As fiber optic transmission technology matures, its development interestingly tracks that of the millimeter and microwave technology of 30 to 40 years ago. Since return loss and multipath interference were found to be of concern in those systems, it is not surprising to learn that reflections present in fiber optic spans can cause similar problems in the single-mode fiber optic transmission systems of today. Analogous to the return loss issue, a single reflection point can reflect light back into the transmitter and disrupt the laser source's spectrum. Multiple reflection points meanwhile, can cause delayed signals that interfere with the transmitted signal, much like the interference caused by multiple signal paths. This interference appears as noise at the receiver, and we will refer to it as multiple-reflection noise (MRN). Several systems have been investigated for sensitivities to reflection induced noise. The results of these system measurements are summarized and the various system dependencies are categorized. By placing reflectance requirements on the individual components placed in the fiber optic transmission span and requiring system performance to have a tolerance to specified reflectance levels, the effects of reflection noise on fiber optic transmission systems can be minimized.

In the past optical time domain reflectometers have been used on an as needed basis for fault location. High dollar capital investments have been sitting on the shelf waiting to be pulled in the case of an emergency. On the other hand, down time of a trunk system directly affects revenues of the operating company. Today OTDR's can be put to work in automated remote site fiber monitoring systems to significantly reduce the down time. This paper describes several approaches to automated monitoring on active and dark fibers. Most other conventional automated monitoring systems are sensitive to a c4ange in the environment (pressure, isolation, humidity, etc.).¹

An effective measurement method for investigating the transmission characteristics of an optical fiber segment using automated backscattering techniques is described in this paper. Three different Optical Time Domain Reflectometers (OTDRs) are employed in the manual and automated modes of operation to characterize several single mode fibers. Length dependent optical loss measurements are made on several precision-wound fiber optic bobbins (containing many layers) by setting the distance markers at the start and end locations which bound the desired fiber segment. Discrepancies are observed in the data acquired using the manual vs the automated procedure. Apparent round-off errors and inaccurate measurements are observed when operating the OTDR in the automated mode due to the elimination of the operator's observations and adjustments. Algorithms are written to automatically frame the fiber segment of interest by monitoring the actual OTDR marker settings and adjusting the marker setting request(s) when the fiber segment boundaries are violated by the OTDR. This novel framing technique ensures consistent and precise measurements. The technique can provide assistance in using OTDRs in the automated mode of operation.

This paper describes an active characterization technique that generates backscatter signatures to measure the performance of optical time-domain reflectometers (OTDRs). These signatures can be used to test an OTDR's loss accuracy, dynamic range, spatial resolution, loss resolution, and receiver recovery time.

The inadequacy of the stuck-at fault model has been well aired and documented.1 ,2 All studies agree that this model does not reflect the physical failures of real devices,3 principally because such failures do not exhibit a 1:1 mapping onto the logic domain.2 ,4 Circuit layouts which are based on stick diagrams do however reflect the physical domain in sufficient detail (see Fig. 1) to enable both structural defects, together with shorts and opens in metallic and non-metallic domains, to be detected and located. The author has proposed the adoption of a novel method which processes information obtained from a scanning laser beam reflected from a surface profile. Scanning may be of a raster nature over the surface, or follow a suitable path search along metal lines. The latter search type has been simulated in PROLOG. Such a topological approach to the testing problem offers a test structure for exploitation using a laser beam probe technique. In this paper the theory of reflectivity is described as it relates to the test method, and the results presented are based upon reflectance measurements obtained by raster scanning.

Optical fibers developed for sensor applications often exhibit high optical losses. While these losses may be entirely acceptable for the design application, efforts to improve fiber performance still require loss quantification and the identification of optical loss mechanisms. Existing fiber loss characterization techniques (e.g., optical time domain reflectometry) are generally only appropriate for low loss (communication grade) fibers and may give little or no information on loss mechanisms. At the Pacific Northwest Laboratory, techniques have been developed for the characterization of high-loss (tens of dB per meter) optical fibers which allow discrimination between scattering and absorptive losses. Two techniques are presented. In the "differential scattering" method, differential fiber scattering losses are acquired over the length of the fiber to obtain a scattering loss coefficient. This information combined with a conventional total fiber loss measurement (e.g., using the "cut-back" method) allows inference of the absorptive part of the fiber loss. In the second method, "scanning aperture" characteriza-tion, the fiber is scanned by a moving aperture to yield curves of differential and integral scatter intensity versus length. This curve not only provides corroboration of the previously acquired differential scattering data, but also points out high-loss regions in the optical fiber. Both methods will be fully described. Experimental data on representative fibers will be presented.

