If you insist on looking at every cell block individually, you can probably accelerate the
configuration process a little by scripting up a template for a .CSV file, editing in all the
appropriate labels/offsets and maybe even gauge positions?? in a real editor with a real
keyboard, and then crossload that into Torque.
I've also pointed out the two figures that are much easier to deal with, lowest cell voltage
and highest cell voltage. You can even graph just those two figures together while putting
the pack through various forms of abuse, to try and ferret out potential (heh) issues under
high load. Data comes back a lot faster because it's only two OBD2 queries instead of 98,
assuming Torque is as stupid about multiple queries/responses under the same PID as the
other "pro level" packages. Maybe SoulSpy etc are better about aggregation?
The queries, btw, are CAN ID 7E4, mode 22, pid 0101. Returned byte X / 50 gives you max
cell voltage, Z / 50 gives you minimum. Resolution is 0.02 volt. Reorder the fields however
is appropriate for your app. Figures for *which* cells are the specific high/low ones can
also be had via the Y and AA parameters, and in theory should bounce around all over in
a well-balanced pack, but on mine, never more than 0.02V different that I've ever observed,
those high/low N tend to settle on a small handful of values like 1 and 84. That probably
has more to do with the internal scan order that the BMS does, not actual cell variance
so far. I'd be interested in what other folks see as their own min/max N, too.
_H*