Showing papers in "Electronics and Communications in Japan Part Ii-electronics in 1988"

Vector finite‐element solution of anisotropic waveguides using novel triangular elements

Abstract: A new triangular element is proposed in the vector finite-element solution of a dielectric waveguide. In the conventional finite-element expression based on the three magnetic field components, each component is treated independently by continuous elements. The elements used in this paper are subject to the condition that the longitudinal field components are continuous between adjacent elements and the transverse components have tangential components continuous at the element boundary. In this method, as many zero eigenvalues as the number of unknowns of the longitudinal components are generated. However, all the remaining nonzero eigenvalues correspond to the physical modes and no spurious solutions typical in the conventional vector finite-element solutions are generated. The present elements are suited for representation of the forced boundary conditions, and the analysis region can easily be applied to an arbitrary waveguide. To investigate the validity and effectiveness of the present method, it was applied to a hollow cylindrical waveguide, a dielectric loaded rectangular waveguide, an anisotropic refractive index dielectric waveguide, and a ferrite loaded square waveguide.

Electromagnetic wave absorbing properties of carbon‐rubber doped with ferrite

Abstract: When ferrites are used for electromagnetic wave absorbers in the VHF and UHF bands we note their excellent properties such as their wider dynamic frequency bandwidth and their thinner thickness than that of lossy dielectric absorbers, which are due to their magnetic losses. However, there exists an upper frequency limit beyond which the frequency bandwidth of ferrite absorbers becomes narrower than that of lossy dielectric absorbers. In this paper we added ferrites to carbon-rubbers and examined their characteristics as electromagnetic wave absorbers. We found through experiments that the relative bandwidth is widened more than twice by adding ferrites of only relative weight ratio 0.2 to rubber when the relative weight ratio of carbon to rubber is 1.1. Although we made measurements in the frequency range of 5 to 6 GHz, we expect the same effectiveness at higher frequencies.

Analysis of electron wave reflectivity and leakage current of multi‐quantum barrier: MQB

Abstract: A multi-quantum barrier (MQB) is proposed which is made of super lattices. The electron wave reflectivity and the leakage current of the MQB structure for an electron energy are calculated theoretically by means of a simplified model. The MQB structure studied here is a multiple heterostructure made of a combination of periodic structures with different well widths and barrier widths. The reflected waves from each interface are intensified by interference. The energy values at which the reflection is intensified are changed slightly by changing periodic structures. By combining these periodic structures the reflectivity is enhanced toward unity up to the energy higher than the classical barrier height. It is found from the calculations that the effective potential height of the MQB structure is higher by about 30 percent in the GaAs/AlAs and by about 50 percent in the Ga0.47In0.53As/InP than the classical barrier height. It is expected that the MQB structure can be used in high-temperature operation light-emitting devices and short-wavelength light-emitting devices.

Finite‐element solution of three‐dimensional resonator problems. Novel rectangular parallelepiped elements

Abstract: New rectangular parallelepiped elements are proposed for analyses by finite elements for problems of three-dimensional resonators. Earlier methods of finite-element analysis confirmed the generation of nonphysical spurious solutions which are mixed with physical solutions. New first-order and second-order rectangular parallelepiped elements are proposed and the formulations of these finite elements are derived. One of the characteristics of these new elements is that tangential components are used for the electromagnetic fields on the surface of the elements as an unknown parameter. As concrete examples, they are applied to a cubic resonator and a rectangular resonator which contained a dielectric material. It was confirmed that except for a few zero eigenvalues, all nonzero eigenvalues corresponded to resonant modes. Specifically, if new rectangular parallelepiped elements are used, it is shown that no generation of spurious solutions occurs with nonzero eigenvalues.