Making fiber loss measurements is one of the prime functions of an Optical Time Domain Reflectometer (OTDR). But the literature on the subject lacks the characterization of the signal processing algorithms used for this purpose. In this paper, we establish a theoretical upper bound on the measurement accuracy for a given noise level. We also analyze and characterize how sampling rate, number of samples used for calculation, and fiber loss affect accuracy. This gives a yardst-ick to measure the. performance of an algorithm or an instrument. If the performance is far from the theoretical limits then one can expect to improve it by alternative means; otherwise it is as good as can be achieved. We use three fiber loss measurement techniques to gather results on measurement accuracy: (i) line-fit to log data , (ii) autoregressive modeling, and (iii) polynomial-fit to unlogged data. We then compare the measured accuracy of the three methods with the theoretical bounds for different noise levels. This clearly shows the circumstances under which their accuracy falls short of the theoretically predicted value. In those situations an improvement is possible with some other method. Finally, we illustrate how alternative techniques can not improve accuracy when it is already as good as the theoretical limit.

Presently, a standard field fiber optic attenuation test, such as FOTP-171, requires one precision reference cable for each fiber size (50 p, 62.5μ, 85μ or 100μ) and for each connector type to be tested. Depending on the test method used, a launch fiber of each size may also be needed. We propose a test method which uses one universal reference cable to replace the 4 reference (and possibly 4 launch) cables presently required. In addition to the obvious cost and logistics saving, test results show that the consistency and repeatability of loss measurement are improved. Results correlating the proposed new method to standard test methods (substitution loss and insertion) are presented.

Acknowledgement: The assistance and support of the MICOM Army Missile Command is gratefully appreciated. An analytical model of the arc fusion splicing process for high silica fiber optic filaments is being developed. This model facilitates systematic investigations on the effects of splicing control variables on the optical and physical splice properties. The analytical model is briefly reviewed along with experimental results used to validate the model. Modifications made to improve the model's fidelity are discussed along with their impact. Results of parametric studies focusing on thermal properties of the model are given along with analysis of the effects of residual and imposed stresses. Plans for continued model improvements and further analysis are given.

There are many techniques currently used for measuring the fundamental Mode Field Distribution (MFD) of a single-mode optical waveguide. There are also many definitions of field width which can be used for characterising the measured MFD. These definitions are key parameters for describing the propagation, and other properties of any optical waveguide. A theoretical study of the evaluation of the various field width definitions from measurement data (containing noise) is described. From this study, the relationship between both accuracy and repeatability of evaluation of the definitions of Mode Field Radius (MFR) and the measurement Signal to Noise Ratio (SNR) for the various measurement techniques has been found. Tables and graphs of data are presented which indicate the necessary measurement SNR for a required accuracy of evaluation of a desired definition of MFR.

Optical fiber technology is finding wide ranging applications in many areas including communications and sensors. In applications, such as "Fiber to the Home" (FTTH), large quantities of fiber optic components will be used and subjected to environments not usually encountered in "Central Office" applications. In Figure 1 a typical "POTS/Video" subscriber loop topology is shown. Telephone service (POTS) will operate in a bidirectional mode over a single fiber at one wavelength and video will be broadcast over a second wavelength to the subscriber. From the figure it can be seen that the link for a single subscriber requires two 3 dB couplers at wavelength 1 and two Wavelength division multiplexers operating at wavelengths 1 and 2. If interactive video, such as Video-Phone, is required an additional two 3 dB couplers, operating at wavelength 2, will be needed.

Long wavelength semiconductor lasers have been extensively developed for use in long distance telecom-munications networks [1]. For applications which do not require propagation over long distances, short wavelength GaAs based lasers remain a viable option for the optical source. Using this material system, quantum well lasers have been developed which show improved performance in threshold current, linewidth, and power over bulk 3-D lasers. In this paper, we describe a measurement system designed to examine spectral dynamics in semiconductor lasers and present representative results from GaAs single quantum well lasers.

Streak cameras offer advantages over conventional methods for making picosecond measurements free from distortion. The operation and calibration of electro-optic streak tubes is reviewed. The use of high speed streak cameras in the study of laser diodes, pulse compression in fibers, electro-optic switching and optical fiber characterization is covered.

Two techniques have been used for determining the recombination lifetime in semiconductor lasers: A capacitance bridge at 10 MHz, or a network analyzer in the range from 45 to 245 MHz. The latter, although more tedious than the former, has the advantage of allowing detection and inclusion of parasitic contributions. The AlGaAs ridge graded-index separate confinement heterostructure lasers studied here are well described by a simple equivalent circuit consisting of an ideal laser and a series resistance. Below threshold, the inverse lifetime squared exhibits a linear dependence on the forward current, as expected from simple theory. The measurements yield non-radiative lifetimes of 2.5 ... 5 ns. The bimolecular recombination constant B = 0.24 ... 0.56 x 10^-10 cm3/s is smaller than the bulk values; possible reasons are discussed.

