Fiber opic is faster than Cat 6

Status
Not open for further replies.
6

6985741

Member
Location
Owensboro Kentucky
Occupation
Electrician
arguing again with someone who says cat 6 is faster than fiber optic. I argue fiber runs at the speed of light with more gigs and they say it has to much light resistance.
 
Start with the speed of electromagnetic waves in copper cable is less than that in glass or plastic fiber (usually 65-75%). Then add high-frequency roll-off in the copper cables, which limits the information transfer.

The add "Cat 6A performance is defined for frequencies up to 500 MHz—twice that of Cat 6" where most single-mode fiber is limited to the speed of the driving/receiving optics ("state-of-the-art DWDM optical systems can span thousands of kilometers at 10 Gbit/s, and several hundred kilometers at 40 Gbit/s"); try that with CAT6....

"Light resistance"??? Do they mean attenuation? That's why high-power drivers exist.
 
I don't think it makes a whole hill of beans of difference for most people. Fiber can be substantially faster than copper but most times the bandwidth is attenuated by the network anyway either coming in from wherever or by the computer that it connects to.
 
Fiber is faster and goes longer distances. We wouldn't be spending all the time, money, and hassle splicing on glass if copper would work.

For short (<300ft) runs I believe you can get GigE (1000MB) to work on cat6, maybe even cat5 for shorter drops.

Fiber is also nice if you think ground isn't going to stay the same between sites. Like between the shop out back and the
main house, especially in areas with a lot of lightning.

For the rest of the local, in building stuff, it's probably a waste of money.

T3
 
6985741 said:
arguing again with someone who says cat 6 is faster than fiber optic. I argue fiber runs at the speed of light with more gigs and they say it has to much light resistance.
Click to expand...
Apples and oranges. The speed at which a waveform moves through a medium is a different consideration from how high a frequency that waveform can be. When you compare "speed" of communication through fiber vs copper, it is the latter you are talking about, not the former. The speed of waveform propagation is pretty much the same for copper and fiber.
 
zbang said:
Start with the speed of electromagnetic waves in copper cable is less than that in glass or plastic fiber (usually 65-75%). Then add high-frequency roll-off in the copper cables, which limits the information transfer.
Click to expand...

Starting base for speed is pretty much the same for either I believe. How that speed is effected by the medium it is traveling through could possibly have huge impact. The speed of light of 186k miles per second I assume is in a vacuum like space. When it hits our atmosphere it likely slows down though not enough for us to notice without use of some specialized instruments.
 
Having checked a few things....
While the speed of light is a vacuum is ~300M m/s, it looks like most optical fibers have a velocity/propagation factor (based on the refractive index) of 70-72%, which puts that in line with many PE-insulated coax cables. OTOH, CAT6 is more like 65%, so it's still slower.

And the high-frequency "attenuation" of fiber is at least a couple orders of magnitude above that of CAT6; I put quotes on that because losing the highs isn't as much of a problem as that even in single-mode fiber the light can take multiple paths and arrive at the far end at slightly different times, thus smearing sharp rise/fall times of a signal.

So which is faster? Depends on what "faster" means. For 1G-baseT and 10G-baseT, the fiber will be marginally faster over lengths supported by the Ethernet signalling.
 
zbang said:
So which is faster? Depends on what "faster" means. For 1G-baseT and 10G-baseT, the fiber will be marginally faster over lengths supported by the Ethernet signalling.
Click to expand...
The signal might be faster, but if either end can't support the bandwidth that the fiber can, it just does not matter much.

I am kind of a fan of fiber for certain applications such as distance where copper does not work all that well. But for most applications, no one will notice any real difference.
 
ggunn said:
Apples and oranges. The speed at which a waveform moves through a medium is a different consideration from how high a frequency that waveform can be. When you compare "speed" of communication through fiber vs copper, it is the latter you are talking about, not the former. The speed of waveform propagation is pretty much the same for copper and fiber.
Click to expand...

I agree.
In a data communications channel, bandwidth, data rate, and latency (of which propagation delay in the medium is often only a small part) are separate parameters describing its performance. I think that the OP was considering the data rate capability of different media like cable and fiber when he said "gigs".

The ultimate data rate performance that can theoretically be achieved over a single communications channel is given by the Shannon channel capacity C in bits/Hz: C = B log2(1 + S/N) where B is the channel bandwidth in Hz and S/N is the signal-to-noise ratio of the channel.
https://en.wikipedia.org/wiki/Shannon–Hartley_theorem

The higher the S/N that is available, the more different levels can be used to represent each data symbol and therefore send more bits per symbol without significant errors due to the noise. For example, if there are 16 levels available such as in 16-QAM (quadrature amplitude modulation), then each data symbol will convey log2(16) = 4 bits of information. Each of the 16 symbol choices in 16-QAM is a phasor with a different amplitude and phase.
https://www.osapublishing.org/getImage.cfm?img=dTcqLmxhcmdlLG9lLTI0LTI2LTMwMjk2LWcwMDQ
https://www.osapublishing.org/oe/fulltext.cfm?uri=oe-24-26-30296&id=356794
If it was only binary (i.e., two levels per symbol) then it would only convery one bit for each symbol that is sent.

In some data applications latency is much more important than the average data rate. For example, in high frequency trading:
https://www.exegy.com/2020/02/ultra-low-latency-trading-infrastructure/
 
Last edited:
Might be more of a superposition argument.
Classical says if you push an electron in on one end, an electron must pop out the other end. This is on the magnitude of instantaneous.
Light is a wave, it travels with a "known" finite speed.

Don't confuse speed with bandwidth. Light & glass has many more benefits than electrons & copper.

If you want fast signal, Rf is the place to be.
 
If you were running a server connected to Wall Street computers the propagation speed would make a difference to you, but for most other folks it's a non issue.
 
Qubit computing throws off all these old ideas of slow processing. ;)
Many magnitudes faster, which brings data transmission speeds back into view.
 
wb9vxy said:
Fiber is faster and goes longer distances. We wouldn't be spending all the time, money, and hassle splicing on glass if copper would work.

For short (
Fiber is also nice if you think ground isn't going to stay the same between sites. Like between the shop out back and the
main house, especially in areas with a lot of lightning.

For the rest of the local, in building stuff, it's probably a waste of money.

T3
Click to expand...

Fiber is also cheaper for decent length runs.

DWDM is indeed massive volume but let’s break it down to reality since DWDM is very expensive. A standard LC transceiver gives you 1 Gbps. But you can add a simple prism and buy 8 pairs if transceivers are different wavelengths to get 8 Gbps in each direction using what is called CWDM. Plus that’s if you run out of fibers. Installed cost for say a bundle of 12 is about the same as 2. So that gives you initially 6 Gbps with the ability to expand to 48 Gbps using CWDM and if you can afford that much DWDM will be even more attractive.

Not sure why plastic fiber even matters. It is only used for short runs for some proprietary controls equipment. All high volume/speed traffic uses either multi mode or single mode fiber. Multi mode is cheaper but range limited, 1 Gbps only and cannot use CWDM or DWDM. Single mode as mentioned can go for up to 40 km with good quality splices and terminations and high power transceivers. That’s why phone companies use it.

As mentioned at some point the biggest problem
Is switches. Effectively most phone companies have effectively unlimited bandwidth these days. The main limitation is in switches big enough to handle getting traffic on and off their fiber networks. And they originally pulled enormous amounts of fiber, 100 strands or more. There is an entire business in trying to utilize “dark fiber” (unused strands) for smaller local ISPs and cable companies,
 
Status
Not open for further replies.
Top