This thesis presents a procedure developed to study the effects of gain bandwidth on the magnitude performance of Leapfrog Filter Realizations. This procedure relies heavily on software developed for IBM compatible computers. Specific examples are carried out for second, third, and fourth order filter realizations. While this thesis deals solely with the magnitude characteristic of Leapfrog Filter realizations, the procedure itself is readily adapted to other applications such as magnitude, phase, or time domain performance of many different circuits.

Electro-Static Discharge (ESD) problems in the semiconductor industry are exacerbated by the lack of information about the magnitudes of these events and their locations. This thesis describes an integrated device, similar in structure to an EEP-ROM cell, capable of measuring the magnitudes and polarities of electrostatic discharges. A quantitative formulation of the transfer characteristic of the device as a function of the design parameters and the subsequent layout, manufacturing and evaluation of this detector was obtained. A chip was fabricated which included experiments to test the presented mathematical model and its extensive range of sensitivity, as well as several structures to evaluate the use of this detector in industrial environments. Extensive use of computers helped in the process design and in the simulation of the non-linear model for the operation.

We study the problem of acoustic wave propagation in a cylindrical borehole possessing a finite number of transverse discontinuities. We model the field behavior through Green's function techniques. We formulate an integral equation whose solution will enable us to solve for the acoustic field everywhere within our structure. We investigate asymptotic forms to speed the numerical convergence of our solution. To solve the integral equation we employ both the method of moments and the low frequency approximation. We study the reflection coefficient in the time and frequency domains. Finally after presenting solutions for the one and two fracture case, we generalize our analysis for many fractures.

The effects of ionizing radiation on the breakdown voltage of p-channel power MOSFETs were examined through two-dimensional simulation. The response of a reverse-biased n+-p junction to positive oxide-trapped charge, Not, is examined in detail, and analytical expressions for its characteristics are derived. These results provide insight into the breakdown performance of p-channel power MOSFETs in ionizing radiation environments, whose performance was found to be very different from corresponding n-channel power MOSFETs. Insights gained through analysis of p-channel devices suggest a termination/isolation scheme, the VLD-FRR, that will enhance p-channel device reliability in radiation environments. Two introductory chapters, which also serve as literature reviews, are provided. The buildup of Not in thick oxides and breakdown voltage of the power DMOS transistor are both reviewed, with attention focused on p-channel devices in both cases. Finally, suggestions for future work are given.

Power Metal Oxide Semiconductor Field-Effect Transistors (MOSFETs) are widely used in power supplies for spaceborne electronic systems. Spacecraft are routinely subjected to ionizing radiation. Ionizing-radiation-induced charges alter the operation of the MOSFET. The mid-gap technique allows separation of the parameter shifts into components due to oxide trapped charge and interface trapped charge. However, non-linearity in the subthreshold curve results in an approximately 20% underestimation of the interface component of the threshold shift. A new characterization technique is presented based on the relation between gate charge and gate-source voltage. The results obtained using the gate-charge method agree with theoretical calculations.

A finite-element formulation for the analysis of shielded lossy planar transmission lines is presented. The formulation determines the propagation characteristics of a planar transmission line and avoids spurious non-physical modes with non-zero divergence by requiring the field to be divergenceless. Singularities in the field at perfectly conducting corners are accounted for by using singular basis functions. In addition, a set of entire-domain basis functions are developed to reduce the number of unknowns. A quasi-TEM formulation is also presented to provide a comparison with the full-wave analysis. Numerical results are provided to characterize the propagation characteristics of a microstrip transmission line. In particular, their dependence on the thickness of the microstrip and losses in the substrate is examined. In addition, the propagation characteristics of symmetric coupled microstrip transmission lines of finite thickness and the use of artificial magnetic side walls with stripline geometries are analyzed.

The purpose of this thesis was to investigate the effects of pre- and post-oxidation RCA cleaning on thin gate oxides through the study of breakdown voltages at several current levels. The RCA Clean consisted of SC1 (NH₄OH in H₂O₂), followed by an optional dilute HF dip, followed by SC2 (HCl in H₂O₂). SAS analysis of the experiments revealed the following results: an HF dip was not found to affect the performance of the pre-oxidation clean, SC2 was harmful to the oxide when performed after oxide growth, and the role of solution age was not clear. In addition, an interaction was found between SC1 and SC2 performed after oxide growth, with SC1 partially alleviating the harmful effects of SC2 when performed prior to SC2. This and other data led to the conclusion that SC2 was selectively etching the oxide.

This thesis presents a new technique, the don't-care propagation method, for logic design verification and functional error location in which a gate-level implementation of a circuit is compared with a functional-level specification. In this method, test pattern set which is generated to detect single stuck-line faults in the gate-level implementation, are used to compare the gate-level implementation with the functional-level specification. In the presence of logic design errors, such a test set will produce responses in the implementation that disagree with the responses in the specification. In the verification phase of the design of logic circuits using the top-down approach, it is necessary not only to detect but also to locate the source of any inconsistency that may exit between the functional-level specification and the gate-level implementation. The proposed technique determines the region which contains the function error. This method has very high resolution that the region usually contains a single gate or a line and therefore, reduces the time required for debugging by the designers.

This thesis describes an implementation of real-time simulation and control in the DEVS-Scheme environment. The plant is described by discrete event models developed within the event-based control paradigm. A model of the controller is employed to validate its design against a model of the plant. The same model is then migrated over to actual operation by interfacing it to a programmable sensor/actuator interface unit. A system entity structure is employed to generate both the simulation and execution versions of the controller. This methodology is supported by extensions to the DEVS-Scheme simulation environment which facilitate its use for real-time control. As an example, an intelligent controller is developed to control temperature and pressure of an oxygen production prototype system which converts carbon dioxide to oxygen. Such a system is eventually intended to operate autonomously on Mars.

The modules for processing of interconnection specifications which are provided in an input file of circuit analyzer were developed. A software package for creation of appropriate computer models on the basis of topology and library of network components was also constructed and implemented using SPICE program. A computer program, called LSPICE, has been developed by combining SPICE numerical techniques and the modal analysis of coupled transmission lines. This program can be used for simulating the transient response of networks containing coupled lossless transmission lines. Several example networks have been simulated using this program. The results have been compared with those generated by the circuit simulator program called UANTL. The LSPICE has several advantages over UANTL and SPICE in simulating the transient response of network containing coupled lossless transmission lines. A description of LSPICE and a summary of the results of numerical experiments are included.