Email alert
1990 Vol. 12, No. 1
Display Method:
1990, 12(1): 1-6.
Abstract:
A simple and practical rectangular transform and the vector quantization technique are described. A method for image data compression by combining the rectangular transform with vector quantization technique is presented. By using this method, the dimension of vector coding and the size of the corresponding codebook are reduced; the vector encoding process is picked up; and also the coding rate is reduced. The results of simulation show that this coding method is very effective in the field of image coding.
A simple and practical rectangular transform and the vector quantization technique are described. A method for image data compression by combining the rectangular transform with vector quantization technique is presented. By using this method, the dimension of vector coding and the size of the corresponding codebook are reduced; the vector encoding process is picked up; and also the coding rate is reduced. The results of simulation show that this coding method is very effective in the field of image coding.
1990, 12(1): 7-13.
Abstract:
Interval and recency rank coding, which are invented by Elias, are good adaptive source coding schemes for independent source. The upper bound of coding efficiency of the two schemes is shown to be that of probability rank coding, and the concept of interval and recency rank coding is extended to relative sources.
Interval and recency rank coding, which are invented by Elias, are good adaptive source coding schemes for independent source. The upper bound of coding efficiency of the two schemes is shown to be that of probability rank coding, and the concept of interval and recency rank coding is extended to relative sources.
1990, 12(1): 14-21.
Abstract:
It is well known that 1IR filters require very few computations, comonred with FIR filters, in order to meet stringent magnitude specifications when the phase distortion can be tolerated. An approximately linear phase, however, can be also obtained with the IIR filter by making use of a technique without increasing complexity. Based on a certain number of attenuation zeros in the pass band, a new approach is developed for the design of polyphase wave digital filters with exact magnitude responses and Chebyshev approximation of the desired phase responses. The minimum number of attenuation zeros is estimated, and two examples are given.
It is well known that 1IR filters require very few computations, comonred with FIR filters, in order to meet stringent magnitude specifications when the phase distortion can be tolerated. An approximately linear phase, however, can be also obtained with the IIR filter by making use of a technique without increasing complexity. Based on a certain number of attenuation zeros in the pass band, a new approach is developed for the design of polyphase wave digital filters with exact magnitude responses and Chebyshev approximation of the desired phase responses. The minimum number of attenuation zeros is estimated, and two examples are given.
1990, 12(1): 22-31.
Abstract:
An automatic tuning low frequency LC low-pass notch filter is developed. The generalized immittance converter (GIC) is used as a tunable inductor.The notch frequency of the filter can track the control frequency (the frequency of power supply or interference voltage) automatically. This filter can be used to eliminate the power supply noise and to prevent the preamplifier of measurement or data aquisition system from saturating caused by the power supply interference. Experimental results show that the frequency tracking error is less that 0.5Hz in the range from 45Hz to 55Hz, and the notch depth of the filter, i.e. series mode rejection ratio, is 30-40 dB.
An automatic tuning low frequency LC low-pass notch filter is developed. The generalized immittance converter (GIC) is used as a tunable inductor.The notch frequency of the filter can track the control frequency (the frequency of power supply or interference voltage) automatically. This filter can be used to eliminate the power supply noise and to prevent the preamplifier of measurement or data aquisition system from saturating caused by the power supply interference. Experimental results show that the frequency tracking error is less that 0.5Hz in the range from 45Hz to 55Hz, and the notch depth of the filter, i.e. series mode rejection ratio, is 30-40 dB.
1990, 12(1): 32-37.
Abstract:
A uniform GTD formula for the TM plane wave diffraction by a perfect conducting wedge loaled with an impedance surface on the illuminated side is given. This formula is used to compute the back-scattered RCS of some rectangular flat plates coated with lossy dielectric. The results computed here are in fair agreement with the experimental values.
A uniform GTD formula for the TM plane wave diffraction by a perfect conducting wedge loaled with an impedance surface on the illuminated side is given. This formula is used to compute the back-scattered RCS of some rectangular flat plates coated with lossy dielectric. The results computed here are in fair agreement with the experimental values.
1990, 12(1): 38-46.
Abstract:
By searching the inverse transformation of the conformal mapping combined with the method of the graphical approximation, a more effective approach is presented for the analysis of many transmission lines. This method is expounded with an example of the coaxial system consisting of a N-regular polygon which is concentric with a circle.
By searching the inverse transformation of the conformal mapping combined with the method of the graphical approximation, a more effective approach is presented for the analysis of many transmission lines. This method is expounded with an example of the coaxial system consisting of a N-regular polygon which is concentric with a circle.
1990, 12(1): 47-56.
Abstract:
Analytical expressions are given for computing the magnetic potential and characteristic functions produced by non-radial deflectors. These expressions are useful for designing the deflectors such as used in the color picture tubes and the electron beam lithography system. The computing results are in agreement with the measured values.
Analytical expressions are given for computing the magnetic potential and characteristic functions produced by non-radial deflectors. These expressions are useful for designing the deflectors such as used in the color picture tubes and the electron beam lithography system. The computing results are in agreement with the measured values.
1990, 12(1): 57-64.
