Monday, October 11, 2010

HOW OFC WORKS?

Fiber-optic cable employs photons for the transmission of digital signals across a strand of ultrapure silica (or plastic in some cases). Optical systems operate in the infrared light range. Photons pass through the cable with negligible resistance. The silica is so pure that, according to Michael Coden of Codenoll Technologies Corporation, a 3-mile-thick window made of the purified silica would give you the same view as a 1/8-inch-thick glass window. As optical system requirements increase, the purity of the cable becomes even more important. The best cable from Corning and Lucent is free of all but the smallest trace elements of iron and other metals, as well as hydroxyl ions, which are present due to water molecules in the cable.

A fiber-optic cable guides light from end to end. A signal is injected in one end by an LED (light-emitting diode) or by semiconductor lasers. LEDs can generate signals up to about 300 Mbits/sec. Lasers can generate signals in the multi-gigabit/sec range. LEDs are used for short-distance optical links such as enterprise backbones while lasers are used for longer distance networks. Lasers are also capable of the higher power levels needed for long-haul backbone links.

Lasers produce light in "windows" of the near infrared range as listed in the following table. A window is an infrared range that is optimized for optical transmissions. The ITU recently defined the following spectral bands in order to clarify the terminology that is used for fiber-optic systems. These definitions are meant for classification purposes only.

Band

Descriptor

Range (nm)

O band

Original

1260 to 1360

E band

Extended

1360 to 1460

S band

Short wavelength

1460 to 1530

C band

Conventional

1530 to 1565

L band

Long wavelength

1565 to 1625

U band

Ultralong wavelength

1625 to 1675

Figure 1 illustrates the structure of fiber-optic cable. The core is the transparent silica (or plastic) through which the light travels. The cladding is a glass sheath that surrounds the core. The cladding acts like a mirror, reflecting light back into the core. The cladding itself is covered with a plastic coating and strength material when appropriate.

Figure 1

Figure F-2 illustrates the different paths that light will take through a fiber cable, depending on the type of cable. If the core is wide and there is a sharp transition between the core and the cladding, high-order rays will follow reflected paths while low-order modes will follow straighter paths. This causes some of the rays to arrive at the source at a different time-a phenomenon called modal dispersion, as discussed later. Dispersion limits cable distance. Graded-index cable can help resolve the problem, or a small core can be used as in the single-mode cable. An animation that demonstrates the way that light travels through fiber cable is at the Bell Labs Web site given later.

Figure 2

Plastic core cable is used for short-distance applications, such as airplane wiring and wiring for some buildings. The remainder of this topic covers silica-based fiber-optic cable.

Silica cable is categorized as either multimode cable, which is used for short-distance connections (LANs, campus networks, and short-distance metro networks), or single-mode cable that is used for long-distance (cross-country networks and intercontinental submarine links). The light source for the former is typically an LED, while the light source for the latter is a laser. Long-haul fiber-optic cable is often bundled with anywhere from 100 to 800 fibers per cable.

  • Multimode cable This cable has a relatively large diameter core (50 to 80 microns) and a total diameter of 125 microns. Step-index multimode cable has an abrupt change between core and cladding while graded-index multimode cables has a gradual change between core and cladding. The former is limited to about 50 Mbits/sec, while the latter is limited to 1 Gbit/sec. With graded cable, the amount of refraction gradually drops outward from the core. Light travels faster in a material in which refraction is lower. This causes the light traveling through the outer material to travel faster than light at the core. The end result is that all the light tends to arrive at the same time. But this fix still has distance limitations.

  • Single mode (monomode) cable The single-mode cables are the most important cables for long-haul use (carrier and Internet core). The cable has a small core (7 to 10 microns) that forces the light to follow a more linear single path down the cable, as opposed to the multipath reflections of multimode cable. However, another form of dispersion, called chromatic dispersion, is a problem (as discussed later). Lasers are the usual light source. This cable is the most expensive and hardest to handle, but it has the highest bandwidths and distance ratings.

Cable specifications list the core and cladding diameters as fractional numbers. For example, the minimum recommended cable type for FDDI (Fiber Distributed Data Interface) is 62.5/125-micron multimode fiber-optic cable. That means the core is 62.5 microns and the core with surrounding cladding is a total of 125 microns. The cladding diameter must be the same when joining cables because connectors typically use the cladding diameter as a guide for aligning the core.

  • The core specifications for step-index and graded-index multimode cables is typically 50, 62.5, or 100 microns. The cladding diameter for step-mode cable is 125 microns.

  • The core diameter for single-mode cable is typically 7 to 10 microns with a cladding diameter of 125 microns.

The ITU has defined a series of recommendations that describe the geometrical properties and transmissive properties of multimode and single-mode fiber-optic cables. The four most important recommendations are listed here:

  • ITU G.651 Covers multimode graded-index fiber-optic cable having a 50-micron nominal core diameter and a 125-micron nominal cladding diameter.

  • ITU G.652 Covers single-mode NDSF (non-dispersion-shifted fiber). This cable constitutes most of the cable that was installed in the 1980s. Transmissions take place in the 1,310-nm range where there is minimal signal dispersion. Dispersion causes signal problems over long distances, as described later. G.652 cable supports the following distances and data rates: 1000 km at 2.5 Gbits/sec, 60 km at 10 Gbits/sec, and 3 km at 40 Gbits/sec.

  • ITU G.653 Covers single-mode dispersion-shifted optical fiber cable. The cable is designed in a way to "shift" the region where dispersion is minimized to the 1,550-nm wavelength range. At this range attenuation is also minimized, so longer distance cables are possible.

  • ITU G.655 Covers single-mode NZ-DSF (nonzero dispersion-shifted fiber) cable, which takes advantage of dispersion characteristics that suppress the growth of four-wave mixing, an effect that is harmful to WDM (wavelength division multiplexing) systems. NZ-DSF supports high-power signals and longer distances, as well as closely spaced DWDM (dense WDM) channels at rates of 10 Gbits/sec or higher. Lucent True Wave is an example of this cable. It supports the following distances and data rates: 6000 km at 2.5 Gbits/sec, 400 km at 10 Gbits/sec, and 25 km at 40 Gbits/sec.

G.655 is the latest development in fiber-optic cable. In particular, G.655 is optimized for WDM and long-distance cable runs such as submarine cables. It uses dispersion to advantage. Dispersion can help reduce an effect called four-wave mixing (FWM), which occurs in DWDM systems when three wavelengths mix in such a way to produce a fourth wavelength that overlays and interferes with the original signals.

With DWDM, a single fiber can potentially carry thousands of lambda circuits. A lambda is a specific subwavelength of light within one of the windows of light. It provides all the capabilities of an individual circuit. Lambdas are set up using frequency division multiplexing. Think of each lambda as a specific color of infrared light transmitting at 10 Gbits/sec or more. An optical multiplexer divides the available spectrum on the cable up into many individual lambdas. For example, the Avanex PowerMux can put over 800 channels on a single fiber with spacing between channels of 12.5 gigahertz. With the potential of thousands of lambdas per fiber, it is practical for carriers to lease entire optical wavelengths to businesses. See "Optical Networks."

An alternative to DWDM are new optical modulation techniques that boost the capacity of existing cables. Kestrel Solution's Optical FDM combines FDM (frequency division multiplexing), DSP (digital signal processing), and optical modulation to improve performance on existing fiber cables, especially in metropolitan areas where low-quality fiber has been installed (due to short distances) and in SONET systems. Optical FDM gives full access to the total bandwidth of the cable.

