Vision Alignment for Semiconductor Device Testing Handlers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional mechanical alignment systems are not suitable for precise alignment of semiconductor devices, especially with finer pitches, and increasing the number of actuators to address this issue is costly and inefficient for testing multiple devices simultaneously.
Innovation Solution
A vision alignment system using a first camera to view the contactor position, a pick-and-place handler, a second camera to view the device position, alignment plates with guiding inserts, and a set of actuators to correct position offsets, controlled by a processor that combines and processes image results to align multiple devices effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional mechanical alignment systems are used, then alignment can be achieved within certain manufacturing ranges, but precision is insufficient for finer pitches
Solution Approach 1:
The patent replaces conventional mechanical alignment systems with a vision-based alignment system. Instead of relying on mechanical actuators and mechanical alignment features, the system uses cameras to capture images of alignment marks, processes these images to determine device positions, and uses this information to control the placement of devices. This substitution enables higher precision alignment beyond the capabilities of mechanical systems while maintaining manageable system complexity through software-based solutions.
2Manufacturing precision
If the number of actuators is increased to align each device individually, then alignment precision improves, but system cost increases dramatically
Solution Approach 1:
The patent implements a universal alignment system where a single set of actuators and a common coordinate system serve multiple devices simultaneously. Instead of having dedicated actuators for each device, the system uses one set of actuators that can be controlled independently for each device through software, allowing the same hardware to perform alignment functions for multiple devices without requiring additional physical actuators for each device.
Solution Approach 2:
The patent merges the alignment control functions for multiple devices into a single integrated system. By combining the control of multiple devices under a common coordinate system and using a single set of actuators that can be independently controlled, the system reduces the total number of actuators required while maintaining the ability to align each device precisely. This consolidation reduces system cost while preserving alignment precision.
3Adaptability or versatility
If more actuators are added to handle multiple devices, then alignment capability improves, but the space required for the system increases
Solution Approach 1:
The patent employs a universal actuator system that can control multiple devices through a single set of actuators. Each actuator can be independently controlled to adjust the position of different devices, allowing one actuator set to serve multiple alignment functions. This multi-functionality enables the system to handle multiple devices without requiring additional actuators for each device, thereby minimizing the space required for the system.
Data Source
AI summary
A vision alignment system and method is provided. The vision alignment system includes one or more grouped alignment plates with guiding inserts configured to receive multiple devices, and groups of three actuators, configured to actuate the alignment plates to correct the position offsets of multiple devices as a group. The position offsets between the device and contactor are determined by a device-view camera during runtime and a contactor-view camera during calibration time. The vision alignment system also includes a pick-and-place handler, configured to transport devices.