ABSTRACT
A technique is presented that uses total-loss data from a conventional monochromator-based spectral-loss measurement set to obtain not only the Rayleigh and wavelength-independent scattering of a fiber but also the UV-band-edge absorption. That absorption causes the total-loss data to deviate significantly from a straight line when the data are plotted vs lambda on a 1/lambda(4) scale. Including the absorption term accounts for the deviation from a straight line over an extended wavelength region (0.66-1.16 microm) and materially affects the A and B coefficients obtained. For one fiber examined in detail A went from 1.30 to 1.13 microm(4) dB/km and B from 0.22 to 0.32 dB/km when the UV absorption was included. This technique also eliminates the problem of negative B coefficients and allows processing variables that affect absorption to be separated from those that affect scattering.
ABSTRACT
In this paper we describe the effect of different launching conditions on the loss measured with a calorimeter in short (~1-m) lengths of multimode optical fiber. A qualitative description of the relation between the energy distribution in the fiber and various excitation variables is given, and many examples are presented that illustrate the effect that launch conditions have on the fiber loss. The use of launch-dependent loss phenomena to diagnose loss mechanisms in optical fibers is discussed. In particular, loss plots are shown for two germania-doped silica-core fibers that indicate a strong absorption loss mechanism located deep within the fiber core.
ABSTRACT
A technique is presented for rapidly measuring Delta of optical fibers with profiles closely approximating n (r) = n(o) [1 - Delta(r/a)(alpha)], the Delta << 1 limit of the two-parameter Gloge-Marcatili profile. Error analysis indicates an accuracy of about +/-9%. The method is as follows: Using a Michelson interferometer together with a microscope, the transverse optical path through the center of a fiber immersed in index matching fluid is set equal to the optical path through the fluid. This is accomplished by varying the temperature of the fluid. The value of Delta can then be determined from a knowledge of the refractive index of the fluid inferred from temperature measurements. Good agreement is shown between Delta values measured by this technique and those measured by-two other interferometric techniques. The possible use of this method to determine alpha, or to determine the actual profile if it does not follow the assumed shape, is discussed.
ABSTRACT
We present a technique for designing holo-graphically fabricated structures in photoresist. A method is given for obtaining a suitable initial resist thickness. The etch depth (Deltat) vs exposure (E) characteristic is determined experimentally for a fixed development time (T). The characteristics for other values of T are found using the linearity of Deltat with T. Whole families of grating profiles can then be generated by using a polynomial fit to the exposure characteristics. For example, using a single holographic exposure plus an optional uniform preexposure, a family of grating profiles is obtained by varying the parameter T. A desired profile can thus be realized by a suitable combination of initial resist thickness, exposure, and development time. Residual resist layers may be eliminated by this method thereby minimizing scattering and other losses. Experimental verification of the various features of the model is given.
ABSTRACT
In this paper we discuss the experimental verification of an empirical photoresist model presented previously. Photoresist grating depths predicted by the model are found to be in good agreement with those obtained from diffraction efficiency measurements. Total rates of field decay alpha(T) obtained from a theoretical analysis of the photoresist grating coupler predicted by the model are also in good agreement with values measured by two independent experimental techniques. These results indicate the usefulness of the model in fabricating periodic devices in photoresist.