Cable Performance Characteristics

Fiber cable has certain characteristics that limit its performance. Cable from different manufacturers may exhibit variations in these characteristics. The primary performance limiters are attenuation and dispersion.

Attenuation, EDFAs, and Raman Amplifiers

Attenuation is signal loss over distance. Think of the light pulses as loosing their energy and flattening out as they travel down the cable. High attenuation will result in errors at the receiver. Attenuation is not a problem for metro-area networks, but it puts distance limitations on long- haul backbone networks. Three types of devices may be used to overcome attenuation:

  • Electronic regenerator This device regenerates signals by first converting optical signals to electrical signals. The electrical signal is regenerated, converted back to optical, and injected back into the fiber. Regenerators are too inefficient for modern high-speed optical networks due to electrical regeneration requirements. On WDM systems, each wavelength requires its own opto-electric amplifier, an expensive proposition if there are many wavelengths.

  • EDFA (erbium-doped fiber amplifier) This device is an amplifier rather than a signal regenerator. It directly amplifies optical signals without a need to do the optical-to- electrical conversion. An EDFA contains a short strand of erbium-doped fiber and two signal inputs. One input is the optical signal that needs to be amplified. The other is light from a pump laser that excites the erbium atoms so that they give up photons that amplify incoming optical signal photons as described later. Thank Albert Einstein for this "trick" of light magic.

  • EDFA/Raman amplifier combination Raman amplification is an add-on component that enhances EDFA optical amplifiers. At each EDFA amplifier, a Raman pump injects high-power laser light into the fiber in the opposite direction of the source signal. The injected photons boost the optical signal where it is needed most-at the far end of the laser signal where it is experiencing the most attenuation. Raman amplification can create signal gain of up to 10 dB, which allows longer distance cable runs. It also allows optical networks to obtain transmission rates as high as 40 Gbits/sec.

EDFAs are crucial to WDMA systems because a single amplifier boosts all the wavelengths simultaneously. With older electrical regenerators, one regenerator is required for each wavelength, meaning that an entire stack of regenerators is needed at each regeneration point.

The 1,550 range (C band) is often called "erbium window" because the energy level of erbium ions is close to the energy level of photons in the C band. Through the process of stimulated emission, erbium can be coaxed into releasing energy that amplifies light in the C-band. The process is as follows. The EDFA amplifier pumps photons into the erbium-doped cable. Erbium atoms absorb photons, which cause electrons to jump to a temporary excited state. When an electron decays, it releases a photon that is absorbed by the signal photons. Thus, the optical signal passing through the cable is amplified without any electrical conversion.

Dispersion

Another characteristic is dispersion, which is the broadening of a light pulse as it travels down the cable. Excessive dispersion will make a signal difficult to read by the receiver. When an LED or laser sends light into a multimode fiber, a range of wavelengths of light is present. Some of those wavelengths travel at different speeds than others. The effect is to distort waveforms, which can cause errors in reading the signal at the other end of the cable. Graded-index cable is designed to minimize the delay of the slower wavelengths.

There are four types of dispersion:

  • Material dispersion Variations in the refractive properties of the cable cause signal dispersion.

  • Modal dispersion Occurs in multimode cable. Light takes different paths through the cable with light on some paths having a longer travel time than others. Graded cable balances this effect.

  • Chromatic dispersion This occurs because some wavelengths travel through a medium faster than others. The longer the cable, the worse the effect, and the harder it is to read the signal.

  • Waveguide dispersion This dispersion occurs in single-mode fiber due to the difference in the speed of the signal between the core and the cladding. It causes chromatic dispersion.

G.652 cable, used in most commercial systems, takes advantage of the 1,310-nm window where chromatic dispersion is minimized. This window is often called the zero-dispersion point-it is the range where chromatic dispersion is minimized because the waveguide dispersion cancels out material dispersion.

Long-haul carriers, on the other hand, have higher bandwidth and distance requirements, so G.653 and G.655 "dispersion-shifted" fiber operating in the C band is preferred. The C band is used for DWDM systems, which support many closely spaced channels at data rates of 10 Gbits/sec and higher. Two other bands are now being used to boost capacity and distance: the 1,460 nm to 1,530 nm S band and the 1565 to 1625 L band. A newer approach employs soliton technology, which can be used to create a cable system that stretches nearly halfway around the world. See "Soliton."

Note that it is possible to support DWDM on many older fiber cables. Standard single- mode fiber will support DWDM at lower data rates. Some older dispersion-shifted fibers were not able to handle DWDM, but these cables may be made to act like nonzero dispersion cable by using wavelength above and below the 1,550-nm window.

Corning and Lucent are the major providers of long-haul cable. Lucent's TrueWave and AllWave cables are made of single-mode nonzero dispersion fibers that support all the wavelength windows. TrueWave is specifically designed for optically amplified, high-powered long-distance DWDM networks operating in both the C band and the L band. Both cable types are manufactured with a patented purification process to remove water molecules in the core, thus allowing wider spectrum usage.

Corning's LEAF is a single-mode NZ-DSF fiber designed for DWDM systems. It combines low attenuation and low dispersion with an effective area that is 32 percent larger than non- NZ-DSF fiber. This allows more power to be pumped into the network over more channels without nonlinear effects that create noise, distort signals, and degrade performance. It can operate at 10 Gbits/sec or higher using high-output-power EDFAs.

Corning's MetroCor fiber is a single-mode NZ-DSF cable optimized for short-distance metropolitan usage. It does not require the powerful lasers that are required in the long-haul environment and so helps reduce the cost of implementing metropolitan fiber networks.

For more information about metropolitan optical networks, see "FTTH (Fiber to the Home)," "MAN (Metropolitan Area Network)," and "PON (Passive Optical Network)."

Using of OTDR

Fiber Optic Cable Plant Link Loss Budget Analysis

Loss budget analysis is the calculation and verification of a fiber optic system's operating characteristics. This encompasses items such as routing, electronics, wavelengths, fiber type, and circuit length. Attenuation and bandwidth are the key parameters for budget loss analysis.

Analyze Fiber Optic Link Loss In The Design Stage
Prior to designing or installing a fiber optic system, a loss budget analysis is reccommended to make certain the system will work over the proposed link. Both the passive and active components of the circuit have to be included in the budget loss calculation. Passive loss is made up of fiber loss, connector loss, and splice loss. Don't forget any couplers or splitters in the link. Active components are system gain, wavelength, transmitter power, receiver sensitivity, and dynamic range. Prior to system turn up, test the circuit with a source and FO power meter to ensure that it is within the loss budget.

The idea of a loss budget is to insure the network equipment will work over the installed fiber optic link. It is normal to be conservative over the specifications! Don't use the best possible specs for fiber attenuation or connector loss - give yourself some margin!

