Paralleling cells before Self discharge and capacity testing

Oz18650

Member
Joined
Oct 2, 2017
Messages
209
I have a large number of cells that are all the same model.
I am also keeping them in groups as they were extracted as I have the manufacture dates of the packs.
Apart from testing theTP4056modules I have, I was considering putting each of the cell groups in a holder all in parallel for a while after charging, so they are all exactly the same voltage before they are separated and put away for self discharge testing. This way I only need to do 1 initial voltage check for all the cells in a group and drops in voltage may be more easily apparent.
I think paralleling the cells should also mean that they are exactly the same state of charge before capacity testing and should make capacity testing slightly more accurate.
What do people think of this idea?
 
The Herd will hide the issue cells. better to leave them single for a month before retesting for V drop
 
You should never paralell any cells that havent lived all their life before any tests.

Its important that the tested pack is identical. You dont save any time packing them up before testing. I have had many many maaaany packs where 1 cell out of 6 or 2 or whatever is causing the issues and rest is fine.
 
The idea was not to skip any testing.
It was also not to conduct SD testing with the cells in parallel (which would hide problem cells).
It was to parallel cells to equalise voltages of all cells before separating all cells to wait and see if any self discharge.
That way they all start at the same voltage before waiting.
If some start at 4.2v and others start at 4.16v after charging with different TP4056 chargers, it is more difficult to compare end voltages after waiting.
 
Thats doable but one extra step. Self discharging show up very quickly as long as they are fully charged. Ie above 4.1V in my world.

Its not rocket science in this case and due to how far off some equipment is you need to use the same methodology on all tests performend. No use doing one test more precise than some other :)
 
There are a few benefits that I think may come from doing this.
1. I am not going to be able to commit to leaving cells for a set time. If they all start at the same voltage it will be easier to spot how they change from one another over time.
2. Despite the testers being inaccurate, the errors when charging and capacity testing are cumulative.
I think if a group of cells all start at the same state of charge, testing will be more accurate than if the cells start off at different voltages.
For example: If you have 4 x 2000mah cells where 2 are charged to 102% of correct full capacity and 2 are charged to 98% of full capacity.
The slightly overcharged cells would have 2040mah and the undercharged cells would have 1960mah.
Let's say we have testers that are accurate to +/- 5%.
For the cell holding 2040mah, the tester could show up to 2040 +5% which is 2142mah or down to 2040 -5% which is 1938mah.
For the cell holding 1960mah the tester could show up to 2058mah or down to 1862.
For these 4 cells you could get test results between 1862mah and 2142mah.
Now, if those same 4 cells were put in parallel to balance out to 100% charge, each would be holding 2000mah.
2000mah +/- 5% gives a range of 1900mah to 2100mah.
The range has improved from
1862mah to 2142mah a range of 280mah
To
1900mah to 2100mah which is a range of 200mah.
This difference will affect how cells are discarded or kept based on state of health compared to new and how cells are included into balanced packs.
 
Oz18650 said:
I was considering putting each of the cell groups in a holder all in parallel for a while after charging, so they are all exactly the same voltage before they are separated and put away for self discharge testing.
The idea is bad because cells in bad shape will influence the result of cells in good shape.
Just charge them to maximum and then do the capacity test without influencing them in any other way.
 
Overmind said:
Oz18650 said:
I was considering putting each of the cell groups in a holder all in parallel for a while after charging, so they are all exactly the same voltage before they are separated and put away for self discharge testing.
The idea is bad because cells in bad shape will influence the result of cells in good shape.
Just charge them to maximum and then do the capacity test without influencing them in any other way.
How will cells in bad shape influence the result of cells in good shape?

re "Just charge them to maximum", what is maximum?
I am using TP4056 and they do not all charge the same, so the results are already affected by the difference in charging.


peremodaw said:
Hi buddy, did you give a try or anything as such? following up from long, excited to know the result.

Regards,
Sloth Smithhttps://plex.software/ https://tutuappx.com/. Vidmate
My testing is only just starting. I will be trying this idea out.


Maybe my idea would have been clearer if the subject was "Balancing cells before self discharge and capacity testing"?
 
If 2 cells are charged up to 4.2V, and one of them is a slow discharger, then the weak one will pull the good one down. Now you have 2 cells that are voltage drained, but you don't know which one is the leaker. You'll have to charge them back up and let them sit again to see which one drains.

