Functional-level analogue macromodelling with piecewise linear signals

An analogue network macromodelling technique oriented towards functional-level simulation of both analogue and mixed analogue-digital networks is presented. The signals are assumed to be piecewise linear (PWL) waveforms. A class of nonlinear inertial building blocks is introduced for modelling. The proposed macromodels are accurate in their timing behaviour and computationally efficient, since an explicit algorithm to obtain the waveforms is used based on a PWL approximation of original smooth time responses. Practical macromodels of particular functional analogue units, such as an amplifier or a voltage comparator, are derived in detail and their performance is compared with SPICE estimates. A macrosimulation example of an A/D network is also included