The best way to illustrate calculating a loss budget is to show how it's done for a 2 km multimode link with 5 connections (2 connectors at each end and 3 connections at patch panels in the link) and one splice in the middle. See the drawings below of the link layout and the instantaneous power in the link at any point along it's length, scaled exactly to the link drawing above it.

fiber optic cable

Fiber Optic Cable Plant Passive Component Loss

Step 1. Fiber loss at the operating wavelength

Cable Length 2.0 2.0

Fiber Type Multimode
Singlemode
Wavelength (nm) 850 1300 1300 1550
Fiber Atten. dB/km 3 [3.5] 1 [1.5] 0.4 [1/0.5] 0.3 [1/0.5]
Total Fiber Loss 6.0 [7.0] 2.0 [3.0]

(All specs in brackets are maximum values per EIA/TIA 568 standard. For singlemode fiber, a higher loss is allowed for premises applications. )

Step 2. Connector Loss
Multimode connectors will have losses of 0.2-0.5 dB typically. Singlemode connectors, which are factory made and fusion spliced on will have losses of 0.1-0.2 dB. Field terminated singlemode connectors may have losses as high as 0.5-1.0 dB. Let's calculate it at both typical and worst case values.

Connector Loss 0.3 dB (typical adhesive/polish conn) 0.75 dB (TIA-568 max acceptable)
Total # of Connectors 5 5
Total Connector Loss 1.5 dB 3.75 dB

(All connectors are allowed 0.75 max per EIA/TIA 568 standard)

Step 3. Splice Loss
Multimode splices are usually made with mechanical splices, although some fusion splicing is used. The larger core and multiple layers make fusion splicing abut the same loss as mechanical splicing, but fusion is more reliable in adverse environments. Figure 0.1-0.5 dB for multimode splices, 0.3 being a good average for an experienced installer. Fusion splicing of singlemode fiber will typically have less than 0.05 dB (that's right, less than a tenth of a dB!)

Typical Splice Loss 0.3 dB
Total # splices 1
Total Splice Loss 0.3 dB

(All splices are allowed 0.3 max per EIA/TIA 568 standard)

Step 4. Total Passive System Attenuation
Add the fiber loss, connector and splice losses to get the link loss.


Best Case TIA 568 Max
850 nm 1300 nm 850 nm 1300 nm
Total Fiber Loss (dB) 6.0 2.0 7.0 3.0
Total Connector Loss (dB) 1.5 1.5 3.75 3.75
Total Splice Loss (dB) 0.3 0.3 0.3 0.3
Other (dB) 0 0 0 0
Total Link Loss (dB) 7.8 3.8 11.05 7.05

Remember these should be the criteria for testing. Allow +/- 0.2 -0.5 dB for measurement uncertainty and that becomes your pass/fail criterion.

Equipment Link Loss Budget Calculation: Link loss budget for network hardware depends on the dynamic range, the difference between the sensitivity of the receiver and the output of the source into the fiber. You need some margin for system degradation over time or environment, so subtract that margin (as much as 3dB) to get the loss budget for the link.

Step 5. Data From Manufacturer's Specification for Active Components (Typical 100 Mb/s link)

Operating Wavelength (nm) 1300
Fiber Type MM
Receiver Sens. (dBm@ required BER) -31
Average Transmitter Output (dBm) -16
Dynamic Range (dB) 15
Recommended Excess Margin (dB) 3

Step 6. Loss Margin Calculation

Dynamic Range (dB) (above) 15 15
Cable Plant Link Loss (dB) 3.8 (Typ) 7.05 (TIA)
Link Loss Margin (dB) 11.2 7.95

As a general rule, the Link Loss Margin should be greater than approximately 3 dB to allow for link degradation over time. LEDs in the transmitter may age and lose power, connectors or splices may degrade or connectors may get dirty if opened for rerouting or testing. If cables are accidentally cut, excess margin will be needed to accommodate splices for restoration.

NOTE: Many techs forget when doing a loss budget that the connectors on the end of the cable plant must be included in the loss budget. When the cable plant is tested, the reference cables will mate with those connectors and their loss will be included in the measurements.

Related Fiber Optic Components Links
Types of Fiber Optic Cables
Fiber Optic Loss Budget Standard
Fiber Optic Connector Types
Troubleshooting Fiber Optic Transmitters & Receivers
Fiber Optic Couplers & Splitters
Fiber Optical Amplifiers

Fiber Optic Cable

Fiber Optic Cable

  • Uses light instead of electricity to transmit signals
  • Fiber is glass or plastic
  • Light source is laser or LEDs
  • Can transmit huge distances compared to copper cable
  • No EMI, hard to intercept

Fiber optics

Single Core.

A data transmission medium consisting of glass or plastic fibers. Light-emitting diodes (LEDs) or lasers send light through the fiber to a detector, which then converts the light back into electrical signals. Fiber-optics offer immense bandwidth, as well as protection from eavesdropping, electromagnetic interference, and radioactivity. Fiber-optic cabling employs one or more optical fibers, consisting of a central glass or plastic fiber, glass cladding, and a plastic outer sheath. Fiber-optic cable carries information as light instead of electricity, and can carry much more information over greater distances than copper cabling.

Optical principles
Fiber can transmit light a long distance because it keeps it in a channel. This is accomplished by surrounding the inner fiber with a cladding that has a different refractive index. This keeps the light bouncing along from side to side until the end.



Fiber optic cable
Figure 21: The structure of fiber optic cable

Single-mode cable
Single mode cable only has one light beam in it. The cable has a small diameter core and is made of the purest glass. A laser is usually the light source and the cable can extend several kilometers without a repeater.

Multimode cable
Multimode cable has greater capacity because several light beams are sent down it. The beams enter at slightly different angles and therefore don’t interfere with each other. The core is wider than used in single-mode and often made of fine plastic. Instead of lasers, the light source is usually light emitting diodes (LEDs). The system has greater capacity and is less expensive than single-mode. On the other hand, multimode doesn’t have the same reach, often restricted to hundreds of meters instead of kilometers. For this reason, multimode is usually used inside buildings, whereas single-mode is used between buildings and long distances. If the cable needs to be extended, optical repeaters are used.

The conductors in the cable only carry signals in one direction, therefore two conductors are used, each carrying signals in the opposite direction. The two cables are connected together with Kevlar webbing.

Pros and cons
Fiber optic cable has the longest range and should be used when long distance is a requirement. It has extremely high bandwidth, in the Gbps range. It is also immune to electrical interference, nor does it give off any of its own. This would be a factor in dangerous environments where an electrical spark could set off an explosion, say in an explosives factory. Finally, it is extremely difficult, but not impossible, to tap into a fiber optic system.

On the negative side, fiber optic cable is the most expensive to install. The cable and connectors are more expensive plus the installation requires highly skilled craftsmen. This is particularly true when two fibers need to be spliced together.

Connectors used for fiber cable are SC connectors which are a push and pull connector and ST connectors which are a bayonet mount.

