Battery sorting

humbleONE

New member
Joined
Jul 22, 2020
Messages
9
After spending many months testing around 1,000 batteries, I sorted them into mAh test results, from 2400+ right down to "1011 (the crappy batteries from tool packs.)" The lower mAh batteries I'll use in torches toys whatever, midrange place back into 2 laptops and 2 scooters I have and the rest for powerwall use. I am now checking the voltages after they've been sitting for several months and also internal resistance. Question is, how should I sort for optimum safety and power in powerwall use. My thoughts are sorting to 'those that retained perfect voltage' after sitting and then sorted again 'by resistance' thinking that maybe the mAh doesn't matter so much although I'm keeping them in the original test batches.... ie: I checked first (at random) all the 1800-1899 mAh lot and then sorted by retained voltage (some I'll use lost .01v and some up to .03v) and those that lost .2v or more I moved to another batch I'll use for other less critical stuff. I'm guessing I'll get less "heaters" and it'll prove safer sorting this way for a powerwall application, any thoughts Pete or anyone else on this.
 

ajw22

Well-known member
Joined
Nov 16, 2018
Messages
531
checking the voltages after they've been sitting for several months [...] use lost .01v and some up to .03v

"several months" can mean a lot of things, but I'd accept cells down to perhaps 4.10v after 3 months, and as low as 4.05v after 6 months. The self-discharge varies with the cell type and storage temperature, so it's often a judgment call what voltage is acceptable. Also, your charger may not charge exactly to 4.20v to begin with, and your multimeter may not be that accurate either.
But if you have more than a dozen cells, there should be very obvious divergence in voltage between the damaged and self discharging cells, vs the more or less good ones.

For safety reasons, I strongly advise not to use self-discharging cells for anything. Discharge safely and then straight into the recycling bin.
 

Wolf

Well-known member
Joined
Sep 25, 2018
Messages
1,321
Question is, how should I sort for optimum safety and power in powerwall use
thanks, question still unanswered
Sorting for optimal safety and power in a powerwall is your question.
The best and safest way would be the same cells from the same manufacture, same date code, same IR.
Obviously in your case that is not the case so lets take your 1000 cells and see which ones would make the grade.
The criteria I use to determine if a cell is worthy and "safe" to be used in my powerwall is IR, Age, Capacity.
30 day SD of >0.05V is problematic. If the cell passes a 30 day SD test with 0.05v drop, it should maintain that voltage for at least 6 month with no more than a 0.03V drop. Most good cells will maintain their 30 day voltage for a considerable amount of time. >120 days.

In a frankenstein (cells of all types) pack you should pick the cells with the lowest IR <55mΩ but yet >40mΩ and a mAh spread between cells of ≈400mAh. Depending on the actual Amp draw on the pack that can make a huge difference. I assume you are building a 7s100p pack.
I would also shy away from mixing chemistries but that is determined by your PW needs.
If you really want some more concrete answers to your unanswered question please give us some more detail on what you are planning to use this
"Powerwall" for. What kind of draw are you expecting at max. What would your 'normal charge /discharge rates be. Then we can make a better recommendation to what we consider '"optimum safety" in my humble opinion.

Wolf
 

Dr. Dickie

Member
Joined
Sep 23, 2020
Messages
80
If you have many more cells than you need.
Then, what I plan to do is, start with the highest capacity cells, take the ones with the lowest IR. Work my way down in capacity taking the low IR.
If I don't have enough cells when I get down to my lowest acceptable cap, go back to the highest cap and take the next higher IR level.
Continue until I have the number of cells I need, but them into the Rpacker program to sort for even capacity of packs.
Of course, all my cells are the same manufacture, date, etc.
 

Wolf

Well-known member
Joined
Sep 25, 2018
Messages
1,321
Of course, all my cells are the same manufacture, date, etc.
Cheater
My first pack was all Frankenstein. That's where you learn about IR, dates, and chemistry
My 2nd, was all same cells as you LGGBM261865 easy peasy
My 3rd is going to be all the same LGFBM161865 dito
and my 4th will be all same cell. You learn and get the same chemistry and same cells if possible.

Wolf
 
Last edited:

humbleONE

New member
Joined
Jul 22, 2020
Messages
9
Sorting for optimal safety and power in a powerwall is your question.
The best and safest way would be the same cells from the same manufacture, same date code, same IR.
Obviously in your case that is not the case so lets take your 1000 cells and see which ones would make the grade.
The criteria I use to determine if a cell is worthy and "safe" to be used in my powerwall is IR, Age, Capacity.
30 day SD of >0.05V is problematic. If the cell passes a 30 day SD test with 0.05v drop, it should maintain that voltage for at least 6 month with no more than a 0.03V drop. Most good cells will maintain their 30 day voltage for a considerable amount of time. >120 days.

In a frankenstein (cells of all types) pack you should pick the cells with the lowest IR <55mΩ but yet >40mΩ and a mAh spread between cells of ≈400mAh. Depending on the actual Amp draw on the pack that can make a huge difference. I assume you are building a 7s100p pack.
I would also shy away from mixing chemistries but that is determined by your PW needs.
If you really want some more concrete answers to your unanswered question please give us some more detail on what you are planning to use this
"Powerwall" for. What kind of draw are you expecting at max. What would your 'normal charge /discharge rates be. Then we can make a better recommendation to what we consider '"optimum safety" in my humble opinion.

