Please take a look at the sensor board, the upper part is made of FPC (two-layer board), and the lower part is the data acquisition hard board (two-layer board).
Description of the problem: The board reserved an LED without welding. The iic communication and serial communication of the sensor are normal when the sensor is powered on for the first time. After the LED is welded, it can be used normally for the first time, but the AMS1117-3.3 chip is very hot, and the LED is not bright when it is powered on again, the iic address scan is missing, and the serial port cannot be connected. The output voltage of AMS1117-3.3 is normal when measured with a multimeter.

 

Please check for short circuit and broken traces in your setup.

 

I also suspect a shorted circuit. Quite probably two outputs tied together.

AND I do not follow the logic path in such a collection of isolated blocks.
To be more specific and detailed, that sort of representation is not useful in signal flow analysis as I do it.
I understand that some CAD and simulation software produces those drawings rather than circuit drawings. So this is not a personal critical remark, just an unfortunate reality.

 

No response in two weeks! Is the problem solved??
AND, Really, following a signal in such a non-connected set of blocks would be a serious challenge.
Besides that, I did not see the LED in any of the circuits.