@carlsbad2 "i think you meant to @ the OP".
No, I really meant you! Mind you, half the time the @ does not actively link for me, witness this post.
You had written: "ethernet input is only for ethernet. Try to put a processed signal into it and the DAC will not know what to do with it".
Frankly, the second part is nonsense! There is an international standard called the Open Systems Interconnection (OSI) model which defines seven layers of communication, each building on the lower layers. At the bottom is the Physical Layer, then comes the Data Link layer. These two layers are where Ethernet plays and are the ’how’ of moving data packets.
The higher levels are Network, Transport, Session, Presentation, and Application layers, which is where the ’what’ is being transmitted is agreed.
The internet uses a similar, but much less rigorously defined, model with four layers and does not really define the lower levels where Ethernet plays.
Unlike USB, Ethernet is in effect a broadcast technology where no device has a controlling role. Any device can transmit whenever it likes, provided two conditions are met. The first is that it can detect no other transmission when it starts broadcasting, and it can detect no other transmission when it stops. If collisions are detected, it must wait before attempting to re-transmit. This is why Ethernet cannot guarantee timing.
For collision detection to always work, Ethernet messages must be sufficiently long that two cannot get scrambled in the middle of the network without the scramble being detected at the far ends. The maximum separation and the speed of transmission determine the minimum message length, which is about 500 bytes. Given so much message, the Ethernet designers did not scrimp when deciding how long to make an Ethernet address - they specified 6 bytes which is 48 bits - a number about 65,000 times bigger than the miserable 32 bits Internet Protocol version 4 struggles with. So every Ethernet device ever made has an absolutely unique address, known as the Media Access Code or MAC.
Physically, Ethernet has evolved since the original near-rigid ’garden-hose’ coaxial cable which needed to be drilled at defined points 2-metres apart to tap for a new connection. This was quickly replaced by thinner, more flexible coaxial cable then with twisted pairs or fibre. Whereas the original cable could connect many devices, subsequent cables were point-to-point with the networking connections provided by repeaters, bridges, routers / switches and gateways. Neatly, these hardware devices represent different layers of the OSI model, from the bottom up!
Starting with the actual data to be transmitted, each layer from the Application down tops and tails that data with metadata pertinent to that layer. Ethernet provides the final wrapper and is totally unconcerned about any metadata other layers have added. It is the other layers that define what the data represents and how it should be handled.
Provided the DAC and the sender use the same higher level protocols, they will understand what is being transmitted and how to handle it.
