Extraction of causal time-domain network parameters from their band-limited frequency-domain counterparts using rational functions

Network parameters of a lumped element device are usually given in a limited frequency band of interest, or in an operation frequency range. To include them in time-domain simulations, they need to be converted to their causal time-domain correspondences. However, direct conversion with a simple technique such as the Fourier transform often leads to noncausal time-domain network parameters. Even when causal parameters are found, convolutions with them in the time domain are usually too computationally inefficient to carry out for a simulation with a large number of time steps. In this paper, a technique that extracts the causal time-domain parameters using rational function approximations is proposed. It results in time-domain parameters which are not only causal but also of exponential form in time. Consequently, the convolutions with them can be performed in a recursive manner. Numerical examples are provided to verify the effectiveness.

[1]  Jiazong Zhang,et al.  FDTD analysis of active circuits based on the S-parameters [microwave hybrid ICs] , 1997, Proceedings of 1997 Asia-Pacific Microwave Conference.

[2]  Peter K. F. Kuhfittig,et al.  Introduction to the Laplace transform , 1978 .

[3]  E. K. Miller,et al.  Model-based parameter estimation in electromagnetics. I. Background and theoretical development , 1998 .

[4]  S. Grivet-Talocia,et al.  Package macromodeling via time-domain vector fitting , 2003, IEEE Microwave and Wireless Components Letters.

[5]  D. Pozar Microwave Engineering , 1990 .

[6]  C. Durney,et al.  Extending the two-dimensional FDTD method to hybrid electromagnetic systems with active and passive lumped elements , 1992 .

[7]  A. Semlyen,et al.  Rational approximation of frequency domain responses by vector fitting , 1999 .

[8]  A. Semlyen,et al.  Fast and accurate switching transient calculations on transmission lines with ground return using recursive convolutions , 1975, IEEE Transactions on Power Apparatus and Systems.

[9]  Shuiping Luo,et al.  Iterative methods for extracting causal time-domain parameters , 2005 .

[10]  T. Brazil,et al.  Forcing causality on S-parameter data using the Hilbert transform , 1998 .

[11]  K. Murakami,et al.  Transient analysis of a class of mixed lumped and distributed constant circuits , 1988, 1988., IEEE International Symposium on Circuits and Systems.

[12]  Ernest S. Kuh,et al.  Transient simulation of lossy interconnects based on the recursive convolution formulation , 1992 .