Harvested cells analysis.

Wolf

Moderator
Joined
Sep 25, 2018
Messages
2,004
Alright I have recorded 2711 cells in my spreadsheet and on my way to 3000 plus. Where it will end no one knows. It may not end at all although once I get my 14s200p wall built it will probably slow down some. :D

So I have been having fun with excel and have built a "Cell Calculator" that pulls data from my main sheet and filters it according to mAh per cell Then refilters to >80% capacity.
In other words enter the minimum mAh per cell required in your pack and it will spit out the results of how many cells that have been tested that have that resultorgreater. It also calculates the average mAhs of all the cells and then can multiply that by the number of p which will give you the Ah of that pack. You can adjust the max cell V, # of strings and then calculates the kWh of the whole pack.
So I did some scenarios with 2000mAh to 2300mAh entered.
I may change my mind now and go with as low as 2100mAh cells. First of I have a lot of 2200mAh rated cells that are testing at >80% of capacity mostly much better 90%+ but that is in the cutoff area of <2200mAh.
I have also added the IR calculations as a side benefit for those that are interested.
I can say this much as my IR investigation continues, I can just about guess what the capacity of a given cell is going to be just by the initial IR measurement alone. But I digress.
Here are the results of 2000mAh to 2300mAh pack calculations. Interesting to see that between theminimum per cell from2000mAhto2300mAh the total increaseis only 1.73kWh.
So with this information would you reduce your cell requirements to 2000mAh or as I possibly plan to 2100mAh or would you hold out for the jackpot ofall at least 2200mAh? I am itching to get started on my cell packs but I also want the best possible cells. So the dilemma continues.

image_waeiri.jpg

image_gyqbfd.jpg

image_qscuur.jpg

image_uahcny.jpg

image_iuoajg.jpg

image_aohmfa.jpg

image_iwatwc.jpg


Wolf
 
Dallski said:
I would hold out for 2200mAh. If you have the patience to take apart and test 2711 cells, then you have the patience to wait to build a minimum 2200mAh wall!
He He yes patience I know. I do have that so no worries.
This question just kind of came up in my mind as I was testing lots and lots of Samsung ICR18650-22* series cells. I went through my sheet and found that a high majority of themare in the high 80s and up in capacity percentage. I still have a lot more to test in packs I haven't even broken apart yet. So these cells are relatively very good if not excellent and probably have a lot of life left. Even just dropping to 2150mAh I pick up 147 cells and at 2100mAh I pick up a whopping 285 cells. Also I only loose .29kWh and .57kWh respectively.
Do you think that percent of capacity remaining trumps tested mAh?
I mean I won't be using cells below 80% of capacity anyway so cells with high 80% and up look very attractive still limiting them to at least 2100mAh.
I certainly can use these cells in some other project.
Wolf

image_ejhsls.jpg
 
I would have to agree with Dallski. You've come this far, might as well see it through.

Love the spreadsheets.

Edit: Although if you are using cells that still have that much life, I would lean towards remaining life over actual capacity. Barely.
 
Church1182 said:
I would have to agree with Dallski. You've come this far, might as well see it through.

Love the spreadsheets.

Edit: Although if you are using cells that still have that much life, I would lean towards remaining life over actual capacity. Barely.
LOL yes I know (barely) it is a dilemma isn't it

Wolf
 
Are you taking into account tester correction factors? I know you have tested all your testers for accuracy. I'm saying that your 2200mAh cells could be 2000mAh in real life depending on the tester, as you know.

Also, if you limit yourself to 2200mah+ you probably won't be using those Samsung 22Fs or any other cells that had an initial capacity of 2200mAh or less. Those tend to be either older cells, or budget cells that might only be rated for 300 cycles, as I just recently discovered.

Of course, I'm not saying those other cells are useless, I'm just saying if you have a steady supply of cells, hold out for a more energy-dense powerwall. You can always use those sub-2200mah 80%+ cells in a parallel pack.

As for capacity vs remaining life, I would have said that remaining life would be more important. But I'm starting to think that the cells would just equalize over time.
 
Dallski said:
Are you taking into account tester correction factors? I know you have tested all your testers for accuracy. I'm saying that your 2200mAh cells could be 2000mAh in real life depending on the tester, as you know.

