[engolling]

Hello, I think we fully understood the problem today.
If the battery protection circuit was active and the battery is off. Then the TP5100 give some pulses and tries to precharge the battery.
This pulses activate the xl6009 which has to regulate its output voltage in the first Moment.
If this happens all the time the output capacitor gets charged, worst case about 45V.
This only happenes when there is no load. When the Weatherduino is connected all the time this can not happen. Moreover it depends also on the Variation of the input capacitor of the XL6009. That is why laulau got another voltage than Heini.
This is probably the reason I did not notice this in my first wired setup described in the first post because it did not happen during testing.

I do not think it happenes with the XL6009 with voltmeter because the voltmeter is load enough to prevent the charging of the capacitor. I did not notice such behaviour during my tests.

There are some measures to prevent this happening.
1. Adding 400uF capacitance at the input of the XL6009
3. Adding a resistor as dummy load at the output.

Sorry for the inconvenienve but I did not notice this could happen because it is also depending on the deviation pf the modules.

UPDATE: I just tried the same with the PCB with voltmeter and here I can not reproduce this behaviour.


Best regards