2

In buildings, it is common for the same cable (coax) to be used for pay-TV and the Internet.

It makes sense that I can transmit several TV channels on the same cable, since it can be modulated by the frequency and sends them. However, in Internet communication, data must be sent and received. There is no cable for the transmission and another for reception. So how is Internet communication on a coaxial cable?

Peter Mortensen
  • 1,676
  • 3
  • 17
  • 23
mrlucasrib
  • 121
  • 1
  • 3
  • 5
    Have you tried looking this subject up? The idea of stack exchange is for questions which can't be answered by simple web searches. – Chris Stratton Mar 02 '19 at 15:47
  • 8
    It's not entirely simple, but the answer is https://en.m.wikipedia.org/wiki/DOCSIS – pjc50 Mar 02 '19 at 15:52
  • 2
    Lucas, the same as for cable can be said for wireless communication... – Marcus Müller Mar 02 '19 at 15:55
  • 1
    @Chris, I do not consider this question trivial because I asked two professors, one PhD in engineering, and another master. one could not answer me and another said that there is a multiplexing by time (ie, the current would stop and wait to be able to change the direction.) what for this new information or it is wrong, or who answered me. – mrlucasrib Mar 02 '19 at 19:50
  • @pjc50 I already researched this source, however it talks about protocols, not about specific things of electromagnetism. – mrlucasrib Mar 02 '19 at 19:51
  • @Marcus in my understanding are different things, for me it is plausible electromagnetic waves traveling through space, but what puzzles me is like in the same thread, there are currents going in opposite directions. For a passing current it needs a potential difference. similarly, it is as if in two boxes of water at different heights, connected by a pipe, the water rises the barrel to the highest box – mrlucasrib Mar 02 '19 at 19:53
  • 1
    @Lucas, I did not say the question was trivial, but that it could be answered a simple web search. It is the *search* which is simple, not the *search results*. And part if the point is that what you can find there is more comprehensive than would fit in a response here. – Chris Stratton Mar 02 '19 at 20:05
  • 2
    @ChrisStratton I think "just google it" is unhelpful especially if you admit that the answer is difficult to understand. – pjc50 Mar 02 '19 at 20:33

3 Answers3

8

Energy moves through the cable in both directions simultaneously. Just as different video signals are modulated on different channel frequencies, incoming and outgoing data streams are modulated on different carrier frequencies, and pass each other without interference.

AnalogKid
  • 16,865
  • 1
  • 13
  • 25
  • If the current can travel both ways at the same time, why is this not common? For example in ethernet cables, USB cables, etc. one wire is used for Tx and another for Rx. Is there any reason for this? PS: At school, an understanding is created that the ordered movement of electrons generates a current. But if they are going in opposite directions will not they collide? – mrlucasrib Mar 02 '19 at 19:27
  • 1
    @Lucas depending on the standard, both USB(USB 2 and lower) and ethernet (1000 base-T) use a bidirectional interface. In USB 2's case, it's half duplex so only the root or device can transmit at one time. In Ethernet's case some special processing is used to subtract the transmitted signal from the data so the other signal can be received. – C_Elegans Mar 02 '19 at 20:01
  • 2
    Think about what happens when two people try to talk at the same time. Do their voices actually collide? To make it even clearer, say it is a young child and their father, so we might say different frequency bands are being used. – Chris Stratton Mar 02 '19 at 20:02
  • 1
    @Lucas the reason many interfaces use separate rx and tx lines is because of cost. Adding the extra processing and circuitry to be able to separate the two signals on one wire makes the hardware cost more. Granted, using separate lanes makes the cable cost more, but that tends to be less of an issue for how long the cables are (ethernet is pretty long, but USB, SATA and PCIe are not). Compare this to how much cable the cable company needs to lay, and it's a bit more reasonable that they would favor expensive hardware and cheap(er) cable – C_Elegans Mar 02 '19 at 20:04
  • 4
    It is very common. The telephone was invented in 1876. – AnalogKid Mar 02 '19 at 20:09
  • 4
    @Lucas as with most electronics questions, ignore the electrons - they will confuse you. The real signal carrier is the electromagnetic fields. It's much easier to understand it as a radio signal that has been contained in a tube. – pjc50 Mar 02 '19 at 20:32
  • But @ChrisStratton, to make the analogy complete, what happens if a human being's ears were located inside their mouth? Any signal on our eardrums would be dominated by our own speech. Similarly, if we have a voltage sensor at our end of the coax and at the same time, impose our own signal at our end of the coax, we will only hear own own signal. – Mark Jul 09 '21 at 22:32
  • Mark - not at all correct. For bi-directional cable TV and internet signals on a single coax cable, the signal types and directions are separated into distinct frequency bands. Bandpass filters prevent local outgoing signals from "dominating" what is being received. Hybrid circuits did this in the late 1800's (yes, that is an 8), decades before the invention of the vacuum tube. – AnalogKid Jul 10 '21 at 00:23
3

Radio frequency communications operate the same when travelling through a coaxial cable as they do through open-air. They are just shielded from outside interference (called ingress) and leakage (called egress). As such, signals of differing frequencies can co-exist, with each travelling in different directions.

Amplification, however, is a different story. Since amplifiers work in only one direction, the incoming and outgoing signals need to be separated when amplification is necessary. This is performed by a device called a diplex filter, which is sort of like a splitter/combiner that splits/combines based upon the frequency of the signal. In legacy CATV systems, downstream signals were generally about 50 MHz (around the bottom of analog channel 2) and up, while upstream signals were from around 5 MHz to 40 MHz.

An amplifier assembly would (basically) consist of a diplex filter on one end separating the two frequency ranges, followed by an amplifier for each frequency range oriented in opposite directions, and then a second diplex filter to merge the two frequency ranges to its original full spectrum signal.

Peter Mortensen
  • 1,676
  • 3
  • 17
  • 23
Hitek
  • 852
  • 4
  • 10
  • Thanks for editing for readability and adding the link, Peter! – Hitek Mar 03 '19 at 19:18
  • THANK YOU, I was searching everywhere to find out whether electromagnetic waves could be conducted through copper or not. So there _are_ two ways to transmit data, either by modulating voltage in a circuit (requires two wires) or inducing an electromagnetic signal, just like how light (which are higher frequency electromagnetic waves) conducts through glass fibers! – Lamp Dec 01 '20 at 20:51
2

Internet over CATV is called DOCSIS (https://en.wikipedia.org/wiki/DOCSIS).

It uses several channels separated by frequency for downstream and upstream. Think about FM radio. How can you have several channels on radio? They just use different frequencies.

This is called "Frequency-division multiplexing" (https://en.wikipedia.org/wiki/Frequency-division_multiplexing)

Here is an article that covers it: https://volpefirm.com/docsis101_rf-fundamentals/

user996142
  • 121
  • 1