Fast Passivity Assessment for $S$-Parameter Rational Models Via a Half-Size Test Matrix

摘要

Rational models must be passive in order to ensure stable time domain simulations. The assessment of passivity properties is usually done via a Hamiltonian matrix that is associated with the state-space model, allowing precise characterization of passivity violations from its imaginary eigenvalues. The calculation of eigenvalues can be time consuming for large models as the matrix size is equal to twice the number of model states. In this paper, we derive for $S$-parameter models a new test matrix which is only half the size of the Hamiltonian matrix. This leads to savings in the eigenvalue computation time by a factor of nearly eight. The new test matrix takes into account that the model is symmetrical, in pole-residue form. Its application is demonstrated by three examples: a microwave filter, a package, and a synthetic model.