Automatic Tuning Algorithms and Statistical Circuit Design

223 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1982.In this dissertation two topics are investigated the first of which is automatic tuning algorithms for active filters. Here the problem is that in order to meet response specifications these filters usually must be tuned or adjusted, preferably by computer automation if the production level is high. Three generalized tuning algorithms which have recently appeared in the literature are comparatively reviewed on the basis of their architecture, computational complexity, and effectiveness. Furthermore, a method is presented for the tuning resistor and frequency selection problem, a problem relevant to all three methods. Several statistical simulation examples enhance the presentation. The second topic is statistical circuit design where the emphasis is on Monte Carlo techniques for yield estimation and yield maximization. Several techniques for achieving variance reduction in the yield estimates are discussed. A quadratic approximation model is set up for the circuit and is used to provide an extrapolated yield approximation technique which is extremely effective and efficient in approximating and maximizing the yield along a search direction. Several examples demonstrate the yield maximization process.U of I OnlyRestricted to the U of I community idenfinitely during batch ingest of legacy ETD