e-book Optoelectronics and Fiber Optic Technology

Free download. Book file PDF easily for everyone and every device. You can download and read online Optoelectronics and Fiber Optic Technology file PDF Book only if you are registered here. And also you can download or read online all Book PDF file that related with Optoelectronics and Fiber Optic Technology book. Happy reading Optoelectronics and Fiber Optic Technology Bookeveryone. Download file Free Book PDF Optoelectronics and Fiber Optic Technology at Complete PDF Library. This Book have some digital formats such us :paperbook, ebook, kindle, epub, fb2 and another formats. Here is The CompletePDF Book Library. It's free to register here to get Book file PDF Optoelectronics and Fiber Optic Technology Pocket Guide.

Optoelectronic devices are electrical-to-optical or optical-to-electrical transducers , or instruments that use such devices in their operation. Optoelectronics is based on the quantum mechanical effects of light on electronic materials, especially semiconductors , sometimes in the presence of electric fields. Important applications [3] of optoelectronics include:. From Wikipedia, the free encyclopedia. Branch of electronics involving optics. Use CE mark is European Union member to selling a product Trade Assurance The supplier supports Trade Assurance — A free service that protects your orders from payment to delivery.

Gold Supplier Gold Supplier is a premium membership for suppliers on Alibaba. Members are provided with comprehensive ways to promote their products, maximizing product exposure and increasing return-on-investment. Product Categories. Fiber Optic Cable. Fiber Optic Media Converter. Passive Fiber Optic Components. CATV Equipment. Company Profile. Company Overview Company Capability. Business Performance.

Additional Information.

Optoelectronics--where photons meet electrons

Trade Shows. In the latter, the pulses are all of the same width but the pulse rate changes to differentiate between logic '1' and logic '0'. The inverting input is used to supply the modulating drive to the LED or laser while the non-inverting input supplies a DC bias reference. By modulating a subcarrier it is then possible to create an analog transmission channel which could be suitable for transmitting video and frequency division multiplex FDM signals as well as digital signals. Several analog signals for instance, video channels can be combined to form an FDM signal, which can then be used to frequency modulate the electrical subcarrier.

This method would, however, reduce the dynamic range of the channels and a far better method would be to frequency modulate a number of carriers with video signals and then to multiplex these on a frequency division basis. Another disadvantage is associated with the fiber itself as, owing to the multipath propagation properties of step-index and graded-index multimode fibers, the signal is greatly distorted. Nowadays, therefore, the only practical type of modulation that can be used is intensity power modulation whose main advantage is that the optical source can be directly modulated by the drive current and the photodiode will convert the received optical power directly into an electrical photocurrent.

There are two main types of intensity modulation: 9 direct intensity modulation; and 9 premodulation with subsequent intensity modulation. Direct intensity modulation Direct intensity modulation is a very simple process, which relies upon the optical signal and the electrical signal being proportional.

Provided that the transmitter and the receiver characteristics are exactly the same, a high linearity can be realized, but as this can only normally be achieved in small sections, its use is restricted mainly to analog transmission although it is sometimes used for transmitting signals over short distances. The design is again simple and this method is particularly suitable for transmitting video signals over short or medium distances.

The usable bandwidth of an intensity power undulated optoelectronic transmission path is determined by: 9 the modulation bandwidth of the optical transmitter; 9 the dispersion effects in the fiber e. The ratio of the power of the two is called the 'signal-to-noise ratio' SNR. SNR is a measure of the quality of a signal and is usually expressed in decibels dB. This noise can be minimized by utilizing a device such as a transimpedence amplifier to reduce the effective time constant and diode capacitance, which limits the receiver frequency response.

Although using an avalanche photodiode can reduce this effect, avalanche diodes themselves produce noise and so there is a limitation to their use. Dark current noise is normally reduced in optical receivers by using a low pass filter. This is particularly noticeable at junctions and splices. Signal-to-noise ratios of around 40 dB can be expected with gainguided laser diodes see Chapter 4.