Direct core observation method using thermal radiation of silica fibers with dopants

Abstract: Various direct core observation methods have been developed for directly observing the core itself of a single-mode optical fiber such that the axial alignment is performed and the splicing loss can be evaluated. In the conventional methods, special devices are needed such as an He-Cd laser and a collimated illuminating optical source. In this paper, the optical fiber core can be observed directly by thermal radiation when Ge-doped optical fibers are spliced by glow discharge heating. Also, this phenomenon can be used for splicing. When a Ge-doped fiber is placed near the glow discharge heating region, the core shines in contrast to the clad. This is because the absorbing loss of the core is larger than that of the clad and hence a stronger thermal radiation occurs from the core. Further, near the wavelength of 420 nm, Ge luminescence exists due to the ultraviolet (UV) light from the glow discharge. The rate of the luminescence of the total radiation from the core is at most one-twentieth the thermal radiation in a steady-state splicing situation. Further, the apparent core location is expanded by the amount of refractive index due to the lens effect of the clad. Hence, the true location of the core must be calibrated.

Two-dimensional numerical analysis of gaas MESFET with a p-buffer layer

Abstract: To suppress the threshold voltage shift of GaAs MESFET's in a shorter channel region, a two-dimensional numerical analysis of the FET structure with a p-buffer layer is carried out. As a result of the analysis, it is clarified that the insertion of the p-buffer layer is effective in preventing the carrier penetration in the substrate. In analyzing the electrical properties as a function of the buffer layer thickness, it is found that the substrate current which flows beneath the p-channel layer, decreases with an increase in the buffer layer thickness and that the threshold voltage shift is suppressed with the increase. As the buffer layer thickness increases, the transconductance increases. Moreover, as the buffer layer thickness increases, the gate-substrate capacitance increases. Therefore, to obtain a high cutoff frequency, it is necessary to optimize the buffer layer thickness.

Proposal for a fiber‐optic endlessly rotatable fractional wave device and its applications to lightwave technologies

Abstract: In the fields of single-mode fiber-optics, polarization control is an important technique to control undesirable polarization fluctuations. This paper presents a new fiber-optic fractional wave device which can endlessly and continuously rotate to compensate for polarization fluctuations. A design method is described and factors determining the performance of the device are presented. Characteristics of fabricated devices are described and their applications are introduced. This fiber-optic fractional device utilizes birefringence induced in a bent fiber. The unique feature is that the fiber is coiled between two rotaing parallel plates with only loose support. This structure makes it possible to achieve a compact endlessly rotatable device. For design comparisons, a figure of merit relating to the characteristics of the bent fiber is proposed. The shape of the fiber bending changes according to the rotation of the device. The effect of this transformation on the characteristics of the device is studied theoretically. Experimental studies are carried out on the effect of coating materials, on fiber durability, and on its characteristics during high-speed operation. Half-and quarter-wave devices for 1.52 μm wavelength are fabricated with losses and birefringence accuracies better than 0.2 dB and ±5%, respectively. A polarization controller realized by utilizing this device is used in a 400 Mb/s optical FSK transmission over 270-km fiber. It also is used in an LD spectrum width measuring instrument.

Design of microwave combiner with waveguide ports

Abstract: A microwave power combiner with waveguide ports is proposed and its design method is described. It can combine a large number of inputs at the same time and its voltage tolerance is high. On the side wall of a circular TM0m0 (m = 2, 3, 4,…) mode resonator, rectangular waveguides are connected as the input terminal while a circular waveguide is connected at the bottom surface of the resonator as the output port. The output is transformed into a rectangular waveguide via a TM-TE mode transformer. Slits on the side of the output circular waveguide prevent propagation of the modes that do not contribute to the power combining. In the design method, the Y parameters are computed from the equivalent circuit seen from the output side. Thus, the matching condition is obtained from the image parameters. Actual dimensions of the iris for adjustment of the coupling between the input and output waveguides and the resonator can be found from this procedure. The insertion loss of a 14-GHz four-way combiner with a TM020 mode resonator was 0.25 dB.

