Distortion Contribution Analysis by combining the Best Linear Approximation and noise analysis

The non-linear performance of analogue electronic circuits is crucial during the design phase, while circuit simulators only give measures about the distortion generated by the total circuit, leaving designers clueless about the source of the problem. Distortion Contribution Analysis (DCA) is a simulation-based analysis technique that determines the distortion generated in the sub-circuits and shows their contribution to the total distortion of the circuit. DCA can be used to efficiently decrease the distortion generated by a circuit, because it points the designer to the origin of the problem. Recently, a DCA based on the Best Linear Approximation (BLA) has been introduced as alternative to the Volterra-based techniques. However, a major drawback of the current implementation of the BLA-based DCA is its limitation to single-input single-output frequency response functions to model the behaviour of the sub-circuits. This approach ignores the input and output impedance of the stages, and hence introduces errors. In this paper, an extension of the BLA-based DCA is proposed which uses a MIMO port representation of the sub-circuits. Combining the port representation with a multi-port noise analysis allows the analysis of non-linear circuits without adapting the modelling of the sub-circuits.