Saturday, June 26, 2010

FABRICATION AND MEASUREMENTS OF FIBER OPTIC

Optical Fiber Fabrication

&

Measurements




Contents

1. Optical Fiber Fabrication

1.1 Fiber Materials

1.1.1 Glass Fibers

1.1.2 Halide Glass Fibers

1.1.3 Active Glass Fibers

1.1.4 Chalgenide Glass Fibers

1.1.5 Plastic Optical Fibers

1.2 Fiber Fabrication

1.2.1 Double Crucible Method

1.2.2 Vapor Phase Oxidation

1.2.3 Preparation Of The Preform

1.2.4 Outside Vapor-Deposition

1.2.5 Vapor Phase Axial Deposition

1.2.6 Modified Chemical Vapor Deposition

1.2.7 Plasma Chemical Vapor Deposition

1.2.8 Fiber Drawing

1.3 Fiber Coating

1.4 Optical Cable design

2. Measurements

2.1 Test Equipments

2.1.1 Optical Power Meters

2.1.2 Optical Attenuators

2.1.3 Tunable Laser Sources

2.1.4 Optical Spectrum Analyzers

2.1.5 Optical Time-Domain Reflectometer (OTDR)

2.1.6 Multifunction Optical Test Systems

2.2 Attenuation Measurements

2.2.1 The Cutback Technique

2.2.2 Insertion-Loss Method

2.3 Dispersion Measurements

2.3.1 Intermodal Dispersion

2.3.2 Time-Domain Intermodal Dispersion Measurements

2.3.3 Frequency-Domain Intermodal Dispersion Measurements

2.3.4 Chromatic Dispersion

2.3.5 Polarization –Mode Dispersion

2.4 Eye Patterns

Optical Fiber Fabrication

1.1 Fiber Materials

There are many materials available for use in fiber fabrication. Only a few meet the special requirements of optical fibers:

1) The material must allow us to make long, thin, and flexible fibers.

2) The material must be transparent at a particular wavelength in order to guide light efficiently.

3) Physically compatible materials that have slightly different refractive indices for the core and cladding must be available.(The index of refraction or IOR is a measure of the speed of light in a material).

4) Finally, we must have a material that is cheap and abundant.

Two materials that are commonly used and meet all of these requirements are plastics and glass.

1.1.1 Glass Fibers

The largest category of optically transparent glasses from which optical fibers are made consists of the oxide glasses. The most common of which is silica (SiO2). Glass composed of pure silica is referred to as silica glass, fused silica, or vitreous silica. The glass is made by fusing mixtures of metal oxides, sulfides or selenides. Some of its desirable properties are a resistance to deformation at high temperatures, a high resistance to breakage from thermal shock because of its low thermal expansion, good chemical durability, and a high transparency in both the visible and infrared regions of interest to many fiber optic systems.

In order to produce two similar materials having slightly different indices of refraction for the core and the cladding, fluorine or various oxides are commonly added to the silica. These dopants can be classified into two basic groups: dopants, which increase the IOR, and dopants, which decrease the IOR. For example, B2O3 and fluorine dopants decrease a material's IOR, while GeO2, P2O5 will increase a material’s IOR as shown in the Figure 1. The list below shows some various dopants used to make the core and cladding of various fibers.

Core

Cladding

SiO2

B2O3-SiO2

GeO2-SiO2

SiO2

P2O5-SiO2

SiO2

                
               
               
               

Figure1: Refractive index as a function of dopants materials and their concentration

1.1.2 Halide Glass Fibers

The second types of glass used are halide glass fibers. It has been found that fluoride glasses have extremely low transmission losses at wavelengths in the range from 0.2 to 8 μm. Fluoride glasses belong to a general family of halide glasses in which the anions are from elements in-group VII of the periodic table, namely fluorine, chlorine, bromine, and iodine. The material that researchers have concentrated on is a heavy metal fluoride glass, which uses ZrF4 as the major component. Several other constituents need to be added to make a glass that has moderate resistance to crystallization. ZBLAN is the material forms the core of a glass fiber .to make a lower refractive index glass, one partially replaces ZrF4 by HaF4 to get ZHBLAN cladding. Theoretically, the minimum attenuation for these materials is estimated at 0.001dB/km.

1.1.3 Chalgenide Glass Fiber

In addition to allowing the creation of optical amplifier, the nonlinear properties of glass fibers can be exploited for other applications, such as all optical switches and fiber lasers. Chalgenide glass is one candidate for these uses because of its high optical nonlinearity and its long interaction length. Chalgenide glass fibers, contains arsenic, germanium, phosphorus, sulfur, selenium, or tellurium. Theoretically, the minimum attenuation for these materials has been estimated at 1dB/m.

1.1.4 Active Glass Fibers

Incorporating rare-earth elements (atomic numbers 57-71) in to a normally passive glass gives the resulting material new optical and magnetic properties. These new properties allow the material to perform amplification, attenuation, and phase retardation on the light passing through it. Two commonly used materials for fiber lasers are erbium and neodymium.

1.1.5 Plastic Optical Fibers

As mentioned earlier, plastics are sometimes used in place of glass optical fibers. Plastic fiber is traditionally used in place of glass for short distances (up to 100m) and abusive environments where its mechanical strength makes it superior. For example, plastics can be used in medical applications and in some sensors where only shorter fiber lengths are needed. In addition, the mechanical flexibility of plastic allows these fibers to have large cores. These factors permit its use in inexpensive, economically attractive systems. The following are examples of some of the compounds used in plastic fibers:

· a polysterene core/methyl methacrylate cladding.
· a polymethyl methacrylate core/copolymer cladding.

Comparing the plastic fibers to their glass counterpart, the transmission spectrum of the two is similar. Plastic fibers are lighter and lower in cost than glass. However, plastics are less widely used because of their high attenuation in comparison to their glass counterpart. Another problem is plastics limited operating temperature range.

1.2 Fiber Fabrication

There are two basic techniques used in the fabrication of optical fibers:

1) Fibers can be drawn directly from melts of silica in crucibles.

2) Vapor phase oxidation.

1.2.1 Double crucible method

The double crucible method can be used to make both silica and halide glass fibers. The technique is simple and straightforward. One glass rod is made from silica powders for the core and one for the cladding. The rods are then used as feedstock for each of two concentric crucibles. The inner contains the molten core, while the outer contains the cladding. In a continuous process, the fiber is drawn from the molten state. The disadvantage of this method is the possibility of introducing contaminants during the melting process. Figure 2 illustrates the double-crucible drawing process.

Figure 2: Double-crucible fiber drawing process

1.2.2 Vapor phase oxidation

With direct drawing, it is difficult to get pure and homogeneous fibers; therefore this method is not commonly used. The vapor phase oxidation processes have proven to be more successful. These processes are usually done in two steps:

1) The first being the preparation of the preform.

2) The second being the drawing of the fiber.

1.2.3 Preparation of the preform

preform

Figure 3: Preform

As shown in the Figure3 a perform is a cylinder of silica composition 10 to 25 mm in diameter and about 60 to 120 cm long. This preform consists of a core surrounded by a cladding with a desired refractive-index profile, a given attenuation, and other characteristics; in other words, this is a desired optical fiber, but on a much larger scale.

The main reason a preform is prepared is to have a "drawable" material that is clean, low in OH concentration, low in metallic-ion contaminants, and inexpensive. Many techniques have been developed to prepare these preforms. Some common commercially used methods are Outside Vapor-Deposition, Modified Chemical Vapor Deposition, Vapor Phase Axial Deposition, and Plasma Chemical Vapor Deposition. They differ mainly by the way the soot is deposited.

The preform is made by vapor-phase oxidation, in which two gases, SiCl4 and O2, are mixed at a high temperature to produce silicon dioxide (SiO2):




Silicon dioxide, or pure silica, is usually obtained in the form of small particles (about 0.1 µm) called "soot." This soot is deposited on the target rod or tube. The depositing of the silica soot, layer upon layer, forms a homogeneous transparent cladding material. To change the value of a cladding's refractive index, some dopants are used. For example, fluorine (F) is used to decrease the cladding's refractive index in a depressed-cladding configuration.

