On Definition of Operator o for Weakly Nonlinear Circuits

For the first time, operator o appeared in the literature on weakly nonlinear circuits in a Narayanan’s paper on modelling transistor nonlinear distortion with the use of Volterra series. Its definition was restricted only to the linear part of a nonlinear circuit description. Obviously, as we show here, Narayanan’s operator o had meaning of a linear convolution integral. The extended version of this operator, which was applied to the whole nonlinear circuit representation by the Volterra series, was introduced by Meyer and Stephens in their paper on modelling nonlinear distortion in variable-capacitance diodes. We show here that its definition as well as another definition communicated to the author of this paper are faulty. We draw here attention to these facts because the faults made by Meyer and Stephens were afterwards replicated in publications of Palumbo and his coworkers on harmonic distortion calculation in integrated CMOS amplifiers, and recently in a paper about distortion analysis of parametric amplifier by H. Shrimali and S. Chatterjee. These faults are also present in some class notes for students, which are available on WWW-pages.