Browsing by Issue Date, starting with "1992-01-30"

Now showing 1 - 20 of 90818

FORMATION OF SELF-ALIGNED COBALT SILICIDE IN NORMAL FLOW NITROGEN FURNACE
(1992-01-30)
YANG, FM
;
CHEN, MC
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
Cobalt can be completely transformed into cobalt silicide during annealing in a normal flow nitrogen furnace above a certain temperature which depends on the thickness of the cobalt film deposited on the silicon substrate. For a 400 angstrom cobalt film this temperature is 500-degrees-C and for a 1500 angstrom cobalt film it is 650-degrees-C. Loading the wafer into furnace tube is a critical step to prevent the cobalt from oxidation before silicidation in the annealing process. Lateral growth of silicide is not observed using this silicidation scheme. The cobalt oxide together with the underlying cobalt over the silicon oxide can be removed by hot 3HCl + H2O2 + 3H2O solution if the annealing is performed at a temperature below 650 and 900-degrees-C for a cobalt film thickness of 400 and 1500 angstrom respectively. The optimum condition for practical silicide formation is defined. The products and microstructure over the silicon and silicon oxide are characterized by X-ray diffraction spectra, Auger electron spectroscopy depth profiles, sheet resistance measurements and scanning electron microscopy inspection.
REDUCTION OF RESIDUAL-STRESS BY PARALLEL HEAT WELDING IN SMALL SPECIMENS OF TYPE-304 STAINLESS-STEEL
(1992-02-01)
CHOU, CP
;
LIN, YC
;
交大名義發表
;
機械工程學系
;
National Chiao Tung University
A new method designated 'parallel heat welding' has been developed for the reduction of residual stress in small specimens. It is carried out using a pair of parallel heating torches attached one either side to a welding torch as a movable heat source during welding. Six different welding conditions were applied to an autogenous gas tungsten arc welded Type 304 stainless steel. During welding, the thermal cycles of four locations in the weldment were recorded to investigate the action of thermal stress and to evaluate the effect of the parallel heat sources on sensitisation. In addition, an optical microscope was used to observe the degree of sensitisation. The experimental results show that the maximum principal residual stress and parallel welding direction stress can be reduced by 21-32% when the conventional welding process is replaced by parallel heat welding. The equilibrium temperature of the welded plate can be increased from 45 to 240-degrees-C. Also, based on the exposure time recorded at temperatures in the range 600-850-degrees-C, and metallographical observations, it was found that parallel heat welding has no influence on the degree of sensitisation of austenitic stainless steel weldments.
MATHEMATICAL MORPHOLOGY ON L-IMAGES
(1992-02-01)
HSUEH, YC
;
交大名義發表
;
資訊工程學系
;
National Chiao Tung University
;
Department of Computer Science
In most applications, morphological operations are considered as unary operations, each associated with a structuring element. Due to the problem of grey-level overflow, the associated structuring elements of morphological operations on grey-level images (functions from R(n) or Z(n) to [0, m], where m is a fixed positive number) are limited to 'flat' or 'non-flat' structuring elements. However, neither flat or non-flat structuring elements are grey-level images. In this paper, we propose the notion of l-images, based on a clc-monoid (complete lattice-ordered commutative monoid), as a unified representation of binary images (functions from R(n) or Z(n) to {0, 1}), signals (functions from R(n) or Z(n) to the set of extended real numbers R*) and grey-level images. Then we generalize morphological operations to l-images such that the associated structuring elements can be arbitrary l-images. As consequences, we obtain several concrete expressions for dilations and erosions on grey-level images such that the associated structuring elements are also grey-level images. Therefore, the grey-level overflow problem is solved with no limitations on structuring elements. Furthermore, we show that the duality property of dilations and erosions on l-images holds if the underlying clc-monoid is self-dual.
DISTRIBUTED ERBIUM-DOPED FIBER AMPLIFIERS WITH STIMULATED RAMAN-SCATTERING
(1992-02-01)
WEN, S
;
CHI, S
;
光電工程學系
;
電信研究中心
;
Department of Photonics
The effect of stimulated Raman scattering (SRS) on the distributed erbium-doped fiber amplifier with 1.48-mu-m pump wavelength and 1.55-mu-m signal wavelength is numerically studied. The contributions of the gains from SRS and the pumped erbium ions doped in the fiber core are shown. It is found that SRS enhances both the gain and signal-to-noise ratio of the amplifier except the signal-to-noise ratio may be decreased by SRS for high initial signal power.