This paper describes some measurements of self-pulsation frequency and visibility, carried out on commercially available 0.811m self-pulsating lasers. For a given laser, operating CW, the self-pulsation frequency is shown to depend principally on the optical power and to be relatively independent of laser temperature and life; the coherence length envelope appears to vary approximately inversely with the optical power.

A high-power, short wavelength, computer-controlled laser diode characterization tester was developed to measure the specialized electro-optical characteristics required for optical data storage applications. The tester design goals were achievements of high accuracy, repeatability, stability, and throughput of all the measured characteristics. To accomplish these goals, optimization of thermal, mechanical, and electro-optical factors of the tester were required. The laser diode used for this work was packaged with a back-facet photodiode in a TO-5 can, and was mounted on a specialized electro-satic discharge (ESD) protected, grounded heatsink, temperature-controlled, brass block mount. The laser diode fixture was mounted on computer-controlled micropositioning stages for fast and repeatable measurements. The data was acquired by a digitizing oscilloscope and sent to an IBM-PC AT on an IEEE-488 bus. The tester capability allows the following measurements: First, the continuous wave (CW) optical power, forward voltage, and back-facet diode sensitivity are measured. An optical power limiter and a maximum current setting are used as a precaution. Second, the peak lasing wave-length in the optical spectrum are measured at a given low and high power. Third, the parallel and perpendicular profiles of the far-field pattern are also measured at a given low and high power. For each profile, the full-width at half-maximum (FWHM), pointing direction, and relative ripple magnitude are calculated using a Gaussain least-square fit to the data. Fourth, the astigmatic length and its shift from low to high power, and the near-field profiles are measured. Fifth, the polarization ratios are measured for a given numerical aperture objective. In these measurements, the collimating and focusing optics are corrected for the operating wavelength of the laser diode. Sixth, the temperature dependence of all the above characteristics is measured, and different important parameters such as the characteristic temperature and wavelength temperature sensitivity are calculated. The repeatability and accuracy of each of the measurements are statistically analyzed using control laser diodes.

The increasing interest in fiber optic communication and sensor systems utilizing interferometric detection of optical fields has made accurate determination of the light field's coherence parameters more important. We have developed a computer based measurement system which allows for video frame grabbing of Michelson interferometer output beams. Key attributes of this system are that it uses a standard non-dedicated microcomputer (Macintosh II) and relatively inexpensive commercially available software and frame acquisition hardware. Video frames are acquired and simply processed allowing fringe contrast to be easily determined, from which coherence length may be found. While the system described here is capable of performing optical contrast measurements of any field illuminating the video sensor array, its primary use in our facility has been to examine coherence of laser diodes that have been coupled into various types of single and multimode optical fiber. In addition, the system has been used to determine laser diode coherence length as the diode was subjected to various biasing conditions. Such measurements will be presented as will a complete description of the entire system.

The requirements for laser diode characterization are continually increasing in sophistication and accuracy. Simple optical power versus drive current curves are not adequate to characterize laser diodes for the variety of present and future applications from telecommunications to data gathering and optical processing. Presently available integrated test sets provide opto-electronic, spectral, far-field, and near-field characterization of laser diodes in the visible and near-infrared regions. Measurements on connectorized and optical fiber pigtailed devices are readily performed. Equipment for evaluation of astigmatism, coherence length, optical noise intensity and polarization ratios will be available in the near future. Important goals in the development of all laser diode characterization equipment are precision, reliability, traceability to NIST (NBS), versatility for graceful upgrading to encompass emerging device requirements and fully-integrated, user friendly software to assure high throughput from a smooth gathering and flow of data from logically sequenced tests.

This paper presents results of environmental tests performed on high frequency packaged, polarisation maintaining fibre-pigtailed Ti:LiNb03 phase and Mach Zehnder intensity modulators. Most of the environmental tests carried out conform to the UK Defence Standard 07-55 and the test requirements specified by the European Space Agency (ESA) for components and equipment employed in space applications [1]. The modulators were subjected to vibration, shock and temperature cycling, to assess ruggedness of package and quality of fibre pigtailing technique. Stability of Ti:LiNbO3 devices in relation to pyroelectric and buffer layer drift is presented. Results of some preliminary work on the effect of moisture and outgassing of the epoxy used for fibre pigtailing are presented.