Authors

S. Grivet-Talocia, I. S. Stievano, F. G. Canavero, I. A. Maio

Abstract

This paper is focused on broadband macromodeling of complex interconnect structures characterized by many ports, like connectors, packages, via arrays. To be characterized, such structures require a large set of data either from measurement or full-wave simulations. Therefore, the generation of a macromodel starting from such large dataset is a critical task when a high accuracy over an extended bandwidth is required. Another critical aspect in the macromodeling process is the enforcement of passivity. A non-passive model is hardly usable for practical analyses since it may lead to unstable behavior depending on the termination networks. This paper tries to overcome these difficulties by splitting the port responses into separate small subsets that are macromodeled independently. Each partial macromodel is constructed from transient responses using a Time-Domain Vector Fitting algorithm. Then, the partial macromodels are assembled into a global macromodel for the entire structure, which is finally tested for passivity. If non-passive, compensation is achieved by spectral perturbation of associated Hamiltonian matrices. Automatic synthesis of SPICE equivalents can finally be performed, thus obtaining circuit representations that are ready to be used for accurate and efficient system-level signal integrity assessments. Some examples illustrate the high degree of accuracy of the proposed approach.

Citation

  • Journal: EMC Europe 2004, International Symposium on Electromagnetic Compatibility
  • Year: 2004
  • Volume:
  • Issue:
  • Pages: 1–6
  • Publisher: IEEE
  • DOI: 10.23919/emc.2004.10805774

BibTeX

@inproceedings{Grivet_Talocia_2004,
  title={{A Systematic Procedure for the Macromodeling of Complex Interconnects and Packages}},
  DOI={10.23919/emc.2004.10805774},
  booktitle={{EMC Europe 2004, International Symposium on Electromagnetic Compatibility}},
  publisher={IEEE},
  author={Grivet-Talocia, S. and Stievano, I. S. and Canavero, F. G. and Maio, I. A.},
  year={2004},
  pages={1--6}
}

Download the bib file