Title :
Improving the efficiency of passivity compensation schemes via adaptive sampling
Author :
Grivet-Talocia, S.
Author_Institution :
Dip. Elettronica, Politecnico di Torino, Italy
Abstract :
A new algorithm for the enforcement of passivity for large lumped macromodels is presented. An accuracy-controlled adaptive frequency sampling combined with an iterative multishift Arnoldi scheme leads to a fast computation of the imaginary Hamiltonian eigenvalues, which are perturbed until passivity is reached. Up to two orders of magnitude speedup factors are achieved with respect to previous formulations.
Keywords :
eigenvalues and eigenfunctions; integrated circuit modelling; iterative methods; accuracy-controlled sampling; adaptive frequency sampling; adaptive sampling; imaginary Hamiltonian eigenvalues; iterative multishift Arnoldi scheme; lumped macromodels; passivity compensation; Computational efficiency; Eigenvalues and eigenfunctions; Frequency; Impedance; Information systems; Iterative algorithms; Iterative methods; Sampling methods; Scattering; Signal design;
Conference_Titel :
Electrical Performance of Electronic Packaging, 2005. IEEE 14th Topical Meeting on
Print_ISBN :
0-7803-9220-5
DOI :
10.1109/EPEP.2005.1563745
