Date of Award
Winter 2019
Project Type
Thesis
Program or Major
Electrical and Computer Engineering
Degree Name
Master of Science
First Advisor
Michael J Carter
Second Advisor
Nicholas J Kirsch
Third Advisor
Richard A Messner
Abstract
Electrical data communication links less than 10 meters long are increasing link rates at a steady pace. As link rate increases, timing tolerances become more important to ensure low bit error rates (BER). Ensuring high performance links in scenarios with large amounts of crosstalk requires the characterization of transmitter output jitter (TOJ). Some jitter models are incapable of effectively separating Gaussian random jitter from jitter caused by crosstalk. To overcome this, IEEE 802.3-2018 Clause 92 100GBASE-CR4 defines TOJ tests using the dual-Dirac (��-��) model. Additionally the 100GBASE-CR4 TOJ test definition separates out components of jitter by selecting isolated edges in a test pattern to be sampled. Applying the ��-�� model to specific edges in a test pattern is not widely present in other jitter test methodologies. In this thesis the IEEE 802.3-2018 Clause 92 TOJ tests are implemented, and issues related to measurement time, captured data size, and measurement accuracy are addressed. A set of measurements were taken of a signal generator with a set of expected worst case jitter components applied to the signal. These measurements are explored to validate the implementation and examine limitations of the test definition.
Recommended Citation
Willis, Paul Alan, "Efficient Characterization of Transmitter Output Jitter Components in 100GBASE-CR4 Ethernet" (2019). Master's Theses and Capstones. 1334.
https://scholars.unh.edu/thesis/1334
Waveform raw data used in convergence test
20190122_TOJ_histograms.zip (13871 kB)
TOJ raw data present in appendices