Why use optical fiber, anyway? We use optical fiber to transmit all types of data and communications signals over all distances, short or long. Because the amount of data and communications signals used for business and personal use has grown exponentially in recent years, sending high amounts of signal has become necessary for modern life. In 1987, for example, an average home used about 3000 Hz worth of bandwidth (basically one telephone line). Now, it's not uncommon for the same home to use two 56 KHz modems transferring data; 14,400 baud for fax transmissions, several hundred MHz of cable television, and 3000 Hz of voice, all at the same time. This situation is often more dramatic for businesses.
Here are a few of the standards that high-end, quality fiber should adhere to:
*** also see www.fiberoptic.com
The most common fiber test tool is an "OTDR" (Optical Time Domain Reflectometer). These babies are EXPENSIVE (a Fluke OTDR meter costs 32 Grand). An OTDR is an instrument that analyzes the light loss in an optical fiber in optical network trouble shooting. An OTDR injects a short, intense laser pulse into the optical fiber and measures the backscatter and reflection of light as a function of time. The reflected light characteristics are analyzed to determine the location of any fiber optic breaks or splice losses.
There are many other fiber test meters as well, which are less expensive, and perform a multitude of functions, such as continuity testing, BERT (Bit Error Rate Tester), etc..
BER of Fiber Systems
The bit-error ratio is usually given by an exponential number such as 10-6. This means that on average, one error occurs for every million pulses sent. Typical error ratios for optical fiber systems range from 10-9 to 10-12 (localarea networks require a BER as low as 10-12). The BER can be measured by repeatedly transmitting and receiving a suitable length of a pseudo-random bit sequence (PRBS) data. The necessary optical power to achieve a given signal-to-noise ratio, and therefore a certain BER, is called the receiver’s sensitivity.
Advantages of Fiber
Our need for more signal transmission capacity is urgent and increasing. The communications networks built of copper wire simply cannot keep up. Yes, copper is reliable and served us well for decades, but it cannot operate well at high signal transfer rates. For data transfer rates of above 50 million bits per second (Mb/s), you need special systems copper wiring. Above 150 Mb/s, even the best copper wiring is questionable. Optical fiber has many advantages - here we list them:
This is a measure of the time required for a signal or pulse to travel through a medium. The following table lists the typical propagation delay times in ns/m for various media.
Weight - A 100 meter coaxial cable transmitting 500 megabits of data per unit time is about 800 percent heavier than a plastic fiber cable and about 700 percent heavier than a hard-clad silica cable, all of equal length and transmitting at the same data rate.
Diameter - fiber cables are a tiny fraction of the diameter of copper cables, as shown below:
High Speed - Optical fiber can handle transmission rates many times that of copper. At the time of this writing, systems operating at 40 billion bits per second (gigabits per second, or Gb/s) over a single fiber are common, and we are not really sure how high our bit rates will be able to go. We are certainly nowhere near the limit yet.
EMI (Electro-Magnetic Interference) Immunity - because optical fiber cable is nonmetallic, it cannot emit or pick up electromagnetic interference (EMI) or radio frequency interference (RFI), each of which is a problem with metallic conductors.
Cross-Talk Immunity - with side-by-side copper wires, the signals in each wire create magnetic fields which partially transmit the signal to the adjacent wire. This is called cross-talk. But since Fiber does not use elctricity, it has no cross-talk.
No Short Circuits or Bad Grounds - unlike electrical copper cables, fiber-optic cables have no grounding or shorting problems. This is also an important feature when installing communications wiring in hazardous environments. Finally, optical fiber cabling causes no sparking or excessive heat-even when broken.
Security - the security of optical fiber is also far superior to that of copper wire. Electronic bugging depends on electromagnetic monitoring. Because optical fibers carry light rather than electricity, they are immune to bugging. In order to plant a bug on an optical fiber cable, you have to physically tap the cables. However, this is easily detectable because the signal diminishes and error rates increase.
The Components of a Fiber Cable
You've probably heard the terms "light tube" and "conduit for light" in reference to an optical fiber. These terms might lead you to think there's some type of hole in the middle of an optical fiber. This is not the case. Looking at a typical optical fiber cross section helps explain its operation.
