Dynamic PCB Testing Optimization via Yield Threshold Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The testing process for mass-produced printed circuit boards (PCBs) is excessively time-consuming and labor-intensive, leading to high costs due to the current testing methodologies.
Innovation Solution
A smart testing system that utilizes a method to adjust test items dynamically based on historical data and yield thresholds, allowing for the comparison of 'pass' yields and automatically optimizing the testing process by determining which objects need full testing, sampling testing, or re-testing, thereby reducing overall testing time and labor costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If full testing is performed on all PCBs, then testing completeness is improved, but testing time and labor cost increase
Solution Approach 1:
The patent applies partial action by performing full testing only on a subset of PCBs (those that fail the quick test or are randomly selected), while performing reduced testing on the majority that pass the quick test. This resolves the contradiction by maintaining adequate testing coverage without universally applying exhaustive testing to all units.
Solution Approach 2:
The testing process is segmented into two distinct stages: a quick test phase and a full test phase. This segmentation allows the system to rapidly filter PCBs that clearly pass, deferring comprehensive testing only to those that require it, thereby reducing overall testing time while maintaining completeness for critical cases.
2Reliability
If full testing is performed on all PCBs, then testing completeness is improved, but labor cost increases
Solution Approach 1:
By applying partial testing only where necessary (failures and random samples) rather than to all PCBs, the patent significantly reduces labor costs associated with manual inspection and testing while maintaining adequate quality assurance through targeted full testing of critical cases.
Solution Approach 2:
The labor-intensive full testing is segmented and applied only to specific subsets of PCBs, separating it from the automated quick testing phase. This segmentation reduces overall labor requirements by limiting expensive manual testing to only those units that truly need comprehensive inspection.
3Loss of time
If quick test with reduced items is performed, then testing time is reduced, but testing accuracy deteriorates
Solution Approach 1:
The patent segments the testing process into a quick test phase with reduced items for rapid screening, followed by a full test phase for comprehensive verification. This segmentation allows the quick test to efficiently filter passing units without requiring full accuracy, while the full test phase ensures testing accuracy for units that need it.
Solution Approach 2:
The quick test applies partial testing with reduced items appropriately - not too little to be useless, but not excessive full testing to all units. This balanced partial action maintains adequate accuracy for screening purposes while achieving the goal of reduced testing time for the majority of PCBs.
Data Source
AI summary
A method for testing mass-produced PCBs and other electronic components more efficiently, the method includes setting testing parameters based on historical test data and a target decision index, obtaining a first specified number of the target objects to have the full test, and calculating a first yield based on the current test result. The method determine whether the first yield is less than the first yield threshold yield, and obtaining a second specified number of the target objects from the remaining target objects to have the full test, and calculate a second yield when the first yield is larger than or equal to the first yield threshold value. The method further determine whether the second yield is less than the second yield threshold value according to a second comparing command and select some of the remaining target objects to have a sampling test.