Periodic Jitter Identification via Transition Length Chi-Square Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automatic test equipment (ATE) faces challenges in accurately identifying periodic jitter in signals, which can lead to false device failure indications and hinder efficient testing of devices under test (DUTs), especially at high data rates where direct BER measurement is impractical.
Innovation Solution
A method involving statistical analysis using chi-square testing to determine the level of periodic jitter in signal transitions, where transition lengths are analyzed to differentiate between random and periodic jitter, enabling identification and potential source isolation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional measurement techniques are used to measure jitter, then the testing process is simple, but periodic jitter cannot be accurately identified leading to false failure indications
Solution Approach 1:
The patent segments jitter analysis into two distinct components: random jitter measurement using traditional techniques and periodic jitter detection using chi-square statistical analysis. This segmentation allows each type of jitter to be measured with appropriate methods, improving overall measurement accuracy while maintaining system manageability
Solution Approach 2:
The patent introduces chi-square statistical analysis as an intermediary mechanism between raw transition length measurements and periodic jitter identification. This statistical intermediary processes the measured data to distinguish periodic patterns from random variations, enabling accurate periodic jitter detection without requiring complex dedicated hardware
2Measurement precision
If direct BER measurement is performed, then the most accurate quality metric is obtained, but the testing time becomes impractically long at high data rates
Solution Approach 1:
Instead of performing complete direct BER measurement which requires collecting all bits over extended periods, the patent applies partial action by measuring jitter characteristics (transition lengths) over a shorter interval and using statistical analysis to infer quality metrics. This provides sufficient accuracy for periodic jitter detection without the time penalty of full BER measurement
Solution Approach 2:
The patent changes the measurement parameter from direct bit error counting (BER) to transition length analysis with chi-square statistical evaluation. This parameter transformation enables faster measurement by analyzing temporal characteristics of signal transitions rather than counting errors over long periods, significantly reducing testing time while maintaining diagnostic capability
3Reliability
If periodic jitter is not identified, then the testing process remains simple, but test results become unreliable due to false failure indications
Solution Approach 1:
The patent implements feedback through chi-square statistical analysis that continuously evaluates transition length distributions and provides feedback on the presence of periodic jitter. This feedback mechanism enables the system to distinguish between random and periodic jitter components, ensuring reliable test results by identifying conditions that would otherwise cause false failure indications
Solution Approach 2:
The patent replaces complex hardware-based periodic jitter filtering mechanisms with statistical analysis software. By substituting mechanical or hardware complexity with computational statistics (chi-square analysis), the system achieves reliable periodic jitter identification without adding significant physical device complexity
Data Source
AI summary
Identifying periodic jitter in a signal includes identifying transition regions of the signal, where the transition regions correspond to regions of the signal where the signal changes between different levels, determining lengths of the transition regions; and performing a statistical analysis that is based on the lengths of the transition regions in order to obtain a value indicative of a level of periodic jitter in the signal.