Development of a high‐reliability flip‐chip packaging reinforced by resin

Abstract: A high-reliability chip-on-board (COB) technique using the flip-chip method to mount a driver LSI on a soda-glass substrate has been developed. By forming a flip-chip structure in which both the chip and the gap between the chip and the glass substrate are filled with resin, the thermal fatigue life can be one order of magnitude longer than that of a device with an exposed flip-chip structure. The extension of the device life is due mainly to the relaxation of stress to the device by matching the thermal expansion coefficients of the resin and the solder. Furthermore, using the resin-filled flip-chip structure, the breakage of glass at the pedestal section is minimized. The COB package is highly reliable and durable against humidity.

Very high‐speed ROM using bipolar/CMOS technology

Abstract: It is a common practice to realize the internal control of a logic LSI by means of a microprogramming method. A high-speed operation of this logic LSI requires a highspeed ROM (micro ROM) which stores this microprogram code. This paper proposes a new circuit configuration derived from a combination of bipolar and CMOS technology for a high-speed micro ROM. In this configuration, the memory cell array is made of MOS transistors while the array periphery consists of a bipolar/CMOS composite circuit in which a high driving capability and a large gm of the bipolar transistor are used so that a high degree of integration and a high speed are attained simultaneously. To confirm the high-speed characteristic of the proposed circuit, a micro ROM of 2 k words x 64 bits has been fabricated with the 2 μm Hi-BiCMOS process mixing the bipolar and CMOS structures. The test results of the fabricated device have a minimum read-out cycle time of 16 ns which is more than twice that of a CMOS circuit having an identical configuration.

Distributed fluid sensor using eccentrically cladded fibers

Abstract: To increase the sensitivity of distributed gas or liquid sensors using the evanescent wave of optical fibers, methods such as making the fiber single-mode and reducing its cladding thickness or removing it completely, can be considered. However, the core of the single-mode fiber is only a few micrometers in diameter and, therefore, it becomes very difficult to preserve it without the cladding layer. On the other hand, increasing the cladding thickness will reduce its sensitivity. In this paper, the eccentrically cladded fiber is proposed as a means to increase the sensitivity without the need to reduce its diameter. The theoretical analysis and its experimental verification are carried out. As a result, it is shown that if the eccentricity is increased, the sensitivity will not decrease remarkably, even at large cladding diameters. Further, as experimental evidence of liquid detection, optical losses of a few decibels per centimeter were observed by attaching a liquid to the fiber. This indicated the possibility of highly sensitive distributed-type sensors. Because of the good agreement of theoretically estimated radiation losses of W-type fibers with experimental results, it is concluded that eccentrically cladded fibers can be designed to exhibit the desired sensing characteristics.

A multiple‐ISFET integrated with CMOS interface circuits

Abstract: To achieve the simultaneous measurements of ion concentrations with high accuracy using multiple ISFETs, the integration of multiple ISFETs with signal processing circuits has been attempted. To achieve process matching in the fabrication of the ISFETs and additional circuits, and to lower the power consumption, CMOS devices have been used for the signal circuits. The integrated circuits were composed of four ISFETs, a constant current source, a buffer amplifier, and a multiplexer. To measure ion concentration at high speed with a high degree of accuracy, each ISFET contained a source follower and a buffer amplifier to achieve an independent operation. Only the output signal was switched by a multiplexer. The average sensitivity of each ISFET was 54.4 mV/pH for the range of pH 1 to 10. The response time for an increase of the pH by 1 was about 1 s. The time needed to switch the output signal was 200 μs, which is three orders of magnitude faster than ion response speed. As a result, it was found that this integrated sensor structure was suitable for measuring the response of multiple ISFETs at high speed. This paper describes the sensor structure, the circuit design and its response characteristics.

