Parameter estimation in strongly nonlinear circuits

The combination of the generalized Volterra approach to compute the steady-state output of strongly nonlinear systems with the maximum likelihood estimator (MLE) developed by J. Schoukens et al. (see ibid., vol.IM-37, p.10-17, 1988) yields a powerful tool to estimate the parameters in strongly nonlinear circuits. As a result, it is possible to determine system characteristics that cannot be measured directly or are difficult to obtain. The latter is illustrated by means of an inverting amplifier built around an operational amplifier causing slew-induced distortion. Two different models are used to represent the operational amplifier. The first considers only one nonlinearity, namely a saturating current source characterized by two parameters, but can only describe symmetric slew-induced distortion. The other model uses two diodes and results in a six-parameter model capable of addressing the asymmetric case. By comparing the results obtained for these approaches with measurements on the actual circuit, the capability of the identification technique for strongly nonlinear systems is demonstrated.<
>