DocumentCode
870970
Title
Characterisation of crosstalk in terms of mean value and standard deviation [multiconductor transmission lines]
Author
Bellan, D. ; Pignari, S.A. ; Spadacini, G.
Author_Institution
Dipt. di Elettrotecnica, Politecnico di Milano, Italy
Volume
150
Issue
6
fYear
2003
Firstpage
289
Lastpage
295
Abstract
A probabilistic model is developed for the characterisation of crosstalk sensitivity to random fluctuations of conductors in a bundle cross-section. A uniform and lossless three-conductor transmission line (TL) is used for the definition of generator and victim circuits, with general loading conditions. A novel circuit model for crosstalk is developed, in which the victim circuit embeds all the interference effects due to the presence of the generator circuit. Such effects are represented by a voltage and a current lumped noise-source and a distributed-parameter passive two-port. Under the weak-coupling assumption, and by assuming random fluctuations of the generator wire in the TL cross-section, closed-form expressions for the mean value and variance of crosstalk voltages are derived, valid both at low-frequencies and in the TL standing-wave region. These closed-form crosstalk estimators extend previous results, limited to electrically short TLs. All the theoretical developments are checked via a numerical repeated-run analysis.
Keywords
coupled circuits; crosstalk; distributed parameter networks; electromagnetic compatibility; multiconductor transmission lines; statistical analysis; two-port networks; EMC; TL standing-wave region; conductor random fluctuation crosstalk sensitivity; crosstalk interference; distributed-parameter passive two-port network; electromagnetic coupling; generator circuits; multiconductor cables; multiconductor transmission lines; multiwire cables; statistical analysis; victim circuits;
fLanguage
English
Journal_Title
Science, Measurement and Technology, IEE Proceedings -
Publisher
iet
ISSN
1350-2344
Type
jour
DOI
10.1049/ip-smt:20030968
Filename
1262428
Link To Document