Tunable laser using sheet of dye-doped plastic fibers

Abstract: The development of the fiber-type laser is desired for simple construction of the optical system for several fiber sensor applications. Therefore, in this paper the fabrication of several types of dye-doped plastic fibers and the laser properties have been studied. When a sheet of dye-doped fibers was used as an active element and pumped by a pulsed N2 laser, the tunable intense radiation was generated over a wide wavelength region. By using two types of dye-doped fiber sheets, almost the entire visible region could be covered.

NMR fresnel transform imaging using a scannable nonlinear field gradient

Abstract: A new NMR imaging method is proposed, where the data with contributions from the whole imaging area are examined, and the spin distribution image is reconstructed by computer processing. The method uses a weak scanning nonlinear gradient field which is an extension of the magnetic focusing method. The experimental result is described. It is shown that the detected signal is given by the Fresnel transform of the spin distribution, and a solution is obtained by a convolution integral. Based on the theory, imaging experiments were performed using a phantom, realizing a clear spin density distribution image. In the experiment, a magnetic field of B0 = 0.0192 Tesla was used, which is one order of magnitude weaker than the field in the usual NMR-CT. The image was obtained with the resolution of approximately 1 mm by collecting the data for 32 s to 8 min. Some discussions are made on the spatial resolution of the method and the application to a large-scale system, indicating that the method is highly promising.

Hi‐bicmos 32‐bit execution unit

Abstract: The applicability of the Hi-BiCMOS (High Performance Bipolar CMOS) technology to the 32-bit execution unit has been researched to improve the function and speed of such processors as minicomputers CPUs. The HiBiCMOS technology enables MOS devices with performance comparable to pure CMOS LSIs and bipolar devices with performance comparable to pure bipolar LSIs to be combined on the same substrate. Device integrations comparable to CMOS LSIs have been achieved by the utilization of macrocells composed mainly of MOS circuits. Attempts have been made to speed-up the macrocells by adding Hi-BiCMOS circuits. In this paper, the feedback-type smallamplitude precharge circuit is proposed and fabricated. This circuit affords precise bus voltage. Also, the carry propagation circuit that uses the bipolar sense amplifier to achieve the machine cycle of 25 ns is proposed. Consequently it is confirmed that the precharge circuit operates two times faster and the carry propagation circuit 3.3 times faster than pure CMOS circuits. Thus, one of the promising candidates for the ultrahigh-speed one-chip processors is shown.

Fabrication of polycrystalline silicon thin‐film transistors by ion shower doping technique

Abstract: In the fabrications of polycrystalline silicon thin-film transistors (TFTs), doping of impurities is often carried out by ion implantation. In this study, dopings of phosphorus and boron were carried out by an ion shower doping technique, where the rf plasma of H2-diluted PH3 or H2-diluted B2H6 gas with a magnetic field was used for the ion source. To form the source-drain regions using this technique, the doping characteristics such as sheet resistivity and impurity profiles in polycrystalline Si were investigated. Using the conditions obtained experimentally, n and p channel Si gate polycrystalline Si TFTs were fabricated on a fused silica substrate by using a CMOS process. It was also confirmed that the conventional photoresist can be used for this process. For the n-channel transistor, field-effect mobility μ = 3 cm2/V·s, threshold voltage Vth = 5 V, and on-off current ratio ION (VD = 6 V, VG = 10 V)/IOFF (VD = 6 V, VG = 0 V) ≌ 106 were obtained for W/L = 12 μm/6 μm. In the p-channel transistor, the characteristics were similar to those obtained by ion implantation. The ion shower doping technique made it possible to fabricate polycrystalline Si TFTs with a CMOS structure, using a simple apparatus. By using this technique, a large area process will be possible.

