diy solar

diy solar

A gifted horse... best way to match 40 used Calb 180ah cells in 8s5p

CapnGil

New Member
Joined
Sep 20, 2020
Messages
40
A gift horse... best method to match used LiFePo4 Calb 180ah cells?

Summary; I was gifted 46 Calb 180ah cells that came off a dock yacht; owner upgraded to entire Victron package. My vessel runs on 24v, so I set them up in 8s5p (using 40 cells) with each bank having its own Daly 200a BMS. They are charged by six 285w solar panels in 3s2p configuration, feeding a Growatt 3000w hybrid solar charger. The cells were given to me individually so I initially did a mid charge to 3.3v each and parallel connected all of them for two days to equalize them and then matched them up by internal resistance: I thought it was a good way to match them.
However, after a few days the voltage have strayed between cells by more than .2 volts.

So my questions are as follows:

1. What would be the best way/method/procedure to match up these cells in groups of 8? Top balancing, bottom balancing, capacity testing (thats gonna take a while), etc....

2. Upon doing so, how could I manage/maintain the voltage between battery banks, once matching of individual cells are done?

Any suggestions would greatly be appreciated.

Be well
Gil
 
My vessel runs on 24v, so I set them up in 8s5p (using 40 cells) with each bank having its own Daly 200a BMS.
However, after a few days the voltage have strayed between cells by more than .2 volts.
First, label your cells so you can keep track of them with notes.

It sounds like thru your use and testing you have the data to rank the 8 cells in each 8S battery. The weakest charge the fastest (highest voltage at cutoff) and discharge the fastest.

I would put the strongest 8 cells into one battery. Maybe start with the strongest of each of the 5 batteries, combined with the next 3 cells you determine to look the strongest (not scientific but its reasonable).
Next 8 strongest together in 2nd 8S battery and so on, creating 5 8S batteries, until you find the weakest 6 to be spares.
Thru continued use, you will find strongest and weakest in your 8S packs and can promote stronger and demote weaker between 8S batteries until you get matched sets of 8. Find the strongest of your spares and swap that in and out of the weakest pack's weakest cell.

You will eventually (hopefully) get reasonably matched cells within each 8S battery. You will also have 5 different capacity batteries that should work together if wired in parallel, in a balanced manner.

Maybe someone has something better, like capacity testing each cell. I was using my cells to run a greenhouse so i did not have the luxury (or patience or tester) to bench test them all. I think its a sound method of ranking your cells.

Taking great notes with labeled cells is vital.
 
First, label your cells so you can keep track of them with notes.

It sounds like thru your use and testing you have the data to rank the 8 cells in each 8S battery. The weakest charge the fastest (highest voltage at cutoff) and discharge the fastest.

I would put the strongest 8 cells into one battery. Maybe start with the strongest of each of the 5 batteries, combined with the next 3 cells you determine to look the strongest (not scientific but its reasonable).
Next 8 strongest together in 2nd 8S battery and so on, creating 5 8S batteries, until you find the weakest 6 to be spares.
Thru continued use, you will find strongest and weakest in your 8S packs and can promote stronger and demote weaker between 8S batteries until you get matched sets of 8. Find the strongest of your spares and swap that in and out of the weakest pack's weakest cell.

You will eventually (hopefully) get reasonably matched cells within each 8S battery. You will also have 5 different capacity batteries that should work together if wired in parallel, in a balanced manner.

Maybe someone has something better, like capacity testing each cell. I was using my cells to run a greenhouse so i did not have the luxury (or patience or tester) to bench test them all. I think its a sound method of ranking your cells.

Taking great notes with labeled cells is vital.
Thanks M. Sandals. I was doing as stated, but usage and whether I had to combine banks for higher loads changed watt usage, so I couldn't rely on my findings.

That said I bought a EBC-A40L Capacity tester and 25 days from now I should know which cells to group up.

Thanks again.
Be well
Gil
 
That said I bought a EBC-A40L Capacity tester and 25 days from now I should know which cells to group up.
Ya, it takes a long time.

