Consider also simply using the physical-layer of Ethernet, using the same connectors, pair identification and pin assignment, and pulse transformer isolation at the transmitting end and the receiving end.
Oh, wow. I didn't know there would actually be circuitry inside an Ethernet connector. Transformeres, LEDs and the impedance matching in one package. Unfortunately, I think I can't use this approach.
I actually posted something about this project some time ago but didn't have time to work on the project. The network calbe is supposed to carry both the digital signal and the power supply for the receiving device. That is one twisted pair for the differential signal, one for GND, one for +12VDC and one for -12VDC supply rail. I guess can't use Ethernet connectors with built in transformers to carry DC through the other twisted pairs.
But I do get the idea. If I were to use some transciever or transmitter/reciever pair and a 75Ohm resistor to match the cable resistance, I should be ok, right?
1. What is the signal?
The Signal is an ADAT digital audio signal. 12.188Mbit/s is the bitrate. I have an optical TOSLINK input in the transmitting device. It gives me an output signal between 0,5V (Low) and 4,5V (High). I want to take the output of this optical reciever, make it differential and send it to another device that has an ADAT decoder/receiver.
2. What does the signal need to connect between, and for how long? PCBs within a single enclosure? Between enclosures in a lab? Between enclosures in an industrial setting? Between buildings or further?
I want to connect two seperate devices. Seperate enclosures, 10m-20m maximum distance. In the same room or between two rooms. All of that happening in a music studio environment.
3. What needs to be done about protection, filtering, error correcting, etc.? This will depend on answers to #2 (a long cable run is likely to pick up more interference and surges than one within a lab or enclosure), but also depends on how reliable and robust you need it. Likewise, how much noise the system is allowed to radiate is an important spec.
Puh. Hm. I have no idea. The reason that got me thinking about a differential/balanced connection was to not flood my studio with electromagnetic fields from the digital signal cable. There is lots of analog circuitry running through cables in the room. most of it is balanced/differential though.
Thanks for the answers so far. Looking through those transceiver ICs now.