1-way Simplex Fiber 2-way Duplex Fiber (Pair)
There are three concentric layers to an optical fiber. Light pulses only through the glass core of the fiber. The cladding (which is a different type of glass) serves as a barrier to keep the light within the core, functioning much like a mirrored surface. The buffer layer (sometimes called the coating) has nothing to do with light transmission and is used only for mechanical strength and protection.
As you can see, light does flow down the center of a fiber like water through a pipe. We could even say that the fiber is a "virtual" tube. Light stays in the center of the fiber, not because there is a physical opening there, but because the cladding glass reflects any escaping light back to the core
These "light tubes" are very thin strands of ultra-pure glass. The dimensions of typical fiber components are:
The core is one density of glass, the cladding is a second grade of glass, and the coating is a plastic.
To suitably protect our glass fibers, we package them in cabling. It's a misconception that fiber cables are fragile. Many people believe if you drop a fiber cable on a hard surface, it will shatter-after all, it's made of glass! But, this is a myth. The fiber itself (not the cable, but only the thin fiber) is fairly fragile, although it's surprisingly flexible and will not break easily like sheet glass. (Actually, it's several times as strong as steel, but because it is so thin, it can be broken fairly easily.)
Fiber cables, however, are not fragile at all. In fact they're often more durable than copper communication cables. Optical cables encase the glass fibers in several layers of protection.
The first protective layer is the coating we mentioned earlier. The next layer is a buffer layer, which is typically extruded over the coating to further increase the strength of the single fibers. This buffer can be of either a loose tube or tight tube design. Most datacom cables are made using either one of these two constructions. A third type, the ribbon cable, is frequently used in the telecommunications work and may be used for datacom applications in the future. It uses a modified type of tight buffering.
After the buffer layer, the cable contains a strength member. Most commonly, it's a Kevlar fabric, the material used in bulletproof vests. The strength member protects the fiber and also carries the tensions of pulling the cable. (You should never pull fiber cable by the fibers themselves.)
After the strength member comes the outer jacket of the cable, which is typically some type of polyethylene or PVC. In many cases, however, there will be additional stiffening members, which also increase the cable's strength and durability.
Important field-installed components As with copper cabling, you have field-installed components used to make a working datacom system.
Fiber sizing. The size of the optical fiber is the outer diameter of its core and cladding. For example, a size given as "62.5/125" indicates a fiber having a core of 62.5 microns and a cladding of 125 microns. We don't typically mention the coating in the size, because it has no effect on the light-carrying characteristics of the fiber.
The core is the part of the fiber that actually carries the light pulses for transferring data. It consists of plastic or glass. The core sizes of joined fibers must match. Larger cores have greater light-carrying capacity than smaller cores, but may cause greater signal distortion.
The cladding sets a boundary around the fiber, so light running into this boundary reflects back into the cable. This keeps the light moving down the cable, keeping it from escaping. Claddings can be glass or plastic, and they always have a different density than that of the core.
Coatings are multiple layers of ultraviolet curable acrylate plastic. This adds strength to the fiber, protects it, and absorbs shock. These coatings come between 25 microns and 100 microns in thickness. You can strip coatings from the fiber (which you must do for terminating) either mechanically or chemically, depending on the type of plastic.
Manufacturing optical fiber is a difficult and complex process. In general, the process entails three parts:
1. The manufacture of a preform. This is a cylinder of glass from which the optical fiber will be made. It's generally about 3 ft long and 1 in. wide. The preform has a physical makeup identical to the final fiber (including both core and cladding), except it's much wider and shorter.
2. Pulling the fiber. The manufacturer heats the preform very precisely to pull a thin strand of glass off of one end. Variances in heating and pulling tension control the diameter of this strand. This is the optical fiber, which contains both the core and cladding.
3. Cooling, coating, and winding. Once the fiber is pulled off of the preform, it's carefully and slowly cooled, covered with the final coating, and wound onto reels.