Observation of elliptical motion in the torsion coupler of ultrasonic motor using optical heterodyne interferometer

Abstract: An elliptical motion in the torsion coupler of an oscillator-type ultrasonic motor has been observed noninvasively by means of an optical heterodyne interferometer. Studies of the elliptical motion in the torsion coupler are important for the research on improvement of the efficiency and optimum design of an ultrasonic motor. However, to date, only microscope observation has been carried out and hence accuracy is not as high as desired. In this paper, an optical heterodyne interferometer based on a frequency-stabilized transverse Zeeman laser is used for detection of the Doppler shift frequency by an ultrasonic vibration so that the vibration waveform of the torsion coupler has been regenerated. The oscillations of the rotor contact plane and side plane of the end of the beam have been measured individually by an identical optical system. From their amplitudes and the phases, the elliptical loci have been computed. In another method, the oscillations of the two surfaces have been measured simultaneously with two interferometer systems and their elliptical relations have been displayed in real time on an XY oscilloscope. Oscillations at each point on the beam and the variations of the elliptical shape versus the driving frequency of the ultrasonic motor have been observed successfully by the second method.

Influence of DC bias on nonlinear distortion of contacts

Abstract: When the nonlinear distortion generated at the metal contact is measured by the two-frequency method, spectroscopic measured results ranging from the second- to sixth-order distortions can be obtained. Depending on the type of contact and the contact condition, the characteristics of these distortion spectra are different. It was reported previously, together with the measurement technique, that this measurement can be used as a measure for finding the physical phenomena on the contact surface. In the present study, the second- and third-order distortions are considered to be representatives of the even- and odd-order distortions for the purpose of finding the cause of the distortion. A dc bias is applied to the contact and the bias dependence of the distortion voltage is studied. As a result, it is found in most contacts that the second-order distortion voltage is affected strongly and becomes a minimum at a certain bias point. This indicates that at this bias point the diode characteristics which are the cause of the even-order distortion vanish and the nonlinearity is almost perfectly symmetric for the currents in both directions. Especially, in this measurement, the bias dependence is observed even if glossy metals are contacted. Hence, this is a method for finding the surface phenomena for a thin-film thickness from the electrical characteristics.

Technology of cable transfer system for error‐free digital transmission

Abstract: At present the transfer or exchange of subscriber cable in the telecommunication network is performed by temporarily cutting off the circuit with the customer's consent. For a communication circuit requiring 24-hr real-time service, such as data transmission, however, a cable transfer method must be developed which can handle a live circuit and prevent an error in the information during transmission. From such a viewpoint, this paper considers the cable transfer technique which improves the traditional cable transfer process, and presents the design condition in relation to the transmission quality of the communication circuit. First, a cable transfer method is proposed which can be applied to a live circuit without affecting the telecommunication, such as the interruption or the noise due to transfer. In this method, resistors are inserted into existing and new circuits at two working points for the cable transfer, and the resistances are varied in a pair. The method can be applied to the code transmission circuit in ISDN, in the cable transfer for the subscriber line. Second, an identification method is proposed, which can be applied to the unbalanced circuit, such as the security alarm circuit. The method utilizes the transmission signal at each working point, and compares the measured values. The idea can be applied to the general balanced lines to identify the line with a high speed.

Elliptical micro‐fresnel lenses fabricated by electron‐beam writing technique

Abstract: Elliptical micro-Fresnel lenses can exhibit astigmatic characteristics; and hence, an astigmatic optical system made of a conventional combination lens can be made much thinner by their use. The authors have developed an electron-beam writing system which has a function suitable for fabrication of thin-film micro-optical elements. The function of this writing system has been expanded so that elliptical micro-Fresnel lenses can be fabricated. It has been confirmed that the fabricated lenses have a good sawtooth cross section. For the focusing characteristics of the lens, the eccentric ratios and the observation position dependence have been investigated theoretically and experimentally. The obtained focused spots almost reach the theoretical diffraction limit in both the major and minor directions. The intensity distribution of the focusing spots agree with the theoretical prediction. Hence, the fabricated lenses have focusing characteristics in agreement with the designed ones. Compared to the conventional microlens, these elliptical Fresnel lenses have a size smaller by one order of magnitude and their focusing efficiency is as high as 72 percent. Hence, it is expected that these lenses are new elements for micro-optics.

