Does the transmission medium affect the speed of a signal? For instance does light traveling through a fiber cable get a bit from $A \implies B$ faster than copper can transmit a bit the over the same distance?
Asked
Active
Viewed 1.7k times
11
-
Yes, but why do you want to know? Light can travel around the entire Earth in 1/10th of a second. The latency of transmission down the cable is often irrelevant compared to processing delays or bandwidth concerns. – endolith Dec 15 '10 at 22:18
-
I think the speed of signal in copper is related to rms speed of electron that is significantly less than speed of light. – S L Dec 16 '10 at 05:33
-
1@endolith : The latency of transmission throught the atlantic is enough to be (1) measured with standard computer https://physics.stackexchange.com/q/357/ , (2) have an impact on some video games http://en.wikipedia.org/wiki/Ping_%28video_gaming%29 (3) have an impact on high frequency trading http://arstechnica.com/tech-policy/news/2010/09/first-nyclondon-cable-in-a-decade-promises-sub-60ms-latency.ars – Frédéric Grosshans Dec 16 '10 at 10:56
-
2@explorex: no: the speed of information in cables is the speed of waves inside them, but not the speed of electron, which is order of magnitudes faster. Like the wave of the see can travel much faster than the water below them – Frédéric Grosshans Dec 16 '10 at 10:59
-
@Frédéric: Yes, but that latency is due to chains of processing delays, as I said, not from the speed of propagation of EM waves. – endolith Dec 16 '10 at 15:16
-
2@endolith: The speed of light in fiber is ~ 8 in/ns (0.2 m/ns). Certainly the processing cascade is significant (usually the largest part of the ping time, but the transmission delay is enough to measure. That is because 0.1 s is forever in computer clocks. – dmckee --- ex-moderator kitten Dec 16 '10 at 17:20
-
@Frédéric I do understand the speed of signal is the speed of waves inside the cable. I was merely implying that the waves are the flow of electrons in the cable after all (because it is impossible to transfer signal as heat). and certainly the rms speed of electrons is in the order of 10*6 or something. while the speed of light in the optical fiber is at the speed of light. so .. correct me if i am wrong. – S L Dec 16 '10 at 18:15
-
@dmckee: Yes, it's enough to measure, but it's not very significant. Time it takes wave traveling at 2/3c to travel 12500 km from NY to Mumbai: 63 ms. Actual ping time to 202.54.1.30: 230 ms – endolith Dec 16 '10 at 18:44
-
1@endolith : NY to Mumbai and back 2×63 = 126 ms : thant already 50 % of the 230 ms ping time you mention. And the cable does not take the shortest route. Among the link I gave in my previous answer, there was a project to build a shorter cable between London and NY for high frequency trading. – Frédéric Grosshans Dec 17 '10 at 09:23
-
2@explorex: According to http://en.wikipedia.org/wiki/Electric_current#Drift_speed , the rms speed of elecrons in copper wire is of the order of mm/s . Really much slower than signal speed ! – Frédéric Grosshans Dec 17 '10 at 09:29
-
@Frédéric now i do understand my mistake. I realized that signal is an voltage or electric field rather than frequency of alternating current. But if I put one end of wire here and other other half way around the world, my question is how fast will it take to reach the signal. I guess electric field travels at the speed of light. – S L Dec 18 '10 at 05:16
-
@explorex : the actual signal propagation is depending on the permittivity of the medium (see @conqenator's answer), but the order of magnitude is usually the speed of light. – Frédéric Grosshans Dec 20 '10 at 17:38
-
@experimentX ""now i do understand my mistake. I realized that signal is an voltage or electric field rather than frequency of alternating current"" No, this is still totally wrong, The "speed" of the alternating current is as high as the speed of alternating electrical field. Those are coupled as Maxwells equations say, the drift speed of electrons is something different. There were enough comments on this now! – Georg Apr 02 '11 at 18:24
-
@Georg still i don't understand. I haven't been here for pretty much time. I guess i won't be back untill I join masters. We will discuss later. – S L Apr 03 '11 at 05:36
-
@FrédéricGrosshans: This Wikipedia entry is wrong--- the speed of conduction electrons is about 5000m/s, the Fermi velocity, the estimate of mm/s is from an obsolete classical model, the Drude model. The speed of signals down the wire is a significant fraction of the speed of light. – Ron Maimon May 24 '12 at 00:28
-
1@endolith I just made a measurement of roundtrip time between two low-end PCs separated by a switch. In my case this was around 200us. That is certainly mostly processing delays, because the total length of wire being traveled in my case was less than 100m. Those 200us would involve 4 separate processing delays, so the processing delays would be only be 50us each. In that time light can only travel 15km. So if the average length of cable between hops was more than 15km, the speed of light would have been the limiting factor to me. On the internet backbone, cables are longer than that. – kasperd May 12 '15 at 08:11
4 Answers
7
The medium does affect the speed of propagation, but I don't have detailed information on the technologies currently in use.
