Fibre Optic cable vs Copper wire

[labview1958]

Is the fibre optic cable cheaper than copper wire for transmission? I believe there is still loses in a fibre optic cable. As the critical angle of glass is 42 degrees, thus about 50% of the signal would be lost. Is that true?

 

[berkeman]

Is the fibre optic cable cheaper than copper wire for transmission? I believe there is still loses in a fibre optic cable. As the critical angle of glass is 42 degrees, thus about 50% of the signal would be lost. Is that true?

Not sure about the relative costs per meter, but your loss number is not right. Read about single-mode fiber. The losses are much lower.

 

[PhanthomJay]

I can't directly answer your question, but miles and miles and miles of fiber optic cable have been or are being installed by telephone, cable tv, electric, and internet provider companies, etc, using light pulse signals, . I don't see any copper (using electric current) being installed. On a material cost basis, fiber probably costs more, but does a lot more. So its more than material cost, it's greater realized income. Mind you, this response is from a non expert in the communications field.

 

[elect_eng]

Is the fibre optic cable cheaper than copper wire for transmission?

Generally, yes? Although, it depends on the data rate and distance involved. Short distance and lower data rates are more conveniently handled and probably cheaper with a copper solution.

Note that USB cables on computers are copper based because the distance is short; however, trans-atlantic telecommunications is done with fiber because the net cost is much much cheaper.

I believe there is still loses in a fibre optic cable. As the critical angle of glass is 42 degrees, thus about 50% of the signal would be lost. Is that true?

There are losses in optical fiber, of course, but the loss is much less than copper.

Note that the critical angle depends on the fiber type. There can be issues of coupling light into optical fiber due to the acceptance angle. However, this is generally not a big issue. This is because lasers and LEDs are available with high optical power, coupling methods achieve 50 % to 90 % coupling efficiency, and once the light is coupled into the fiber, it transmits with very much lower loss than copper.

The exact value of the loss depends on the wavelength of light used and the type of fiber used. The most common fiber type is made from silica glass. For these, typical communications wavelengths are 800 nm, 1300 nm and 1550 nm, all in the infrared range. The low loss window is 1550 nm with a loss of about 0.2 dB/km. Note that this equates to loss of half the light after 15 km. There is another transmission at 1300 nm with a local minimum loss of about 0.4 dB/km. The loss at 800 nm is significantly higher, but still quite good.

 

[labview1958]

Is trans Atlantic cable done with fibre optics? Trans Atlantic cable was laid before the age of fibre optics. Why change it? The actual manufacturer price of the fibre optic and copper cable per metre. Which is cheaper?

 

[Anti-Meson]

Fibre optics is cheaper and yes the trans-altantic cable has been laid several times and currently fibre optics are used.

 

[mheslep]

Fibre optics is cheaper ...

No, the cost comparison depends on the length of the cable, the bandwidth required, and the environmental conditions. High temperature FO transceivers, for example, are expensive.

 

[berkeman]

No, that depends on the length of the cable and the bandwidth required.

Links guys. My gut says fiber is more cost effective for BW/cost, but I haven't looked for supporting links...

 

[Anti-Meson]

No, that depends on the length of the cable and the bandwidth required.

What is up with you? The OP asked the question is fibre optics cheaper, in the case of the transatlantic or any distance at which laying cables is needed. Fibre optics is preferable.

 

[berkeman]

What is up with you? The OP asked the question is fibre optics cheaper, in the case of the transatlantic or any distance at which laying cables is needed. Fibre optics is preferable.

I'm only guessing, but for short distances and lower BW, UTP is cheaper overall. I agree that that's probably not the OP's question however.

 

[mgb_phys]

Is trans Atlantic cable done with fibre optics? Trans Atlantic cable was laid before the age of fibre optics. Why change it?

Fibre can carry vastly more data.
The early copper cables could carry telegrams at a few bits/second, modern fibre can carry several billion bits/second.
The fibre cable is much smaller and cheaper to make and lay, especially if you are trying to lay 1000s of km of the stuff. It's also extremely low loss so fewer amplifiers and much less power is needed.

This long, but utterly fascinating, account explains all about the wonderful world of international telecoms fibre http://www.wired.com/wired/archive/4.12/ffglass_pr.html

 

[Anti-Meson]

It's also extremely low loss so fewer amplifiers and much less power is needed.

I believe that was the primary driving force behind the decision to lay new cables with the additional pressure that more data would be needed to transferred in the future.

 

[mheslep]

What is up with you? The OP asked the question is fibre optics cheaper, in the case of the transatlantic or any distance at which laying cables is needed. Fibre optics is preferable.

The OP didn't specify a distance at all, nor bandwidth, nor environmental conditions. You assumed 'transatlantic' or 'laying' cable distances and gave a blanket answer. Cables have to be terminated in connectors, and driven by a transceiver. IF the cable is relatively short, such as in network LAN, then the connector & transceiver cost make up a significant portion of the cost. Even the cheapest FO ST connectors are more expensive installed than a comparable Ethernet RJ-45 jack. If the BW is high enough, say greater than 5-Gbit, so that copper has to move from twister pair to coax, then the balance tips back in favor of FO again. Include the cost of the transceiver and FO looks much worse than even 1Gbit-Ethernet, again for short distances.