A NEW ACCESS-CONTROL METHOD USING PRIME FACTORIZATION
(1992-02-01)
HWANG, JJ
;
SHAO, BM
;
WANG, PC
;
管理科學系
;
資訊管理與財務金融系註：原資管所+財金所
Based on prime factorisation in number theory, a new but simple and efficient protection method for access control is proposed. This method has several advantages on performance over previous methods in the literature. Especially, a file/user can be added to or removed from the system without much effort. Excellence of this method is more pronounced for those systems where files are accessible to only a few users.
INVESTIGATION OF THERMAL COEFFICIENT FOR POLYCRYSTALLINE SILICON THERMAL SENSOR DIODE
(1992-02-01)
YEH, CF
;
LIN, SS
;
YANG, GP
;
交大名義發表
;
電控工程研究所
Thermal sensors utilizing p-n junction diodes built into the polycrystalline silicon layer are developed. The effective factors influencing the thermal coefficient of polycrystalline silicon thermal sensor diodes are clarified. It is found that diodes with a light boron dose have a larger thermal coefficient than those with a heavy boron dose. The plural p-n diodes connected with more cascade series number have a larger thermal coefficient than ones with less cascade series number. The periphery of the thermal sensor seems to have little effect on the thermal coefficient. In addition, the thermal coefficient will become higher as a larger forward current is selected. The thermal coefficient of the sensor with a boron dose of 1 x 10(15) cm-2 will reach a saturation value of -2 mV/K when the forward current is above 2-mu-A.
ANALYSIS OF DISPERSION AND SERIES GAP DISCONTINUITY IN SHIELDED SUSPENDED STRIPLINES WITH SUBSTRATE MOUNTING GROOVES
(1992-02-01)
WU, LK
;
CHANG, HM
;
交大名義發表
;
傳播研究所
;
National Chiao Tung University
Dispersion and series gap discontinuity of shielded suspended striplines (SSL's) on Duroid substrate (epsilon(r) = 2.22) are analyzed using the finite-difference time-domain method (FD-TD). Numerical accuracy of better than 0.15% is achieved when the FD-TD is used to calculate the effective dielectric constant (epsilon(reff)) of an air-filled rectangular coaxial transmission line. Data obtained for the frequency-dependent epsilon(reff) of uniform SSL's, and both scattering and equivalent circuit parameters of various series gap discontinuities, are then presented. In general, presence of sidewall mounting grooves causes a nearly frequency-independent small reduction in epsilon(reff). On the other hand, proximity effects of the housing are found to be more important. It is demonstrated that: 1) for a given housing, epsilon(reff) decreases first with an increase in strip width before increasing again when strip is wider enough to interact strongly with the sidewalls and 2) reducing sidewall spacing causes an increase in epsilon(reff). For the gap discontinuity, coupling across the gap is stronger for wider strips and/or narrower gap width. Irregular transmission behavior is also found when strip is wide enough to interact strongly with the sidewalls.
BIT-SLICED MEDIAN FILTER DESIGN BASED ON MAJORITY GATE
(1992-02-01)
LEE, CL
;
JEN, CW
;
交大名義發表
;
電控工程研究所
;
National Chiao Tung University
There are arithmetic problems for the hardware realisation of bit-level median filtering algorithms. A design of a majority gate which is composed of output-wired inverters is proposed. The area and time complexities are better than the digital and analogue designs now available. This circuit is applied to a median filter design which is based on majority selection, the computation problems are thus avoided. It is a bit-sliced architecture with constant cycle time. Window shapes can be arbitrarily changed through mask-and-set modules. A median filtering system for two-dimensional image processing is presented. A binary majority gate is also an essential element in decision-making circuitry which is applied in fault-tolerant computing systems, artificial neural networks or related applications.
NUMERICAL-ANALYSIS OF THE LOOPING EFFECT IN GAAS-MESFETS
(1992-02-01)
LO, SH
;
LEE, CP
;
電子工程學系及電子研究所
;
奈米中心
;
Department of Electronics Engineering and Institute of Electronics
The looping effect in the I(D)-V(D) characteristics of GaAs MESFET's on semi-insulating substrates has been studied using a two-dimensional numerical analysis. Both the transient and the steady-state behaviors of the looping phenomenon are simulated and discussed. Peak voltage- and frequency-dependent behaviors of the looping effect are analyzed. The I(D)-V(D) loop is due to the difference in the distribution of ionized EL2 concentration when the drain voltage rises and falls because of the trapping process of EL2's. The output conductance is also found to be frequency-dependent and is explained by the frequency-dependent modulation of the potential barrier height at the channel/substrate interface due to the drain-voltage variation.
