Stability is one of the most critical design concerns for analog and mixed-signal integrated circuits. Loss of stability would produce performance degradation or even catastrophic failures. At the analog functional block level, this project addresses the significant challenges brought by the existence of large numbers of feedback loops constructed by design or formed through layout parasitics. New loop-based stability metrics will be developed to fill in the important gap left by the existing concepts of phase and gain margins and single-loop based stability analysis methods. Computationally efficient methods for identifying unstable loops in extracted large analog designs will be developed. Another source of stability challenge arrives as increasingly complex digital controls are integrated for purposes such as performance boost and variation tolerance. Inclusion of these controls, however, drastically complicates the understanding of design stability, which will be addressed by developing control-theoretical methods for rigorous stability analysis and new circuit design techniques to ensure stability. Significant stability design challenges also exist in networks of analog circuits that interact with each other through complex distributed loads. For such large networks, this project will develop theory and methods that support localized stability checking and design assurance. To bring the theory and methods developed in this project to practice, circuit-level design techniques that provide guaranteed stability with minimum performance degradation will be investigated. This will entail systematic tradeoff analysis that sheds light on the interactions between the proposed stability metrics and other design specifications, development of novel compensation/control schemes and circuit topologies that ensure stability with minimum overhead, and development of design methodologies that support this new design practice. Broader Impacts The focused theoretical investigation and engineering research of this project will provide excellent educational opportunities to students. Research participation from undergraduate students and students from underrepresented groups will be attracted. Undergraduate and graduate curriculum will be benefited from the expected results of this project, which will also be disseminated in the research community and industry. The impact of this work to society is broad and significant. Analog and mixed-signal circuits become increasingly ubiquitous and are employed in many systems including systems-on-chips. The design of analog and mixed-signal circuits is confronted by many challenges as design complexity grows. One of them is stability. The ability in ensuring design stability and performing stability-constrained design optimization will help build better and more robust circuits and may also create new markets of electronics. Finally, the generality of the targeted work is likely to contribute to the advancements in stability analysis and design in other science and engineering disciplines.