 

[mheslep]

I'm only guessing, but for short distances and lower BW, UTP is cheaper overall. I agree that that's probably not the OP's question however.

Agreed.

 

[mheslep]

Links guys. My gut says fiber is more cost effective for BW/cost, but I haven't looked for supporting links...

Cables to Go Ethernet 100Base-TX cable - RJ-45 - M - Stranded wire 5 ft: $2
http://www.cablesondemand.com/category/LC-LC9/product/FO-LCX2SIMP00/URvars/Items/Library/InfoManage/FO-LCX2SIMP00.htm" : $13.73

 

[Anti-Meson]

mheslep & OP:

How about cost? Isn't that the bottom line for copper vs. fiber? Well, fiber prices continue to fall while copper prices (and the more sophisticated hardware needed to support high bandwidth transmission) rise. The cable plant is a wash - comparing Cat 5 and multimode fiber - both are about the same price by the time you consider everything (including fiber testers at $995 and copper testers at $4,000-6,000 and going up for Cat 6, 7 etc.)

From http://www.thefoa.org/tech/fo-or-cu.htm

by the Fibre Optic Association a international non-profit educational organization.

 

[Anti-Meson]

Cables to Go Ethernet 100Base-TX cable - RJ-45 - M - Stranded wire 5 ft: $2
http://www.cablesondemand.com/category/LC-LC9/product/FO-LCX2SIMP00/URvars/Items/Library/InfoManage/FO-LCX2SIMP00.htm" : $13.73 (to be fair I should have quoted a duplex FO cable which is more)

Referencing a commercial website and comparing two dissimilar products is pathetic. Has it not crossed your mind that they can charge whatever they want for a fibre optic cable because they KNOW people will believe it is superior and are therefore likely to buy it, so why not take a huge profit on that item. Basic business.

 

[mheslep]

The all fiber LAN article has some interesting points, but its hype laden:

So why are 99% of all desktops connected with copper? The comfort factor. Installers and customers are both more comfortable with that old familiar copper wire (even if what they install has little in common with their simplistic perception!) and they like the fact that they save $100 or so on every desktop with copper.

Yes, they're all simple rubes who don't know any better, waiting for enlightenment from the FOA.

 

[Anti-Meson]

...and they like the fact that they save $100 or so on every desktop with copper

This may be true for computers, but for telecoms, using fibre optics is substantially more cheaper than using copper.

 

[mheslep]

Referencing a commercial website and comparing two dissimilar products is pathetic.

No, but wild assertions are.

Has it not crossed your mind that they can charge whatever they want for a fibre optic cable because they KNOW people will believe it is superior and are therefore likely to buy it, so why not take a huge profit on that item. Basic business.

I for one don't believe its superior for all needs, and I buy FO cable by the km from the lowest bid when I need it. You're stacking assertion on top of assertion.

 

[Anti-Meson]

No, but wild assertions are.

I for one don't believe its superior for all needs, and I buy FO cable by the km from the lowest bid when I need it. You're stacking assertion on top of assertion.

What I am saying is that your referencing, to put it bluntly, is rubbish. Take consideration when using websites to provide evidence for your arguments. Think critically.

 

[elect_eng]

Is trans Atlantic cable done with fibre optics? Trans Atlantic cable was laid before the age of fibre optics. Why change it? The actual manufacturer price of the fibre optic and copper cable per metre. Which is cheaper?

All modern trans-ocean telecommunications is done with fiber optics. The old copper cable is essentially obsolete. The change was made because the bandwidth of fiber is so great. Each fiber has teraHertz of bandwidth and the fiber is 0.25 mm in diameter, hence allowing many separate fibers to be included in one cable.

When talking about ocean cables you need to understand that the cable itself is an engineering marvel. It includes steel strength members, protective outer coatings, copper power conductors, in addition to the fibers themselves. It must be able to withstand great depths, shark bites (no kidding!), temperature extremes and the stresses of laying the cable.

Given the above, you can see that the cost of the fiber itself (of the order of $.1 per meter) is a small part of the cable's material cost, manufacturing cost and installation cost. In addition, the number of repeaters required with copper is prohibitive in terms of their number, required electrical power and cost. Nowadays, simple and inexpensive optical amplifiers can be used (with fiber) to directly amplify the light at 1550 nm wavelength. This again reduces cost and complexity of a repeater and allows very long spacing between repeaters (many tens of km, even 100 km).

When talking about trans-ocean telecommunications, making a comparison between copper and fiber is like comparing cell phones to the old telegraphs. And, that is no exaggeration.

 

[mheslep]

This may be true for computers, but for telecoms, using fibre optics is substantially more cheaper than using copper.

No, again the answer depends on distance and required BW. Certainly for long distances and high capacity, fiber is cheaper. For short distances, copper UTP is usually cheaper. Add to the reasons I have detailed above (terminations and transceiver costs): power requirements at the end of subscriber loops.

