Matt Young

Optical fiber index profiles by the refracted-ray method (refracted near-field scanning)

This paper has two primary purposes. First, it provides an elementary description and tutorial overview of the refracted-ray method of measuring fiber index profiles. Second, it presents new results concerning the theoretical foundation, the linearity and precision, and other aspects of the method. In particular, we find that index differences may be measured to 5% or better and conclude by showing ~3% agreement with another laboratory and good agreement with numerical aperture measurements performed by participants in an interlaboratory comparison.

Mode-field diameter of single-mode optical fiber by far-field scanning

I use the direct far-field method to measure the mode-field diameter of a single-mode fiber with an expanded uncertainty of 30 nm, with a coverage factor of 2. For a step-index fiber with a mode-field diameter of approximately 9 mum, the major sources of uncertainty are nonlinearity in the electronics, angular errors and scattered light in the apparatus, and the polarization and noncircularity of the mode of the fiber. The paper concludes by showing an inconsistency in the derivation of the far-field expression for mode-field diameter.

Two-dimensional index profiling of fibers and waveguides

We have constructed a two-dimensional refracted-ray scanner that can resolve index-of-refraction increments of approximately 4 x 10(-5). This resolution is an order of magnitude finer than the uncertainty of the measurement. The scanner can be adapted to evaluate either fibers or planar waveguides. The two-dimensional scan and the high precision allow visualization of features, such as deposition layers, that are difficult if not impossible to see in conventional one-dimensional scans.

Optical Fiber Geometry: Accurate Measurement of Cladding Diameter

We have developed three instruments for accurate measurement of optieal fiber cladding diameter: a contact micrometer, a scanning confocal microscope, and a white-light interference microscope. Each instrument has an estimated uncertainty (3 standard deviations) of 50 nm or less, but the confocal microscope may display a 20 nm systematic error as well. The micrometer is used to generate Standard Reference Materials that are commercially available.

Spatial light modulator for texture classification

This paper describes a hybrid computer-optical processor devoted to the analysis of texture. Textures are displayed on a spatial light modulator and their power spectra are calculated optically by a Fourier optical technique. A video camera and a computer with a frame digitizer process the power spectra. We define a multidimensional feature space and associate each texture with a point in this feature space. After a training set, the system can distinguish several textures. This hybrid computer is a step toward real time texture classification because of the nearly instantaneous optical Fourier transformation.

Scratch-and-dig standard revisited

The scratch standard (MIL-O-13830A) is a cosmetic standard effected by a visual comparison with a set of secondary standards that are in turn evaluated by comparison with a set of master standards. Both manufacture and certification of the secondary standards are somewhat unreliable. This paper shows that they can be classified according to the relative power scattered at a relatively small angle and describes experiments with etched gratings that have the appearance of scratches but diffract light into a broad peak between 5 and 10° off the axis of the incident beam. Some prototypes have been classified both by comparison to the master standards and by a photoelectric measurement; agreement between the two methods is good. Such gratings, used as the secondary standards, should display less intersample variation than scribed or other artifacts. The paper concludes by presenting evidence that the original primary standards have been stable over a long time.

Profile inhomogeneity in multimode graded-index fibers

Profile parameters (g) of several multimode graded-index fibers have been measured. It was found that g may vary azimuthally by +or-0.15 or more in fibers for which the average value is between 1.8 and 2.2. >

Linewidth measurement by high-pass filtering: a new look

Earlier workers have noticed that high-pass filtering produces a sharp dark line in precisely the location of the geometrical image of an edge. They proposed using this fact as an aid in measuring linewidth in microscopy but found that the other edge of the line caused significant error. In this paper, I examine that error as a function of normalized linewidth and normalized spatial-filter width and find that it may be limited to ±5% or so, provided that the spatial filter subtends between 0.25 and 0.3× the numerical aperture of the objective and that the linewidth exceeds about twice the resolution limit.

Robust regression applied to optical-fiber dimensional quality control

To minimize coupling losses when low-cost connectors are used to mate optical fibers, manufacturers need to maintain tight control of fiber dimensions. The video-microscope, or gray-scale, method is the most frequently used technique on the manufacturing floor for measuring the geometric parameters of the cleaved end of a telecommunications fiber. We present a method for performing optical-fiber dimensional quality control that allows for end face damage and accounts for the special structure of measurement errors in fiber edge points calculated from gray-scale images. The new approach adheres to the industrial standard test procedure by fitting an ellipse to the edge table to obtain geometric measurements. But, to create high-breakdown resistance to outliers, a data filter based on the least median of squares criterion is used.

Video microscope with submicrometer resolution

We have constructed and evaluated a video microscope with a 150- x 150-microm field of view for performing measurements of optical fiber geometry. The microscope consists of a frame transfer video camera, condensing and filtering optics, a 40x, 0.65 N.A. microscope objective, and frame digitizing electronics. Using simple digital algorithms, we measure distance with a random uncertainty of approximately 40 nm across the full field of view, but width measurements suffer from a systematic error between 0.1 and 0.2 microm.