High level fault modeling of analog circuits through automated model generation using Chebyshev and Newton interpolating polynomials

Abstract

With the help of automated model generation (AMG), high level modeling (HLM) of analog circuits is able to provide useful speedup and acceptable accuracy compared with standard SPICE-level circuit simulation. Unfortunately, this is not the case for high level fault modeling (HLFM) and high level fault simulation (HLFS). This is still a critical issue that industry is facing in reducing analog testing cost. We present a novel algorithm using a fusion of Chebyshev and Newton interpolating polynomials (CNIP) in nonlinear state-space (ss) termed AMG-CNIP for HLFM in analog circuits. It is written in MATLAB and the hardware description language (HDL) VHDL-AMS, respectively. The properties of AMG-CNIP are investigated by modeling nonlinear transmission line circuits using transient analysis. Results show that the AMG-CNIP models can handle both linear and nonlinear fault simulations with reasonable accuracy, and simulation speedup is achieved compared to standard SPICE-level simulations.