Cells for rebuilding cordless tool batteries

Hi, so I have been getting into rebuilding power tool batteries, especially 5ah Dewalt batteries. The cells that are in them are inr18650-25r samsung with a capacity of 2500mah.

Question 1. When I test every cell, is a threshhold of at least 80% capacity to be used in rebuilding another battery a good threshold?
Question 2. If a 2500mah cell tests less than the 80% capacity, but tests at say 70%, is it alright to put that cell in a 4ah tool battery which uses 2000mah cells, or will it not last the same?

I would be grateful for advice on this. Thanks

You generally want to mix cells, that have a close "% of capacity" left. Cause that number will degrade pretty much the same across all cells.
If you mix a 3500mAh, that has only 2100mAh left (~60%) and a brand new 2200mAh cell together, the 60% capacity left cell will degrade faster because its towards its end of life and will fail, before the 2200mAh cell will fail.

Lipower said:

Question 1. When I test every cell, is a threshhold of at least 80% capacity to be used in rebuilding another battery a good threshold?

Click to expand...

80% capacity is serviceable.

Lipower said:

Question 2. If a 2500mah cell tests less than the 80% capacity, but tests at say 70%, is it alright to put that cell in a 4ah tool battery which uses 2000mah cells, or will it not last the same?

Click to expand...

Agree with @Oberfail but its not just capacity (80% is serviceable) - its also whether the cells have enough discharge amps so the tool runs normally and for a reasonable amount of time. If you use low-amp cells in a hi-amp application - one of the things that can happen is the voltage will pull waaaaaay down as you turn on the tool... say 4.1v pulls down to 3.7v and then it will go 3.7v/3.6v/3.5v/3.4v (cutt-off) within a short amount of time. Maybe minutes of use instead of hours. In this scenario is not a capacity issue overload on the discharge specs of the cells.

Assuming the cells match the discharge requirements, the tool runs 'normally'. The 80% or even 70% capacity will likely work for a while - especially if the tool use is like mine where my drill is discharged over several weeks - maybe 1 recharge every month or 2 weeks type of thing. At a 24cycles/year rate - the capacity won't drop that fast

Bottom line - here's nothing wrong or dangerous with what you're attempting. Give it a try and let us know how it works

Just don't replace the power tool cells with ones that are used in laptops, or similar low-current draw devices. They won't have the amps delivery capability.

Ok thanks for the advice, what I am hearing is that the older a cell, the faster its capacity decreases. Am I correct?

P.S. As I will try to sell some batteries, I better not try mixing really old cells with newer, less worn ones, even if the present capacities are the same but the original capacity of the older cell was higher than the original capacity of the newer cell.

Lipower said:

Ok thanks for the advice, what I am hearing is that the older a cell, the faster its capacity decreases. Am I correct?

Click to expand...

Kinda, the more capacity percentual to its original capacity it has lost, the earlier the cell will fail.

Lipower said:

Ok thanks for the advice, what I am hearing is that the older a cell, the faster its capacity decreases. Am I correct?

Click to expand...

What range of ages are you thinking of?

In my experience, this generalized statement isn't necessarily true. After weeding out bad ones thru a good test process - I have cells in my powerwall that are 4yrs old and some are 12yrs old and they both kinds seem to perform similarly with respect to capacity loss.

The journey/storage of a cell over the years - e.g. extreme temps? extreme charge? - can affect things. A good test process can weed out ones more deeply affected.

OffGridInTheCity said:

What range of ages are you thinking of?

In my experience, this generalized statement isn't necessarily true. After weeding out bad ones thru a good test process - I have cells in my powerwall that are 4yrs old and some are 12yrs old and they both kinds seem to perform similarly with respect to capacity loss.

The journey/storage of a cell over the years - e.g. extreme temps? extreme charge? - can affect things. A good test process can weed out ones more deeply affected.

Click to expand...

I guess I wasn't thinking of ages of cells so much as percentage of original capacity that the cell has. For example, putting a 2000 mah cell that has 90% capacity with a 2500mah cell with 60% capacity.

Lipower said:

putting a 2000 mah cell that has 90% capacity with a 2500mah cell with 60% capacity.

Click to expand...

You can do that. However, the 60% is a bit too low. As mentioned earlier, 80% is pretty much a good cut off point. However, that doesn't mean that the 60% cell is worthless. Just use those cells in less critical roles, like torches/flashlights and such. Things that don't normally pull a lot of current and are used periodically.

The biggest thing to note is that you just want each parallel group of cells to be as close as possible the same "count" and same "capacity" as the other packs. This keeps the amp load balanced across all cells and helps to keep the voltage equal across all packs. These two things are pretty much necessary to help increase life cycle count.
If the counts are off, say you have a pack that has 10 cells, and another that has 15 cells, and you pull 20A, then the pack with 10 cells will experience 2A per cell, whereas the 15 cell pack will experience 1.3A per cell. Not only does this effect the cells lifecycle through extra wear, but it can also lead to voltage imbalance as the 10 cell pack will deplete faster than the 15 cell pack. Even if the overall capacity is the same.
This is because a cell that is loaded with higher current will expend its energy non-linearly faster as amps go up. For example, a 2000mAh cell @ 1A could take 2hrs to deplete. But ran at 2A, it might deplete in 50mins, not 1hr.

As a practical example - In my powerwall I have 6 separate batteries of 14s in parallel. That's 84 packs that range from 88p/pack (2950mah/cell) to 120p/pack (2166mah/cells) - but each 14s battery has the same p in the 14packs.

The capacity ranges of the individual cells at build time were 85% or higher, mostly 90% or higher of the 'new spec' - 300mah max difference between cells in a pack / battery.

The powerwall is large enough overall compared to the load that the average charge/discharge per cell <250mah - well below the 'standard' specs of any cell in the battery bank.

One could go to larger extremes - but I offer these metrics as ones that are working well and I'm at 1,207 cycles on the oldest battery.