Optical Fiber Testing - Report Example

Polarized mode dispersion test PMD400 is a software-controlled instrument used in measurement of polarized mode dispersion. The instrument uses a desktop computer and is controlled using SOFTKEYS and a MENU through either a mouse or a keyboard (Figure 6). Standard telecommunication uses a single mode fibre, but there are as a matter of fact, two orthogonally polarised modes which are supported using waveguide structure. The dual polarized models travel with varied phase velocities and group velocities. This gives rise to PMD and the basis of operation of PMD400. The machine uses interferometric technique; a method that uses a polarised source of light and an interferometer (receiver). The test fibre is excited using a linearly polarised source of light that conventionally has a 1310nm or 1550nm LED as well as a polariser effective over the region of operating wavelength. The typical extension ratio of the car is above 30dB and meets the recommended >20dB. Additionally, spectral shapes’ sources are approximately Gaussian to make certain there is a smooth autocorrelation function which can be deducted from measured fibre result. The sources’ spectral width directly affects the minimum measurable PMD. Conventionally, 0.065ps. values can be achieved at 1550nm and is slightly lower at 1310nm. This produces a minimal accurate PMD measurement in the range of 0.1 to 0.15ps. PMD 400’s receiver has a Michelson interferometer that is implemented primarily in fibre although the variable optical path is attained in air through movement of a mirror over a range estimated as approximately 55mm. As mirror is set in motion, varying amplitude fringes are viewed by detector diode. The detected fringes’ enveloped is used in provision of PMD information. When the interferometer’s two arms are of same length, interference fringes’ amplitude attain optimal values. The interferometer’s zero time delay position is determined using the central autocorrelation peak. Where the fibre linking the source of light to interferometer has polarization mode dispersion, the additional delay will make the interferogram to broaden in proportion. Pulse broadening can be normalized for fibre length manually entered at the beginning of a user’s measurement. In fiber optic communications, the acceptable limit for instantaneous digital group delay is defined as a tenth of the bit length, or 10 ps.. for 10 Gbps.. SONET per SDH. To convert this figure to a PMD limit requires knowledge of the system tolerance for crossing DGD limit, or outage probability. The DGD provides an estimated limit based on the worst-case probability. In short, it determines the acceptable DGD (instantaneous) and acceptable bit error rate for a system, before an outage. For 10 GigE, in 99.999 percent of the cases, the system will not accept a DGD above the limit, which leads to outage probability, or a 3.7 DGD per PMD ratio. Therefore, it is shown that with a DGD maximum of 19 ps.. for 10 GigE the PMD limit is 19 per 3.73 = 5.094 ps.., or approximately 5 ps.. Comparing 10 Gbps.. SONET per SDH transmission limits exhibits a tighter dispersion tolerance for 10 GigE., as the IEEE standard defines. Two primary reasons are responsible for this also apply for chromatic dispersion: Forward Error Correction (FEC) is less robust compared to the one applied to 10 Gbps. SONET per SDH Acceptable outage probability is low The probability that a DGD value will exceed a given value is directly related to the PMD ratio, often called the Safety (S) ratio. It is illustrated that the limit for PMD is much tighter for 10 GigE than for 10 Gbps. SONET per SDH Cut-off frequency test Photon kinetics PK-2210 is a high performance Optical Fiber Analysis System with measurement capability for optical fiber and cable. The system offers high-speed characterization of spectral loss in both single-mode and multi-mode fibers. The systems employs unique fiber preparation as well as signal processing techniques, both of which, deliver measurement performance and testing throughput needed by high volume fiber, cable as well as component manufacturers. The system has an option for measurement of mode field diameter as well as multi-mode numerical aperture measurements. The wavelength range for the system is 1000 – 1600 nm1 and 800 – 1600 nm2 for single-mode and multi-mode respectively. Its measurement range3 and measurement range1 are >58 dB and >44 dB respectively while the repeatability fiber is < 0.005 dB RMS. The device selectively transmits lights of varying wavelengths while blocking others. Passage is either for purely long wavelengths (long-pass), or purely short wavelengths (short-pass), or a band of wavelengths, whereby both longer and shorter wavelengths (band-pass) are blocked. Pass-bands can be narrower or wider depending on the device. It is important to note that the cut-off wavelength single-mode fiber wavelength and is the shortest wavelength for which, solely, the primary mode propagates across the fiber. Its measurement is a necessity to steer clear off degradation of modal noise of optical signals a single-mode fiber carries. The measurement system applies an internationally recognized transmitted test method. It involves launching of spectrally selected radiation from a grating monochromator to a few meters of the fiber being tested. Cut-off wavelength A single 60mm diameter fiber is used to suppress higher modes and measurement of transmitted power is redone (Figure 8). The measurements ratio the two as a wavelength function is used in obtaining of the fiber’s cut-off wavelength. In essence, determination of cut-off wavelength of a single-mode fiber involves establishing wavelength above which the power transmitted via abruptly drops. The power length and bending effects differ on varying fibers based on whether they are either, matched-clad or depressed-clad. Either