Regardless of which charger you use, check the voltage when pulled off the charger and put it in a bin of like cells. So if the charger only charged to 4.18V, then have a bin for 4.18V cell. If another charged to 4.22, then have a bin for 4.22V. That way you can easily check how fast a cell discharges.
However, regardless if a cell charged to 4.18 or 4.25V, even a good cell will try to settle down to about 4.15-4.18V. This is why you don't want to test for SD only a day after charging. It could still be settling.
The way I test for SD's is charge all cells, at least 4.18V. Let them sit for a month or so (for me, I've actually had cells for 6 months or longer) and then test them again. If their voltages have dropped below 4.15 - 4.10V they get a second charge/test. If they drop below 4.10V, they get put into the SD box. After a month or two, these cells should not drop below 4.10V if they are healthy cells. Even 4.15V is suspect, which why those cells get a second run/test just to make sure there wasn't an anomaly or they just needed to be woke up.

Just don't parallel connect all cells during the resting phase. It is perfectly fine to parallel them during the charging phase as long as your charger can handle the load required.
 
Why not charge them paralel? If your psu is up to the job!


image_hqpood.jpg


image_fkjihs.jpg


image_hezbkf.jpg


:cool:
 
Nice little parallel contraption. I do a fair share of parallel charging myself, or to allow cells to equalize with each other before putting them in series.

I do like parallel charging them removing them and allowing them to rest. I just mark the box what voltage they all came off at and its interesting to see how some drift from others. The self discharges become a little more obvious using that method for me.
 
I parallel charge 36 at time before sorting into capacity. Much easier to see self dischargers when they all start at exactly the same voltage.
 
@Oz18650 to maximum means to the official 4.2V.
If a cell cannot reach that charge level, or overheats a lot while doing so, it's dying anyway and should be excluded from the top lots.
 
Overmind said:
@Oz18650 to maximum means to the official 4.2V.

This is kind of the point of this whole thread.
The very commonly used TP4056 and/or clones do not charge to 4.2 exactly.
They are a bit all over the place.
You can not just charge 10 cells uaing 10 tp4056 boards and then put them away knowing that they are all at 4.2v
Some will be over and some will be under, so when checking later for self discharge, you don't know what they started at without binning based on start voltage, or recording starting voiltages some other way.
I was suggesting balancing sets of charged cells so they all start waiting for SD testing at the same voltage.
Parallel charging could alao work.
Has anyone checked how consistent the opus charge voltage is?
 
Opus is basically the same. Question is what DO you gain from running 0.01v 1% accurazy before SD?

I bet nothing. :)

For instance the capacity is also all over the place. As Long as its randomized i bet IF you just guessd the Numbers and say lets have Them above 4.1v it should be ok. With that Said i have of course checked so all My opused atleast Charge to 4.1ish
 
Oz18650 said:
Overmind said:
@Oz18650 to maximum means to the official 4.2V.

This is kind of the point of this whole thread.
The very commonly used TP4056 and/or clones do not charge to 4.2 exactly.
They are a bit all over the place.
You can not just charge 10 cells uaing 10 tp4056 boards and then put them away knowing that they are all at 4.2v
Some will be over and some will be under, so when checking later for self discharge, you don't know what they started at without binning based on start voltage, or recording starting voiltages some other way.
I was suggesting balancing sets of charged cells so they all start waiting for SD testing at the same voltage.
Parallel charging could alao work.
Has anyone checked how consistent the opus charge voltage is?
You don't need to charge to exact value. charge up to 4.1+ as most TPs actually do and then charge the remaining with the actual tester.
 
Use a buck converter.
Or order ONE batch + 10% of those tp/tc,s and sort them with a high uf (F) capacitor (to be quick)to match them and to make sure they all end at the same voltige...
I loaded them per 100 at the same time at the same speed to the same voltige in one go.
That experiment was better than 30 tp/tc's sorted out.

4 x https://www.aliexpress.com/item/32872607362.html?spm=a2g0s.9042311.0.0.27424c4dJaFzQX
Connected them to my Philips psu's.
Someone like me can not have enough of them.
max they will do is 11,25A ish

https://www.aliexpress.com/item/32945969098.html?spm=a2g0s.9042311.0.0.76a24c4dykVCjt
Probes are very easy to solder.
 
Back
Top