Since you're working on a protocol for data input into the CA anyway, how about this?
I got an idea when I saw this thread's post:
http://www.endless-sphere.com/forums/viewtopic.php?p=669709#p669709
where any BMS (or other device that can read cell voltages) could use a serial protocol to periodically send the voltages both pack and per-cell to the CA for display, if the CA is set to that mode.
If it is in the "cell-level" display mode, the bottom line turns into a bargraph, with the top line still having two other numeric values (for myself I would choose speed and either realtime amps or realtime watts, depending on what I was testing, occasionally realtime Ah, and possibly very occasionally other data).
The bargraph could display up to five cells per "character", with up to seven levels per cell, with the bottom dot representing 0V and top dot representing full-scale for whatever cell type it is. The BMS would send the scaling info first in each "frame" of data (containing the values for all cell groups in the pack), then the full pack voltage, then the most negative cell thru all the middle cells up thru the most positive cell.
If there are enough fewer cell groups than 1/2 of the total number of vertical bars possible with the screen width, then each character can be used for only two cells, with the center column of each character always left blank, and only the two leftmost and rightmost columns of a character used for the cell bargraphs, so it is even more apparent that they are each separate cell.
It could also be done that way for even higher cell counts, using both rows of the screen for bargraphs, but for myself I need the speed data onscreen all the time regardless (well, most of the time).
The update speed would probably be pretty limited, so it might not be useful for watchng realtime sag during short power bursts, but it would be helpful in pack diagnostics on longer hauls with high drains, or during charging.
Of course, until someone then develops, makes, and sells the device that can monitor the pack's cells (doesnt' ahve to do anything else) and send that data via this protocol to the CA, it's not a useful function. But if the ability was there, I would guess that someone would make the other device soon enough.
I periodically see requests for a single device to do what the CA does plus cell-level readouts in some easier-to-interpret manner than a list of voltages, so....
I recommend using the bottom dot for 0V rather than simply whatever the cell-level LVC should be, because it can also help diagnose a pack with broken balance wires (assuming the BMS isn't able to detect that and cut off discharge or charge, apparently a common enough issue).
I wouldn't expect it to be implemented anytime soon, if ever, but it would enable a small part of what the OP of the thread I linked to appears to want--a way to have cell-level voltages affect the operation of the bike. This might allow the CA to become a kind of BMS, moreso than it already can be (at pack level), by letting it monitor cell-level activity and reduce power demand to reduce voltage sag on dangerously low cells, and cutoff discharge entirely once a cell is too low.
The actual idea as-posted couldn't, but it's easy enough to extend the idea so that a cell-level LVC can exist in the CA, and it checks the data in the stream against that and anything below that LVC results in whatever action is programmed into it by the user. (or picked from the options).