un-cabled or cabled single mode fibers can be used in measurement. The test method used in this case describes the test equipment input optics, mode filters, as well as cladding-mode strippers’ required for its completion. Additionally, the test fiber is loosely supported in single-turn with constant radius of 140 mm. The launch as well as detection conditions are unaltered while scanning over a range of wavelengths. The wavelength range scanned in the test covers the expected cut-off wavelength. Chromatic dispersion test EG&G Fiber Optics chromatic dispersion300 uses a calibration method resulting into in a 0.2-nm accuracy in zero-dispersion wavelength measurement (Figure 3). The single-mode optical fiber allows light to travel with a propagation group delay proportional to the length of the fiber length and is a wavelength function. The wavelength variations imply that the sources of light that are not perfectly monochromatic are time-dispersed. This is known as chromatic dispersion. The system has finite spectral line width, and hence dispersion imparts pulse spreading as a result, limiting maximum data rate which the system can support (Figure 4). The previous studies of optical fibers chromatic dispersion temperature dependence have been focused on the variations of chromatic dispersion slope at zero dispersion wavelengths. The study has shown that the dispersion variation with temperature is related with the zero dispersion wavelengths and with the dispersion slope at the zero dispersion wavelength, and that this second term cannot be ignored in general. For the studied fibers, it is shown that this second term contribute with around 2.5% - 16.7% to the variations of the dispersion with temperature. Similar reasons for a tighter dispersion tolerance with 10 GigE also apply to chromatic dispersion tolerances with a limit of 1550 nm of 738 ps. per nm compared to 1176 ps (Figure 7). per nm for OC-192 per STM-64 10 Gbps. for a non-return- to-zero (NRZ) coding format. Obtaining a shorter reach for 10 GigE can be achieved if transmission using a G.652 fiber with a typical chromatic dispersion of 17 ps. per nm per km is considered. In the SONET per SDH case, the maximum distance before regeneration or compensation is about 70 km, where it goes down to 40 km for 10 GigE transmissions. Dispersion rises both with length of fiber used and spectral width of optical source used. It is measured in ‘picoseconds per nanometer kilometer’ or (ps. per nm.km).  That is, where a communication link has a dispersion of 2ps per nm.km, and is 2kms long, and used alongside a source with a line-width of 5nm, duration of pulse increases by 20 picoseconds by the time it gets to the detector. Dispersion further increases along wavelength and in many fibers, changes from -ve to +ve between 1300nm and 1550nm transmission windows.  The wavelength at which value of dispersion passes via zero is referred to as ‘lambda zero’ (λ0) (Figure 6).  The dispersion shifted fibers are specially designed such that their lambda zero takes place within 1550nm transmission window. Once an operator has tested the fiber span for chromatic dispersion, the overall ps. per nm may not be needed for the coherent signal. Nonetheless, the overall ps. per nm may be used with the fiber length found by single-ended dispersion analyzer to calculate the ps. per nm per km at 1550 nm. It can also be used with the dispersion slope and the lambda zero to determine fiber type. For instance, a large service provider deploying 100GigE coherent detection signals (Figure 5). Blocking probability is defined as resource blocking probability. If nodes are equipped with wavelength conversion then the light-paths can be assigned on different wavelengths on the links of the route. This lowers blocking probability and optical network behaves as conventional circuit switched network. Nonetheless, the converters increase the cost of the network if implemented. Further, if nodes are unable have wavelength conversion, then the network is called all-optical network, and the network suffers from the wavelength continuity constraint (Figure 9). In the high speed networks where the data rates are of the order of Gb/s, there can be many impairments (including the non-linearity) in the physical fiber link which affect the quality of the signal. Hence as the signal goes from the source to the destination signal will degrade drastically and might become unacceptable at the destination end. In this case bit-error rate (BER) is the chosen metric. This results in another type of blocking called physical layer blocking. Results Figure 5 shows a typical time-delay and dispersion curves. Single-mode dispersion data is obtained from measurements of the relative time-of-flight of light signals at various wavelengths. Light from an amplitude-modulated light-emitting diode (LED) is selected by a monochromator and launched into the fiber under test. The relative time of flight is calculated from the phase delay of the detected signal after passing through the fiber under test, referenced to the system phase response. Dispersion is measured on a full length of fiber. On the other hand, figure 7 employing various signal processing techniques, in order to generate a summary of the achievable ranges of performance specifications for these devices as interferometric sensing units. These interrogators sample optical intensity in the wavelength domain, and as a result they lend themselves naturally to an interferometric approach. Figure 9 shows a range of wavelengths falling within the range of measurement equipment’s. The cut off frequency as indicated is 1.204. This is the wavelength at which the mode ceased to propagate or at least propagated at a fundamental mode. At this wavelength, the power transmitted through the fiber abruptly ceased. Figure 17 not only presents plain results of using the first