Wolf
Thanks Wolf, appreciate the useful feedback... I am not a high power user, I have (on my bill) my last 3 months of use @ just 51Kw for the whole period, rough estimate is therefore just 15Kw/Mth meaning 1/2 - 3/4Kw/Day, not a real power user lol I will definitely be going completely off grid, no need for it really, plus I have a couple of nice washing machine motors to make for windy days without sun. I'm pretty sure my batteries will remain mostly at full charge with just 1,530 Watts of solar feeding them, and unsure what a wind driven charge could be, haven't rewired the fisher paykel windings yet to test. I have a 12V solar charge controller from an Australian company here, wasn't cheap but I bought that coz I had a 12V 1500W/2500Wpk inverter which I have to repair (it LOOKED to be working when I hooked it up, but it was only the fans), an ebay buy from over 6mths ago I never tested (looked brand new lol lesson learned!) My setup will definitely be a frankenstein as I bought about 400 second hand cells from another seller in Queensland, many already resleeved so I have no clue what they are. 550+ will be identical salvaged vacuum LG batteries and they will definitely go together in packs, which I assume will be the best idea.
 

humbleONE

New member
Joined
Jul 22, 2020
Messages
9
Sorting for optimal safety and power in a powerwall is your question.

Depending on the actual Amp draw on the pack that can make a huge difference.

Wolf
YEAH a HUGE difference!
Experience is a big teacher and found out an idea I had is definitely not viable, at least not with cheap holders.

I made up a "battery bank" of 20 cells, (ones I won't be using in a powerwall coz they're below acceptable specs), and used the little black plastic chinese holders you see people on youtube using to make banks for charging. Decided to use this little bank and run my air pump from it as I had a trailer with a flat tyre and as I did I felt the batteries with the back of my fingers and it worked nicely I can say and no problems with heat in the cells. After about 3 minutes I noticed a little smoke coming from the top of one of the holders and promptly pulled the positive end of that cell from the holder. The cell itself wasn't hot but the holder had started slight meltdown at the spring end, seems this battery had more amps passing than it's neighbors, then another in a different "group" started doing the same.... I stopped the pump and checked everything, springs now 70% shorter on holders, 3 in fact. I did consider buying holders and making up packs this way instead of soldering or welding but now I'm glad this lesson happened this way. Lesson learned, and not with a major fire burning down a shed or something ;) Jehu Garcia's (on yootoob) cell holders must be immensely more capable of handling higher current, but he does use 7 batteries in each bank.

I ran 5 cells parallel joined to a thick wire at the back -then series- to the base of the next 5, etc etc for 4 "groups" of cells creating a bank that when low would be low a min 14.2V and high 16.8V at the ends.

I think the pump drew far too many amps from the 20 cells as configured and using the holders this way is a huge no-no, even thick wires don't help as the weak spot becomes the spring, heating faster as it is thinner metal and higher resistance than the connecting wire. I think these cheap holders are ok for charging use as not much current is passing to do that but as a holder for high current outwards - seriously DANGEROUS and shouldn't be considered period!

With safety the biggest concern unless I can get Jehu's holders cheap enough, spotwelding on nickel strip is now all I would consider. I think a localized high current situation could even cause solder to melt and disconnect (or worse, pool and short) the hot cell.

So I guess I've answered my own question about mixing cells, whereas amps is the biggest consideration, ie; not throwing in any say 1500mAh cells with 2400mAh together as failure will be imminent in that local area. I'll now consider I should match cells as close as possible in "rating" together as per what my tester said.

Anyone with knowledge of an Aussie supplier of holders as least as good as Jehu's, or a reasonably priced nickel strip would be appreciated.
 

daromer

Moderator
Staff member
Joined
Oct 8, 2016
Messages
5,468
Its cheaper just to get a proper spot welder and you get alot better setup in current handling. Holders in all its glory but how often do you plan to change out the cells? Holders are made to swap cells out on daily basis. Powerwall should be rigid packs that lasts years.
 

Wolf

Well-known member
Joined
Sep 25, 2018
Messages
1,321
Powerwall should be rigid packs that lasts years.
Exactly. Also any supplemental Battery you build should also be "rigid" and properly put together. I do understand the ease of having slip in battery holders and how easy it can be to replace a "bad" battery. Two things to consider when using these type of holders. (1) The holder itself is a source of resistance as @humbleONE found out. and (2) when using these types of holders in a powerwall it is an admission (maybe not intentional) that you are not sure of the quality of the cells you are using in your battery or powerwall build.

Proper powerwall and for that matter of fact any lithium based battery build needs to follow a set of rules. Primarily for safety and of course longevity. These 2 criteria are accomplished by a carefully constructed battery. Know your cells and know what you are building. I am by no means one of the founding members of this group nor do I have the years of experience the likes of @hbpowerwall, @daromer, @OffGridInTheCity, etc.
but there is one thing no one can deny and that is that I know my cells. I have done and published much in depth research on cell "health" and what constitutes a good cell, how cells behave with each other and a host of other information.
I believe my original spreadsheet of 6225 tested cells is the only one in existence (at least I haven't found another one) that gives a good picture of what constitutes a good cell and a bad one. https://drive.google.com/file/d/1NujY1eO6MKwGrpyEm185m6vpkMdb_Gp9/view?usp=sharing
So tooting my own horn is not something I usually do ,but I for one want everyones battery/powerwall build to be a successful and frustration/trouble free experience. What method you use to sort and qualify your cells is your choice. I obviously have my way and consider it the best.;) It is also the most time consuming as I touch each cell from the beginning to the final commitment to the pack/battery 4 times recording each measurement. I just want to build the best battery I can.

In the final showdown though,
Know your cells

Wolf
 
Last edited:
Top