The soot for the core material is made by mixing three gases: SiCl4, GeCl4, and O2 which results in a mixture of SiO2 and GeO2. The degree of doping is controlled by simply changing the amount of GeCl4 gas added to the mixture. The same principle is used for doping other materials.

Since deposition is made by the application of silica layers top one another, the manufacturer can control the exact amount of dopants added to each layer, thus controlling the refractive-index profile. Figure1 explains the role of several widely used dopants.

The vapor-phase oxidation process produces extremely pure material whose characteristics are under the absolute control of the manufacturer.

1.2.4 Outside Vapor-Deposition

Outside Vapor Deposition (OVD), also called the "soot process", was first developed by Corning Incorporated. This fiber was the first to have a loss of less then 20dB/km. The three main steps involved are laydown (Figure4-a), consolidation (Figure4-b), and drawing. In the laydown process, several materials such as SiCl4, GeCl4, BCl3 and O2 are allowed to react in a hot flame to produce soot (Figure 5). The soot is in turn deposited into a rotating ceramic rod known as a mandrel. Initially the core material is deposited, followed by the cladding. The soot builds up on the rod, and layer-by-layer, a cylindrical preform is built up. In preparing the preform, many characteristics like glass composition, refractive index, and the dimensions of the core and cladding can be controlled. Next, the deposited preform is removed from the rod and placed in a consolidation furnace. This high temperature furnace removes any water vapor that may be in the preform. The resulting product is a solid, dense, glass blank. Now the blank is ready to be placed in a draw tower where a continuous fiber is made into a strand. It is during the draw process, discussed in more detail below, that the hole in the tube collapses and a perfectly symmetrical fiber is formed.

Either step or graded index performs can be made.

Figure 4: Two phases of the OVD process: (a) Laydown; (b) consolidation.


Figure 5: Outside Vapor-Deposition

1.2.5 Vapor phase axial deposition

Another similar method of producing preforms is vapor phase axial deposition (VAD). In VAD, the SiO2 particles are formed in the same manner as in OVD. The particles are deposited onto the end of a glass rod, which is in turn attached in the upright position to a motor. A porous preform is grown axially as a pulling machine rotates the rod upward. The preform is then made into a solid rod by zone melting. The preform is ready to be drawn into a fiber. This method is done completely inside a closed deposition chamber. It therefore has the advantage of a clean environment. Also, there is no central hole created, as is the case for OVD. Finally, VAD has proven to be cost effective because the preform can be made in continuous lengths. This method is shown in Figure 6.

VPAC

Figure 6: Vapor phase axial deposition

1.2.6 Modified Chemical Vapor Deposition

Modified Chemical Vapor Deposition, or MCVD, is a process that was developed at AT&T Bell Laboratories in the 1970's. It was so successful that AT&T used it to mass-produce optical fiber in the 1980's. It provided a simple and straightforward means of manufacturing high-quality optical fibers. MCVD differs from OVD in that the deposition occurs inside of a fuse-quartz tube instead of on the outside, hence the term modified. The process works as follows: Reactants (SiCl4+ O2) are introduced at one end of the rotating tube while an exhaust is located on the other end. A burner that traverses back and forth along the length of the tube sinters the deposited SiO2 particles to a clear glass layer. When the desired thickness is achieved, a valve is closed to stop the flow of reactants. As one might expect, the flow of reactants and the speed of the traversing oxy-hydrogen burner have to be closely monitored using a video camera. Finally, the temperature of the burner is increased so that the rod collapses onto itself resulting in a solid preform. An advantage of this method is that it is an inherently clean process since the reaction occurs inside of the tube. According to The Journal of Lightwave Technology, this method produces the highest quality product under factory conditions. Other advantage this method has is the ability to mass-produce the fiber quickly under various design requirements while using the same equipment. This method is shown in Figure 7.

MCVD

Figure 7: Modified Chemical Vapor Deposition

1.2.7 Plasma Chemical Vapor Deposition

Scientists at Phillips Research invented plasma Chemical Vapor Deposition, or PCVD. Clear glass instead of soot is also deposited inside a silica tube as is done in the MCVD process. Non-isothermal plasma in the microwave frequency range (2.45GHz) is used instead of a torch or flame. The plasma makes the reaction proceed at about 1000 to 1200°C. This results in very thin layers being deposited inside the tube. Although this method allows us to grow layers at relatively low temperatures, the deposition rate is rather slow in comparison to other methods. This method is shown in Figure 8.

Figure 8: Plasma Chemical Vapor Deposition

1.2.8 Fiber Drawing

Figure 9: Schematic of a typical drawing process.

The major steps of a typical drawing process are shown in Figure 9. The preform is put into a draw furnace, where the bottom tip is heated to melting. This molten piece now starts to fall, forming a fiber with a 125-µm outer diameter. Diameter-monitoring equipment controls the actual fiber diameter by changing, if necessary, the rate of drawing that is executed by a tractor assembly. A coating applicator applies a coating over the cladding. Concentricity-monitoring equipment controls this parameter. The coating is cured by ultraviolet lamps or some other heat source. The coated fiber is then wound onto ready-to-ship reels.

From this oversimplified description you might get the impression that drawing a fiber is very simple. It is not. It took many years of intensive, costly research and development efforts and the brainpower of thousands of scientists and engineers to make this process commercially possible. Critical to the process, of course, is the rate of drawing. And herein lies a dilemma. On the one hand, the slower the drawing, the better the manufacturer can control fiber quality. On the other hand, the faster the drawing, the more fiber one can produce in a given amount of time. It's clear, then, that there can be no "best" rate in an industrial operation, where both quality and productivity are demanded. So one has to find an acceptable compromise. Each of the processes described here has its own advantages and drawbacks associated with the drawing rate at which it runs. To give you some sense of how widely the numbers vary, the draw rate can run from 200 m/min to 2,000 m/min. To reach such a speed, all rotating parts of the drawing mechanism must be manufactured to extremely tight tolerances and the tension level to which the optical fiber is subjected must be controlled with a high degree of accuracy.

Even though Figure 9 shows only diameter-measuring equipment, the manufacturer actually measures and controls many of the fiber's characteristics during the fiber-draw stage, when some serious problems can crop up. Among them are internal problems, such as bubbles, contamination, and discontinuity, and external problems, like neck-downs, lumps, and flaws. Control of the external problems is very important because they can weaken a fiber. This is why manufacturers not only detect them constantly but also compare their size with threshold values.

If you analyze Figure 9 closely, you will find many problems that need to be resolved at each particular step and for the entire process.

Both stages of fiber manufacturing are fully automated and are performed in a clean, climate-controlled room. Obviously, the manufacturers use high-precision measuring equipment to automatically control each step of the fabrication process. For example, preform analyzers measure the critical characteristics of the optical-fiber preform. Also, specific measurement systems control fiber geometry, the refractive-index profile, and the coating geometry.

1.3 Fiber Coating

Requirements

As you know from the previous sections, an optical fiber—a core surrounded by cladding—has to be coated. Figure 9 shows that manufacturers coat on-line. A liquid polymer is applied by the coating applicator as a fiber passes through the coating line. This liquid is solidified by heat or ultraviolet curing.