and second methods in fiber testing but also presents useful trends to take into consideration when selecting an appropriate method between the two. Initially, the ratio of rejection is higher for the first method as compared to the second method. This however gradually changes as the number of fibers increase. Most evidently, the rate of rejection increases for both as the number of fibers increase. However, the increment rate is higher in the second method as compared to the first and ultimately with high number of fibers, the ratio of rejection is higher in the second method compared to the first method. In essence, the more the number of fibers, the higher the rate of rejection is for both the first and the second method. When CD and PMD are used one after the other, the results used, although showing a trend similar to the case in Fig. 17, marginally differs in terms of the rate of rejection. The results shown in Fig. 18 and Fig. 19 shows a trend where as the number of fibers increase, the ratio of rejection also increases. The increase gets even more pronounced as the number of figures becomes larger. Conclusion Then findings raise and respond to an important question: “What is the most accurate and capable measurement technique for links?” The performance and reliability of the cabling infrastructure within the data center and in premises applications are of paramount importance. For optical networks, it is critical for all network stakeholders to have accurate knowledge of performance of permanent links deployed in a network. It is further important to assure that links deployed present a warrantable solution when measured against standards. Another issue responded to is the lots of time and cost that can be saved if contractors are aware of the proper measurements that need to be made, understands how to make the measurements appropriately, has appropriate tools, keeps the tools in good condition, calibrates them regularly and knows how to efficiently use them.  OTDRs always require a launch cable for instruments to settle down after reflections from the high-powered test pulse overloads to the instrument (Figure 1). However, OTDR method does not measure loss of connector on the far end. Adding a cable at the furthest end permits measurement of the loss of entire cable, although it negates the big advantage of OTDR, that it gets measurements from only a single end of the cable. The accuracy of testing long single-mode fibers with several splices depends on lots of factors, evidently the number of fibers being one of them as the results show. Additionally, since fibers are long, fiber attenuation is an important part of measured loss. Additionally, OTDRs relies on fiber backscatter for creating measurements, so any variations in fiber backscatter at a splice leads to higher loss in a single direction and lower loss in another direction. Conclusion Amongst the many advantages of fiber optics is its ability to allow long distance high-speed communications. Long wavelengths attenuation is extremely low. While fibers can be spiced through fusion with literally no loss and high-powered lasers as well as fiber amplifier regenerators mean long distances can be easily attained, over extremely long distances, new aspects in fiber performance are of extreme importance. Chromatic dispersion caused by light of varying wavelengths, and polarization mode dispersion, resulting from fiber polarization are important and limit fiber links. As evidenced, chromatic dispersion is a product of single-mode glass fibers transmission of light of varying wavelengths at varying speeds.  Notably, as pulse moves down the fiber, light of longer wavelength moves substantially faster and the pulse spreads. Evidently, Polarization mode Dispersion and Chromatic Dispersion influences transmission quality of high speed fiber optic networks and as a result is important in ensuring proper network performance through characterization of the two parameters in deployment or upgrade of high speed transmission. The testing approach, just like other testing menthods, involved testing as various wavelengths using various discrete sources of varying wavelengths. The method uses phase delay and requires access to both ends of the fiber and a second fiber for synchronization of test instruments at each end. Polarization mode dispersion (PMD) is somehow sophisticated and is a phenomenon where light travels in a medium in form of a wave with elements at right angles. PMD involves each component of polarized light travelling at varying speeds resulting into dispersion. As mentioned earlier, PMD magnitude in a fiber is expressed as the difference known as differential group delay (DGD) (Δτ (“delta Tau”)). It is evident that PMD is as a result of birefringence of fiber and is influenced by material birefringence and waveguide birefringence, same as chromatic dispersion although it is more sophisticated. PMD causes pulse broadening and/or jitters in received electrical signal and causes errors in signal reception. Nonetheless, PMD testing is not easy, reproducible, or accurate testing method. Uncertainty in measurement can be as high as 20%. All the uncertainty of PMD measurements has effect of making comparisons between tests. Variations are especially high on tests of older fiber-links. Chromatic Dispersion (CD) that causes pulse broadening depending on wavelength, and to Polarization Mode Dispersion (PMD) that causes pulse broadening depending on polarization. Excessive spreading will cause bits to “overflow” their intended time slots and overlap adjacent bits. The receiver may then have difficulty discerning and properly interpreting adjacent bits, increasing the Bit Error Rate. To preserve the transmission quality, the maximum amount of time dispersion must be limited to a small proportion of the signal bit rate, typically 10% of the bit time. Read More