Hi there,

Here below is my first iteration of the circuit circuit, which worked perfectly without any issues:










Now I’ve designed an updated version of the same circuit, but I suspect there’s a grounding issue. When the board is connected via USB (either through UPDI and GND or RS232), it works fine. However, when it’s powered only by DC, the ADC gets stuck if I touch the board’s ground with a soldering iron. I’ve found that this reliably replicates the problem, although the root cause, possibly ESD or something similar, is still unclear.







I’ve been scratching my head trying to figure out what in the new design could be causing this, as the old circuit didn’t have this problem. The old circuit was connected to an external display transistor drier board, identical to how the transistors are integrated into the new design. Unfortunately, I don’t have the schematic of that external board handy.

If anyone has any ideas or suggestions on what to check or modify to improve this circuit, I’d really appreciate it!

New PCB layout will be in the comments as i can only attach 10 files.

 

At a first glimpse I can't see a connection between GND and AGND neither by resistor nor by ferrite bead.

 

It's telling you that there's a track missing on U4 between pin 15 and ground. At the bottom of the screen Kicad will tell you the number of unrouted tracks. Don't stop tracking that pcb until it reads zero!

 

Now I saw the connection. Ignore post #3

 

It is difficult to follow a system that is chopped into small blocks like this, AND, the exact details of the malfunction are unclear. But my guess at a cause is an island ground that should not be a island.