After the fiber is pulled, the manufacturer applies a protective coating just after the formation of the hair-thin fiber. The coating provides protection and prevents the ingress of water into any fiber surface cracks. The coating consists of a soft inner and harder outer coating. The thickness of the coating varies between 62.5 mm and 187.5 mm.
You can strip these coatings by mechanical means. You must remove them before splicing fibers or fitting them with connectors.
Fiber Connectors (for more info - see Optical Fiber Connectors)
Fiber connectors are used to make nonpermanent connections at fiber ends. Because they have such as small diameter, you must hold optical fibers rigidly in place and accurately align them to mate with other fibers, light sources, or light detectors. Advances in design and technology make connector installation easy today. But, it wasn't always this way.
Many use connectors as canceling fixtures for temporary non-fixed joints allowing them to be "plugged-in" and disconnected many times. Since no connector is ideal for every situation, manufacturers developed this has lead to the development a variety of styles and types.
Connector compatibility usually exists between manufacturers (one company's ST connector can be used interchangeably with another's). Adapters are available in either sleeve connectors or patch cords to allow coupling of different types of connectors. Although no single connector is best for every application, listed below are the currently popular connectors.
As the fiber-optic field began to develop, one of the biggest mechanical problems was how to permanently fix the fibers at their ends. The first fiber connectors were difficult to install. They used a variety of glues, ovens, and long, difficult polishing methods. Since then, things have drastically changed. Although terminating a fiber is not yet as easy as installing a coaxial cable connector, it's far, far easier than before. In fact, you can terminate a fiber in about half the previous required time, and the process continues to get easier as time goes on. Within a few years it should be quite simple.
When installing a fiber system (the whole system of optical fibers is often called a cable plant), you must test it to verify its performance. Basically, you're making sure light will pass through the system properly. There are three types of optical testing:
Continuity Testing - a simple visible light test. Its purpose is to make sure the fibers in your cables are continuous (unbroken). You do this with a modified type of flashlight device and the naked eye. It takes only a few minutes to perform.
Power Testing - accurately measures the quality of optical fiber links. As shown in Fig. 3, a calibrated light source puts infrared light into one end of the fiber and a calibrated meter measures the light arriving at the other end of the fiber. You measure the loss of light in the fiber in decibels.
OTDR (Optical Time Domain Reflectometer) Testing - this is a device that uses light backscattering to analyze fibers. Basically, the OTDR takes a snapshot of the fiber's optical characteristics by sending a high powered pulse into the fiber and measuring the light scattered back toward the instrument. You can use an OTDR to locate fiber breaks, splices, and connectors as well as to measure loss. The OTDR method may not give the same value for loss as a source and power meter due to the different methods of measuring loss. However, the OTDR gives a graphic display of the status of the fiber under test. Another advantage is it requires access to only one end of the fiber. As useful as the OTDR is, it's not necessary in the majority of situations. Also, it's quite expensive. Even when they are necessary, many installers prefer to rent rather than purchase them.
Fiber Optics Technical Concepts
In addition to the things we've covered so far, you must understand several technical concepts.
Attenuation - this performance characteristic is the measure of weakening of an optical signal as it passes through a fiber. In other words, it's a measure of signal loss. Attenuation in an optical fiber is a result of two factors: absorption and scattering.
Absorption - the absorption of light and its conversion to heat by molecules in the glass. Primary absorbers include residual deposits of chemicals used in the manufacturing process to modify the characteristics of the glass. This absorption occurs at definite wavelengths. (Remember, the wavelength of light signifies its color and its place in the electromagnetic spectrum.) It's determined by the elements in the glass and is most pronounced at the wavelengths around 1000 nm, 1400 nm, and above 1600 nm.
Scattering - the largest cause of attenuation. It occurs when light collides with individual atoms in the glass and is knocked off its original course. Fiber optic systems transmit in the "windows" created between the absorption bands at 850 nm, 1300 nm, and 1550 nm wavelengths, for which lasers and detectors can be easily made.
Networks - to communicate between several pieces of equipment (for example, between 20 different computers in an office), you have to connect them together. To do this, you must:
There are many types of networks, each with their own strengths and weaknesses. In fact, you've probably heard of them: Ethernet, 10base T, FDDI, ATM, and Token Ring. These are simply different methods of connecting computers together.
