Color,Original Capacity,Initial V,Initial IR,Charging V,Initial Charging A,Initial Charger used,End V,End IR,Capacity Tester used,Capacity Results Test 1,Capacity Results Test 2,Capacity Results Test 3,...

docs.google.com

 

[Wolf]

OK final update the Sony green cell suck analysis has been completed.

These cells where tested from a low Vof .918 to a high V of 3.73.
Charging was done using various methods as listed in the chart.
Cell # 44 to 46 were beaten up with the SKYRC tester with various charging parameters mostly mA charging cut off and mA discharging cutoffs.
I overcooked them on test 9 and exceeded the cells recommended cutoff mA charge and discharge specs and it shows in the following tests.
The SKYRC was set to cut off everything if temperature went above 45 C but it never did. Got close but not close enough.

My conclusion is that they are not bad cells.
One thing of note is that the US18650GR G8 has a high IR from the get go and is no indication of a bad cell.
One oddity I did find is that after a nice slow wake up charge the IR dropped pretty dramatically an anomaly I have not found in other manufactures cells.
If I go with the standard IR readings of most other cells (My own conclusions coming in a report soon) I would bin them for sure. On the other hand I don't think I will or would use them in a powerwall with other cells. I just think the IR difference is too much to pair them with sub 75 mΩ cells.
As a group of just sony cells I would have absolutely no problem with them. If I had 3000 of them I would build them into a powerwall without a worry.
As 20 or 30 cells in a diversified cell combination of 200 I am not so sure I would do that.
One more note the cells do like a small rest between charges it seems that if I let the cells rest for 8 to 12 hrs and then discharge them their consistency seems to be better.
The Min Max of these cells through 11 cycles was at the highest 12.8% and at the lowest 1.18% (other than the cell I beat up )
so not so bad and they certainly did not lose 100mAh after each test.
So here is the chart and as always like it, leave it, or lump it. comments always welcome and required.
If you want the sheet uploaded in xlsx let me know.
Wolf

 

[Oz18650]

Thanks for all the effort in testing and sharing.
While i have not used the info much yet, I am sure it will be very interesting and useful once my testing and data collection gets under way.
I am sure your findings week be useful to many other people too.
My initial plan was to dispose of all low voltage cells. This had been revised as a result of discussions about how low voltage cells can be recovered.
I am sure many other people week have gained a lot from this testing, data collection and sharing too

 

[100kwh-hunter]

Tested also some sony, no problem found.
Only the ir puzzles me, it is not consistent. :huh:
The ir test was every time on the same voltige, exact at 3.45
Even some cells improved in soh and capacity, but that is not strange.
Ir on some cells was going down and on some cells was going up? :huh:

Those were my tests, nothing or no one can beat The Wolf!

Thanks for the head up, a warend person counts for two!

---

sorry, the v3, v6 and the nc1 10 of each, 5 tests on each

 

[kc8adu]

nice demo.
if there are any particular cells with a high percentage of duds it would be older red sanyo's not sonys.
this is my experience.