The dyadic Green s function for the rectangular waveguidee filled with multisoc-tor isotropic dielectric is derived by the method of mode expansion and stattering superposition. The derived expressions can be extended easily so as to be applicable to the cases of other types of cylindrical waveguide, i.e. circular waveguide, and elliptic waveguide.

This paper presents a method to derive the dyadic Green s function of a loaded rectangular waveguide by image method. In the calculation of the DGF, it is simplified by using the integral transformation and replacing the multi-infinite summation with a single one. As an example of the DGF s application, the scattering field calculation of a metal sphere resting on the broad wall of the loaded rectangular waveguide is given by mo ney ethod. Results of calculations agree well with both data of experiments and those published.

The time domain radiation characteristics of dipoles and V antennas are obtained by solving the time domain electric field integrodifferential equation with the time-stepping approach. The effects of positions of the feeding source, the amounts of distributed loading along the antennas and the lengths and angles of V antennas on the input urrents and far field radiation characteristics are analyzed. The results obtained show that the appropriats amounts of distributed loading will yield a far field waveform that exhibits a "time limited" characted, that the deviation of the feeding source will reduce the maximum instant field intensities, and that in order to obtain the maximum instant field intensities in the certain direction, the optimum design for the parameters, such as length and angle, of V antennas is needed.

A formula of input impedance of a circular loop antenna in spherical cavity surrounding by conductive medium is obtained. The definition of radiation performance of the antenna in conductive medium is given. The formula of radiation resistance of the antenna is derived. The results of numerical calculation show that the two formulas converge very fast, so it is feasibe to take the first term or the sum of first two terms only. The relationship between above mentioned characteristics and the parameters, such the radius of spherical cavity and others, is discussed.

By means of logic symmetric relation, single neighbouring logic path for N dimension Boolean ordered set is solved. A new method of determining any logic neighbouring subset in limited dimensions is given. Its results are intuitional and realizable for computer.

A new method for analysing linear feedback networks is derived by determining the loop impedance matrix and generalized cofactor. This method is used to study deeply and tystematically network functions, sensitivities, return difference and null return difference, as well as return difference matrix and null return difference matrix. A set of formula with strong regularity which are suitable for CAA is obtained. The method can be used to analyse non-planar networks. It is more suitable for networks composed of RLCM and current controlled voltage sources which is excited by independent voltage sources.

In order to obtain uniform exrnosure in variable shaped electron beam lithography, the beam current density and edge resolution on the target must not change for different spot shapes and sizes. The key to the goal is the anpropriate design of shaping deflectors. A linear and rotation compensation approach is presented. Values of linear and rotation compensation factors versus the distances between electron source image and centers of deflectors are measured on an experimental electron beam column with variable spot shaping. The experimental results are in good agreement with the calculated ones.

The analysis method used in the previous paper by the authors is extended to the magnetic deflectors with ferromagnetic shield. The shape of the deflector winding may be arbitrary. It may be radial or non-radial, saddle, toroidal or fan-shape, etc.. A set of new expressions for the deflection field is suggested for the case where the windings are spaced by a gap from the shield. The integral equations are solved by Gauss-Chebyshev quadrature. The computation results are checked with the values measured from a deflector of CRT COTY GE-14''.

Physio-chemical reactions in emissive materials of the barium scandate dispenser cathode during the sintering process are investigated by X ray analysis with the aid of an X-ray diffractometer using CuKα radiation. The emissive materials of barium scandate are the compounds resulting from the reaction and sintering of BaO-CaO-Al₂O₃-Sc₂O₃ with a specified proportion. During the course of the investigation, it became clear that the formation of different compounds is very dependent on the firing temperature.The results of experiments indicate that the phases present at 1000--1600℃ are the compounds of Ba-Al-O system, Ba-Sc-O system and CaO; at 1300--1500℃ a solid solution of Ba-al-Sc-O system is found, which is formed from Ba-Al-O and Ba-Sc-O system by their mutual solubility and belongs to a quasistable state of thermodynamics. the solid solution is first formed to be an orthorhombic unit cell for about 5BaO-Al₂O₃-Sc₂O₃ with a=9.725(2)Å, b=8.698(3)Å, c=6.152(3)Å, and finally a tetragonal unit cell for about 4BaO-Al₂O₃-0.5Sc₂O₃ with a=14.4996(19)Å, c=5.0265(8)Å, and CaO is in the form of free state.

A quasi-direct solution (QDS) for the finite difference equations has been developed, which is basel on the sparse and pivot-dominant properties of the matrix and is the combination of the direct method and the iterative method. The principle and the features of QDS method have been discussed. A comparison has been made with the successive over-relaxation (SOR) method. It is shown that the QDS method requires less computer memory than the direct method, covnerges faster than the SOR method and is suitable for calculating the finite difference equations of electromagnetic and magnetostatic problems.

It is of both theoretical and practical importance to reduce the storage and CPU time of moment methods by utilizing the geometrical and physical features of the scatterer. An unified approach based on the group theory was presented to deal with the EM scattering from symmetric and anti-symmetric structures.

A new numerical method for obtaining the input reflection coefficient circle of a linear reciprocal two-port network is presented. Its advantages are that the principle is simple and the operation time is short. Its calculated results are nearly the same as those given by the Ref. [2] and [4] .

A new simple method in which arbitrary transfer function matrix is realized with active multiterminal port network instead of computer simulating is proposed. The active mul-titerminal port network is carried on through state feedback and state cross output for RC ladder network. The main process is to determine coefficients of each adder. Finally, an example is given for illustration.

A K-line location algorithm for buillding block cells in LSI/VLSI is presented. When the relative positions of rectangular cells are given, there are 2ⁿ states according to thetwo orientations of a cell. It is proved that to find the optimum solution from 2ⁿ states canbe reduced to calculate the n states in k-line algorithm. So the algorithm is very effectiveand can be used with association for cluster method in BBL placement. Under certain conditions, this method can also be used to pesudo BBL placement directly.

The response of the majority carrier traps in the semiconductors p⁺-n junction is analysed in detail under the low-injecting conditions. It points out that the effects of the majority carrier injecting and captured by the traps may happen if the forward bias voltage is larger enough but limited within the low-injection range. Yhen the reverse bias voltage is smaller so that the response of the tail region is not neglected than of the whole response region, then the effect of the minority carriers injecting and captured will be observable in the DLTS. It can be used to measure the C_p/C_n.

The contact characteristics and the interdiffusion of p-InP/Ag-Zn/Mn system are investigated by four-probe method and Auger electroscopy. The surface morphology of the heat treatment sample is observed by SEM. The specific contact resistance (e_c) which is function of the alloy temperature and time is also given. The (e_c≈(2-4)×10^-4Ω·cm at 400. The Ag-Zn/Mn system already is used in the InGaAsP/InP Edgeemitting LEDS, R_s≈4-6Ω. The results shown that the Ag-Mn material is stable and reproducible.

The principle of the spin polarized effect of photoelectrons emitted from NEA GaAs surface which is irradiated by circularly polarized light is described in detail. The preparation of NEA GaAs surface and apparatus used also are mentioned. A photoelectron beam with sensitivity of 8 μA/mW and polarization of above 35% is obtained on the NEA GaAs (100) surface in the activation with caesium and oxygen. It is found that the first maximum emission occurs as the GaAs surface is caesiated in a monolayer of 50% to 60% and unstable emission is caused by desorption of caesium.