Bandwidth - the range of signal frequencies or bit rate at which a fiber system can operate. Basically, it's a measure of the amount of signal able to be put through a fiber. Higher bandwidth means more data per second; lower bandwidth means less signal.
Total internal reflection. Optical fiber functions well for signal transmission because of the principle of total internal reflection. Here's how this works.
When light goes from one material to another of a different density (the index of refraction), the light's path will bend. This is why you can see the bottom of a clear pond at the edge you're standing at, but when you look across the pond, you can see only a reflection of the other side. At a certain angle, light will not pass through the surface, but bounces off.
Optical fiber uses this phenomenon to bend the light at its core/cladding boundary and trap the light in its core. By choosing the material differences between the core and cladding, you can select the angle of light at which total internal reflection occurs.
Numerical aperture. The selected angle defines a primary fiber specification: the numerical aperture (NA) of a fiber. NA designates the angle (called the angle of acceptance). This is the angle beyond which the light rays injected into an optical fiber are no longer guided. Instead, they will pass through the core/clad boundary and be lost.
Fibers with higher NAs will accept a wider range of light paths. (The technical term for paths is modes.) And because there are so many modes, the signal distorts. So, a fiber with a higher NA will have increased signal distortion and be able to carry less signal. We can then say high NA fibers have less bandwidth while low NA fibers have greater bandwidth.
Index of refraction (IR). The IR of a material is the ratio of the speed of light in vacuum to that in the material. In other words, the IR is a measure of how much light slows down after it enters the material. Since light has its highest speed in vacuum (approximately 300,000 kilometers per second), and slows down whenever it enters any medium (water, plastic, glass, crystal, oil, etc.), the IR of any material is greater than one.
For example, the IR of a vacuum is 1 while the IR of glass and plastic optical fibers is around 1.5. Water has an IR of around 1.3. Light signals moving through an optical fiber travel at considerably less than the speed of light. The "speed of light" is its speed in a vacuum; not in all materials.
Pulse spreading. Signal distortion comes from two primary causes:
The colors of light through the fiber (chromatic dispersion) and
The path the light takes as it moves through the fiber (modal dispersion).
Both of these reasons have the same final effect: distortion of the signal by pulse spreading.
Notice the digital signal input into the fiber is square. As the signal travels down the fiber, it distorts and begins to spread. Pulse spreading is not a loss of light. In fact, as much light is leaving the fiber as entering it. However, the light signals distort. If the pulses spread too much, they will be unintelligible to the receiver. Let's see why both color and path cause this pulse spreading.
Dispersion - there are two potentially confusing terms you'll come across in your readings: Chromatic dispersion and modal dispersion. In both of these terms, the word "dispersion" refers to the spreading of light pulses until they overlap one another. This distorts and causes the loss of the data signal.
Chromatic dispersion is signal distortion due to color, and modal dispersion is signal distortion due to path. Dispersion is not a loss of light; it's a distortion of the signal. Thus, dispersion and attenuation are two very different and unrelated problems. Attenuation is a loss of light; dispersion is a distortion of the light signals.
Chromatic dispersion. This phenomenon (pulse spreading due to the colors of light sent through the fiber) occurs because of one fact: Different colors of light (which we also call different wavelengths) travel at different speeds in an optical fiber.
For example, if you send two different wavelengths (colors) of light into a long fiber at the same time, one will reach the far end before the other.
Here, it's obvious the time difference between the wavelengths arriving at the end of a long fiber would spread a data pulse.
Because of chromatic dispersion, you should use light sources that put out only one color of light. Many of the newer lasers do this well. And even though these lasers are more expensive than LED light sources, they often cause far less chromatic dispersion. Most use LED-type sources only for shorter runs, where higher chromatic dispersion is manageable.
Good laser sources have a narrow spectral bandwidth, putting out light within a 1-nanometer (nm) range. (A nanometer is .000000001 meter.) So, the light output from a 1550 nm laser will be within a range of 1549.5 nm and 1550.5 nm.