A NOVEL TECHNIQUE FOR LOW-THRESHOLD AND HIGH-POWER INGAAS/GAAS STRAINED-LAYER 0.98-MU-M BURIED HETEROSTRUCTURE LASER FABRICATION
(1992-02-01)
LIOU, DC
;
CHIANG, WH
;
LEE, CP
;
CHANG, KH
;
LIU, DG
A novel fabrication technique has been developed for InGaAs/GaAs strained-layer buried heterostructure lasers. Dielectric masks and Zn diffusion are not required in this technique. This novel fabrication process is much easier than the conventional approach and yields excellent laser results. A low threshold of 3 mA and high-power operation for lasing wavelength of 9800 +/- 20 angstrom have been achieved with graded index separate confinement heterostructure devices using this novel technique.
A NEW VARIABLE FRACTIONAL SAMPLE DELAY FILTER WITH NONLINEAR INTERPOLATION
(1992-02-01)
LIU, GS
;
WEI, CH
;
交大名義發表
;
電控工程研究所
;
National Chiao Tung University
This paper presents a finite impulse response (FIR) filter that can synthesize any fractional sample delay by nonlinear interpolation technique. Analytically closed-form solutions for the tap weights of such an FIR filter and their frequency responses are presented.
THE NEUTRON STREAMING TERM IN A GENERALIZED COORDINATE SYSTEM
(1992-02-01)
WAY, YS
;
WAY, YM
;
電子物理學系
;
Department of Electrophysics
The vectorial form of the neutron streaming term used in the Boltzmann equation has been derived. With the aid of this vectorial form, the explicit expression of the neutron streaming term in a generalized coordinate system can readily be obtained. Examples of applications in the orthogonal curvilinear and the elliptical-elliptical toroidal coordinate systems are given for illustration.
RELAXATION BY THE HOPFIELD NEURAL NETWORK
(1992-02-01)
YU, SS
;
TSAI, WH
;
資訊工程學系
;
Department of Computer Science
The relaxation process is a useful technique for using contextual information to reduce local ambiguity and achieve global consistency in various applications. It is basically a parallel execution model, adjusting the confidence measures of involved entities based on interrelated hypotheses and confidence measures. On the other hand, the neural network is a computational model with massively parallel execution capability. The output of each neuron depends mainly on the information provided by other neurons. Therefore, there exist certain common properties in the relaxation process and the neural network technique. A mapping method that makes the Hopfield neural network perform the relaxation process is proposed. By this method, the neural network technology can be easily adapted to solve the many problems which have already been solved by the relaxation process. An advantage of this is that the relaxation process can be performed in real time since the Hopfield network can be implemented by conventional analog circuits. Experimental results are given to demonstrate the feasibility of the proposed method by performing the image thresholding operation on the proposed neural network.
A SIMULATED ANNEALING APPROACH TO THE NETWORK DESIGN PROBLEM WITH VARIATIONAL INEQUALITY CONSTRAINTS
(1992-02-01)
FRIESZ, TL
;
CHO, HJ
;
MEHTA, NJ
;
TOBIN, RL
;
ANANDALINGAM, G
The equilibrium network design problem can be formulated as a mathematical program with variational inequality constraints. We know this problem is nonconvex; hence, it is difficult to solve for a globally optimal solution. In this paper we propose a simulated annealing algorithm for the equilibrium network design problem. We demonstrate the ability of this algorithm to determine a globally optimal solution for two different networks. One of these describes an actual city in the midwestern United States.
EFFECTS OF RAPID THERMAL ANNEALING ON THE FORMATION OF SHALLOW P+N JUNCTION BY IMPLANTING BF2+ IONS INTO THIN METAL-FILMS ON SI SUBSTRATE .1. THIN TITANIUM FILMS
(1992-02-01)
JUANG, MH
;
CHENG, HC
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
High-quality Ti-silicided shallow p+ n junctions have been fabricated by implanting BF2+ ions into thin Ti films on Si substrate and subsequent silicidation/drive-in by rapid thermal annealing (RTA) or conventional furnace annealing (CFA) under proper implant and anneal conditions. For both the RTA and CFA techniques, annealing temperatures higher than 800-degrees-C degrade the junction formation because of more severe dopant confinement within the silicides and more serious diffusion of knock-on Ti into junction regions. The high-dose implant greatly enhanced the dopant activation and thus improved the junctions. The high heating rate for RTA caused an immediate formation of Ti-B compounds at high temperatures, while CFA considerably promoted the drive-in efficiency because of its low heating rates and long annealing times. Hence, CFA yielded better low-bias rectifying characteristics than RTA due to larger dopant activation. However, CFA caused much worse high reverse-bias characteristics. A rapid increase of reverse currents with bias voltage was observed for the CFA-treated samples, indicating a severe Ti penetration due to long annealing times.