Well yes and no as far as tester correction factors. As you know I have done some meticulous testing of the "Testers" and am continuing to do so.
The Zanflairs are out of the picture as way to high of results constantly +8.5% to +10%.So I am left with the OPUS, LiitoKala and Foxnovo.
The OPUSes I have seem to be a max of 4% high. The LiitoKalas closer to +5% max and the Foxnovos -3.5%. I have 3 LiitoKalas,3OPUS,4 Foxnovo and 1 SkyRc. 1 LiitoKala and 1 Foxnovo are dedicated to the Sony Green Cells "Suck" experiment as well as the SkyRc every so often.

I mean it is easy enough to write a column in excel and take ~5% off of the OPUS and LiitoKala results and then sort with that.
In the meantime I will continue my quest, as you suggest, for the 2200mAh and up holy grail as I am fortunate enough, as you say, to have a steady supply of batteries.
I think I am going to purchase 4 more brand new 2200mAh cells to run through my testers and see what the results are again. It is always a good reality check to be able to say yes I do have confidence in my testers and know the approximate deviation plus or minus.
At one point in time or another there has to be a "Reference" pieceof equipment that is the standard and at this point in time it is the SkyRc. I do find it very consistent and really close to the cells (Brand new cells) advertised capacity especially if following the manufacturer's spec sheet on how they tested the battery and under what circumstances to get the results they advertise. I also just ran my 4 Samsung INR18650-25R test cellsthrough an iCharger X6 (yes 1 at a time) and the results were very good not quite -3% but very consistent.

image_aqgwdo.jpg


Wolf
 
@Dallski a 2000+mAh cell is still a good cell no matter where between 2000 and 2200 it actually situates.

Exception to this is testing high current power cells at low discharge values (Like 10,20A tested at 1A or under). If any cell has capacity under the specs at a low current test it will be unusable in a high current power tool. In this case having 2000 instead of 2200 means the cell must be no longer used for power applications.

@Wolf are you selling some of your cells or you plant to stack large quantities and used them in various projects of your own ?
 
Overmind said:
@Dallski a 2000+mAh cell is still a good cell no matter where between 2000 and 2200 it actually situates.

Exception to this is testing high current power cells at low discharge values (Like 10,20A tested at 1A or under). If any cell has capacity under the specs at a low current test it will be unusable in a high current power tool. In this case having 2000 instead of 2200 means the cell must be no longer used for power applications.

@Wolf are you selling some of your cells or you plant to stack large quantities and used them in various projects of your own ?

The chemistry of high current cells is mostly in the INR and IMR chemistrygroup. You won't find many if at all any in laptop battery packs. Power Tool packs yes.
I have 80 INR18650-13Q I pulled from a bunch of tool packsthat have very lowIR as this chemistry hasthe low resistance of manganeseand the high energy of nickel. I will more than likely build a ~12V pack with those cells and it will havegood high drain capability without much temperature.


The cells I am working with mostly areICR chemistry which most of them (again the exception proves the rule) have a manufacturer's "standard" discharge rate of ~400mAto 600 mA usually .02C of rated capacity. That is also the discharge rate that "most of these batteries" are ratedat for their mAh results.
That is not to say that a discharge rate of .5C or even 1C wont produce close to the desired results. But more heat is involved.
More than likely I will keep the INR chemistries and high drain capable cells out of my packs as it will influence the IR of the whole pack by reducing it to an artificially low level. as most of the INR cells have an IR of <30m? and the standard ICR cell chemistry usually runs around 50m? to 80m?.

As far as selling my spare cells certainly. I do have a personal pride rule though! I will give away all cells that I have that test out withless than 80% of rated capacity on the OPUS, LiitoKala, and Foxnovo. The rest with over 80% of rated capacity I will sell when the time comes. The buyer has the advantage to be able to look up the cell # on my google drive "Harvested Cell Analysis.xlsx" Excel file and under the "IR & V Before Capacity Test" Sheet find the cell and see what its history was.

Wolf
 
Wolf, if you are already tracking which tester and slot you use to get each test result, I would suggest using correction factors in an additional column to get "true capacity." That way you are not including a 2150 cell and rejecting a 2250 cell if that makes sense. Also, another option for you if you want to include cells down to 2000mAh is to increase your pack size. You can easily add 20P to your 14S 200P by including those 285 cells between 2200 and 2000mAh.

