Efficient Metastability Characterization for Schmitt-Triggers

TL;DR

This paper introduces novel methods for accurately and efficiently characterizing the metastable behavior of Schmitt-Triggers, addressing a gap in existing research and enabling better design optimization.

Contribution

It presents new approaches for metastability characterization of Schmitt-Triggers, compares their accuracy and runtime, and demonstrates their applicability to various circuit implementations.

Findings

Some designs exhibit problematic metastable behavior.

Proposed methods are accurate and efficient.

Approaches are extendable to other circuits.

Abstract

Despite their attractiveness as metastability filters, Schmitt-Triggers can suffer from metastability themselves. Therefore, in the selection or construction of a suitable Schmitt-Trigger implementation, it is a necessity to accurately determine the metastable behavior. Only then one is able to compare different designs and thus guide proper optimizations, and only then one can assess the potential for residual metastable upsets. However, while the state of the art provides a lot of research and practical characterization approaches for flip-flops, comparatively little is known about Schmitt-Trigger characterization. Unlike the flip-flop with its single metastable point, the Schmitt-Trigger exhibits a whole range of metastable points depending on the input voltage. Thus the task of characterization gets much more challenging. In this paper we present different approaches to determine the metastable behavior of Schmitt-Triggers using novel methods and mechanisms. We compare their accuracy and runtime by applying them to three common circuit implementations. The achieved results are then used to reason about the metastable behavior of the chosen designs which turns out to be problematic in some cases. Overall the approaches proposed in this paper are generic and can be extended beyond the Schmitt-Trigger, i.e., to efficiently characterize metastable states in other circuits as well.