25th element

No, it's not a Luc Besson movie. I wouldn't mind if it was, they're great, but it's a cell. I now have 24 cells and a 87.6 volt charger. That makes 3.65 volts per cell as the charging voltage. CALB recommends 3.6, but with bottom balancing you want to leave a bit more headroom to protect the cells. With 25 I'd get to 3.5 which is ideal. Hence the 25th element. If I can't find one used I'll have to order. The postage will be one fourth of the whole price, but it cannot be helped.

At 3.5 I will sacrifice some SOC (State of Charge) in order to make the cells last longer and also make them stay better balanced. Right now three of my 24 cells can get up to 4 volts while charging which is way too much. They drop back into line quickly after the charge ends, but it still makes me uncomfortable. The extra cell needs to find a place somewhere, but I'm sure I'll figure it out. I should also see a couple of kilometers more range, which never hurts.

Something I realized about charging. It's just a voltage per cell. It's not a target. It's just means to an end. Sure, it won't be exceeded, but it's really the current that counts. How you should charge these LiFePO4 cells is with constant voltage and a constant current until at the end when the current should gradually drop until the pack is deemed full enough. It's quite possibly the only "BMS" you need. At least for the charging phase. Of course you should make sure you don't charge when it's too cold, or too hot for that matter, and make sure your charger is reliable enough not to exceed the set voltage and also ends the charge when due.

As for protecting the battery when discharging it's really where the magic of bottom balancing happens. If you've carefully drained your cells to the same limit below three volts before using them they should all also return to that same voltage on each and every full discharge. For a bit of additional protection I've set my controller, the AXE7245, to a low voltage limit of 60 volts. That's 2.5 volts per cell. It should be quite improssible to kill any cells this way. At least if a cell dies despite these precautions it really was a bad cell and should be put away anyway. No amount of active BMS could have saved it.

That's all there is to it really. Some human-based BMS logic is needed not to do anything stupid and keeping an eye on the cells for any physical defects, but we're just fine this way. If I was building a car with a larger battery pack, and I'm not saying I wouldn't be, someday, and especially if it was for someone else, I might consider some electronics to keep an eye on the minimum and maximum voltages. Perhaps battery pack temperature as well. Just to be safe and catch a few situations that might occur when you have more voltage and the cells are hidden out of sight. On a motorcycle, there's just no need.