So you used jumpers and it works. ?
Are you sure when you plug it in the header is not backward.
And when you jumper your hooking it right side up.
But when you Plug it in you maybe rotating the board.

 

Thanks for the input everyone! So I'm pretty sure that there are no components in contact when the 2 are mated, and I'm 100% sure that nothing's plugged the wrong way (I've checked thousands of times at this point now haha). I recently tried with shorter cables (as seen in the picture attached), where I'm using 4-pin rectangular housing socket connectors. The board that is connected to the bricklet in the picture is the exact same board that I've posted in this thread, just a lot smaller. The smaller cable with the 4-pin rectangular connector doesn't work, but as before, the longer breadboard jumpers work fine. I'm starting to think it might be an issue with the connectors that might not be made for the specific pin headers on the bricklet, but this doesn't make too much sense to me.

I'll keep messing around with my new and old boards I designed. I'll try wiring directly to my PCB or install pin headers instead of the socket.

 

Other things I've tried:

1. Tried different rectangular connectors. Doesn't seem to make a difference with the shorter cable. Longer breadboard cables still work fine.
2. Powered both the transmitter and the receiver with 5V using Tinkerforge's step down power supply bricklet, as seen in the attached picture. Noise appeared to get even worse.
3. I tested the transmitter and receiver individually on my PCB with the socket connector that connects directly to the bricklet, and it looks like it is the transmitter that is sending noisy info. The receiver is working fine.
4. Soldered wires directly into the PCB, as seen in the other attached picture. This appears to work fine, which is interesting. I wonder why this would be the case.

I have a redesigned PCB coming that has an internal rectangular cutout that allows me to use wires or breadboard jumpers to connect the bricklet to my PCB (as last resort). Still would like to figure out what this problem is, though...

 

Just thinking out loud, you could have either a damaged or defective fiber optic transceiver. If you have changed everything else, but reused the same transceivers, then that *could* be your issue.

 

Possible, but that won't explain why longer cables work OK.

@ryan1 is the issue symmetrical? Is Tx corrupted in both directions?

 

Thanks for the input everyone. Forgot to send an update two weeks ago.

I just gave up on the transmitter/receiver I had and used a different product. Followed the application circuit in that datasheet and everything worked fine. Should've switched products earlier haha

For anyone who is interested in using fiber optic transceivers with a UART port or who is encountering the same problem, just use these: https://www.digikey.com/product-det...nd-storage/TODX2355(F)/TODX2355(F)-ND/7100242