An approximate multiconductor transmission-line (MTL) model is developed for predicting the radiated susceptibility of bundles composed of twisted-wire pairs (TWPs) running above a ground plane, and illuminated by a nonuniform electromagnetic field. Through the use of averaged per-unit-length parameters and sampling of the incident electromagnetic field, the nonuniform MTL is modeled as an equivalent uniform MTL including common-mode (CM) voltage sources accounting for fieldto-wire coupling effects. The proposed model is able to predict CM currents/voltages induced in the terminal loads of every TWP in the bundle. In particular, it is shown that the CM noise significantly differs depending on the position of the considered TWP in the bundle cross section. Comparison of results versus full-wave simulations validates accuracy and computational efficiency of the proposed model.Index Terms-Common mode (CM) and differential mode (DM), field-to-wire coupling, multiconductor transmission-line (MTL) model, twisted-wire pair (TWP).

No abstract

magnetic modeling still represents a challenging problem as concerns prediction of crosstalk and field-to-wire coupling. Indeed, deterministic modeling based on the representation of specific and arbitrary cable geometries would be an exercise in futility, due to the high sensitivity of the induced noise to different random configurations of the bundle. Consequently, statistical models are more suited, since they are based on the description of model parameters and results in terms of random variables and their moments (e.g., mean, standard deviation, etc.) [1]-[3]. Though the complexity of the involved electromagnetic problem prevents analytical closed-form solutions, the numerical Monte Carlo method (that is, repeated-run simulations carried out on several random samples of the bundle) is a common viable approach. Particularly, repeated-run analysis requires two fundamental tools: 1) a method to generate a physically sound geometry of random-bundle samples; 2) a fast solution method, so to optimize the computational burden associated with several simulations [4]. These aspects are here treated with a specific Manuscript

No abstract

No abstract