Abstract:
A new phenomenological model for fitting the illuminated I-V curves of a-Si:H PIN solar cells is proposed, in which the relation between photocurrent and applied bias voltage is used. The I-V curve of the a-Si:H PIN solar cells were measured and analyzed by IBM personal computer and the Marquardt mathematic method. The experiments show that the fitting results were excellent. The fitting error was less than one of other models and the model parameters could be modified automatically, so that reproduction and reliability of experiment data were good.
A new phenomenological model for fitting the illuminated I-V curves of a-Si:H PIN solar cells is proposed, in which the relation between photocurrent and applied bias voltage is used. The I-V curve of the a-Si:H PIN solar cells were measured and analyzed by IBM personal computer and the Marquardt mathematic method. The experiments show that the fitting results were excellent. The fitting error was less than one of other models and the model parameters could be modified automatically, so that reproduction and reliability of experiment data were good.
1990, 12(1): 65-68.
Abstract:
A sixteen tree method of data compressing of bilevel image is described. This method has high efficiency, and no information loss during compression; and it is easy to realize.
A sixteen tree method of data compressing of bilevel image is described. This method has high efficiency, and no information loss during compression; and it is easy to realize.
1990, 12(1): 69-73.
Abstract:
An efficient method for the optimization of linear network is presented. The computation cost in circuit optimization mainly depends on the simulation of network. In general, the simulation of a linear network needs to solve high dimension linear equations. An important characteristic in circuit optimization is that the number of independently tunable parameters is small. In term of the property of linear networks, the circuit is described by a multiport network in present method, and the hybrid matrix is established. The dimension of the equations to be soloved is the same as the number of optimization parameters in objective function evalutions, so this method is very efficient.
An efficient method for the optimization of linear network is presented. The computation cost in circuit optimization mainly depends on the simulation of network. In general, the simulation of a linear network needs to solve high dimension linear equations. An important characteristic in circuit optimization is that the number of independently tunable parameters is small. In term of the property of linear networks, the circuit is described by a multiport network in present method, and the hybrid matrix is established. The dimension of the equations to be soloved is the same as the number of optimization parameters in objective function evalutions, so this method is very efficient.
1990, 12(1): 74-79.
Abstract:
Based on the second kind of Green s identity, a boundary integral equation for arbitrary cross-section waveguide is transformed to a system of linear homogeneous algebraic tquations by means of expansion of boundary bases and using the eigenfuncuons of a fictitious regular boundary as weighting functions, which corresponds to less algebraic equations than BEM and simpler coefficients than the modified BEM. The numerical results for some typical metallic waverguides are given by using the method of eigen-weighted bounlary integral equation, and they are exact enough with fast convergence.
Based on the second kind of Green s identity, a boundary integral equation for arbitrary cross-section waveguide is transformed to a system of linear homogeneous algebraic tquations by means of expansion of boundary bases and using the eigenfuncuons of a fictitious regular boundary as weighting functions, which corresponds to less algebraic equations than BEM and simpler coefficients than the modified BEM. The numerical results for some typical metallic waverguides are given by using the method of eigen-weighted bounlary integral equation, and they are exact enough with fast convergence.
1990, 12(1): 80-82.
Abstract:
A frequency tripler covering 210-270 GHz frequency band with a peak efficiency of 5.8% is lescribed. The maximum output power occurs at input power of 30-50 mW. The frequency tripler consists of an input waveguide WR-12, an output waveguide WR-4 and a coxial low pass filter between them.
A frequency tripler covering 210-270 GHz frequency band with a peak efficiency of 5.8% is lescribed. The maximum output power occurs at input power of 30-50 mW. The frequency tripler consists of an input waveguide WR-12, an output waveguide WR-4 and a coxial low pass filter between them.
1990, 12(1): 83-88.
Abstract:
A new method of analysing the high-frequency behaviour of the deflection system of oscillotrons is described. By analysing the tr ansit-time-different model of the deflection system of oscillotrons under unit-step function voltage, either the sweep trace of the electron beam or calculating formulae of the rise time can be derived.
A new method of analysing the high-frequency behaviour of the deflection system of oscillotrons is described. By analysing the tr ansit-time-different model of the deflection system of oscillotrons under unit-step function voltage, either the sweep trace of the electron beam or calculating formulae of the rise time can be derived.
1990, 12(1): 89-92.
Abstract:
The expression for the optimum output electrode depth of MCP is given. At optmum operation of MCP, the energy distribution curve of the output electron (EDOE) only shows a low energy sharp main peak, its FWHM is reduced to lower limit, and MCP can operate up to its upper spatial resolution limit. The properties of the EDOE from MCP are explained successfully.
The expression for the optimum output electrode depth of MCP is given. At optmum operation of MCP, the energy distribution curve of the output electron (EDOE) only shows a low energy sharp main peak, its FWHM is reduced to lower limit, and MCP can operate up to its upper spatial resolution limit. The properties of the EDOE from MCP are explained successfully.
1990, 12(1): 93-99.