The main function of the coating is to protect the fiber from any external damage. But a closer look reveals that the coating has several other major functions:

  • Adhesion: The coating, obviously, has to stick firmly to the glass surface of the fiber.
  • Ability to be stripped: To connector a fiber, the coating has to be stripped. The stripping force has to be very small to facilitate handling of the fiber during installation. This force has to be stable in any environment—dry or wet. The range of the stripping force is between 1.4 N and 4.2 N.

If you think that the above two functions seem to conflict, you are right. It's a manufacturing problem optical fiber producers have to live with today. The conflict typifies the kinds of problems fiber makers have yet to resolve.

  • Toughness: This quality is necessary to provide enhanced abrasion protection and to enable fiber handling and cabling without loss of strength. Toughness is gauged by elastic (Young) modulus testing, which determines whether a coating is soft or hard.
  • Moisture resistance: The coating is the line of defense protecting the fiber from moisture. This characteristic, in a sense, determines a fiber's aging and stability properties. Moisture resistance is determined by measuring the increase in attenuation during the fiber's exposure to water.


Many other parameters characterize coatings but these are the important ones to know. The point is simply this: The coating is a critical component of an optical fiber. It determines bending sensitivity, abrasion resistance, static-fatigue protection, and many other important properties of the fiber.

We've considered the role of the coating from the internal standpoint, in other words, how the coating protects the fiber. But there is another important consideration: how the coating works with a fiber cable. From this vantage, a coating should be smooth enough to be put inside a tight buffer and strong enough to protect the fiber from undergoing changes in its optical properties.

Solutions

To meet these complex requirements, manufacturers developed many coatings, which vary in design, materials used, curing processes, and so on. The overall point to bear in mind is this: The coating, along with its fiber, is designed for a specific application. For example, a tight buffer and a loose buffer apply different forces to the fiber, thus requiring different coating characteristics.

There are two main types of design: single-layer and double-layer coatings. In the single-layer design, the manufacturer tries to satisfy a wide range of requirements by choosing the proper material and coating thickness, among other characteristics. In the double-layer design, the inner layer is a soft coating. It provides good adhesion and cushions the fiber. The hard outer layer protects against an adverse environment, including abrasion.

The outer diameter of a coated optical fiber ranges from 245 µm to 900 µm.

Armed with this new knowledge, you will now have a better understanding of the manufacturers' data sheets.

1.4 Cable Design

There are two distinctly different methods used to protect the optic fibers:

  1. Loose tube.
  2. Tight buffer designs.

Loose tube designs to be used externally. Tight buffer designs to be used within buildings.

Some important mechanical properties must be considered in cable design:

  1. Maximum allowable load on the cable: since this factor determines the length of cable that can be reliably installed. In copper cables the wires themselves are generally the principal loads bearing members of the cable, and elongations of more than 20 percent are possible without fracture. On the other hand, extremely strong optical fibers tend to break at 4 percent elongation, fiber elongations during cable manufacture and installation should be limited to 0.1-0.2 percent.
  2. Fiber brittleness: since glass fibers do not deform plastically, they have a low tolerance for absorbing energy from impact loads. Hence the outer sheath of optical cable must be designed to protect the glass fibers inside from impact forces. In addition, the outer sheath should not crush when subjected to side forces, and it should provide protection corrosive environmental elements. In underground installations, a heavy-gauge-metal outer sleeve may be required to protect against potential damage from burrowing rodents, such as gophers.

In designing optical fiber cables, several types of fiber arrangements are possible and a large variety of components could be included in the construction, As shown in the Figure 10 and Figure 11.

Figure 10: Tight Buffer construction

36-fiber

288-fiber

Figure 11: Loose tube construction

Measurements

The following figure shows some of the relevant test parameters and at what points in WDM link they are of importance.

2.1 Test Equipments

2.1.1 Optical power meter

Optical power measurement is the most basic function in fiber optical metrology.

Some form of optical power detection is involved in almost every piece of light wave test equipment. Hand held instruments come in wide Varity of types with different levels of capabilities. Multiwavelength optical power meters that use photodetectors are the most common instrument for measuring optical signal power levels. Usually, the meter outputs are given in dBm (where 0 dBm=1 mW) or dBμ (where 0 dBμ=1μW).

Figure 12 shows a compact hand held Model FOT-90A fiber optic power meter from EXFO. In this versatile instrument, various photodetectors heads that have different performance characteristics are available .for example, using a Ge detector allows a measuring range of +18 to –60 dBm in the 780 to 1600 nm wavelength band, whereas an InGaAs detector allows a measuring range of +3 to –73 dBm in the 840 to 1650 nm wavelength band.

An RS-232 interface together with application software allows a user to download the measurements and view, export, or print them in either tabular or graphical form.

Figure 13 shows another hand-held tester that also contains optical sources so it can do more sophisticated optical power measuring .For example this instruments can function as a power meter, an optical-loss tester for automatically measuring loss in a fiber in two directions at two different wavelengths, an optical return-loss tester for measuring the quality of optical patch cords, a visual fault indicator for locating breaks and failures in a fiber cable.

Figure 12: Optical power meter (Model FOT-90A, from EXFO)

Figure 13: Multipurpose test instrument (Model FOT-920A, from EXFO)

2.1.2 Optical Attenuators

In many laboratory or production tests, the characteristics of high optical signal level may need to be measured. If the level is very high, such as a strong output from an optical amplifier, the signal may need to be precisely attenuated before being measured. This is done to prevent instrument damage or to avoid overload distortion in the measurements. Optical attenuators allow a user to reduce the optical signal level. For example up to 60 dB (a factor of 10^6) in precise steps at a specified wavelength, which is usually 1310 or 1550 nm. The capabilities of attenuators range from simple tape-cassette-sized devices for quick field measurements that may only need to be accurate to 0.5 dB, to laboratory instruments that have an attenuation precision of 0.001 dB.

2.1.3 Tunable laser Sources

Tunable laser sources are important instruments for test that measure the wavelength response of an optical component or link. Figure 14 shows an example from Hewlett-Packard (Model 8168B) that generates a true single mode laser line for every selected wavelength point. The source is an external cavity semiconductor laser. A movable diffraction grating is used as a tunable filter for wavelength selection. Depending on the source and grating combination, typical instrument is tunable over either the 1280 to 1330 nm or the 1450 to 1565 nm band. Wavelength scans, with an output power that is flat across the scanned spectral band, can be done automatically. The minimum output power of these instruments is –10dBm and the absolute wavelength accuracy is typically ± 0.1 nm.

Figure 14: Tunable laser source (Model HP-8168, from Hewlett-Packard)

2.1.4 Optical spectrum analyzers

Optical spectrum analyzers measure optical power as function of wavelength. The most common implementation uses a diffraction grating based optical fiber, which yields wavelength resolutions to less than 0.1 nm. To measure very narrow linewidths for example the 10 MHz linewidth of typical single frequency semiconductor laser-optical analyzers employing homodyne and heterodyne techniques are used. Figure 15 shows a general-purpose optical spectrum analyzer with a typical measurement trace on the display screen.

Figure 15: A general-purpose optical spectrum analyzer (Model HP-70951A, from Hewlett-Packard)

2.1.5 Optical time –domain reflectometer (OTDR)

The long-term workhorse instrument, in optical systems is the OTDR. In addition to locating faults within an optical link, this instruments measures parameters such as attenuation, length, connector and splice losses, and reflectance levels. A typical OTDR consists of an optical source and receiver, a data acquisition module, a central processing unit (CPU), an information storage unit for retaining data either in the internal memory or on an external disk, and a display. Figure 16 shows an example of OTDR.

