Mixed-signal characterization environments for analog to digital converters

With analog blocks frequently representing the biggest risk in single substrate mixed-signal designs, intellectual property (IP) libraries of fabrication-proven and fully-characterized analog cells can offer a low-risk method to decrease time-to-market (TTM), in any group where TTM is the dominant factor of success. While the development of these libraries has received significant research attention, analog and mixed-signal designs are frequently developed from scratch without much benefit from external IP. We believe that a reason for this recalcitrance has been a lack of confidence in the external IP, and specifically the limited access to characterization that has been done to describe the IP. In this paper, we develop methods to characterize IP with a focus on decreasing TTM. Further, we show how these same characterizations yield performance metrics which are useful in calibrating abstract models to shorten the simulation time for large systems, as well as providing performance specifications for cell-level functional verification. The collection of simulation test benches, interpretation of results, and control scripts is our "characterization environment." As an example, we develop a system of simulation libraries to characterize the behavior, sensitivity to parameter variation, and performance of analog library cells, which are combined for the fabrication of a delta-sigma modulator analog to digital converter (DeltaSigma ADC). Finally, we show how the characterization of libraries of analog cells is a key form of IP necessary to promote the common use of external analog IP libraries