Gyrotropic thin-film waveguides with isotropic top layers are discussed using hybrid modes. The fundamental propagation properties of optical waves and TE-TM mode conversion efficiencies are estimated. When refractive indices of isotropic top layers are equal to those of substrates, 50% TE-TM mode conversion efficiency required for non-reciprocal circuits of optical isolators can be realized with gyrotropic films. From a fabricational point of view, these structures are very useful for constructing thin-film optical isolators because no severe control on film thickness is necessary and film growth of amorphous materials on YIG crystal films is possible for top layers. The design data for non-reciprocal circuits consisting of ZnO/YIG/GGG layered structure are presented from the hybrid mode analysis. The thickness of YIG and ZnO are 3.43 µm and 1.03 µm respectively and the length of the device is 2.28 mm. The non-reciprocal circuits of this structure can be experimentally fabricated by RF sputtering method.

In order to find the optimum index profile of the single-mode optical fiber with respect to the cabling loss due to the random bends with the Gaussian correlation function, the relation between the random-bend loss and the index profile is investigated. The formula of the random-bend loss is derived for arbitrary index profile. In the derivation, the loss is calculated as the power of dipole radiation due to the equivalent refractive index perturbation in contrast with the conventional mode coupling analysis. Using this formula, random-bend losses are calculated for four types of index profiles; the convex, parabolic, step, and concave type of index profiles, under the condition that connection losses are same for all types of profiles. As results of the comparison, the random-bend loss is minimum in the case of the convex type of profile and it is decreased down to 40% of that in the case of the step profile.

Simple design curves are presented for linear phase, low-pass impedance matching networks. A parallel RC and an RLC load are considered. Using the design curves, we can easily determine the maximum DC gain K for a given load.

Partial response is applied to an optical fiber transmission system with low fiber loss for the purpose of expanding repeater spacing beyond the range imposed by fiber bandwidth restriction. Detailed theoretical and experimental results on the relation between receiver power penalty and data rate to fiber bandwidth ratio are presented. Duobinary codes are preferable to non-return to zero full response codes, provided that the 3 dB fiber bandwidth is less than 0.6 times the data rate.

A synthesis of high frequency active RC filters using an operational amplifier with single pole characteristics is presented. As a measure of the quality of filters, the pole sensitivity with respect to the gain bandwidth product GB of the amplifier is used and it is shown that the absolute value of the pole sensitivity has a minimum limit. The synthesis is based on the predistortions of Q and center frequency errors caused by the finite GB of the amplifier. The proposed circuit has quite stable temperature characteristics and examples show illustrative differences of the temperature responses between the proposed circuit and conventional ones.

A frequency modulator basically consists of an amplifier and a resonator employing a varactor diode, and the linearity of frequency modulators has been analyzed and discussed looking at the resonance frequency calculated from the capacitance-voltage characteristics of a varactor diode. However, as the oscillation frequency of the frequency modulator increases, it becomes essential for designer to evaluate the influences on the modulation linearity by the phase characteristics of the amplifier. This paper discusses the influences on the modulation linearity by the phase characteristics (the phase slope (or group delay) and the cross frequency fc) and by the capacitance-voltage characteristics of varactor diodes, and the design of a UHF frequency modulator utilizing a hybrid integrated amplifier, is described. It is pointed out that, although the capacitance variation ratio n(log C(V)/log V) is less than 2, the capacitance variation ratio increases with the varactor diode bias voltage so that good modulation linearity can be obtained. Furthermore adopting a push-pull mode of operation, two UHF frequency modulators are connected to obtain a wideband frequency modulator (the bandwidth is f025 MHz). Finally a feedback-type frequency modulator operating at 6 GHz is discussed, as comparison to the result of the UHF frequency modulators.

A carbon granule microphone with carbonaceous electrodes, which is lower in cost and more convenient for mass production than the conventional microphone with gold plated electrodes, was studied as a telephone microphone. A molded piece made from a mixture of synthetic resin and carbon powder can be used for the back electrode and paint made from it can be used to coat the dome electrode. A new microphone, whose sensitivity is as high as a conventional microphone, shows higher resistance. This means that the efficiency of the new microphone seems to be slightly less. However, the sensitivity change in the new microphone under excitation by voice is less than that of the conventional microphone. The new microphone sensitivity dependence on DC input power is almost the same as that of a conventional microphone.

This paper describes a realization problem of infinite dimensional, nonlinear, time-invariant dynamical systems. A necessary and sufficient condition that a given input/output behavior is to be realized by the above system is obtained. The theory of nonlinear semigroups plays an important role in our study.

In this letter, we deal with the minicomputer system developed for imaging the degree of the decrease in a peak frequency of the first heart sound measured at 25 points on the chest surface and the similarity of one. This system provides a useful and powerful means to understand the diagnostic information rapidly.

The improved photoenhancement of the SAW convolution effect in monolithic convolvers with high-resistivity surface layer has been discussed. CdS monolithic convolvers with one or a half acoustical wavelength thick high-resistivity layer formed on semiconductive substrate are prepared in order to check the expected behavior. Due to much reduction of the acoustoelectric loss of SAW, the convolver with a half acoustical wavelength thick high-resistivity layer shows a remarkable photoenhancement, 20 times the value in the dark, when photocarriers are generated uniformly in the high-resistivity layer. Nonlinear capacitance mechanism can also induce photoenhancement of the convolution effect. It is first found that further enhancement of the convolution can be achieved by applying positive pulsed bias to the gate electrode under illumination of a light with high absorption coefficient.