Abstract:
Fine low dimensional modulated fringes (FLDMF) by named in GaAs/ AlxGa1-xAs multilayer heterostructures using transmission electron microscopy (TEM) have been observed. Philips EM400ST and JEOL EM4000EX were used in this work. The FLDMF with equal width were discovered within the buffer layer and also within the two adjacent superlattice layers of GaAs/AlGaAs The width with 9.1 is GaAs fringes, 12 is AlGaAs fringes. Some results from the density analysis of the fringes using microdensitometer are presented. Otherwise lattice image of GaAs/AlxGa1-xAs multilayer heterostructures and analytic results of composition in GaAs/AlxGa1-xAs multilayer heterostructures by using the techniques of energy dispersive X-ray analysis (EDX) have been given.
Fine low dimensional modulated fringes (FLDMF) by named in GaAs/ AlxGa1-xAs multilayer heterostructures using transmission electron microscopy (TEM) have been observed. Philips EM400ST and JEOL EM4000EX were used in this work. The FLDMF with equal width were discovered within the buffer layer and also within the two adjacent superlattice layers of GaAs/AlGaAs The width with 9.1 is GaAs fringes, 12 is AlGaAs fringes. Some results from the density analysis of the fringes using microdensitometer are presented. Otherwise lattice image of GaAs/AlxGa1-xAs multilayer heterostructures and analytic results of composition in GaAs/AlxGa1-xAs multilayer heterostructures by using the techniques of energy dispersive X-ray analysis (EDX) have been given.
1990, 12(1): 100-102.
Abstract:
Depletion Model Selective doped heterojunction transistor is designed and fabricated. Epitaxial modulation loping material were grown by a home-made MBE system. The gate length and width for low noise depletion devices were 1.2-1.5m and 2180m respecipely. The electron mobility of the material is typically 6500cm2/v.s at 300K and 75000 cm2/v.s at 77 K. The sheet electron concentration n, is 91011 cm-2. The transconductance of the depletion Mode levices is 100-130 ms/mm at room temperature. However at low temperature the transconductance is 200 ms/mm.Its noise figure is above 2-3dB.
Depletion Model Selective doped heterojunction transistor is designed and fabricated. Epitaxial modulation loping material were grown by a home-made MBE system. The gate length and width for low noise depletion devices were 1.2-1.5m and 2180m respecipely. The electron mobility of the material is typically 6500cm2/v.s at 300K and 75000 cm2/v.s at 77 K. The sheet electron concentration n, is 91011 cm-2. The transconductance of the depletion Mode levices is 100-130 ms/mm at room temperature. However at low temperature the transconductance is 200 ms/mm.Its noise figure is above 2-3dB.
1990, 12(1): 103-108.
Abstract:
The investigations of carbon films which are coated on Mo and Hf by using RF sputtering method are presented, including the preparation of carbon films, analysis of the composition of them, their secondary emission and thermal radiation properties and mass analysis of gas released from carbon coated Mo and Hf when heated in vacuum. Main results are: (1) carbon coated Mo can absorb oxygen easily as compared with carbon coated Hf; (2) the maximum secondary emission coefficient of carbon coated Mo or Hf is about 0.7; (3) the carbon coated on Mo is used up at 930℃ or so, but the carbon coated on Hf makes littlechange until 1050℃; (4) considerable amount of CH4 is formed as carbon coated Mo isheated in vacuum.
The investigations of carbon films which are coated on Mo and Hf by using RF sputtering method are presented, including the preparation of carbon films, analysis of the composition of them, their secondary emission and thermal radiation properties and mass analysis of gas released from carbon coated Mo and Hf when heated in vacuum. Main results are: (1) carbon coated Mo can absorb oxygen easily as compared with carbon coated Hf; (2) the maximum secondary emission coefficient of carbon coated Mo or Hf is about 0.7; (3) the carbon coated on Mo is used up at 930℃ or so, but the carbon coated on Hf makes littlechange until 1050℃; (4) considerable amount of CH4 is formed as carbon coated Mo isheated in vacuum.
1990, 12(1): 109-112.
Abstract:
A metal-oxide secondary emission film can be deposited on glass, ceramic or semiconductor substrates by thermal decomposition of an organo-metal alkoxide. For instance, MgO (or Al2O3) secondary emission film can be obtained from magnesium (or Aluminium) ethoxide thermal decomposition. In the present method, Al2O3: Mo secondary emission film has been made from thermal decomposition of home-made Aluminium ethoxide and Molybdenum pentachloride. The depositing conditions are T = 450℃ and t=12min for the about 1000 thickness of film to get. The maximum secondary emission coefficient and resistivity of the film are max=3.1, =107-109, cm respectively.
A metal-oxide secondary emission film can be deposited on glass, ceramic or semiconductor substrates by thermal decomposition of an organo-metal alkoxide. For instance, MgO (or Al2O3) secondary emission film can be obtained from magnesium (or Aluminium) ethoxide thermal decomposition. In the present method, Al2O3: Mo secondary emission film has been made from thermal decomposition of home-made Aluminium ethoxide and Molybdenum pentachloride. The depositing conditions are T = 450℃ and t=12min for the about 1000 thickness of film to get. The maximum secondary emission coefficient and resistivity of the film are max=3.1, =107-109, cm respectively.
