Search

Packaging is a well-known barrier to the advancement of microelectromechanical systems (MEMS) for RF applications. To pave the way for the removal of this barrier, we have developed a flip-chip assembly technology to transfer foundry-fabricated MEMS devices from the host silicon substrate to a ceramic substrate. Specifically, posts have been designed and fabricated to assure excellent RF performance by achieving a precise gap between the device and ceramic substrate. In addition, a novel liquid crystal polymer (LCP) encapsulation technology has been developed to protect the RF MEMS device. LCP is a good encapsulation material for nonhermetic packaging because it significantly reduces the packaging cost. We have demonstrated excellent RF performance of variable MEMS capacitors that have been flip-chip assembled and LCP encapsulated. The quality (Q) factors of such capacitors were measured to be higher than 300 at 1.0 GHz. Index Terms--Liquid crystal polymer (LCP), microelectromechanical systems (MEMS) devices, RF packaging, variable capacitor.

Neural-network computational modules have recently gained recognition as an unconventional and useful tool for RF and microwave modeling and design. Neural networks can be trained to learn the behavior of passive/active components/circuits. A trained neural network can be used for high-level design, providing fast and accurate answers to the task it has learned. Neural networks are attractive alternatives to conventional methods such as numerical modeling methods, which could be computationally expensive, or analytical methods which could be difficult to obtain for new devices, or empirical modeling solutions whose range and accuracy may be limited. This tutorial describes fundamental concepts in this emerging area aimed at teaching RF/microwave engineers what neural networks are, why they are useful, when they can be used, and how to use them. Neural-network structures and their training methods are described from the RF/microwave designer's perspective. Electromagnetics-based training for passive component models and physics-based training for active device models are illustrated. Circuit design and yield optimization using passive/active neural models are also presented. A multimedia slide presentation along with narrative audio clips is included in the electronic version of this paper. A hyperlink to the NeuroModeler demonstration software is provided to allow readers practice neural-network-based design concepts. Index Terms--Computer-aided design (CAD), design automation, modeling, neural networks, optimization, simulation.

This paper discusses three emerging trends in the area of millimeter-wave (mm-wave) electromagnetic (EM) simulation and computer-aided design (CAD). Diakoptics techniques for frequency- and time-domain EM-simulation methods are needed for mm-wave circuit design. Development of artificial neural network (ANN) models is likely to result in an efficient use of EM simulators. Possible applications of knowledge-based tools are suggested for initial stages of the design process. Index Terms - ANN modeling, diakoptics, EM simulation, knowledge-based design, mm-wave CAD.

In this paper, electromagnetically coupled (EMC) patch-antenna configurations on two-layered dielectric substrates are investigated. The radiating patch is located on the top of the upper substrate and is excited by feed circuitry underneath it on the lower substrate. The electromagnetic field problem is transformed into a network problem and a multiport network model (MNM) is developed to characterize the three-dimensional (3-D) fields associated with the radiating structure. Antenna characteristics such as input impedance and radiation pattern are computed using the MNM model and the approach is validated by comparison with experimental results. Index Terms - Microstrip antennas, networks.

Accurate and efficient electromagnetically trained artificial neural-network (EM-ANN) models have been developed for coplanar waveguide (CPW) circuit components. Modeled components include: CPW transmission lines (frequency dependent [Z.sub.o] and [[Epsilon].sub.re]), 90 [degrees] bends, short-circuit stubs, open-circuit stubs, step-in-width discontinuities, and symmetric T-junctions. These models allow for circuit design, simulation, and optimization within a commercial microwave circuit simulator environment, while providing the accuracy of electromagnetic (EM) simulation. Design and optimization of a CPW folded double-stub filter and a 50-[Omega] 3-dB power divider circuit using the developed CPW EM-ANN models are demonstrated. Index Terms - CAD, CPW circuits, CPW models, neural networks.

A novel approach for accurate and efficient modeling of monolithic microwave/millimeter wave integrated circuit (MMIC) components by using electromagnetically trained artificial neural network (EM-ANN) software modules is presented. Full-wave EM analysis is employed to characterize MMIC components. Structures for simulation are chosen using design of experiments (DOE) methodology. EM-ANN models are then trained using physical parameters as inputs and S-parameters as outputs. Once trained, the EM-ANN models are inserted into a commercial microwave circuit simulator where they provide results approaching the accuracy of the EM simulation tool used for characterization of the MMIC components without increasing the analysis time significantly. The proposed technique is capable of providing simulation models for MMIC components where models do not exist or are not accurate over the desired region of operation. The approach has been verified by developing models for microstrip vias and interconnects in dataset circuits. A new hybrid ([Delta]S) modeling approach which makes use of existing approximate models for components is introduced and shown to be a more efficient method for developing EM-ANN models. An example of using EM-ANN models to optimize the component geometry is included.

A generalized network model for asymmetrical and inhomogeneous coupled lines has been derived based on normal mode parameters. This model is useful for synthesis of single and multi-layer coupled-line circuits, such as planar baluns and directional couplers. The synthesis procedures are described and have been verified by comparing with analysis results.

Describes fundamental concepts in artificial neural networks for radio frequency (RF) and microwave design. Use of a neural-network computational modules on RF and microwave modeling; Details on the electrical behavior of passive and active components/circuits; Description of a multilayer perception neural network.

Editorial. Comments on some of the papers presented at the 8th Topical Meeting on Electrical Performance of Electronic Packaging (EPEP) held in San Diego, California, on October 25-27, 1999. Role of the EPEP in advancing engineering research; Categories of papers presented; Discussion of methods for extracting models for on-chip interconnects; Analysis of model order-reduction methods for analyzing interconnections.

Presents information as it relates to how the technical program of the 1997 IEEE Microwave Theory and Techniques-Society International Microwave Symposium conference came off. How many sessions made up the technical program; Identification of these sessions; Specification of the themes for these sessions.