A measurement and a classification of arcing phenomena of AgPd alloy contacts in low pressure

Abstract: To understand the arcing phenomena observed when contact is breaking, an apparatus was developed to observe and record arc voltage waveforms. The apparatus is formed by adding a waveform acquisition circuit composed of an A/D converter, memories and control circuits to the conventional contact measurement system. Using this equipment, the voltage waveforms were observed for Ag, Pd or their alloys contacts for 500 operations under various surrounding ambient pressures. The waveforms were analyzed with respect to length, slope and fluctuation. The waveform tendencies were classified by the combination of ambient pressure and alloy ratio. For example, when the Ag content is high or the ambient pressure is high, the transition to gaseous phase becomes smoother with the increase of operation repetitions. When the Pd content increases and the ambient pressure is low, only a short metallic phase arc occurs. Using the measurement and classification, data for establishing the relationship between the ambient pressure and the contact material was acquired. The classification carried out by using contact resistance was similar to that by arc waveforms.

High–power HeCd+ white light laser

Abstract: The hollow cathode HeCd+ laser includes twelve oscillation lines. The oscillation line of 441.6 nm (blue), 533.7 nm and 537.8 nm (green), and 635.5 nm and 636 nm (red) in the visible area can be oscillated concurrently by using wide-band mirrors for a resonator to produce a white light laser. This laser is expected particularly in the image production industry because a single laser tube allows the three primary colors of light to be produced and the green wavelengths are optimum for green emulsion of photographic films. However, it has been difficult to produce the hollow cathode HeCd+ laser of continuous oscillation and high power. We have investigated the high power and output stability of the white light laser for practical purposes. As a result of examining the structure of the hollow cathode as a means for high-power laser, we could realize a high-power HeCd+ white light laser by using a new structure which was an improvement over the conventional flute type. This paper explains the structure of the hollow cathode and shows the dependencies of the output on the Cd vapor pressure, He pressure, input power for the cathode length of 61 or 43 cm, and the noise characteristics for the cathode length of 61 cm.

High stable SAW devices on LST cut of quartz

Abstract: Recently, we discovered an LST cut which utilizes leaky surface waves for a high stable quartz crystal for SAW devices. This paper examines experimentally the properties of the LST cut. Due to an error for the cut angle for the LST cut, zero temperature coefficients are sometimes not obtained. However, it was shown that such coefficients can be obtained by varying the direction of wave propagation within the cut plane. We examined SAW resonators and oscillators in a 180-MHz band. We obtained the temperature characteristics of the LST cut in the temperature range from −20 to 80°C with deviations as small as 15 ppm. The obtained Q was as high as 22,500. One of the features of the leaky surface waves is that they have a large value for the propagation velocity. We took advantage of this feature and manufactured SAW filters for 1.8- and 4-GHz bands. Their characteristics are presented in this paper.

A method to measure position and attitude of a tunneling machine

Abstract: Development of a method to measure the position and attitude of a tunneling machine is required for constructing a curved and long tunnel with a small diameter for communication cables. This paper presents a method to satisfy this requirement. Two coils (“send coils”) are set on the ground to generate an ac magnetic field. Another coil (“search coil”) to detect this magnetic field is set on the tunneling machine under use. This search coil rotates on orthogonal planes, one of which is parallel to the direction of the tunneling-machine progression. The position and attitude of the machine are calculated from the voltage induced in the search coil. The direction of a component of the magnetic field at an arbitrary position (with reference to the send coils) is essential in this method; this was obtained from theoretical (the finite-element method) and experimental values. Appropriate dimensions of the search coil were found by introducing “the height/width ratio” and “the winding ratio.” The detection accuracy of this method is about 20 cm on a horizontal plane, and about 2° in attitude, thereby confirming its effectiveness.