A mirror reflects the light. Theory 71 Figure 2. Although electrical multiplexing is much more economic and considerably less complex than wavelength division multiplexing see later for normal telecommunications links, the use of a wavelength division multiplex for subscriber line networks is advisable due to the greater flexibility provided for the gradual expansion of installed networks.

For instance, in an initial stage, narrow band sources such as telephone and data transmission can be transmitted over a single fiber to the subscriber using bi-directional wavelength division multiplex in the range nm. If, at a later date, visual communication is required, then signals can be transmitted in the nm region over the same fiber by the use of additional optoelectronic transducers. Theory 73 Figure 2. However, optical fibers have the ability to carry multiple channels using technology known as WDM.

Put simply, this uses the same fiber to carry different channels using different wavelengths or 'colours' of light. Several techniques of achieving this exist, such as using a prism as a wavelength dispersion element Figure 2. The devices that perform wavelength division multiplexing and demultiplexing are commonly called couplers.

In the example system shown in Figure 2. This sort of coupler is called a muldex coupler: a contraction of the words multiplexer and demultiplexer. This technology enables signals that are independent of each other, at different carrier wavelengths and close channel spacing to be modulated in a different manner i. This technology is ideal for high volume point-to-point or backbone links with minimal switching and routing requirements.

With current technology, the signal must be converted at switching points along the route from optical to electrical, the header information analysed, then the signal must be returned to the optical domain for correct routing. One side effect of DWDM is that the overall efficiency of the link is reduced due to added noise and the bandwidth restrictions imposed by the electronics. Theory 75 This ensures that only one carrier is actually t r a n s m i t t e d at a n y one instance. The difference between the two is their relative size and optical performance. In Chapter 3 their construction, use and technology are reviewed.

Optical fiber communication systems were originally developed as a means of transmitting high bandwidth signals over long distances. One reason for this was the growing requirement for computers to manipulate and transfer data at ever-faster speeds. Figure 3. The core region's refractive index or optical density is greater than the cladding layer and the boundary between the layers is where the light reflects. The base material used in fibers for most long-distance communication is silica glass which, in its purest form, has the desired qualities of low attenuation, low cost and is readily availability.

The high refractive index of the core can be achieved by adding dopants such as phosphorous, germanium and boron to the silica during manufacture. However, this does not mean that all fibers for all applications are made from silica glass. Single-mode fibers are capable of wide bandwidths for example, up to 40 GHz , and are ideally suited for long haul, low cost and high capacity circuits for the transmission of telephone and cable television etc. Although they have primarily been developed for the nm window, they can also be used effectively with time division multiplex TDM and wavelength division multiplex WDM systems operating in the nm wavelength region.

Instead of the Normally, as much as 20 per cent of the light in a single-mode cable actually travels down the cladding and the effective diameter of the cable is a blend of the single-mode core and the degree to which the cladding carries light. This is referred to as the 'mode field diameter', which can be larger than the physical diameter of the core depending on the refractive indices of the core and cladding. The unfortunate disadvantage of using a single-mode fiber is the difficulty experienced in trying to couple the extremely small crosssectional area of the fiber core to an optoelectronic detector or when trying to splice it to another length of fiber.

If the fiber is subjected to mechanical stress, local discontinuities can be introduced. Curvatures of the fiber involving axial displacements of a few micrometres and spatial wavelengths of a few millimetres result in unwanted light radiance and extra attenuation losses. These are referred to as microbending losses or in the case of macroscopic axial deviation of the fiber from straight line, macrobending losses see Chapter 2.

Optoelectronics and Fiber Optic Technology - O'Reilly Media

To overcome these problems, high intensity laser diodes are frequently used. Another disadvantage of using single-mode fibers is that as the refractive index of glass decreases with optical wavelength, the light Fibers and cables 79 velocity will also be wavelength dependent. Thus the light from an optical transmitter will have a definite spectral width e. There is, therefore, an advantage to using a limited multimode fiber. As the name implies, multimode fibers are capable of propagating more than one mode at a time and they are ideally suited for high bandwidth i.

