Lithography Failure Detection Using Frequency-Band Signal Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current failure event detection methods in lithography systems are time-consuming and inefficient, relying on reproducing issues and using time-domain signal anomalies, which fail to capture faults in specific frequency bands, leading to prolonged downtime and difficulty in identifying root causes.
Innovation Solution
A method involving signal decomposition into frequency components and evaluation against nominal system behavior to identify deviations as failure events, using filters to define frequency ranges and logic operations for real-time diagnostics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If time-domain signal analysis is used for failure detection, then the detection method is simple to implement, but the detection speed is slow and root cause identification is difficult
Solution Approach 1:
The patent segments the time-domain signal into multiple frequency components using Fourier transformation. This allows parallel analysis of different frequency bands, significantly reducing detection time while maintaining ease of implementation through standardized signal processing techniques.
Solution Approach 2:
The patent transforms the analysis from the time domain to the frequency domain, adding a dimensional perspective to signal analysis. This frequency-domain approach enables faster identification of periodic anomalies and root causes without complicating the implementation process.
2Measurement precision
If frequency-domain analysis is applied to detect failures, then detection speed and accuracy improve, but system complexity increases
Solution Approach 1:
The patent divides the frequency spectrum into multiple bands and analyzes each band separately. This segmentation improves detection accuracy by focusing on specific frequency ranges associated with different failure modes, while managing complexity through modular analysis of individual frequency bands.
Solution Approach 2:
The patent applies different analysis thresholds and methods to different frequency bands based on their characteristics. This local quality approach enhances detection precision for specific failure types while keeping the overall system manageable by tailoring complexity only where needed.
3Reliability
If comprehensive signal monitoring is performed across all system modules, then failure detection coverage is improved, but data processing load increases
Solution Approach 1:
The patent extracts only the relevant frequency components from the comprehensive signal data using band-pass filters and Fourier transformation. This extraction approach maintains high detection coverage across all modules while reducing computational load by focusing only on the frequency ranges that indicate failures.
Solution Approach 2:
The patent performs full frequency analysis only on signals that show anomaly indicators, while applying simplified monitoring to stable signals. This partial action approach ensures comprehensive coverage where needed while conserving computational energy in normal operating conditions.
Data Source
AI summary
A method for determining a failure event on a lithography system. The method includes decomposing at least one signal generated within the lithography system into a plurality of component signals, each component signal relating to a different respective frequency range; evaluating at least one of the component signals with respect to nominal lithographic system behavior; and identifying any deviation of at least one of the component signals from the nominal lithographic system behavior as a failure event.