Interesting project.

I'm no expert on schematics, but something doesn't sit right with the coms section. The 5v logic is used to turn on the 2n3906 via a 4k7 current limiting resistor. To prevent a direct short R10 is used, with R8 restricting the current drawn through the transistor. Normal RS232 ports like to use 12v not 5v as shown, so wondered if this could be part of the problem. But then in the linked post it must of worked for the OP who designed the volt meter ?? uhmm.

Just for clarity, I would get an FTDI based 5v USB to Serial TTL converter, that is seen as a normal com port by the PC and see if that works... If so then the com1 port on the PC is not getting a true voltage and corrupting the system. In the original schematic there is a 20v max line. If this was 12v and the emitter of the transistor was connected to this line than that might work as it's closer to the normal RS232 voltage levels, but then again - I'm no expert and wouldn't want you to damage the PC.

One other thing you can try - check the com setting in the serial capture program and set the flow control to none... I found that with this set to any other option in many serial terminal programs you get garbage displayed on the screen.

Disclaimer: you follow any suggestions at your own risk.