LED sources have a broad spectral bandwidth. Many LEDs have a spectral bandwidth of 20 nm. So, the light output from an 850 nm LED would be between 840 nm and 860 nm.
Modal dispersion. This phenomenon (pulse spreading due to the paths of light sent through the fiber), occurs because some paths through a fiber are more direct than others. Look at Fig. 2, below, and you'll see this. Here, one of the light rays goes right down the middle of the fiber while others enter the fiber at angles and bounce from side to side.
You can see how the light ray going down the middle of the fiber has a significantly shorter path, and will reach the far end considerably sooner than the other ray of light. This causes a data pulse to spread.
Modal dispersion is a major factor in determining the design of optical networks; even in the design of fibers.
Optical fibers. (for more info, see Fiber Basics) Let's discuss the three main types of optical fibers the datacom industry uses today.
This type of fiber allows only one ray to transmit down the core. The core is small, usually between eight and nine microns. Because of quantum mechanical effects, the light traveling in the narrow core stays together in packets, rather than bouncing around the core. Thus, single-mode fiber has an advantage: It can handle more signal over far greater distances.
This type of fiber contains many layers of glass, each with a lower IR as you move outward from the fiber's center. Since light travels faster in the glass with lower indexes of refraction, the light waves refracted to the outside of the fiber speed up to match those traveling in the center. The result: This type of fiber allows for high-speed data transmission over a long distance.
Most use multimode fibers with LED light sources, because they are less expensive than the laser-light sources, which most use for single-mode fiber. Graded-index fibers come in core diameters of 50, 62.5, 85, and 100 microns. (One micron equals one millionth of a meter. For comparison purposes, a sheet of paper has a thickness of approximately 25 microns.)
Most people use this type of fiber less than others, because it has a lower capacity. It has a wide core (like the multimode, graded-index fiber). But since it's not graded, the light bounces wildly through the fiber. This results in high levels of modal dispersion (pulse spreading due to path losses).
Optical Data Transmission
As an example, suppose we are sending a telephone voice conversation arcoss a fiber cable. The telephone transforms the voice into an electrical signal. A digital encoder then scans the signal, converting it into binary code (a series of offs and ons). The driver, which activates the LED or laser light source, transmits the "ons" as bursts of light and "offs" as the absence of a light pulse. The light travels through the optical fiber cable until it is received at its destination, amplified, and fed into a digital decoder. The decoder translates the digital signal back into the original electrical signal. Finally, the telephone changes the signal back into sound.
Simplex and duplex transmission. Simplex, half-duplex, and full-duplex define the methods of optical transmission. Simplex and half-duplex systems use only one fiber to communicate, therefore, they are less expensive to build.
The simplex method transmits in one direction, while the half-duplex system can send signals in both directions (but not at the same time). This makes half-duplex similar to a two-way radio.
The full-duplex system uses two fibers to communicate. This allows one fiber to transmit from point A to point B while the other fiber transmits from B to A. Therefore, both ends of a full-duplex system have both transmitters and receivers. Be careful not to transpose the ends of the fibers during installation. Manufacturers include identification methods for fibers used in a full-duplex system, such as color coding or a ridge marking.
Putting two Fiber Ends Together - Splicing vs Connectors
Oddly enough, fiber splicing is a very complex area. Field technicians require years of experience before they can truly master splicing, and they usually command a high salary (there are tons of help-wanted ads for such experts in the Telecom magazines). When installing a splice, you must address two critical factors - the fiber joint must be able to pass light without loss, and the joint must be mechanically secure so that it won't be easily broken.
Connecting and splicing optical fibers is one of the more labor-intensive parts of the installation process. Joining the glass fibers correctly requires time, special tools, and specific skills. All fiber joints must be capable of withstanding moderate to severe pulling and bending tests. In addition, the joints must be Optically sound with low loss. Since the purpose of fiber is to transmit light, the fiber joint must transmit as much light power as possible with as little loss and back reflection as can be designed into the joint. There are two primary ways this is done:
Fusion splicing uses an electric arc to ionize the space between prepared fibers to eliminate air and heat the fibers to proper temperature (2,000DegrF). The fibers then feed in as semiliquids and meld together. A plastic sleeve or an other protective device replaces the previously removed plastic coating. This process generally requires a controlled environment, such as a splicing van or trailer, to reduce the possibility of dust and other contamination. Don't use fusion splicing in manholes because of the possible presence of explosive gasses and the electric arc generated during this process.