The rule of thumb is about 2/3 of the speed in vacuum for either coax cable or fiber. No idea for twisted pair.
-
+1. In the paper "measuring the speed of light using ping" [http://arxiv.org/abs/physics/0201053], there is some discussion on the propagation speed for different cables. – Frédéric Grosshans Dec 15 '10 at 17:51
-
The paper mentioned in the comment above is useful, but the link has a typo. Here is the correct link: http://arxiv.org/abs/physics/0201053 – codingoutloud Aug 04 '12 at 17:25
7
Well, definitely yes, i think. Depending on properties of the cable as well as the Relative permittivity(also measured as dielectric constant) of the material surrounding the copper is a factor.
EDIT: The velocity factor also called wave propagation speed or velocity of propagation (VoP or vP), of a transmission medium is the speed at which a wavefront (of an acoustic signal, for example, or an electromagnetic signal, a radio signal, a light pulse in a fibre channel or a change of the electrical voltage on a copper wire) passes through the medium, relative to the speed of light. For optical signals, refractive index is a similar quantity.
Twisted pair copper cabling are known to transmit signals at speeds anywhere between 40% to 70% of that of light. So clearly, there are a lot of variables that come into play here that vary with material used, methods of cabling etc. See this table for reference.
As for your question, a quick wikipedia lookup for fiber optic cables reveals this..
The typical value for the cladding of an optical fiber is 1.46. The core value is typically 1.48. The larger the index of refraction, the slower light travels in that medium. From this information, a good rule of thumb is that signal using optical fiber for communication will travel at around 200 million meters per second.
Which at first glance, might seem like faster but does not necessarily mean "always faster" than any/all types of copper cables.
Robin Maben
- 1,187
5
Summary: A good rule of thumb is roughly 2/3 the speed of light in a vacuum for all these transmission media. Optic fiber is slower than some and faster than others.
Details:
Electrical energy normally doesn't flow inside of metals. Poynting-flow diagrams show where the energy flows, and the path(s) traveled by the information carried by that energy, although they are a bit difficult to draw.
As conqenator pointed out,
The refractive index ... Typical values for core and cladding of an optical fiber are 1.48 and 1.46, respectively.
which gives a wave propagation speed of 1/1.48 and 1/1.46 of the speed of light in a vacuum (0.67 c and 0.68 c) in these regions of a typical optical fiber.
Various popular cables have a wave propagation speed that ranges from 42% to 95% of the speed of light. In particular, 100 Ohm Cat5e cable has a typical propagation speed of 0.64 c. The 300 Ohm Twin-lead has a typical propagation speed of 0.82 c.
dmckee reminded us that it's usually adequate to approximate all of these as around 2/3 c.
So if you use Cat5e cable, bits will arrive slightly later than bits sent through fiber optic. If you use 300 Ohm Twin-lead, bits will arrive slightly before bits sent through fiber optic.
David Cary
- 1,043
1
Depends on coaxial cable's dielectric RG58 using PVC will have a Velocity of Propagation Factor(VOPF) of .66 Andrew's FSJ1 having foam dielectric has a VOPF of .78 Fiber Optic has a VOPF of .68. A pulse sent down 100' of RG58 will have 154,045pS delay, FSJ1 will have a 130,346pS delay, Fiber will have a 149,515pS delay
James Eler
- 11