These "close-but-not-the-same" module variants are super annoying!
Ideally all the modules in a particular pack are the same.
From a safety standpoint, the LiBCM firmware is far enough along that it won't overcharge/overdischarge mixed cell packs, as long as all cell types can handle being operated between CELL_VMAX_REGEN
and CELL_VMIN_ASSIST
.
Beyond that, though, the LiBCM firmware isn't far enough along that I can recommend mixing cell types. So while it's 'safe' to operate mixed cell packs, I don't recommend it. If you do decide to run different cell types in the same pack, you'll need to set STACK_mAh_NOM
to the LOWEST nominal cell capacity.
In your specific case, the cells are probably the same chemistry and are about the same size... so as long as you change STACK_mAh_NOM
to 47000 (and not 49500) it should probably work just fine.
In a couple months the firmware will support "per-cell SoC and SoH", at which point we can revisit this topic. For now my blanket statement is "don't mix different cell types".
...
FYI: The QTY5 modules I have are all LX68-10C790, 2.07 kWh. That's what I will focus my efforts on. I must limit my focus, as there are many more cell chemistries out there than I have hours to characterize them. All of these tangents are fun to think about, but right now they're preventing me from focusing on core firmware development.