While I was testing and debugging for ADP5062 Overcharging, I noticed that on some units, the recharge cycles are occurring much more frequently than they should. I've attached logs from 2 units showing the battery voltage and charger state once the battery is charged - one (EOC-A) runs a recharge cycle after 22 hours, after a voltage drop of around 61mV, and one (EOC-B) runs a recharge cycle after around 30 minutes, after a voltage drop of only 26mV. Both units were configured identically (read during initial CC period of charging cycle):
|Manufacturer and model ID||0x19|
|VINx pin settings||0x06||500mA|
|Charging current settings||0x1F||400mA/80mA|
|Functional Settings 1||0x05|
|Functional Settings 2||0x07|
|Charger Status 1||0x62||CC|
|Charger Status 2||0xE4||Thermistor ok, Sense > Weak|
|Charger Status Pin||0x00|
|Charger Sys En Pin||0x00|
The batteries being charged are 1900mAh Li-Ion cells, and the capacity measure is read from a gas gauge, which is forced to 100% at the end of a charge cycle (that's the step you can see at 5:45 in EOC-B). I have other logs which show the same pattern of increasingly-regular recharge cycles as EOC-B, but repeated multiple times.
The charger chip doesn't seem to be waiting for the 80mV recharge voltage threshold before starting a recharge cycle. This doesn't seem to match the behaviour specified in the datasheet - what could cause this?