DocumentCode
2179713
Title
Reflectionless Design for Differential-Via Transitions Using Neural Network-Based Approach
Author
Jeng, Yung-Shou ; Guo, Wei-Da ; Shiue, Guang-Hwa ; Lin, Chien-Min ; Wu, Ruey-Beei
Author_Institution
Nat.Taiwan Univ., Taipei
fYear
2007
fDate
29-31 Oct. 2007
Firstpage
143
Lastpage
146
Abstract
Based on the concept of impedance matching, this paper presents a novel design method that can efficiently achieve the least reflection noise for differentia-via transitions. The relation between the effective via-impedance and the physical dimension is well approximated by a three-layer neural network. In use of the full-wave simulation, the time-domain reflectometry waveform is also acquired to confirm the accuracy of the proposed approach.
Keywords
electromagnetic wave reflection; impedance matching; neural nets; printed circuit design; time-domain reflectometry; differential-via transition; full-wave simulation; neural network; printed circuit; reflectionless design; time-domain reflectometry waveform; Acoustic reflection; Artificial neural networks; Capacitance; Circuit simulation; Computational modeling; Geometry; Impedance; Neural networks; Semiconductor device manufacture; Semiconductor device noise;
fLanguage
English
Publisher
ieee
Conference_Titel
Electrical Performance of Electronic Packaging, 2007 IEEE
Conference_Location
Atlanta, GA
Print_ISBN
978-1-4244-0883-2
Type
conf
DOI
10.1109/EPEP.2007.4387145
Filename
4387145
Link To Document