The outer cladding is normally ten wavelengths thick and approximately twice the size of the core diameter. As its name implies, there are multiple paths modes by which a light 'ray' may be propagated. For example, a fiber with a core diameter of The problem with multimode cable, however, is that as some of the modes are longer than others, the pulse of light will be 'spread out' due to the modal dispersion. This causes an effect referred to as 'intersymbol interference', which restricts the distance that a pulse can be usefully sent over multimode fiber.

There are two types of multimode fiber: 9 step index; 9 graded index. This enables propagation along paths that are much longer than the axial route but attenuation and dispersion losses are subsequently increased. G indicates that 'the core diameter should be 50 ]am and the cladding ]am' and enclosing it in a cladding material whose refractive index is only slightly lower than the fiber but nevertheless capable of ensuring that the refractive index changes directly at the core-cladding interface not only will the critical angle be reduced, but losses will also be minimized.

In a simple step index fiber the refractive index of the cladding is typically 10 per cent lower than the refractive index of the core and, provided that the light coupled into the fiber is of a low acceptance angle, total reflection of light rays will occur. It should be noted, however, that although the light velocity will be constant within the core, a ray with a high angle will have a smaller velocity component than those with a smaller reflective angle.

An angle of 12 degrees is usually considered as a typical value of acceptance angle. Light rays passing through the fiber are continuously reflected off the glass cladding towards the centre of the core at different angles and lengths, limiting the overall bandwidth of the cable. The main disadvantage of step index fibers is that the different optical lengths caused by the various angles at which light is propagated relative to the core, causes the transmission bandwidth to be fairly small.

These effects can be avoided by using graded index fiber, and because of these limitations, multimode step index fiber is typically only used in applications requiring distances of less than I km. This can be achieved by carefully manufacturing the fiber so that the light entering the fiber with a narrow acceptance angle can be made to travel more slowly than the light that is being reflected by the modes near the outside of the core.

Varying the path length speed i. A graded index fiber typically transmits roughly modes. Although previously only used for trunk i. With multimode graded index fiber, the core's refractive index gradually decreases from the centre to the outer edge, resulting in a slow bending of the light rather than an abrupt bouncing or 'step' pattern and because of these variations in refractive index, the trajectory of the light is sinusoidal as opposed to zigzag and the light rays follow curved paths.

To Probe Further

The reason for this phenomenon is that the core is, in effect, made up of a central membrane or fiber axis, surrounded by a succession of infinitely thin shells each consisting of a different, but constant, refractive index. From Figure 3. Cladding - i Cladding I Figure 3. Owing to the reduced refractive index at the core edges, the rays on more deviant paths will travel much faster than axial rays and cause the light components to take almost the same amount of time to travel the length of the fiber, thus minimizing dispersion losses.

Light is, therefore, transmitted along a fiber in a multitude of different paths, varying from those parallel to the fiber axis to those approaching the critical angle. Each path at a different angle is termed a 'transmission mode', and the NA of graded index fiber is defined as the maximum value of acceptance angle at the fiber axis. Typical attenuation coefficients of graded-index fibers at nm are 2.

The main advantage of using graded-index fiber is the reduced refractive index at the centre of the core. This is particularly advantageous when trying to couple the tiny emitting area and sharply focused beam of a small LED or Laser diode to a fiber.

  • Participation and Power: Civic Discourse in Environmental Policy Decisions?
  • Optoelectronics--where photons meet electrons | Cabling Installation & Maintenance.
  • Navigation menu;
  • Stik.
  • Curbing Corruption: Toward a Model for Building National Integrity (Edi Development Studies)?
  • An Introduction to Relativistic Quantum Field Theory.

