On combining 1800 to 3500 mAh cells in a pack...

wimpie007

Member
Joined
Apr 8, 2017
Messages
81
Hey everyone,

After testing cells for a little over a year, I've now got 1477 cells >1800mAh or 1287 cells >2000 mAh.

I'm having some doubts, as some of those cells have tested around 3500 mAh or more (Opus inaccuracy) and I'm about to combine them with cells like 1857 mAh or 2015 mAh.
Will this cause issues?

Depending on if I include the 1800 - 2000 mAh ones or not, I'll go for 14S80P or 100P.

Thanks for your insight!
 
Actually no, I have 190 cells between 1800 and 2000 mAh, and 1287 over 2000 mAh.
I didn't explain it too well above, sorry about that.

Those 190 between 1800 and 2000 mAh make the difference in either going 100P or 80P...
 
wimpie007 said:
Hey everyone,

After testing cells for a little over a year, I've now got 1477 cells >1800mAh or 1287 cells >2000 mAh.

I'm having some doubts, as some of those cells have tested around 3500 mAh or more (Opus inaccuracy) and I'm about to combine them with cells like 1857 mAh or 2015 mAh.
Will this cause issues?

Depending on if I include the 1800 - 2000 mAh ones or not, I'll go for 14S80P or 100P.

Thanks for your insight!
Imho you asking only half of the important question.

For the range of capacity included i let the experienced ones here give the guide.

But once you ask, in theory you should take in account the inner resistance as well. Now, i know rarely somebody makes a fuzz about it, and i will write that up just for explanation.
When you have cell(s) in parallel which have different Inner Resistance (IR), then under (heavy) load the cell with lower IR will deliver more current than the other(s), and exhaust her capacity quicker. If you unload before they are empty, they will have a different SOC and therefore a different voltage. But, they are tied together, so the cell with higher SOC will transfer load to the others. That is no real problem, but causes some losses on charge/discharge.

These losses are one possible reason for unbalancing the pack.

I guess, with all our second life cells that happens more or less, in small quantities, in nearly all packs put together.

As i said, nothing to worry too much about, but your question called for this explanation, so you can make your mind up by yourself.
 
I think its fine, if you not to go low or high voltage...
lets say 3.4v/4.1v?
iam using 1900 - 2200.
 
The voltage as such does not matter if you got high or low capacity cells.
Its like Cherry said. Its also about the capacity or lifespan left in the cells
If a 1800mAh cells is 1800 due to being 3000 New it have lost a lot meanwhile a 2000 could have been 2100 new.

So the capacity difference could potentially be that you mix highly used cells with new cells? are you doing that or are they in the middle ?

So basically if you dont stress the packs at all you can rule out most of the IR issues that could come. Then the capacity difference doesnt matter much
If you also can rule out that you mix cells that are EOL with new cells this can also rule out issues. Thats why you see most people using cells with close to same capacity and type. This to hope they are better matched.
 
daromer said:
The voltage as such does not matter if you got high or low capacity cells.
Its like Cherry said. Its also about the capacity or lifespan left in the cells
If a 1800mAh cells is 1800 due to being 3000 New it have lost a lot meanwhile a 2000 could have been 2100 new.

So the capacity difference could potentially be that you mix highly used cells with new cells? are you doing that or are they in the middle ?

So basically if you dont stress the packs at all you can rule out most of the IR issues that could come. Then the capacity difference doesnt matter much
If you also can rule out that you mix cells that are EOL with new cells this can also rule out issues. Thats why you see most people using cells with close to same capacity and type. This to hope they are better matched.

You are so right.
What brings me to the detail that you should not equal capacity, but percentage of New Capacity.... chuckle.
Life can be so difficult.... :D
 
Thanks again for your insights!

I'll double check the reported capacity vs original capacity before using them in my setup.

@Cherry67, I have never checked internal resistance, just because it is so difficult to measure accurately.

Will keep everyone updated once I start my build threat.
 
Its not hard to meassure IR but you need a tool for it. BUT with that said if you discharge and the discharge pans out you know that the IR is low enough for that current right now.

So testing at 1A and max useage later on is 0.5A per cell will render them "ok" for now
 
wimpie007 said:
Thanks again for your insights!

I'll double check the reported capacity vs original capacity before using them in my setup.

@Cherry67, I have never checked internal resistance, just because it is so difficult to measure accurately.

Will keep everyone updated once I start my build threat.
Wimpie, dont misunderstand me, i didnt want to recommend CARING for it, (neither the SOC issue nor the IR), hardly anyone does this seriously in this way here.
I just wanted to give the explanation of the issue, that you can start making your mind up on this for your own understanding.

Regarding measurement accurracy, i have started a thread about this.
First measurements come soon, i can already say an unmodified Opus is like throwing dice on IR. Now, nothing new to the folks, i guess. :D
 
If all Your cells are above 80% You should be fine.

You can safely mix capacities, but if You do so, Pay attention tothe fact thatYou can not draw 100 Amps from a 100P pack without drawing more than 1A from the highest capacity cells.

However, I have found a way to figure out how much load You put on the highest capacity cell, and You can then do the math and keep maximum load below the wanted max.

Example: 100p mixed from 1800 to 2800 mAh. If in this example, capacities are evenly distributed, total capacity will be 230 Ah.

Now we try drawing 100 A. Total time to empty is 2,3 Hours.

(I know the discharge curves are not linear, but this will get us near enough.)

Highest capacity cell is 2800 mAh and to empty this cell in 2,3 Hours You need to draw 1,217 Amps.
Lowest capacity cell is 1800 mAh and to empty this cell in 2,3 Hours You need to draw 782 mA.

In order to limit the drain from the high capacity cell to 1 A The max current You can draw from this pack is 82,1 A.

Another example we have a lot of 1800 - 2300 mAh and 20 really good cells clocking in at 3300 mAh. Now what?

80 x 2150 (average from the low cap cells) 172 Ah plus 20 x 3300 mAh = 66 Ah.

Pack capacity 238 Ah. Almost equal to the first example, but this time the draw is much higher from the 3300 cells.

Time to empty @ 100 A is 2,38 Hours.

3300 mAh emptied in 2,38 Hours will result in 1,38 A.

In order to keep the drain from the high capacity cell to 1 A The max current You can draw from this pack is reduced to 72,46 A

Apart from this reduction in max current, I see no problems in mixing capacities in the banks.

As long as the banks You build have the same capacity, you should be good to go.

Hope this helps a bit, best of luck building Your packs :)

ChrisD
 
IMHO unless you te expecting longish periods of high current draw IR is pretty irrelevant, also mixing a range of capacities is fine.

The thing Im very picky about is getting rid of leakers (self discharging cells), its the major thing I believe that will bring a system undone, fortunately its the easiest test to do accurately and consistently.
 
Back
Top