I'd start testing the 8 cells you suspect are the strongest (slowest to charge, slowest to discharge). You can get those up and running together when you find 8 of your best.

The charging was the slowest part for me. I have 10A hobby charger that i power from a solar charged main battery. Kind of at the mercy of sunshine quite a bit of the time too.
It took me 8 weeks to do 18x 280Ah cells.

Get as many charged full while you are waiting!
 
Ya, it takes a long time.

I'd start testing the 8 cells you suspect are the strongest (slowest to charge, slowest to discharge). You can get those up and running together when you find 8 of your best.

The charging was the slowest part for me. I have 10A hobby charger that i power from a solar charged main battery. Kind of at the mercy of sunshine quite a bit of the time too.
It took me 8 weeks to do 18x 280Ah cells.

Get as many charged full while you are waiting!
Wow a dedicated man.... 280ah at 10a's
The load tester I got does 40 amp and my cells are 180ah, so much faster, but I do have 40 cells, plus 4 extras...

Side note; I bought 5 Neey Active Balancers through Aliexpress and recieved older units, and their dispute center is trash so I'm out 355.00 US Dollars.

I was hoping those would resolve the imbalance issues, but cell matching should better manage imbalance issues.

If my credit card company helps, I'll seek out a reputable seller; I hear Hankzor is good.

Be well
Gil

Ps. I did a 7 day equalization on all cells with busbars (see photo) and then recorded the internal resistance of each cell. I'll group cells by both capacity and internal resistance when I can, with capacity being primary criteria.
 

Attachments

  • 20230904_182938.jpg
    20230904_182938.jpg
    1 MB · Views: 7
I initially did a mid charge to 3.3v each and parallel connected all of them for two days
... and in this state, add a charger, set to 3.5 or 3.55 and let them sit on the charger until current goes below 0.05C.
Just paralleled at 3.3, no matter for how long, absent a charger, accomplished nothing.
 
... and in this state, add a charger, set to 3.5 or 3.55 and let them sit on the charger until current goes below 0.05C.
Just paralleled at 3.3, no matter for how long, absent a charger, accomplished nothing.
Well then I got lucky or the air in my area has magic unicorn pixel dust because after the 7 days all 43 cells equalized to within .009 volts ( see images ), but they were within .5v to begin with.

Come to think about it, it was like 12 days in parallel, not 7, but after soo many days messing with these, who's keeping accurate records of days...lol.
 
Well then I got lucky or the air in my area has magic unicorn pixel dust because after the 7 days all 43 cells equalized to within .009 volts ( see images ), but they were within .5v to begin with.

Come to think about it, it was like 12 days in parallel, not 7, but after soo many days messing with these, who's keeping accurate records of days...lol.
 

Attachments

  • 20231011_112259.jpg
    20231011_112259.jpg
    748.3 KB · Views: 5
  • 20231011_112509.jpg
    20231011_112509.jpg
    754 KB · Views: 5
Well then I got lucky or the air in my area has magic unicorn pixel dust because after the 7 days all 43 cells equalized to within .009 volts ( see images ), but they were within .5v to beg

Equal resting voltage is not equal SoC.
If you connect in series and charge (Use a BMS!) some may run to 3.65V while others have not yet reached that.

If you charge in parallel to 3.55V, or better 3.65V (well above knee of curve) it should put them at very similar SoC.
Then so long as your usage charges lower than that, no runners.

Your busbars look solid, but connect positive lead to one end of one busbar, negative lead to other end of other busbar. While not perfect, it will better balance voltage drop while current is flowing.
 
Except in very high current situations (ie, 1C charge/discharge) the only matching to be done is based on capacity. Individual testing is the go-to for capacity testing.

If their capacities are close to each other (within 2-3%) then it doesn't matter how you gang them. The worse the match is, the more your balancing system will work, but in general they'll be happy.