Another advantage of using graded index fiber is that it is comparatively cheap to produce. Table 3. Nowadays, of course, attenuations of less than I dB km -1, and in exceptional cases as low as 0. Low transmission bandwidth. Short-distance applications e. Modal Dispersion reduced. Very low attenuation factor. Coupling efficiency. Used for long-haul telecommunications i. Basically, there are three separate steps: preparation of the glass preform, forming the glass rod and fiber drawing.

There are three main methods of forming the glass rod: internal vapour deposition, external vapour deposition and vapour axial deposition. By further heating, this hollow rod is collapsed to form a solid cylindrical rod with an internal core and a cladding of exactly the right refractive index.

An outer cladding region tube of commercial silica glass then surrounds the rod. This process is referred to as the deposition drying and sintering process. A steel mandrel is mounted in a lathe and rotated. A burner is then fed with chlorides of silica and the other required dopants and mode to traverse the mandrel.

The dopants oxidize in the oxy-hydrogen flame and are deposited on the mandrel in the form of soot. After all the required layers have been deposited, the ceramic rod is then removed and the Silicon dioxide powder for core and claddino Ceramic substrate rod i 0 Figure 3. This method was originally developed by Corning.

As it grows, the porous glass body is d r a w n off and heated to form a narrow solid rod. One advantage of this process is that the porous rod does not have to be removed. The area of the end face, however, limits the a m o u n t of silicon dioxide that can be deposited and therefore the growth rate.

It is important that this is done at a steady rate to minimize the a m o u n t of diameter variations in the fiber. Care must also be taken to ensure that the surface of the glass is not contaminated by foreign particles, as this will cause microcracks in the fiber. Most manufacturers n o w a d a y s look for a chemically pure glass with less than one foreign atom in silicon atoms. This makes fibers easier to handle, protects them from microbending losses and mechanical damage and is normally about pm in external diameter.

To improve strength, fatigue resistance, resistance to increased hydrogen-induced attenuation and to protect against chemical corrosives such as hydrofluoric acid, a hermetic coating can be applied between the glass and the coating. One of the problems of making a fiber in this manner is an undesirable dip in refractive index produced within the central core region. This is the direct result of the intense heat required for the collapsing and fiber drawing stages because heat causes the germanium dioxide core dopant to be evaporated away from the tube's inner surface.

In multimode fibers, this refractive index dip can cause reduced coupling efficiency, reduced Fibers and cables 89 Figure 3. In single-mode fiber, the dip can result in increased microcurvature sensitivity and changes in the cut-off wavelength or the shortest wavelength at which only the fundamental mode of an optical waveguide is capable of propagation.

In an attempt to overcome these problems, AEG Telefunken developed a far more efficient method of applying the dopant, called full length outside deposition FLOD. This process is shown in Figure 3. In addition to maintaining a virtually constant refractive index throughout the fiber, the FLOD method has the advantage that it does not require such a high operating temperature. In the process, a tubular glass body with concentric regions containing a number of doped glass layers, each having a different refractive index, is produced first.

The glass body is then heated to its softening point and drawn into a glass fiber. During drawing, a partial vacuum is maintained in the hollow centre of the tubular body. The size of this partial vacuum is sufficient to reduce the evaporation of the doping material from the tubular body interior, which effectively minimizes the dip in the index of refraction in the centre of the optical fiber. Initially, optical cladding, consisting primarily of fused silica, is deposited on the inside of the substrate tube. Core material, which may also contain germanium, is then deposited at a reduced temperature to form a diffuse and permeable layer known as a 'frit'.

After deposition of the frit, this partially completed preform is sealed at one end, removed from the lathe and a solution of suitable salts of the desired rare earth dopant e. Over a fixed period of time, this solution is left to permeate the frit. After discarding excess solution, the preform is returned to the lathe to be dried and consolidated. During consolidation, the interstices within the frit collapse and encapsulate the rare earth. Finally, the preform is subjected to a controlled collapse at high temperature to form a solid rod of glass - with a rare earth incorporated into the core.

