Introduction to numerical analysis: 2nd edition
Introduction to numerical analysis: 2nd edition
System-on-Chip Test Architectures: Nanometer Design for Testability
System-on-Chip Test Architectures: Nanometer Design for Testability
| Hi-index | 0.00 |
Monte Carlo methods of statistical analysis are applied to the problems of transistor design and optimization. The experimental tolerances associated with any diffusion process are shown to represent an important factor in the initial design of diffused junction transistors. Many transistor parameters exhibit a substantial degree of sensitivity to small variations in the diffusion process. This is confirmed by a comparison between the theoretical and experimental open-base breakdown voltage, and current gain, for a large number of devices. It is therefore proposed that the design of a transistor be based upon attaining a specified set of electrical characteristics when the device is assumed to be fabricated by a non-ideal diffusion process. An electronic computer has been used in an investigation of the foregoing problem. The investigation shows further that a margin-of-safety must be designed into each electrical parameter of a transistor to assure that the resulting device satisfies a given set of design specifications, even though this margin-of-safety may differ for each parameter. In this paper examples are presented to illustrate the theoretical trade-off between several opposing transistor parameters that exhibit a substantial degree of variability due to a non-ideal diffusion process.