Finite‐element analysis of plane wave diffraction from metallic gratings with arbitrary complex permittivity

Abstract: A method of analysis based on the finite-element method is presented for the problem of plane wave diffraction from metallic gratings with arbitrary complex permittivity. First, the problem of diffraction of a plane wave from the metallic reflected-type grating is analyzed rigorously as a two-media boundary value problem by the finite-element method. Next, a new method is proposed in which an approximate boundary condition using the surface impedance is incorporated into the discretized equations by the finite-element method. Especially, in the method of using the surface impedance approximation, the region to be analyzed is reduced substantially from the one for the two-media boundary value problem, and the computational effort is expected to be reduced significantly. The diffraction characteristics of grooved gratings with sinusoidal, triangular and rectangular profiles are actually calculated and the results are compared with those by other methods and experiments. In the case of the TE wave incidence, it is found that the results by the two-media boundary value problem and those by the surface impedance approximation agree well. On the other hand, in the case of the TM wave incidence, the results are substantially different depending on the grating profile. By comparison of the results with those based on the assumption of a perfect conductor, it is found that the loss in the metal must be considered at a longer wavelength in the case of the TM wave incidence than in the case of the TE wave incidence.

Direct formation of solder bump on Al pad using ultrasonic soldering

Abstract: Although a conventional solder bump formation on Al pad uses an evaporation of Pb and Sn on the underlying metal (Au/Cu/Cr, Au/Pt/Ti, etc.) and reflow, it has the disadvantages of high equipment cost and complex process. By applying an ultrasonic soldering technique, in this work, solder bumps were formed directly and selectively on an Al pad without underlying metal. The sample chip used in the experiment had 118 Al pads which were 1.5 μm thick and 100 μm□, and anticorrosive Sn-37 Pb-5 Ag was used as a solder. By examining the factors affecting the bump formation percentage, an optimum condition was selected [(1) chip preheating temperature: above 280°C; (2) solder temperature: 280°C; (3) distance between chip and ultrasonic horn: ≈0 mm; (4) dip time: more than 8 s; (5) ultrasonic power: more than 135 W]. Using this condition, bumps were formed without damaging the passivation film, which was also the case on 30 μm □ pads with 50 μm pitch without bridging. The bump height was approximately 36 μm on 100 μm □ pads with a variation of ±6 μm. Since the shear force of the bump was on the average 32.5 g and all the breakage occurred within the bumps, a strong adhesion at the interface is guaranteed. The interface contained a compound several 1000 A thick with a 4 : 1 composition of Al and Ag.

Design of fiber holder for optical fiber-ribbon splice

Abstract: A field-usable optical fiber-ribbon holder is proposed. Optical fiber ribbon is fixed by placement between wavy plates. The fiberfixing mechanism is analyzed theoretically and experimentally. Optimum fiber-holder sizes are also investigated. Using this fiber holder, average and maximum fiber-end displacement of 3.8 and 15μm are obtained. Lowloss splicing of SM fiber ribbon is confirmed, with average splice loss of 0.05 dB.

Method of materials design for I-III-VI2 chalcopyrite-type mixed crystal semiconductors

Abstract: Binary compound semiconductors have become extremely, important in the study of semiconductor devices. On the other hand, it is difficult to obtain a large size, highquality single crystal for a semiconductor composed of more than three elements belonging to different groups which is a so-called multinary compound semiconductor. Therefore, such a semiconductor has only been of interest in research for material properties. A chalcopyrite-type compound has been a target of interest as a ternary compound semiconductor. This paper describes the relatins between several properties and mean atomic weight, and the relations between the crystal ionicity and material properties in chalcopyrite-type semiconductors. These factors are used to obtain some guidelines for designing the materials with the desired characteristics in chalcopyrite-type mixed crystal semiconductors. As a result, contour maps of gap energy and lattice constants as functions of mean atomic weight for these materials were obtained. This method of designing materials is confirmed experimentally for I-III-VI2-type mixed crystal semiconductors.