katou wrote:I thought that having each parallel strand about equal was best?
There is a small flaw in that logic. It sounds like you are assuming that you are going to string matched cells in series first, then tie these series strings in parallel. That would result in a nightmare situation where you have many, many balance taps all hooking together in a big mess.
What you actually want to do is sort the cells out into "bins" - call them A, B, and C
Then when you build up your pack, you want to build them in parallel first and series second.
If it were a 20S 6P pack then you would want each 6P chunk of cells to be comprised of A,A,B,B,C,C -> a perfect balanced mix of strong and weak.
You would then tie twenty of these in series and set up one set of balance taps.
The strong will help the weak, the high AH will make up for the low AH, the high impedance will be offset by the lower impedance.
Important? yes... Critical? No -> not if you have a random sampling and you are going massively parallel like I am.
I have 48 cells in parallel - somewhat randomly matched together - 960 cells in total. The prospect of actually trying to characterize 960 cells is just retarded. It makes much more sense to just hook them all up, drive the car, and when I am feeling especially ambitious - run the battery down until the first bank reaches LVC. That is obviously the "worst bank". I can then just replace that bank - or - better yet - I can then find the bank with the highest voltage, remove both banks, and take them into the shop.
Now we are working with a more reasonable number of cells to characterize: 8 banks of 12S2P.
Match the weak with the strong and reinstall
repeat until happy.
Successive approximation is much faster and easier than brute force characterization. Got me on the road ASAP where I can do cell testing with my ass and the gas pedal
The doc is a brave man for sorting all those cells but he was only building a 2kwh pack where as I am in the process of building a 26kwh pack (10kwh running as of today). Just not worth the trouble.
If I were you - I would spend less time fretting about it and just build them up
Get your bike ready for this beautiful summer weather that is heading our way.
The last thing you will want to be doing is characterizing cells while everybody else is posting pictures of girls with big boobs riding ebikes in tiny bathing suits.
Increasing battery voltage and controller current limit will result in a non linear experience