Due to the "welding" process, it's sometimes necessary to modify the fusion parameters to suit particular types of fibers, especially if you have to fuse two different fibers (from two different manufacturers or with different core/cladding structures).
Mechanical splicing is quick and easy. It does not require a controlled environment other than a reasonable level of dust control. The strength of a mechanical splice is better than most fiber-optic connectors, although fusion splices are stronger. Back reflection and loss vary from one type of splice to another.
Equipment investment for specific splicing kits is far less expensive than for fusion splices. Splices are either glued, crimped, or faced. All mechanical splices must use some type of index matching gel or liquid, which is subject to contamination and aging. The splices requiring adhesive glue can become outdated as the glue ages.
Mechanical splices use either a V-groove or tube-type design to maintain fiber alignment. The V-groove is probably the oldest and still most popular method, especially for multifiber splicing of ribbon cable. This type of splice is either crimped or snapped to hold the fibers in place.
Tubular splices, on the other hand, may rely on glue or crimping to hold the fibers together, while a small tube inside ensures alignment.
Faced-type splices are like miniaturized connectors using ferrules and a polishing process.
Fiber Cable Termination and Connector Installation
Installing a fiber connector onto a pigtail or unbuffered fiber is a widely varied process. The three most common ways to accomplishing this task are:
• Epoxy glue with oven cure, then polish
• Hot melt pre-glued, then polish
• Cleave and crimp, no polish.
Although it's the oldest method, epoxy glue is still the most widely used. This process involves filling the connector with a mixed two-part epoxy. Next, you insert the prepared and cleaned fiber into the connector. After curing the epoxy in an oven for the proper time (10 min to 40 min) you scribe and clean the fiber nearly flush with the end of the connector and polish it with a succession of finer and finer lapping papers. Typical polish papers start at 3 microns and go as fine as 0.3-micron grit.
Some techniques use a preloaded connector. The connector is placed into an oven to soften the glue and allow insertion of the prepared fiber. After cooling, the scribe and polish process is the same as the previously described process.
Cleave and crimp connectors, on the other hand, do not need a polish procedure. The connector already has a polished ferrule tip and requires only the insertion of a properly cleaved fiber to butt against the internal fiber "stub." Once in place the fiber connector is crimped to hold it in place.
Each mounting method has advantages and disadvantages: varying from ease of installation to cost per connector to performance qualities.
Terminating single-mode fibers. Terminating single-mode cables generally uses a combination of connector installation and splicing. Since single-mode connectors have fine tolerances, they are generally terminated in a manufacturing lab. Exacting precision for the insertion and contact polishing is better in the controlled environment of the laboratory.
In the field, the assemblies are cut in half and spliced onto the installed backbone cables. Although the splice adds some additional loss and cost, the overall method provides a higher yield and better connection at lower cost than trying to control the termination process in the field.
Losses and back reflection. Whether you join fibers using splices or connectors, one negative aspect is always common to both methods: signal loss. We call this loss of light power at fiber connections attenuation, which you measure in decibels.
Another type of loss is back reflection or reflectance, measured as return loss. This happens because as the light travels through the fiber (passing through splices and connections) some of it reflects back.
Typical allowable splice losses for single mode fiber are 0.0 dB to 0.25 dB, with a return loss or back reflectance of less than 150 dB. In multimode fiber splices, typical losses range from 0.0 dB to 0.25dB with an average of 0.20 dB and return loss of less than 150 dB. In the case of fiber connectors, single-mode allowable connector losses range from 0.05 dB to 0.5 dB per connector (0.1 to 1.0 dB per connection) and return loss typically is less than 130 dB. Multimode connectors have a nominal connector loss of 0.06 dB to 0.7 dB per connector (0.12 dB to 1.4 dB per connection) with a return loss less than 125 dB typical.