To make a rare earth-doped fiber, you need to start with a rare earthdoped preform using the solution doping process. The actual manufacturing process is similar to the other methods previously described. Where this is not available, or short cuts of cable are being used, an 'eye' can easily be locally manufactured, as shown in Figure 3.


In small capacity optical fiber cables, the factory-fitted cap is often transparent see Figure 3. Fibers and cables 91 Figure 3. I mm , increased flexibility and, because the ends can be cut using a hot razor blade, ease of termination. Unfortunately, however, owing to the material's high intrinsic loss, the use of plastic fibers is normally restricted to only a few metres and to environments protected from temperature extremes.

Plastic optical fiber POF , however, offers noise immunity and low cable weight and volume, and is very competitive with shielded copper wire, thus making it suitable for industrial applications, such as those found in factories etc.

  • A People’s History of the World: From the Stone Age to the New Millennium.
  • [Khare, R.P.] Fiber Optics and Optoelectronics(b ok.xyz).
  • Optical fiber - Wikipedia?

Silica glass optical fiber has better light transmission characteristics less loss than POF and silica fiber can tolerate higher temperatures than plastic fiber. On the other hand, POF is more flexible, less prone to breakage, easier to fabricate into special assemblies, and lower in cost than glass fibers. The bending diameter of an optical fiber cable is the minimum permissible diameter to which the cable can be bent without damaging the conductors. The deflecting diameter is the minimum permissible diameter over which the cable can be drawn or guided while being subjected to a high tensile force.

Fibers and cables 93 Figure 3. For example, cables with an outside diameter of up to 30 m m can be bent to mm. Cables with metallic elements, however, are not so flexible as they use wire armouring or have an aluminium-laminated sheath with copper conductors. To meet the overgrowing requirement for heavyduty cables, these ruggedized cables typically have a tensile strength in excess of lb. Normally, and to enable easy identification, each fiber is colour coded with a 2 p m thin film of coating material.

1st Edition

This coating has no effect on the optical properties of the fiber. These thixotropic filling compounds gases, gels, pastes etc. The flammability of the buffer tube must also be insignificant and the tube must be capable of being easily cleaned. The tube and filling compound enable the core and fibers to adjust to a wide range of tensions and they are also impact resistant and laterally strengthened.

Past experience has shown that the wall thickness of the buffer tube should be about 15 per cent of the overall buffer diameter and single-fiber buffers combined in one cable form should ideally have an outside diameter of 1. The cable is often then encased in a rugged, flexible, age-resistant polyethylene or similar jacket which has a relatively low coefficient of expansion over a wide temperature range to allow the cable to be placed in ducts and pulled through in a similar manner to copper cables.

A further Kevlar strength member may also be included as shown in Figure 3. They are small, compact, relatively inexpensive and are by far the most common type on the data communications market. The fiber is loosely laid into a hollow tube which may be filled with a chemically neutral thixotropic gel to prevent ingress of water. Many of these tubes may be laid up together and sheathed to make the final cable.

This type is primarily used for telecommunications applications, major links and inter-building cross-connections. The trouble with this type of cable is the difficulty in providing sufficient protection at junction points. Fibers can, of course, be grouped together in a very similar manner to multipair copper cables and are normally contained in plastic ducts. To ease identification, individual fibers are colour coded. Note: If there are more than 12 fibers in a tube they are usually 'bundled' together in quantities of 12 and held together with a coloured binding yarn.

The pilot tube is red, the directional one green and all the other tubes are natural coloured. Kevlar , which is either laid helically or braided around the fiber coating. This is then surrounded by a tough outer sheath to provide the required environmental and mechanical protection Figure 3. Fibers and cables 99 Figure 3. The nonmetallic construction eliminates problems caused by earth loops and induced voltages caused by lightning strikes.

