basics question for assembling packs and testing cells.

vivi0512

New member
Joined
Jan 17, 2019
Messages
15
Hello guys, always in my outddor light energy storage project.

i have written 2 or 3 small questions that come during my work, maybe some of you can answer that ::)


First concerning the way to build pack, i test all my 18650 from recycled laptop battery.
I have battery's from 1000mah to 2700mah ( LG made exceptionnal work by the way, somecells on a old laptop have awesome capacity ).

My question is : is it true that it isbetter to do :4S5P with exactly same capacity ( for example 20 x 18650 at about 1800-2000mah )
the thing is that i have not 20x sameunits for each capacity.

Can i take a wide range of capacity ( 1600-2000mah )
and do my 4S5P with it, if yes do i need to do 4x2000mah then in parralel4x1900mah then 4x1600mah etc... is that work well ?



2nd question is about rentability of small capacity's cells ( 1000-1200mah ):
if i don't pay the 18650 battery, i ve done a rentability calculation and with 1000mah i can do battery for 90$ / Kwh ( wrap shrink, copper plates, fuses wire, bms, and cell's holder included ) i see people on youtube whosayto don't use cell under 1200mah , but if i m not wrong, it is still profitable to use them if they works no ?


Finally last question is about testing ( wrong category, but better to do 100 different subject on the forum )
I have many used cells who left in laptop during some years, they are under 1V , sometimes 0.25V like that, i put them on my C4 charger, it try to charge them at 1.7V then the voltage of the cell goes up in 4-5second then it changes and charge them at 4.2V..... when i do that the cells are hot ( handle them in hand is possible but still hot ) so to avoid this i charge them in 2 times : 1 hour or less if they became hot quickly then, i let them rest and comeback to ambiant temperature.... then i charge them to full and they works... i do the NOR TEST ( charge, discharge, charge ) and then they are normal , there is a very very small heat ( absolutely like others , and sometimes good capacity , sometimes 700 ( bad ) , 1200 sometimes, 1700-1900....

Question is : Can i conclude it is possible to ressurect a dead cell ? (because on youtube they throw them away and even test it if it is not more than 2V , it seems to be a waste of potential good cells ).
 
1. No not really. Whats the factor is the life left on them. Example the 1200mAh cells most likely are closer to their EOL than the ones at 2Ah if all new was 2.2Ah. So its all about what are their capacity new?

2. Look at #1. a 1Ah cell is more likely to be closer to EOL uf they come out of same pack as cells that have 2Ah.... Personally i wouldnt build of cells that have degraded 50% of its capacity nor smaller ones. Its better to wait.

3. I test all cells. but i bin all heaters. I never give the cells a 2nd chance if they have shown sign of issues... Im the guy that will NOT take a chance if they are going to power my house. Some people do retest or use them but is it worth it? There are no scientific papers out there stating what will happen with them more than the likelyness for failure are larger when they have several 100s of cycles. Is it then worth the chance? Just because 1 or 10 people here have 100 cells that seem to work a 2nd time thats not proof that its 100% safe.
You have to do the choice here..

With that said there are evidence that if you slowly start the cell back and let it sit and next time charge them they have less resistance => less heat generated. Thats a fact.


On my page you can see a flow chart on how I test them:
http://diytechandrepairs.nu/current-solar-layout/
 
Like Daromer says, try to work out the "capacity now" vs "capacity new" for your cells & use ones where now is closer to new.
Use the cell database for details on new capacity.

For Q2, each of your 5p group of cells should = same total capacity.
If you make one 5p with all 1600mAhr cells & another with say all 2300mAhr cells you will have a very unbalanced 4s5p battery
If each 5p was made with say 3x 1600 + 2x 2300 cells it would be OK.
You will have to mix up the best cells in your pack and do some math to get each 5p group as close as you can (see repackr link at site top right side)

For Q3, apparently results & safety is better if you carefully slow charge low volt cells until they reach 3V with only 50mA current.
Don't put them into your C4 charger first, only after slow charge & when = >3V
 
I combine them by capacity. I keep the best ones for premium projects and use junk ones for lesser things like LED emergency illumination.

For the packs I make I keep a tolerance of 100mAh.

That means, I'd use 2100mAh to 2200mAh for the important things and I can pair up fine cells with 1000-1100mAh remaining for usual needs.

I will never combine in a pack cells with very different capacities. That is a waste of good cell capacity and a lower durability of the pack overall.
 
Overmind said:
I combine them by capacity. I keep the best ones for premium projects and use junk ones for lesser things like LED emergency illumination.

For the packs I make I keep a tolerance of 100mAh.

That means, I'd use 2100mAh to 2200mAh for the important things and I can pair up fine cells with 1000-1100mAh remaining for usual needs.

I will never combine in a pack cells with very different capacities. That is a waste of good cell capacity and a lower durability of the pack overall.

I understand well what you are sayingbut in my case

I have more than 50x 2000mah cells
And a little more than 50x 1400mah cells ( nominal 1500 so goods cells )

Can i mix them using repack to built a 14s(MAX)p battery
Or is it better to do a battery with all the 2000mah
And an other battery with all the 1500mah cells.?
Then put the 2 battery in parrallel to charge them because i would have them in the same powerwall with the same PV inverter/battery charger


Electrically it is the same no ?
 
This is really up to "you" what you want to do. There is no real difference between them. Both methods are electrically the same.

The pros of separate strings (batteries) is that the lower capacity unit can be replaced later on with a better one when you get enough cells to build that one. Also, the lower capacity cells if anyone of them become bad, ie a self draining cell, it's easier to narrow down where the problem is.

The con is that you will need two bms units to monitor each string, unless you electrically parallel each strings pack with it's neighbor.
 
Overmind if you have those in series that is guarantee to have uneven packs. Its better to mix them ie same amount of 2Ah and same amount of 1.5Ah in each pack.

If you do want to have one pack of each capacity you need to work out the IR and the total capacity along side with resistance in any busbars and so forth to keep them equal. More work that way ;)
 
I was not talking about using both packs together.

If a single combined thing must be made, then 3x2000mAh pairs can be put in series with 4x1500mAh pairs and we have balance.
 
Of course, but if you separate them, they might age different, so you would at least after some time have uneven packs.
 
Ok i will do separate pack, and try to have the same capacity for the 2 packs,

for example i will put 1s10p ( not 1s in real but for easier calculation ) with 2000mah cells, to have 20.000mah capacity.

and anther pack with 1s14p with 1500mah cells to have closer capacity ( 21.000mah )
 
I think you don't understand what I mean.

It's better to do mixed packs (4sXp with both 2000mAh and 1500mAh), or you make two whole strings (4sXp with 2000mAh and 4sXp 1500mAh) and connect them in parallel (also each connection in between has to be paralleled).

Hope you understand now what I mean...
 
Overmind no dont serie 3x2Ah with 4x1.5ah. That is 3 cells in one and 4 in one and it will cause imbalance in IR and things like busbar style and other factor. Its better to have same amount of 2ah as 1.5ah in each pack that are in series.

You can of course do 1 full string with only 2ah and parallel that one with 1 full string of 1.5ah. but that is the same as i try to explain. Keep the packs in a string 100% even. IN all ways possible.
 
Back
Top