Diced Wafer Inspection via Unique Feature Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing diced wafer inspection methods face challenges in accurately inspecting wafers with non-linear spatial relationships and mechanical inaccuracies, as they require adjusting scanning patterns and aligning with global coordinates, which can be inefficient and prone to errors.
Innovation Solution
A method and system that acquire multiple images of a diced wafer using a predefined scheme, locate unique features, assign die indices, and associate them with reference dice, allowing continuous scanning without stopping to adjust patterns, and calculating coordinate transformations to align with global coordinates, enabling efficient defect detection and registration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional diced wafer inspection methods are used, then defect detection is performed, but the inspection process requires stopping to adjust scanning patterns and align with global coordinates, reducing productivity
Solution Approach 1:
The system performs preliminary actions by pre-calculating the non-linear transformation map from diced wafer coordinates to global wafer coordinates before inspection begins. This pre-computation allows the inspection system to directly use the transformation relationships without needing to stop and adjust scanning patterns during the inspection process, thereby maintaining high productivity while ensuring accurate defect detection
Solution Approach 2:
The system creates a digital copy or model of the non-linear spatial relationships through the transformation map, which represents the distorted grid of diced wafer coordinates. This digital model allows the system to virtually correct for mechanical inaccuracies and non-linear distortions without physical adjustments, enabling continuous scanning at full speed while maintaining measurement precision
2Measurement precision
If scanning patterns are adjusted to account for non-linear spatial relationships, then alignment accuracy improves, but the complexity of the inspection system increases
Solution Approach 1:
The system replaces complex mechanical adjustments and dynamic scanning pattern modifications with a computational approach. By using a pre-calculated non-linear transformation map and applying coordinate transformations during image processing, the system achieves accurate alignment without needing complex mechanical control systems or real-time scanning pattern adjustments, thereby reducing device complexity while maintaining high measurement precision
3Productivity
If continuous scanning is performed without stopping, then productivity increases, but mechanical inaccuracies and alignment errors accumulate
Solution Approach 1:
The system creates a comprehensive transformation map that captures all non-linear spatial relationships and mechanical inaccuracies in a digital model. This digital copy allows the system to continuously scan without stopping while computationally correcting for accumulated errors through coordinate transformation, thereby maintaining both high productivity and reliable alignment accuracy throughout the inspection process
Data Source
AI summary
An inspection system and a method for inspecting a diced wafer. The method includes: acquiring multiple images of multiple portions of the diced wafer according to a predefined image acquisition scheme; locating multiple unique features within the multiple images; and assigning a die index to each die of the multiple dice and associating between the multiple dice and multiple reference dice in response to locations of the multiple unique features and to at least one expected die dimension.