Picture Optoelectronics and Fiber Optic Technology 3. Picture 3. Picture Fibers and cables Note: The diameter of the fiber is an important consideration. Increasing the diameter of the fiber will reduce tolerance problems, but the difficulty of producing a clean fracture surface is increased, fiber flexibility is reduced and costs are increased. For additional protection against moisture cables are also available with a laminated aluminium sheath Figure 3.

Protection is usually achieved by incorporating non-metallic armouring and protective poly- Optoelectronics and Fiber Optic Technology Figure 3. Non-metallic protection has advantages in that there are: 9no problems with grounding or potential equalization; 9 no lightning protection measures are required; 9 buildings can be electrically isolated. Non-metallic armouring includes laminated glass-fiber yarns which also serve as strength elements and a polyamide the hardest plastic used for sheathing cables jacket.

Metallic armouring, however, is still probably considered the most effective protection against rodent damage but it has cost limitations. This is usually some form of metallic armouring such as corrugated steel tape under the outer jacket Figure 3. Fibers and cables Figure 3. Where the mechanical loads are exceptionally high, a further inner jacket can be inserted underneath the steel tape.

In this particular example, two individually coloured primary-coated optical fibers are housed within a gel-filled polyester tube. Up to 12 of these tubes can be helically stranded around a non-metallic central strength member. The spaces between the tubes are then filled with a non-halogenated water blocking compound. The inner sheath is made out of polyethylene whilst the outer sheath is made from a hard grade nylon. Normally the cables are weatherproof and resistant to attacks by rodents and insects.

The technique used, of course, depends not only upon the environment but on the length of the cable to be installed. Today, one of the most practised methods of installing short lengths into highly populated duct systems is to use compressed air to inject an auxiliary rope. This will then be connected to a winch rope, which in turn will be connected to the cable. Ideally, however, a system of ducts should be included in preliminary drawings of all new buildings. The price of the actual duct is very cheap and the building can be constructed so that it has a cost-effective futureproof wiring system capable of meeting changing communication requirements.

Because optical fiber cables are very light, they can be incorporated into overhead power line routes without any additional stress being exerted on to the existing mast Figure 3. In addition, as the optical fiber cable is completely non-metallic, any problems associated with inductive interference are eliminated. The process is very similar to installing normal copper aerial cables and the optical fiber is merely lashed to the phase conductors or ground wires of the overhead power line.

The ideal requirements for this light source are: 9 9 9 9 9 small spot size; fast switching speed; high intensity; monochromatic light; operation at wavelengths between and nm. Depending on the nature of this signal, the resulting modulated light may be turned on and off or may have its intensity varied sinusoidally between two predetermined levels. Figure 4. This conversion of electrical energy into optical energy is achieved by the use of an optoelectronic semiconductor device, such as a light emitting or laser diode.

Although the radiation from a light emitting diode LED is quite safe to the naked eye, there is, of course, a potential hazard with the use of lasers and high radiance LEDs because of the damage that these high intensity point sources can cause through absorption of energy inside the eye's retina. The main requirements for lasers and LED devices is that they should: 9 be small; 9 be rugged; 9 possess a long operating life at least comparable to other components within the system ; 9 have an emission only in the low attenuation region of the optical waveguide; 9 have a high coupling efficiency between semiconductor and optical waveguide; 9 have a large radiating power i.

See Table 4. Although these are more expensive to manufacture, they have the advantage that they have a high breakdown voltage, high electron mobility, are a high speed device and so give a relatively greater bandwidth. An LED is essentially a pn semiconductor that emits light w h e n a forward bias is applied i. The forward bias causes the electrons and holes to move towards one another and cross the depletion area of the junction. Electrons and holes combine and cause light to be emitted and as long as the voltage is applied, the electrons will keep on flowing through the diode and recombination will continue at the junction.

An LED is, therefore, a very efficient semiconductor diode that converts electrical energy into light in a process k n o w n as spontaneous emission. In many semiconductors, the energy produced by recombination is released as heat; however, in materials used in the manufacture of LEDs, the recombination energy is released as a photon of light which can emerge from the semiconductor material.