Figure 17 shows the bases of the OTDR technique. OTDR fundamentally is an optical radar. It operates by periodically launching narrow laser pulses into one end of a fiber under test by using either a directional coupler or beam splitter. The properties of the optical fiber link are then determined by analyzing the amplitude and temporal characteristics of the waveform of the backscatter light.

Universal Test System

Figure 16: Optical time –domain reflectometer (OTDR) (Model FTB300, from

EXFO)

Figure 17: Principle of an Optical time-domain Reflectometer (OTDR)

OTDR Applications:

Figure 18 shows a typical trace as would be seen on the display screen of an OTDR. The scale of the vertical axis is logarithmic and measures the returning (back-reflected) signal in decibels. The horizontal axis denotes the distance between the instrument and the measurement point in the fiber the backscattered waveform has four distinct features:

1) A large initial pulse resulting from Fresenel reflection at the input end of the fiber.

2) A long decaying tail resulting from Rayleigh scattering in the reverse direction as the input pulse travels a long the fiber.

3) Abrupt shifts in the curve caused by optical loss at joints or connectors in the fiber line.

4) Positive spikes arising from Fresenel reflection at the far end of the fiber, at fiber joints, and at fiber imperfections.

Fresenel reflection and Rayleigh scattering principally produce the backscattered light. Fresenel reflection occurs when light enters a medium that has a different index of reflection.

Figure 18: Representative trace of backscattered optical power as displayed on

OTDR screen

Attenuation Measurements:

The optical power at a distance x from the input coupler can be written as

..(1)

here, P(0) is the fiber input power and is the fiber loss coefficient in Km-1

which may be position –dependent . the parameter 2can be measured in natural units called nepers , which are related to the loss in decibels per kilometer through the relationship

…(2)

under the assumption that the scattering is the same at all points along the optical waveguide and is independent of the modal distribution ,the power PR(x) scattered in the reverse direction at the point x is

PR(x) = S P(x) ..(3)

Here ,S is the fraction of the total power that is scattered in the backward direction and trapped in the fiber. thus the back scattered power from the point x that is seen by the photodetector is

… (4)

where is the loss coefficient for the reverse scattered light .since the modes in the fiber excited by the backscattered light can be different from those launched in the forward direction, the parameter may be different from .

substituting Eqs.1,2,and 3 into Eq 4 yields

… (5)

where the average attenuation coefficient is defined as

(6)

using this equation ,the average attenuation coefficient can be found from experimental semi log data plots such as the one shown in Figure 18. For example, the average attenuation between two points x1 and x2 , where x1>x2 is

Where PD is the backscattered power from point X that is seen by the photodetector.

Fiber Fault Location:

The fiber length L (and, hence, the position of the break or fault) can be calculated from the time difference between the pulses reflected from the front and far ends of the fiber. if this time difference is t, then the length L is given by

…(2)

Where n1 is the core refractive index of the fiber. the factor “ 2” accounts for the fact that light travel s a length L from the source to the break point and then another length L on the return trip.

2.1.6 Multifunction optical test systems

For laboratory, manufacturing, and quality-control environments, there are instruments with exchangeable modules for performing a variety of measurements. Figure 19 shows an example from EXFO, which includes a basic mainframe and expansion unit. The mainframe is a microprocessor-based unit that coordinates data compilation and analysis from variety of test instruments. This test system can control external instruments that have RS232 communication capability, and it has networking capability for remote access from a computer. The plug –in modules cover a wide range of test capabilities. Example functions include single channel or multichannel power meter, tunable laser sources, variable attenuator, optical spectrum analyzer, and PMD analyzer (polarization-mode dispersion).

Figure 19: Multifunction Optical Test System (Model IQ-203, from EXFO)

2.2 Attenuation measurements

Attenuation of an optical power in a fiber waveguide is a result of absorption process, scattering mechanisms, and waveguide effects. The manufacturer is generally interested in the magnitude of the individual contributions to attenuation, whereas the system engineer who uses the fiber is more concerned with the total transmission loss of the fiber. Here we shall treat measurements techniques for total transmission loss.

Three basic methods are available for determining attenuation in fibers. The earliest devised and most common approach involves measuring the optical power transmitted through a long and a short length of the same fiber using identical input couplings. This method is known as the cutback technique .A less accurate but nondestructive method is the insertion-loss method, which is useful for cables with connectors on them.

2.2.1 Cutback technique

The cutback technique, which is a destructive method requiring access to both ends of the fiber. Is illustrated in Figure 20. Measurements may be made at one or more specific wavelengths, or alternatively a spectral response may be required over a range of wavelengths. To find the transmission loss, the optical power is first measured at the output (or far end) of the fiber. Then without disturbing the input condition, the fiber is cut off a few meters from the source, and the output power at this near end is measured. If PF and P­N represent the output powers of the far and near ends of the fiber, respectively, the average loss α in decibels per kilometer is given by:

…(3)

Where L (in kilometer) is the separation of the two-measurment points. The reason for following these steps is that it is extremely difficult to calculate the exact amount of optical power launched in to the fiber. By using the cutback method, the optical power emerging from the short fiber length is the input power to the fiber of length L.

In carrying out this measurement technique, special attenuation must be paid to how optical power is launched into the fiber. This is because, in multimode fiber, different launch conditions (different numerical apertures and spot sizes at the launch end of the fiber) can yield different loss values.

Figure 20: Cutback Technique

2.2.2 Insertion loss method

For cables with connectors one cannot use the cutback method. In this case one commonly uses an insertion loss technique. This is less accurate than the cutback method, but is intended for field measurements to give the total attenuation of a cable assembly in decibels.

The basic setup is shown in Figure 21,where the launch and detectors couplings are made through connectors. The wavelength tunable light source is coupled to a short length of fiber that has the same basic characteristics as the fiber to be tested. A wavelength selective device, such as an optical filter, is generally included to find the attenuation as a function of wavelength.

To carry out the attenuation tests, the connector of the short length launch in fiber is attached to the connector of the receiving system and the launch power level P1(λ) is recorded. Next, the cable assembly to be tested is connected between the launching and receiving systems, and the received power level P2(λ) is recorded. the attenuation of the cable in decibels is then

…(4)

This attenuation is the sum of the loss of the cabled fiber and the connector between the launch connector and the cable.

Figure 21: insertion loss technique

2.3 Dispersion Measurements

Three basic forms of dispersion produce pulse broadening of light wave signals in optical fibers, thereby limiting the information carrying capacity. In multimode fibers, Intermodal dispersion arises from the fact that each mode in an optical pulse travels slightly different distance and thus arrives at the fiber end at slightly offset times. Chromatic dispersion stems from the variation in the propagation speed of the individual wavelength components of an optical signal. Polarization mode dispersion arises from splitting of a polarized signal into orthogonal polarization modes, each of which has a different propagation speed.

There are many ways to measure the various dispersion effects. Here, we look on some common methods

2.3.1 Intermodal dispersion

For pulse dispersion to be negligible in digital systems, one of the approximately equivalent conditions should be satisfied:

(1) The fiber transfer function must not roll off to less than 0.5 of its low frequency value for frequencies up to half the desired bit rate.

(2) The rms width of the fiber impulse response must be less than one-quarter of the pulse spacing.