Overmind, I was just referring to the cell topic where Wolf was thinking about upping his minimum cell capacity from 2000mAh to 2200mAh. I don't disagree that a 2000mAh cell could be a good cell. Wolf has 285 cells that are between 2100mAh and 2200mAh. A few mAh here or there can make a big difference in the number of cells that make the cut.
 
You could also set up a trade as lots of users would be happy to use cells in 2000-2200 range
 
Dallski said:
Wolf, if you are already tracking which tester and slot you use to get each test result, I would suggest using correction factors in an additional column to get "true capacity." That way you are not including a 2150 cell and rejecting a 2250 cell if that makes sense. Also, another option for you if you want to include cells down to 2000mAh is to increase your pack size. You can easily add 20P to your 14S 200P by including those 285 cells between 2200 and 2000mAh.

Overmind, I was just referring to the cell topic where Wolf was thinking about upping his minimum cell capacity from 2000mAh to 2200mAh. I don't disagree that a 2000mAh cell could be a good cell. Wolf has 285 cells that are between 2100mAh and 2200mAh. A few mAh here or there can make a big difference in the number of cells that make the cut.

I am tracking which tester but not whichslot. It can be deduced on some binge charging of same cells as I always have the lowest # on the right.
So to say as an example cell 3692 would be the rightmost slot on the Foxnovo and so on following down the line. but that is not always a sure thing.
image_gvrjwh.jpg

Nevertheless I am not that concerned about slot difference as my testing has shown that to be very minimal I mean 4.86% difference between slots at most and that being the LiitoKala as the worst offender. The more I analyse these chargers the more I like my Foxnovos. :)
I will add to this chart another test again of 4all brand new Sony VT4 18650 2100mAh 22A Batteries I just ordered. I will run them through the SKYRC first and then all the others I have. (Except the Zanflair we already know about that one)recording it all. Lets see what happens with that and go from there.
Adding 20p would be an option for sure.

image_osqzyq.jpg


Wolf


Chablis_m said:
You could also set up a trade as lots of users would be happy to use cells in 2000-2200 range

Yes I certainly can but shipping is the big holdback with that.

I already gave away ~1000 sub 80% cells to someone that livesabout an hour away by car.
I will continue to do that and advertise here when I have a large enough lot.As far as the 80% and up sub 2200mAh cells those I will more than likely sell at a reasonable cost at some time in the future.

Wolf
 
@Wolf, for ICRs, all new Samsungs from 2600mAh (26F) above officially support over 1A constant discharge. Same probably goes for other manufacturers too. Older ones from all manufacturers indeed are rated for 4-600mAh.
 
Overmind said:
@Wolf, for ICRs, all new Samsungs from 2600mAh (26F) above officially support over 1A constant discharge. Same probably goes for other manufacturers too. Older ones from all manufacturers indeed are rated for 4-600mAh.

I certainly agree that these cells can discharge at up to 2.0C 5200mA but will only produce 80% of capacity.
According to the spec sheet for 100% capacity the discharge recommendation is still 0.2C 520mA.

image_xuoolz.jpg

https://www.batteryspace.com/prod-specs/ICR18650-26F.pdf

image_whyzsf.jpg


Wolf
 
Good observation. It meas a 1A discharge would need a correction factor since that will measure 97% of capacity.
 
Wolf said:
I will add to this chart another test again of 4all brand new Sony VT4 18650 2100mAh 22A Batteries I just ordered. I will run them through the SKYRC first and then all the others I have.

The Sony US18650VT4 cells have arrived and are now in the SKYRC running their first test.

image_skpion.jpg

They arrived with a V of 3.57 each and 19.8m? / 20.0m?
Thestandard charge rate is 2000mA and thestandard discharge rate is 420mA which is what the SKYRC is doing.
They will then go to each one of my testers and be tested at 500mA C/D.
I will post the results when done.

Wolf
 
Hi Wolf,

I hope you don't mind this copie and question, cous i think, i maybe (don't) understand, you wrote:

The cells I am working with mostly are ICR chemistry which most of them (again the exception proves the rule) have a manufacturer's "standard" discharge rate of ~400mA to 600 mA usually .02C of rated capacity. That is also the discharge rate that "most of these batteries" are rated at for their mAh results.
That is not to say that a discharge rate of .5C or even 1C wont produce close to the desired results. But more heat is involved.
More than likely I will keep the INR chemistries and high drain capable cells out of my packs as it will influence the IR of the whole pack by reducing it to an artificially low level. as most of the INR cells have an IR of <30m? and the standard ICR cell chemistry usually runs around 50m? to 80m?.