The materials used in the manufacture of LEDs e. GaAs LEDs will normally emit light at nearly nm. Adding aluminium can increase the band gap and shift the emission to wavelengths of nm. Other semiconductor compounds for instance, GaAsP will produce wavelengths of nm visible red and InGaAsP the and nm wavelengths.

In practice LEDs are very rarely used in nm systems. There are various disadvantages associated with using an LED as a transmitter. Simple LEDs, such as the one shown in Figure 4. More complex internal structures can concentrate the emission of surface emitting LEDs into a narrower angle Burrus diodes while other configurations can cause the light to be emitted from the edge of the LED in a more concentrated beam. Although this will produce more power, it requires more complicated circuitry to dissipate the extra heat.

In general, the more complex the LED structure, the brighter and more collimated is the emitted light and the faster the LED can be modulated. As the average lifetime of a radiative recombination is only a few nanoseconds, the modulation bandwidth of an LED is limited to only a few hundred megahertz as opposed to laser diodes, which can be modulated to many gigahertz. Another disadvantage of LEDs is their low coupling efficiency and the large amount of inherent chromatic dispersion.

On the other hand, the main advantage of using LEDs in optical fiber systems is that their small core diameter permits an enhanced coupling efficiency of the light into the optical waveguide. Other advantages include: 9 9 9 9 9 simple design; ease of manufacture; simple system integration; low cost; and high reliability. As the light and current characteristics of LEDs are only slightly dependent upon aging and temperature, the design of an LED optical transmitter is comparatively simple and virtually only consists of a driver to modulate the LED. This double heterostructure make-up restricts the production of photons to a narrow concentrated beam because, although they might have sufficient energy to drift across the inner forward biased junction, they cannot overcome the higher barrier potential associated with the outer regions.

In a similar manner to that of the fiber cladding, the outer layers also have a lower refractive index, which helps to confine the emitted photons by internal reflection. By varying the mixture of the junction, different energy gaps can be produced which will provide different beam concentrations. High radiance LEDS, emitting at and nm, are often used as they have an intrinsically high quantum efficiency due to the advantages gained by the double heterostructure of the diode.

How To Talk Fiber Optics - The Language of Fiber Optics

Note: Quantum efficiency- the average number of electrons for each absorbed quantum of light emitted. By designing the diode so that the size of the emitting surface layer is very limited, the emission wavelength can be set to low fiber attenuation and dispersion. This method ensures that a low thermal resistance is maintained and a high radiance therefore a high radiant power is coupled to the fiber.

The low dislocation density less than cm -2 of the n-doped substrate wafer minimizes losses due to non-radiative recombination at crystal defects. The p-region acts as a diffusion mask and the p-n junction is graded to reduce lattice strains and additional losses due to, for example, crystal defects. To improve the radiation output of the diode, the final p-surface is coated with an anti-reflecting layer of silicon nitrate. The emission wavelength of this particular type of diode would normally be about nm, the spectral bandwidth approximately 40nm, and the radiant power would be roughly proportional to the forward current typically t2W at mA.

The average lifetime i. Doping the p-region to a higher level, although facilitating large modulation bandwidths, would reduce the amount of radiant power available. To ensure that the A1GaAs layers have an extremely small transmitting area, the epitaxial layers point downwards. This is sometimes known as the 'upside-down configuration'. The diode is mounted on a thin gold heatsink and soldered to a silicon chip containing a conductor track and insulation layer. The high aluminium content of the two AIGaAs layers ensures that a potential barrier is formed to reduce the a m o u n t of electrons and holes that would otherwise be capable of entering the active region.

Q u a n t u m efficiency depends not only on the crystalline quality but also, to a significant degree, on the doping and composition of the active layer. Such a diode would exhibit a rise time of approximately 15 ns. By using A1GaAs layers the possible reduction of q u a n t u m efficiency due to non-radiative recombination is minimized.