One of our customers has asked us to solve a problem. It seems that of a batch of chargers and batteries we were supplied, most batteries cause the charger CHARGE state to flash at about a 1 second rate. When looking at the charge status pin, the on time varies with battery and temperature and is always shorter with a colder battery As the battery warms up, the ON time gets longer then eventually stays on above a certain temperature. This GOOD temperature varies with the different battery and charger combinations.
When looking around with a scope probe while the chargers are flashing, it would immediately stop flashing when I got to the VFB point. When removing it, it goes back to flashing. Normally, I would minimize any capacitance around any feedback path like this, but I see that the evaluation board seems to have purposefully added capacitance at that feedback point. There are relative large copper areas at the junction of the feedback resistors. That would add capacitance at that point. My estimate would be about 4-5pF. If the reference design of the evaluation board was optimized with 22pF in parallel with the top resistor, then I should probably add 5pF across the bottom resistor to compensate for the lack of capacitance due to the lack of "extra" copper at that point on the design I have.
Another possibility would be to reduce or eliminate the 22pF across the top resistor since the design I am working with has minimal extra copper at that point.
Any thoughts? I will try adding 5 pF in the equivalent position of C31 in the evaluation board and re-run my tests My previous tests were done between about 45 degrees F and about 55 degrees when they all worked OK
