Paralleling Different Cells - REPT and Lishen - Best Practices?

[baipin]

I'm the proud owner of 16 cells; 8 "new" Lishen 272Ah cells (previous owner/friend took good care of them but never used then in a contracted build that fell through) and 8 "new" 280Ah REPT cells. All of these cells were made from about 2020 to 2022 and none have many if any cycles on them.

I need to assemble these 16 cells into a 24V pack used as the house battery for my 4x4 motorhome. Is the best course of action to make two discrete batteries, one with only REPT, the other with only Lishen cells, and then parallel them after their BMSs, or would it be preferable to mix the REPT 280AH (0.022uOhm) and Lishen 272Ah (0.012uOhm) cells into a single 2p8s battery with a single BMS? For balancing, I would use a technique like this: https://diysolarforum.com/threads/parallel-cell-capacity-balancing-pccb-procedure.26301/ and would use flexible busbars to acount for cell terminal height differences.

One of the reasons a single large battery with a single BMS appeals to me, is it allows me to use an economical BMS like the electrodacus with a contactor, which I like for its simplicity, apparent reliability, and ability to talk to other components. Not FETs to worry about. I haven't been paying attention to DIY LiFePO4 stuff for a while though, so are there some better options? Is the QUCC/JBD contactor BMS any good now? I remember years ago, people spoke of some quirky issues with them. Back then at least, having two contactor BMSs for two packs was prohibitively expensive for me, or in the case of the QUCC, folks seemed to run into issues when tthey were paralleled.

For some context about my previous experience paralleling different cell brands...

A while back I had asked about paralleling EVE 280Ah (0.017uOhm) and REPT 280Ah (0.022uOhm) cells. I ended up paralleling both 1p8s batteries at the pack level, each with their own JBD 8s 100A 24V BMS. The pack was only recently comissioned and cycling but seems to be going well, there is one cell in the REPT and two in the EVE that sit a bit low and slightly reduce total capacity but it's negligible. That pack was sold a while ago and the current owner loves it.

That was in 2021

This got me thinking. Have best practices for this sort of thing changed at all? I've read a number of posts on this forum and others, most from several years ago, and could not find any consensus on what option was best to make a 24V back out of 16 cells, half of which are one brand, half another...

Some input would be generously appreciated.

 

[chrisski]

Are these cells just as tall? If one is taller than the other, then busbars won’t sit flat without leveling the cells first.

IME, getting a taller busbar connected to a shorter fuse by busbar is not always an easy task. The bus bar must be flat to both sides of the connection. You can’t even have a hairs worth of separation. For me spacers and washers against wood has worked, but may not for cells.

For me, I am building my first parallel battery 2p8s, and another problem I found was the terminals are so close I had to cut some of the busbar off. Anywhere from a quarter inch to 1/16”.

For a 2p8s configuration took 25 busbars versus 14 for 2 x 8s. I found that busbars can be bought for 8s, but the 2p8s is harder and I had to make my own. It’s taking me many, many hours to cut and tin these.

 

[42OhmsPA]

Is the best course of action to make two discrete batteries, one with only REPT, the other with only Lishen cells, and then parallel them after their BMSs

This is the route I would go. It would also give you redundancy and lower C rates.