 

[FAC]

No, again the answer depends on distance and required BW. Certainly for long distances and high capacity, fiber is cheaper. For short distances, copper UTP is usually cheaper. Add to the reasons I have detailed above (terminations and transceiver costs): power requirements at the end of subscriber loops.

There are more variables still. Scale-economics play into this in a big way.

In large enterprises, for example, due to the distance constraint of copper cables (100 meters) sometimes two or three LAN rooms are required on every floor of a building.

In a skyscraper, supporting copper could lead to using hundreds of LAN Rooms that are equal in complexity toTier II type data centers (the number of LAN rooms on campuses becomes geometrically more numerous depending on how many buildings a campus contains), where each room comes complete with dual air conditioners, UPS, humidification and fire controls, raised flooring and air returns, automated access constrols, etc. ete.

Fiber, on the other hand, has no such distance limitation, hence can be centrally located in only one or two rooms.

 

[mheslep]

..., where each room comes complete with dual air conditioners, UPS, humidification and fire controls, raised flooring and air returns, automated access constrols, etc. ete.

Fiber, on the other hand, has no such distance limitation, hence can be centrally located in only one or two rooms.

I've seen this claim from all-fiber-LAN advocates before, but at least this 'fiber needs no closets' part of it doesn't bare out in practice, at least not from what I've seen in drawing up lab installation plans in existing office bldgs. In my largish office building for instance, all of the building wiring, power and LAN, is in drop ceilings. That seems to be the today's standard practice in generic office buildings. The LAN rooms are broom closet size containing one router, a switch or two, and some punch down blocks. The high capacity trunk lines floor-to-floor are fiber-optic. There is no additional environmental control beyond what's available for offices and desktops, nor is it necessary since if the offices lose power the LAN closets serve no purpose. Data centers are another story of course, often using raised floors and specialized environmental gear, but their bldg scope is much more limited.

Second, I've not yet seen a building network design that eliminates LAN closets with all-fiber, though I'll grant it could reduce the closet count. The idea of running all the fiber lines from a basement center to each office desktop on higher floors, just because fiber can make the hop distance, is still untenable for cable routing reasons. Break out points are unavoidable, in the same way that large skyscrapers require intermediate elevator stops between the ground and top floor: the entire core of the bldg would consumed by elevator shafts. Building cabling also traditionally has the requirement of allowing cheap floor-by-floor modification and additions (e.g. a new office) without requiring entirely new cable pulls over the length of the entire building.

 

[FAC]

Second, I've not yet seen.

Excellent feedback, mh. Thanks. I'll await comments from others.. hopefully there'll be a few, and I'll return to this in a day or two, after I shake these 'gators loose.

 

[FAC]

Break out points are unavoidable, in the same way that large skyscrapers require intermediate elevator stops between the ground and top floor: the entire core of the bldg would consumed by elevator shafts. Building cabling also traditionally has the requirement of allowing cheap floor-by-floor modification and additions (e.g. a new office) without requiring entirely new cable pulls over the length of the entire building.

All corporate spaces are not equally qualified for reengineering. Lease terms, network diameter (number of segments and vertical extensions), the time lapsed since the last cabling refresh, all play a part. Smaller installations may actually be best served by WLAN segments. Assuming that fiber to the workplace is indicated, however, consider the following.

At a minimum there would be between two to four "break out" points within a structure for a variety of reasons: diversity; load sharing; fail over, etc. Ver tall structures would have more. These breakouts would be equi-distanced from the extreme floors served. E.g., an enclosure on the 8th floor would serve floors 1 through 16; an enclosure on Floor 24 would accommodate Floors 17 through 32, say. And so on.

As well, a breakout (optical patch panels) or a splice point (or even a blown fiber distribution panel) could be situated on each floor in lieu of a fully-loaded Tier II equivalent data center enclosure.

Perhaps your elevator metaphor is misplaced. The way I see it, it is precisely because of the 100 meter constraint of copper that you today have such a model in place consisting of a cascading series of bottlenecks three or four levels deep (access-distribution-aggregation- core). Fiber's passivity and insensitivity to distance allows one to forego multiple stages, in other words, and does so with only a minimal footprint.

Rather than get caught up in the same old 1980s fiber vs copper to the desk argument, I think it's important to acknowledge that all three: fiber, copper and wireless will make up the majority of next generation LANs. Wireless WLANs have already entirely supplanted unshielded twsited pair LANs in over 22% of SMBs (according to a recent Network World survey), and even WLANs require wire LAN for backhaul. That backhaul could be over copper to the nearest closet, or it could take a fiber path to a larger enclosure a half-building away. Those who wish to continue burning up power and occupying real estate gratuitously will choose the former.

Copper NICs aren't going anywhere anytime soon, so the last ten to thirty meters closest to the desk will continue to use copper Cate 5e/6A, depending on which topology standard is used: straightaway to the desk or one of several, more-flexible "zone" topologies. Tradeoffs? Sure, there are always tradeoffs, but in this case the main concessions may very well be taking place in wetware (mindset) rather than in hardware or software.

 

[mheslep]

...

Rather than get caught up in the same old 1980s fiber vs copper to the desk argument,

Agreed. Good points.