Other Parts Discussed in Thread: BQSTUDIO
We use this battery monitor in our product, and on numerous devices I've reflashed the IC with a new golden image. A small percentage of the devices have a lower uncompensated full capacity by up to 30% less than Qmax, and a SOC that is not close (too high or too low) to what the average voltage would suggest (as seen over multiple days). I use the uC to reset the battery monitor over I2C, and the data jumps by a lot. Then I reset it again and again in short succession and see more jumps.
Remaining capacity / full charge capacity = SOC (compensated and uncompensated values are similar) at voltage and average current
3885/17794 = 22% at 7.5V and -179 mA (Oldest)
4351/18595 = 24% at 7.59V and 113 mA
18714/18714 = 100% at 7.56V and -179 mA
7204/18656 = 39% at 7.52V and -174 mA
11613/18654 = 63% at 7.48V and -179 mA (Newest)
How is this possible? Charging kicked in for a small time, but the change from 39 to 63 was during continuous discharge. To reiterate, on most devices/ICs the SOC is stable and agrees with voltage pretty well, but on a few cases there is a discrepancy in SOC that changes during reset. Then when I reset again it changes the SOC by a lot. If the first guess at SOC after a reset/power loss or reflash is suspicious, should the device be reset once? Are there conditions where a reset should be avoided?