2.3.2 Time –domain Intermodal dispersion measurements

The simplest approach for making pulse dispersion measurements in the time domain is to inject a narrow pulse of optical energy into end of an optical fiber and detect the broadened output pulse at the other end. Figure 22 illustrates a setup for this. Here output pulses from a laser source are coupled through a mode scrambler into a test fiber. The output of the fiber is measured with a sampling oscilloscope that has a built-in optical receiver, or the signal can be detected by external photodetector and then measured with a sampling oscilloscope. Next the shape of the input pulse is measured the same way by replacing the test fiber with a short reference fiber that has a length less than 1 percent of the test fiber length. This reference fiber can be short length cut from the test fiber or it can be a fiber segment that has similar properties. The variable delay in the trigger line is used to offset the difference in delay between the test fiber and the shorter reference fiber.

From the output pulse shape, an rms pulse width σ as defined in the figure 23 can be calculated by:

… (5)

Where the center time of the pulse is determined from

… (6)

The evaluation of the equation (6) requires a numerical integration. An easier method is to assume that the output response of a fiber can be approximated by a Gaussian described by

… (7)

Where the parameter determines the pulse width, as shown in the figure 23.

Figure 22: Time domain Intermodal Dispersion Measurement.

Figure 23: Definition of pulse shape parameters.

2.3.3 Frequency-Domain Intermodal Dispersion Measurements

Frequency domain Intermodal dispersion measurements yield information on amplitude –versus-frequency response and phase –versus- frequency response. These data are often more useful for system designers than time domain pulse dispersion measurements, especially if equalization techniques are to be performed on the detected signal at the receiver. The dispersion measurements can be made by sinusoidally modulating a narrowband continuous wave (CW) light signal about a fixed level. The baseband frequency response is then found from the ratio of the sine wave amplitudes at the beginning and end of the fiber.

Figure 24 shows an experimental arrangement for finding fiber baseband frequency response. A swept-frequency RF source or a microwave signal source is used to modulate an optical carrier sinsoidally .the optical signal is coupled through a mode scrambler to the test fiber .At the exit end of the fiber, a photo detector measures Pout(f), the output power as a function of the modulation frequency. The input signal is then measured by substituting a short reference fiber for the test fiber, thereby yielding Pin(f).

Comparison of the spectrum at the fiber output with the spectrum at the fiber input provides the baseband frequency response H (f) of the fiber under test:

… (8)

As the modulation frequency is increased, the optical power level at the fiber output will eventually start to decrease. The fiber bandwidth is defined as the lowest frequency at which H (f) has been reduced to 0.5.

Figure 24: Frequency Domain Intermodal Dispersion

2.3.4 Chromatic Dispersion

Chromatic dispersion is the primary dispersion mechanism in single mode fibers. Here, we present one method for its measurements.

Figure 25 shows a setup for chromatic dispersion measurements by the modulation phase shift method. An electrical signal generator intensity modulates the output of a narrow band, tunable optical source by means of an external modulator. After detecting the transmitted signal with a photodiode receiver. A vector voltmeter is used to measure the phase of the modulation of the received signal relative to the electrical modulation source. The phase measurement is repeated at wavelength intervals across the spectral band of interest. Using the measurements at any two adjacent wavelengths, the change in-group delay over the wavelength interval between them is

…(9)

Where λ is the wavelength at the center of the interval, fm is the modulation frequency in MHz, Φ and is the phase of the measured modulation in degrees.

These data points are then plotted to yield the typical curve shown in Figure 25,the dispersion can be calculated by applying the curve fitting equations to the pulse delay data.

Figure 26: Chromatic Dispersion

2.3.5 Polarization Mode Dispersion (PMD)

At least seven different methods have been developed for measuring PMD. Here, we present only the fixed-analyzer method. In this technique, the mean differential group delay is evaluated statistically from the number of peaks and valleys appearing in the optical power as it is transmitted through the polarizer and scanned as a function of wavelength. Figure 26 shows a simple setup using a spectrum analyzer. A typical spectrum analyzer trace showing the transmitted power level as a function of a wavelength is given in Figure 27. Automatic methods using extrema counting and Fourier analysis are used to extract the PMD information from the measurement data. Using extrema counting the expected value of the differential group delay of the fiber (or of any other device) under test can be calculated from the relationship

…(10)

Where and are the beginning and the end, respectively, of the wavelength measurement sweep, Ne represents the number of extrema occurring in the scan, and c is the speed of the light. The dimensionless mode-coupling factor k statistically accounts for the wavelength dependence of the polarization states. The subscript λ on the () terms means that the expected value of the differential group delay is determined over a wavelength span.

Figure 26: Polarization mode Dispersion

Figure 27: typical spectrum analyzer trace showing the transmitted power level

As a function of wavelength.

2.4 Eye Patterns

The eye pattern technique is a simple but powerful measurement method .the eye pattern method measurements are made in time domain and allow the effects of waveform distortion to be shown immediately on an oscilloscope.

Figure 28 shows the basic equipment setup for making eye-pattern measurements.

Figure 28: The basic equipment setup for making eye-pattern measurements

The following information can be derived from the eye-pattern:

(Note: To interpret the eye pattern, consider the simplified drawing shown in

Figure 29. )

Figure 29: Simplified eye diagram showing the key performance parameters

1) Timing jitter (also referred to as edge jitter or phase distortion) in optical fiber system arises from the noise in the receiver and pulse distortion in the optical fiber. If the signal is sampled in the middle of the time interval, then the amount of distortion at the threshold level indicates the amount of jitter. Timing jitter is thus given by

Timing jitter (percent) …(11)

2) System rise time, rise time is defined as the time interval between the point where the rising edge of the signal reaches 10 percent of its final amplitude and the time it reaches 90 percent of its final amplitude.

3) The width of the eye opening defines the time interval over which the received signal can be sampled without error from itersymbol interference.

4) The best time to sample the received signal is when the height of the eye opening is largest. This height is reduced as a result of amplitude distortion in the data signal. The vertical distance between the top of the eye opening and the maximum signal level gives the maximum distortion. The more the eye closes, the more difficult it is to distinguish between ones and zeros in the signal.

5) The rate at which the eye closes as the sampling time is varied (i.e. the slope of the eye pattern sides) determines the sensitivity of the system to timing errors .The possibility of timing errors increases as the slope becomes more horizontal.

6) Any nonlinearity of the channel transfer characteristics will create an asymmetry in the eye pattern .if a purely random data stream is passed through a purely linear system, all the eye opening will be identical and symmetrical.

References

1. Joseph C.Palais . “Fiber Optic Communications “ ,4th Edition. Prentice –

Hall,1998.

2. Gerd Keiser.” Optical Fiber Communication”, 3rd Edition. McGraw-HILL

International Edition , Electrical Engineering Series, New York,2000.

3. Dietrich Marcuse."Principle of optical fiber measurements",1st Edition,USA,1981.

4. Ajoy Chatak, K.Thyagarajan” Introduction to Fiber Optics”.1st Edition.USA,1998.

5. DR.Hideo Fukutomi,"Optical Fiber Cable",1st Edition 1986.

6. Stewart D.Personick." Optical Fiber Transmission System".

7. Charles K.Kao ."Optical Fiber System":Technology,Design and

Applications".McGraw-Hill.1982.

8. Internet