INSTABILITIES IN COUPLED PHOTOREFRACTIVE RING CAVITIES AND SELF-PUMPED PHASE CONJUGATORS
(1992-02-01)
DAI, LK
;
GOU, YS
;
GU, C
;
YEH, P
;
光電工程學系
In the preceding paper, we used the mean-field approximation to formulate the time evolution equations for two coupled unidirectional photorefractive ring resonators. Linear stability analysis based upon the steady-state solutions of the mean-field equations showed that the coupled system may become unstable in the unstable off- and steady-state operation regimes. In this paper, we investigate the temporal behavior and nonlinear dynamics of the coupled system. The results indicate that the coupled system exhibits self-pulsation and chaos in the unstable off-state and steady-state operation regimes. The instability is explained in terms of the competition between the nonlinear gain and loss. Numerical results of the transient equations are presented and discussed. The coupling between unidirectional ring resonators is extended to explain the instability in self-pumped phase conjugators.
ANALYSIS OF BEAM SIZE EFFECT ON THE LASER-INDUCED FREEDERICKSZ TRANSITION AND THE DYNAMIC-RESPONSE IN NEMATIC LIQUID-CRYSTAL (5CB) FILMS WITH A FREE-SURFACE
(1992-02-01)
CHEN, SM
;
PAN, RP
;
電子物理學系
;
光電工程學系
;
Department of Electrophysics
The non-local effect caused by the finite laser beam size in the laser induced Freedericksz transition for the homeotropically aligned nematic liquid crystal film is derived analytically and calculated numerically. The influences on the transition threshold and the dynamic response are discussed. The coupling between backflow effect and the non-local effect is shown to be important for the films with a free surface by the calculated results and by comparing with the existing experimental results.
SOLVING CIRCULAR CURVE USING NEWTON-RAPHSON METHOD
(1992-02-01)
CHEN, CS
;
HWANG, LS
;
交大名義發表
;
土木工程學系
;
National Chiao Tung University
In highway, railway, canal, and pipeline locations, the horizontal curves employed at points of change in direction are arcs of circles. A circular curve can be described by seven principal elements: (1) Radius of the curve; (2) deflection angle between tangents; (3) tangent distance; (4) external distance; (5) middle ordinate; (6) long chord; and (7) length of curve. In conventional surveying, the radius of the curve is given, and the deflection angle between the tangents is measured; all other elements can be calculated. However, these two elements sometimes are unknown in the fieldwork. In this case, a new method must be considered. In this paper, a numerical solution called the Newton-Raphson's method is presented. Using this method, the radius of a curve and the deflection angle between the tangents can be calculated so long as two of the other principal elements are given. To be convenient for actual use, a computer program is included.
THE VIBRATIONAL-MODE SOFTENING IN THIN-FILM CDS
(1992-02-01)
DAI, CM
;
CHUU, DS
;
電子物理學系
;
Department of Electrophysics
Lattice softening of the CdS LO phonon mode was observed by resonant Raman spectroscopy (RRS). This phenomenon was interpreted in terms of the combined effects of grain size and low-dimensionality of the film. As the thickness of the film decreases below 800 angstrom, the influence of the free-surface on the vibrational spectrum becomes more prominent. Higher order harmonics of the LO-mode also are observed. The intensity of the Raman line is enhanced as it approaches the maximum of the fluorescence emission peak.
MODIFIED RECURSIVE LEAST-SQUARES ALGORITHM FOR PARAMETER-IDENTIFICATION
(1992-02-01)
CHEN, YM
;
WU, YC
;
電控工程研究所
;
Institute of Electrical and Control Engineering
A modified recursive least-squares algorithm is proposed for estimating the parameters of a linear system, including ill-conditioned situations. The proposed algorithm is obtained by minimizing the modified cost function in which the initial guess in the cost function of an ordinary recursive least-squares (RLS) algorithm is replaced by the latest estimate. Analysis of the convergence shows that the proposed algorithm has a better convergence rate than the ordinary RLS algorithm. The extension of the algorithm to the time-varying case has also been considered.