This was surprising. To me. I added logic to write the timestamp to the LOG file every time the battery in my EDGE 1030 computer dropped 1% point. It appears that Garmin checks the voltage level and translates that to a % charged every 5 minutes, exactly. Looks like in my case, with the backlight on 80% to accelerate the drain, it dropped 1% about every 7-8 minutes. So about half the time, it would catch a 1% drop at the 5 min check, and about half the time it would take until the next check at 10 mins to see the 1% drop.
Not a big deal... except that explains why my Time To Live estimate jumps around. I need to average it over a longer time period to smooth out the stairstep effect of the 5 min polling.