Outside that, though, you will want to match them in capacity, forming one battery with cells with similar capacity, then the next with cells of a larger capacity.

Another option, if the capacities are significantly off, is to form 2p batteries, one with a high capacity, and one with a low capacity, then form those into the series packs. This way each pair can have a specific capacity that matches the other pairs, and you'll see fewer runners. Pair the 175AH with the 164AH, and the 174 with the 165, etc. With as many cells as you have you should find a good linear progression that would allow you to make batteries of the same capacity you can put in series.

All that said, I echo the point others have made - there is no such thing as mid balancing. There is top balancing, and bottom balancing. Between those two voltages there isn't a strong enough correlation between voltage and state of charge to definitively state that two cells have the same state of charge at that voltage, and are therefore matched.

Also, I strongly recommend the iCharger series of chargers - https://www.progressiverc.com/pages/search-results-page?q=icharger . Not only can they charge at speedy rates, but they can discharge into the power source if it's a battery bank. So you set up one large battery bank, use it as the source, and then charge and discharge the battery under test at 30A (12 hours for you per cell). Little power actually wasted, no need to plug it in and use AC power. Run through a bunch of batteries, make a new battery from the tested cells, and then use that as the source to test the cells you were first used as the source.
 
I recently matched 16, 100ah cells, for a 4P4S configuration.

I matched based on IR, measured by voltage drop.

I top-balanced the 16 cells, and then removed 10% of their capacity (40amps), then broke the pack apart and gave each cell a 0.3C load and recorded the voltage drop (from resting) after 3 mins durration.

Cells with similar voltage drop were identified and grouped together in packs of 4.

I then put all 16 back in parallel and top balanced again ( to 3.55v per cell).

I also put the strongest packs at the beginning and end of the series. Not sure if this is actually advised, but it seemed intuitive to do so...?

So far seems to be working well as all cells packs (4P) are within 11mv after six months. My BMS has only about 150ma passive balance current.

I am not saying this is the only way to do this but it was pretty painless and quick.

MP
 
Great input....
I always say " I'm not Stupid, Just Ignorant, educate me on the subject and I will make better choices ".... smile. So, thanks for the input and although my news is bad (see below) I'm not crying, the cells have carried the electrical loads on my boat as they were so if anything, I should get more out of them once balanced.


That said, I now have two EBC-A40L's (don't ask, Aliexpress screwup) so I'm capacity testing all cells through them for true capacity Ah's, with final Ah capacity test being done on one EBC-A40L to cancel out any tested indifferences. The bad news is that the gifted horse (44 Calb 180Ah cells) are averaging 145Ah; is there such a category as D rated cells....LOL? Bummer, what's that saying, "never look a gifted horse in the mouth".
Screw that, I now say, "shoot the horse and sell it for glue manufacturing" and buy new cells.... LOL


If I may let me present my capacity testing procedures for the groups input, however since I've done 16 to date, I'm gonna have to live with any errors I've made as I'm not starting over...... N-LOL




Pre-test:

Constant Charge all cells to 3.65v. I noted Ah capacity and time it took to reach 3.65v. I created a batch of cells that went through this process for the automatic EBC-A40L to run on the second EBC-A40L



EBC-A40L Automatic Run Mode Capacity Test:

Step One - CCV to 3.65v with 5A and shut down at 2A draw (note all cells were at holding at 3.40ish volts due to previous charge to 3.65v) it took about ten to reach 3.65v before next step activates.

Step Two - rest for 10 minutes. I guess this could be longer, but I did so because of time and thinking it only took a few minutes before cell reached 3.65v. Absolutely no change in cell temp post this stage.

Step Three - Discharge Constant Current to 2.5v at 40A with cutoff at 2A and note Ah Capacity.

Step Four - Wait 5 minutes

Step Five - Constant Charge to 3.3v ..... test done.



Be well
gil
 

Attachments

  • Capture 6.JPG
    Capture 6.JPG
    627.7 KB · Views: 1
Back
Top