Capacity or IR? Which is more important?

[Lt.Dan]

I'm currently testing 48 cells for capacity in prep to put them into separate 16s packs, but i'm questioning which is more important when pairing/matching cells.

Do I try and keep all the cells together with the same (or close to) Internal Resistance? Or do I try and keep all the cells with the closest Capacity?

I also have read @RCinFLA thread (thank you for your information) on Andy's cells https://diysolarforum.com/threads/lfp-cell-testing.61272/ . And now noticing that I'm already 30 cells deep, and I should go back and test with incremental pauses throughout. I am using the same EBC-A40L and have saved the charts of all the cells so far.

Thanks guys.

 

[sunshine_eggo]

Capacity. IR is kinda pass/fail, and they really shouldn't deviate a lot. 10% variation in capacity means 10% loss. 10% variation in cell IR means... not much.

Self-discharge is also a key matching factor for balance retention. Fully charge, let sit for 30 days and then re-charge to full... compare the mAh needed to bring each to full charge.

 

[740GLE]

Also doesn’t IR change as the SOC changes?

 

[Lt.Dan]

Capacity. IR is kinda pass/fail, and they really shouldn't deviate a lot. 10% variation in capacity means 10% loss. 10% variation in cell IR means... not much.

Self-discharge is also a key matching factor for balance retention. Fully charge, let sit for 30 days and then re-charge to full... compare the mAh needed to bring each to full charge.

Can an internal resistance variance cause imbalance in the cells during a charge/discharge cycle? I would assume a lower internal resistance cell would contribute more power than a higher resistance cell, causing it to be imbalanced further? Or does it not work like that? I'm not sure.

Self discharge is a good idea. Is it as important though, if the batteries never rest? If they are constantly being used, I could see this being moot. Or does it give hints to a battery having a shorter life span?

Also doesn’t IR change as the SOC changes?

I would assume you just check at all the same SOC? Thats a good point though, if you have 1 cell with a wildly different IR at high SOC vs low SOC.

 

[sunshine_eggo]

Can an internal resistance variance cause imbalance in the cells during a charge/discharge cycle? I would assume a lower internal resistance cell would contribute more power than a higher resistance cell, causing it to be imbalanced further? Or does it not work like that? I'm not sure.

Not really. The same Ah goes in and out. During a discharge, the voltage is reduced, so you get less Wh. During a charge, the voltage is pushed higher, so it takes more Wh to charge.

Self discharge is a good idea. Is it as important though, if the batteries never rest? If they are constantly being used, I could see this being moot. Or does it give hints to a battery having a shorter life span?

If the batteries are regularly cycled and subject to passive balancing, it shouldn't play a significant role.

would assume you just check at all the same SOC? Thats a good point though, if you have 1 cell with a wildly different IR at high SOC vs low SOC.

IR varies a little with SoC, and it shouldn't be wildly different at different SoC - I mean I don't think that's a thing at all.

If you test IR at roughly the same SoC for each cell, it should be good for comparison purposes.

 

[Lt.Dan]

Got it, so as long as none have an IR that are wildly different, it won't matter much.

Matching based on capacity. Got it.

 

[S Davis]

I think it is important to test IR at the same temperature.

 

[S Davis]

Double post

 

[Lt.Dan]

I should add, this is my process:

Charge at 40a to 3.65v, until 2a.
Wait 30 min
Discharge at 40a to 2.5v
Charge at 40a to 3v, until 2a. (just to get it out of the bottom knee).

This is also testing inside the house, at a constant 76*F. So I should be pretty consistent.

 

[sunshine_eggo]

I should add, this is my process:

Charge at 40a to 3.65v, until 2a.
Wait 30 min
Discharge at 40a to 2.5v
Charge at 40a to 3v, until 2a. (just to get it out of the bottom knee).

This is also testing inside the house, at a constant 76*F. So I should be pretty consistent.

Looks good to me.

Does the 3V to 2A charge take awhile? IMHO, charge from the bottom for a fixed time - say 45 minutes. That will get 30Ah into the cell, which should be close to 10%, and 10% is mostly out of the knee. If that's not enough, figure out how many Ah you're putting into them with your 3V/2A charge, and do a timed charge to that Ah level @ 40A.

 

[Lt.Dan]

Looks good to me.

Does the 3V to 2A charge take awhile? IMHO, charge from the bottom for a fixed time - say 45 minutes. That will get 30Ah into the cell, which should be close to 10%, and 10% is mostly out of the knee. If that's not enough, figure out how many Ah you're putting into them with your 3V/2A charge, and do a timed charge to that Ah level @ 40A.

Getting to 3v is only 5-6ah pretty consistently. I thought about it after I had done a few, but I had already started, and figured I might as well carry it all the way through. I also want to get started with the next one already lol.

 

[Alkaline]

I think self discharge is the most important because this is what affects real world reliability. E.g. if you have a cell that consistenty self discharges no matter how much top balancing you do it will go out of balance.

For these the only solution is do a routine full charge and re-balance.

I also value holding voltage underload or no dippers / runners.

I really don't care about capacity as much as making sure all cells work together properly. E.g. I'd happily trade 10Ah out of the 280AH simply to ensure the cells are in balance.

You will never be caring about capacity in the real world, but you will care if your cells go out of balance or they can not hold voltage underload. It will drive you nuts.

I'm not sure how IR plays into this but I would pay close attention to this.

 

[sunshine_eggo]

Getting to 3v is only 5-6ah pretty consistently. I thought about it after I had done a few, but I had already started, and figured I might as well carry it all the way through. I also want to get started with the next one already lol.

but how long does that take you? My point is at 40A on a timer, that would take 8 minutes.

 

[Craig]

Curious @Lt.Dan on your testing process what are you using to get a 40 amp load? How are you measuring as well?

 

[Lt.Dan]

Curious @Lt.Dan on your testing process what are you using to get a 40 amp load? How are you measuring as well?

Using the EBC-A40L tester. It makes some nice graphs and gives a lot of data.

 

[Craig]

Using the EBC-A40L tester. It makes some nice graphs and gives a lot of data.

Thanks. I have my own that I built some years ago but i have never been able to get that High of an amp draw. So Now I know it cab be done somehow.

 

[Lt.Dan]

Thanks. I have my own that I built some years ago but i have never been able to get that High of an amp draw. So Now I know it cab be done somehow.

What did you build? I'd love to see it

 

[houseofancients]

I should add, this is my process:

Charge at 40a to 3.65v, until 2a.
Wait 30 min
Discharge at 40a to 2.5v
Charge at 40a to 3v, until 2a. (just to get it out of the bottom knee).

This is also testing inside the house, at a constant 76*F. So I should be pretty consistent.

that what i use when testing my cells

 

[740GLE]

Not really. The same Ah goes in and out. During a discharge, the voltage is reduced, so you get less Wh. During a charge, the voltage is pushed higher, so it takes more Wh to charge.




If the batteries are regularly cycled and subject to passive balancing, it shouldn't play a significant role.



IR varies a little with SoC, and it shouldn't be wildly different at different SoC - I mean I don't think that's a thing at all.

If you test IR at roughly the same SoC for each cell, it should be good for comparison purposes.

Somehow I thought at the knee of SOC the Ir climbs a little bit that’s why the tail current falls off. I could be completely wrong.

Thanks for your comment.

 

[Lt.Dan]

but how long does that take you? My point is at 40A on a timer, that would take 8 minutes.

I went to go change the parameters and there is no charge at constant current for X amount of time option. There is only a CV with X cutoff amps. Bummer.