At-Speed Scan Test Pattern Speed-Path Debug
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speed-path debug techniques for integrated circuits are time-consuming and inefficient, as they rely on functional test patterns and traditional logic fault diagnosis methods that struggle to identify all failing signal propagation paths, especially when multiple paths are involved, and often miss critical speed paths due to inaccuracy and pattern independence.
Innovation Solution
The use of two-cycle at-speed scan test patterns to identify and rank suspect speed paths by tracing signal propagation paths, employing logic simulation and unknown 'X' values to model defects, and applying pruning rules to eliminate non-critical paths, followed by ranking techniques based on structural properties and SDF timing information to prioritize paths for modification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If functional test patterns are used for speed-path debug, then the debug process can identify critical paths, but the process becomes time-consuming and requires extensive design experience and knowledge
Solution Approach 1:
The patent changes the test pattern parameters from functional patterns to at-speed scan patterns, which are specifically designed to exercise critical paths at the target operating speed. This parameter change allows automated identification of speed paths without requiring extensive design experience, thus reducing both time loss and maintaining identification accuracy
Solution Approach 2:
The patent replaces the manual, experience-based functional pattern debugging approach with an automated at-speed scan pattern-based system. This substitution eliminates the need for debug engineers to have extensive architectural knowledge and functionality understanding, significantly reducing the time required while maintaining or improving identification accuracy
2Measurement precision
If traditional SLAT diagnosis concepts are applied to at-speed defects, then single defects can be diagnosed, but multiple failing paths cannot be effectively identified
Solution Approach 1:
The patent extends the diagnosis capability from handling single defects (SLAT concept) to handling multiple simultaneous defects and failing paths. The at-speed scan pattern methodology is designed to exercise multiple critical paths concurrently, and the analysis system is adapted to identify all failing paths in a unified manner, providing both precision and versatility
Solution Approach 2:
The patent segments the diagnosis process into distinct phases: pattern application, failure detection, path identification, and ranking. This segmentation allows the system to effectively handle multiple failing paths by analyzing each path independently while maintaining the ability to diagnose single defects with high resolution when applicable
3Reliability
If the number of suspect speed paths is large, then comprehensive coverage is achieved, but debug efficiency decreases due to the large number of paths to investigate
Solution Approach 1:
The patent introduces ranking parameters based on structural properties and SDF timing information to prioritize suspect paths. By changing the selection criteria from exhaustive listing to ranked prioritization, the system maintains comprehensive coverage for reliability while improving debug efficiency by directing attention to the most critical paths first
Solution Approach 2:
The patent implements a two-phase approach: first identifying all suspect paths to ensure comprehensive coverage, then ranking and prioritizing them to focus debug efforts on the most critical subset. This partial action on the ranked list maintains reliability through complete identification while achieving productivity gains through focused investigation
Data Source
AI summary
Speed-path debug techniques based on at-speed scan test patterns. Potential speed paths are identified based upon detected at-speed scan pattern failures and unknown X-value simulation. When the number of identified speed paths is large, the suspect speed paths are ranked.