Work smarter not harder. Cell pairing.

[venquessa]

So I have 8 new cells. I intend to parallel them with the existing 8. Some of that 8 are lower capacity and some higher, maybe a few Ah! It would be wise to pair the lower cells with their higher counterpart.

So I did this to the new cells.


The peak charge was initially assisted by the charger, but completed by the solar. Discharge was "real loads" and real cut off voltage.

By zooming in on that I can get an ordered list of cell voltages, at their visibly widest point ... both at top of charge and at bottom.
Example. Cell 6 is "fecked" on the existing 8. It's always early up and first down. On the new pack, even above you might see that Cell 2 was first over and first under. The reverse pairs come out just as quickly. i have months of data for the existing 8, so I now have my cell pairings

 

[sunshine_eggo]

Nice shortcut.

 

[Ampster]

It would be wise to pair the lower cells with their higher counterpart.

It probably is wise if you have the time. I am not sure what the end result would be since the sum of the Ah capacity would be the same from a pack standpoint. For a couple of years I used what I called a Frankenpack of various Nissan Leaf Modules which had significant differences in capacity and age. I was paralleling seven modules into a 48 volt pack and used conservative values for maximum charge and discharge. I never had an issue and sold them for $20 a module when I replaced them with my current LFP pack.

 

[venquessa]

Putting them in parallel is not the concern. I also bought a bunch of 3Ah shim cells to put in parallel if I need them.

It's capacity matching, as best I can, the in series modules (pairs in my case).

If, for an extreme example I pair cell 6 and cell 2, the total capacity of that pair might only be 195Ah. Pairing cell 8 and cell 4 however the pairs capacity might be 205Ah. The resulting usable "pack capacity" would be the 195Ah+whatever your balancer will do... not much.

By pairing weak with strong, you make modules of as close to equal capacities as you can. Mathematically you are reducing the standard deviation.

One draw back of the short cut. It only considers pairs from the two groups of 8. It could be the case that 2 cells from the SAME 8 would be the best capacity match, but my short cut is blind to that.

 

[Ampster]

The resulting usable "pack capacity" would be the 195Ah+whatever your balancer will do... not much.

The balancer will not do anything to the indivdual cells in a pair but it will try to top balance each pair to the other pairs. You might see a small increase in capacity once the pairs are balanced but the total capacity of each pair will be the sum of each cells capacity in that pair divided by two.

By pairing weak with strong, you make modules of as close to equal capacities as you can. Mathematically you are reducing the standard deviation.

Standard deviation or not the capacity will still be the sum of the cells divided by the number of cells. By reducing the differences at the top, ie. top balancing you may see the same small increase in capacity as noted above. Once that is done your total capacity will still be dictated by the cell or cell pairs at the bottom that reach your low voltage disconnect first and shut down the pack or disconnect the discharge source.

 

[venquessa]

Well, asides having to ream out a few bus bars and make the "fold series link" to join the two halves, it all went well.

Luckily the old 8 were nearly flat when I swapped to the new 8. The new 8 were almost flat when I rebuilt the new combined pack.

I tested each pair delta before I dropped the second bus bar on. 34mv. Not enough to cause an significant current ... verified each time with a spark test. The same, as a precaution on all bus bar links. If it's sparks, stop, triple check. The only thing that sparked was the Heltec Balancer, but it always does that. One "crack" as you connect it as some of it's capacitors discharge or charge suddently to the new voltage.

Everything tighted up, BMS and balancer on it and it works. Cells are still in balance, which is part luck and part just that all cells where pretty much the same voltage anyway.

I'll have to wait to see what the top end balance looks like. 4-5kWh to charge it from the charger is wasteful. I'll let the sun at it for a few days and then top it off with the charger to test top balance.

 

[Ampster]

You will know how well balanced when they are at the top.