A new technique for fabricating surface acoustic wave transducers is proposed. By using this technique, an IDT with oxide loaded in the space between the metal electrode fingers is easily realized. The fabricated transducers had ripple free response as well as the low velocity dispersion.

One of the main loss factors in the optical fibers fabricated by the vapor phase axial deposition process is at present residual OH ions which are originated mainly from an oxy-hydrogen flame. Residual OH ions show an influence on loss increase, especially, in the long wavelength region. OH-ion behaviors in the porous perform, transparent perform and drawn fibers are studied with use of an infrared spectroscopy and fiber transmission measurement. In order to realize low OH content VAD fibers, the dehydration effect of halide vapor on the porous perform is investigated experimentally. Dehydration effect of thionyl chloride on the porous perform is studied by various heat treatment temperature. As a result, residual OH ion content in the drawn fiber was reduced from 30 ppm to 20 ppb and the attained minimum loss of dehydrated fiber was 0.5 dB/km and 0.28 dB/km at 1.3 µm and 1.6 µm wavelengths, respectively. Present investigation clarified that the VAD fibers are applicable to an optical transmission system in the long wavelength band. We compare the loss peak values at the wavelengths of 1.39, 1.24 and 0.95 µm among several optical fibers fabricated by different method. Especially, the VAD fibers dehydrated with SOC12 show abnormal peak ratios at the wavelengths.

A short channel MOS FET with 1-1.5 µm channel and double boron-ion implantation has been designed as a high speed static RAM element. High speed properties and short channel effects are compared among the MOS FETs with 1.5-100 Ωcm substrate and several deep/shallow implant dose combinations. It is found that 40 Ωcm substrate or so with 4/45/31011 cm2 deep/shallow implant dose combination is most desirable.

We reveal that the degree of pattern separation in a two-layer random neural net with inhibitory connection has a favorable property for pattern classification. We also consider the effect of the number of connections between the two layers on the above property.

Any imperfections of a dielectric thin-film waveguide can be mainly classified into the irregular boundary interfaces and the refractive-index fluctuations. In this report we demonstrate a simple method to obtain the scattering losses caused by the two typical imperfections in terms of the coupled power coefficients.

An extraction method of contour lines from a contour map on the market and their restoration method are described. Twin tracing points (clockwise and counterclockwise) are used to trace and extract line components from the map which involves not only contour lines but also block components such as symbols and numerals. To implement the tracing successfully, ending conditions and setting manners of the twin tracing points are dealt with taking various kinds of curves into consideration. By this method, sets of the coordinates of traced contour lines, step numbers of tracing, and moreover, the thickness and length of traced contour lines, the coordinates of their skeleton lines are obtained. Since some contour lines are cut off by the elimination of symbols and numerals, to restore them a finding method of a pair of cut off points which must be connected is also described. For the restoration, an adjacent graph is introduced with an example, and several cases of relations between a graph and set of contour lines are dealt with. A determining algorithm of a pair of connecting candidates is given. Finally, as the results, an original contour map of a mountainous area, extracted contour lines, skeleton contour lines and a connected (by straight lines) contour map obtained through the twin-tracing-points method and restoration method are illustrated.

The pulsed HF chemical laser of the HI/SF6 system can be excited to the sixth HF vibrational level, and the spectra of this system coincide with those of the HF chemical laser of the H2/F2 system. The pulsed HF chemical laser of the HI/SF6 system can therefore be used as an oscillator or a preamplifier for the power-amplifier of the H2/F2 system. The power of a pulsed HF chemical laser of the HI/SF6 system is said to be poor in comparison with other mixed systems, but the use of a Blumlein-discharge circuit to initiate the electric discharge proved to be effective. An electrical efficiency of 3.9% for a pulsed HF chemical laser of the HI/SF6 system was obtained, and both the electrical efficiency and the output energy of this laser system were comparable with those of the HI/SF6 system.

Evaporated Ni thin films being formed on a bulk Ni electrode were eroded by a single break are in resistive circuit. A model of the crater formation was proposed from observations of the eroded area by Scanning Electron Microscope.

This paper describes an effective method for coping with short-line-interruptions in high-speed data transmission, and effectiveness of the technique is shown by the experiments.

A pulsed HF chemical laser of the HI/SF6 system can be used as an oscillator or a preamplifier for the laser power-amplifier system of the H2/F2 system. The HF chemical laser of the HI/SF6 system is predicted to extract more energy than the other mixed systems, but the results of calculations based on a simulation model were not in agreement with experiment. In this paper, our calculation was based on the assumption that the HI molecules dissociated upon the initiation. The pulsed HF chemical laser of the HI/SF6 system is here treated theoretically, utilizing the constant-gain methode by a pulsed chemical laser code. The pulsed HF chemical laser is characterized by the laser output energy, power and pulse width. The calculated results in accordance with the assumption that 1% of SF6 and 25% of HI are dissociated by the electric discharge, are in good agreement with the experimental data for a pulsed HF chemical laser of the HI/SF6 system with the Blumlein-discharge circuit.