An interval algebraic method which can be used to solve the worst-case tolerance design problems of the electroniccircuits are presented. Since it utilizes the linear relation between the variations of branch admitta...
详细信息
An interval algebraic method which can be used to solve the worst-case tolerance design problems of the electroniccircuits are presented. Since it utilizes the linear relation between the variations of branch admittance and port voltage in current, so it can avoid the critical ″sharp″ problem of the interval arithmetic and get a simpler design formula.
Proceedings includes 165 papers dealing with mm-wave IC's, microwave FET's, computer-aided design and measurements, printed circuits, FET applications, six-ports, dielectric resonators, passive components and ...
详细信息
Proceedings includes 165 papers dealing with mm-wave IC's, microwave FET's, computer-aided design and measurements, printed circuits, FET applications, six-ports, dielectric resonators, passive components and networks, solid-state circuits and devices, high-power circuits and systems, ferrite applications, mm-wave solid-state devices, GaAs monolithic circuits, microwave acoustics, phased and active array techniques, low-noise techniques, biological effects and medical application, microwave field and network theory as well as guided wave optics and signal processing. Topics covered include: waveguides, antennas, amplifiers, oscillators, microstrip devices, filters, electroniccircuits including mixers and phase shifters, telecommunication lines, semiconductor devices and diodes, radar circuits, acoustic surface wave devices, radio systems, laser applications, integrated optics and spectrum analyzers. 161 papers are abstracted and indexed separately.
An attempt is made to combine a couple of well-known concepts into a new approach to solve the biasing problem in electronic nonlinear circuitsdesign. Assume that for a particular circuit various large- and small-sig...
详细信息
An attempt is made to combine a couple of well-known concepts into a new approach to solve the biasing problem in electronic nonlinear circuitsdesign. Assume that for a particular circuit various large- and small-signal DC-requirements are specified. At the same time let there be a set of designable parameters. A description is provided of a way to obtain diagnostic information about the feasibility of the design problem as well as a procedure to compute the values of the parameter if such scheme exists.
A new algorithm for minimizing a function subject to box constraints is described. The algorithm is intended for use in a set of programs for automated and interactive design of electroniccircuits. The constrained op...
详细信息
A new algorithm for minimizing a function subject to box constraints is described. The algorithm is intended for use in a set of programs for automated and interactive design of electroniccircuits. The constrained optimization problem is treated through an active set strategy and the particular form of the constraints is exploited to simplify the manipulation of the various matrix operators. Numerical evidence is reported showing the efficiency of the algorithm in the number of function evaluations required to solve well known test problems. Finally, the user-assisted optimization of a current amplifier is presented to illustrate the capabilities of the new algorithm for interactive design.
The design of nonrecursive digital filters based on the Kaiser window function, and the alternative design based on the weighted-Chebyshev approximation due to Rabiner, McClellan and Parks are used together in an impr...
详细信息
The design of nonrecursive digital filters based on the Kaiser window function, and the alternative design based on the weighted-Chebyshev approximation due to Rabiner, McClellan and Parks are used together in an improved procedure for the design of digital differentiators. The new procedure leads to significant savings in the amount of computation, of the order of 15 to 38 percent. In addition, a method is proposed which can be used to design equiripple differentiators satisfying prescribed bandwidth and inband accuracy specifications.
A method for computing the output noise spectrum of sampled - data circuits by means of autocorrelation techniques is presented. The noise at the output of most switched capacitor circuits may be separated into two cl...
详细信息
A method for computing the output noise spectrum of sampled - data circuits by means of autocorrelation techniques is presented. The noise at the output of most switched capacitor circuits may be separated into two classes, namely noise which appears sampled and held at the output and direct wideband noise. Each one of these classes has an autocorrelation function with a specific shape. In this analysis noise contribution of the amplifier and the ″on″ resistance of the MOS switches are considered. Experimental measurements at an exemplary circuit verify the results obtained by this method. Rules for the design of low noise switched capacitor circuits are given.
Consideration is given to fault isolation of linear analog circuits taking the design tolerances on the nonfaulty elements into consideration. A two-step algorithm is proposed. First, an approximate method is used whe...
详细信息
Consideration is given to fault isolation of linear analog circuits taking the design tolerances on the nonfaulty elements into consideration. A two-step algorithm is proposed. First, an approximate method is used where a linear programming technique for isolating the elements most likely to be faulty is presented. Second, a verifying method for the faulty set is presented where the algebraic invariants associated with the faulty set are constructed. Both methods can also be used individually as independent methods for fault isolation.
A new approach to the DC optimization of electrical circuits, which avoids the usual procedure of manually repeated local optimization runs, is presented. Given a range of value for each design parameter, it returns a...
详细信息
A new approach to the DC optimization of electrical circuits, which avoids the usual procedure of manually repeated local optimization runs, is presented. Given a range of value for each design parameter, it returns a list of all the local minima found in the parameter space, and among them the global one. Its advantages are a reduced turn-around time for the circuit designer, comparing to the repeated runs procedure;and also the possibility of comparing the computed local minima, a very useful feature from the engineering point of view.
A new class of Monte Carlo (MC) methods for yield estimation and optimization is proposed and applied to practical problems of statistical design centering and tolerancing in nonconvex acceptability regions. The appro...
详细信息
A new class of Monte Carlo (MC) methods for yield estimation and optimization is proposed and applied to practical problems of statistical design centering and tolerancing in nonconvex acceptability regions. The approach is based on a point, segment, plane, etc. approximations to R//a. General formulae for estimators of yield and its derivatives are developed, based on the MC sampling in subspaces of R**n, and analytical or numerical integral calculations. Four algorithms utilizing the proposed approach were realized and tested on practical problems.
暂无评论