If i understand this correctly, it is a bad idea to mix laptop cells and powertool/ebike cells together in one pack.
Because a laptop cell has an higher IR and "slows down" the powertool cells.
One ?pack? is 3,7 volts and depending on the amount of cells in the pack, 100 in my case would be 220A
So its better to form a pack alone with laptop cells or alone with powertool cells?
And you can mix different packs to a battery/accu (in my case) 24v? or not, and keep each battery from laptop or from powertool?

Thanks in advance, best
 
100kwh-hunter said:
If i understand this correctly, it is a bad idea to mix laptop cells and powertool/ebike cells together in one pack.
Because a laptop cell has an higher IR and "slows down" the powertool cells.
One ?pack? is 3,7 volts and depending on the amount of cells in the pack, 100 in my case would be 220A
So its better to form a pack alone with laptop cells or alone with powertool cells?
And you can mix different packs to a battery/accu (in my case) 24v? or not, and keep each battery from laptop or from powertool?

Thanks in advance, best
100kWh-hunter,
Good question to which I do not know the complete answer to.
The only thing that I have to go on is my understanding of electrical flow and that is electrons will always choose the path of least resistance.
What does that mean in a pack of 100p? I don't know. Logicsays that the lower the resistance the more current will flow through it. If you have cells with a delta of say 50m? 50 cells at30m? and 50 cells at80m? what does that mean electrically? I know that the combined resistance of the pack is going to be 1.1m? (30m?/50=0.6) + (80m?/50 =1.6) / 2 =1.1 how does that affect each individual cell?
I do know that manufactures put the same cells into their packs for a reason and that is that the cells all have the same characteristics. As in mAh, IR and chemistry. So as far as I am concerned we are in somewhat uncharted territory but then again that's the fun of it. I do know that if I put 2 resistors in parallel and one is 3? and the other is 8? more current will flow through the 3? resistor. Electrically the parallel pair of resistors equal 5.5?. Physically they are different.
Most powertool (but not all) are generally lower capacity (as in 2200mAh or below) but very high drain. which requires them to have a different chemistry ( usually IMR and INR)compared to laptop batteries which tend to be higher capacity 2200mAh and up but low drain( usually ICR).
My thinking is not to mix them but don't take my word for it. As for me the "powertool batteries" that I get I will use in separate packs as in 12V battery packs just to have and to hold. You never know when you need to jumpstart a car. :D Till someone convinces me otherwise that will be my plan of attack.

Wolf
 
@ wolf,
Thank you for you replay, your resistance story sound very familiar and thus logical.
But i wont settle with: storage is storage, just like you.
The last pack,(giving by the amount of cells i have harvested now) will be from power tools, cous for the electric flow.
Its 1/3 laptop and 2/3 powertool/ebike.
I think i will end up this first experiment with 1/15 laptop and 14/15 powertool/ebike cells, so no need yet for laptop cells.

The chemistry story i will have to dig in further, when the time is there

Speaking of to jumpstart the car....that buzzer that is warning you, that you left your lights on, is broke, stopt working two months ago.
I had to jump start several time's, i am not kidding!


Almost forgot, sorry.

How much do we actually drain from each cell? One battery containing 7 packs and every pack a 100 cells and must feed 600w an hour at 230v.
Or we built it up, and go to a battery of 21 pack(3x7), each pack contain 400 cells, and it feeds the same 600watt system at 230v.
Btw both examples are in 24v battery configuration.

Or you need to have big power for a short while(one hour), like for the oven, and you have small capacity.
I think: where and what are youre demands, so that is how you are going to assamble your battery's
Until you have enough ofcourse

However i think the high drain cells must be in one pack and at the and of the electric line.
In other words: the laptop cells must be the first one
Correct me if i am wrong: current flows from - to +

My classes where 30 years ago, have to refresh a lot, i did not do a lot with that knowledge in the meantime.

Most of the writing in this post was thinking out loud, hoping for answers,

thanks in advance
 
Back
Top