Ultrasonic Wire Bond Defect Detection Using CNN Signal Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wire bonding inspection methods, including manual macrography and automated image processing, are inefficient and unreliable due to physical damage, camera positioning issues, and environmental changes, and additional inspections require separate equipment and time, lacking real-time analysis of bonding parameters.
Innovation Solution
A wire bonding defect detection apparatus using a convolutional neural network (CNN) for machine learning, which collects ultrasonic bonding parameters, applies wavelet transform to convert data into images, and determines defects through a trained CNN.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual macrography or automated image processing is used for wire bonding inspection, then defect detection is performed, but inspection efficiency decreases and reliability is compromised due to physical damage and environmental factors
Solution Approach 1:
The patent replaces mechanical inspection methods (manual macrography, physical wire pulling) with an ultrasonic-based inspection system that uses acoustic waves to detect defects. The ultrasonic inspection apparatus sends ultrasonic waves through the wire bonding and analyzes the reflected waves to detect defects without physical contact or damage, thereby improving both reliability and efficiency simultaneously
Solution Approach 2:
The patent introduces ultrasonic waves as an intermediary medium to detect wire bonding defects. Instead of directly manipulating or visually inspecting the wires, the system uses ultrasonic waves as a mediator to probe the bonding quality, eliminating the need for physical damage while maintaining high inspection reliability and efficiency
2Reliability
If separate equipment for electrical or mechanical inspection is used during bonding, then wire and solder defects are detected, but additional cost and time are required
Solution Approach 1:
The patent merges the ultrasonic inspection function with the existing wire bonding process by integrating the ultrasonic inspection apparatus into the bonding system. This allows defect detection to be performed concurrently with or immediately after bonding without requiring separate inspection equipment or additional time, thereby maintaining high detection accuracy while eliminating time loss
3Measurement precision
If image processing with cameras is used for inspection, then defects are detected, but the position of camera, resolution, and lighting changes affect reliability
Solution Approach 1:
The patent replaces optical measurement systems (cameras, lighting, positioning systems) with ultrasonic measurement technology. The ultrasonic inspection apparatus measures wire bonding defects through acoustic wave propagation and reflection, which is not affected by camera position, resolution, or lighting conditions, thereby maintaining high measurement precision while improving reliability
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables efficient, non-destructive inspection of wire bonding defects by analyzing bonding parameters in real-time, improving the quality and reliability of battery packs.
Implementation Method 1
a wire bonding process unit configured to perform bonding of the wire to the substrate using ultrasonic vibration
Implementation Method 2
a data generator configured to apply wavelet transform to the collected data so as to be converted into images having a two-dimensional array
Data Source
AI summary
An apparatus and method are disclosed for detecting a bonding defect occurring during an ultrasonic wire bonding process between a battery cell and a busbar connected to each other by ultrasonic wire bonding. A bonding parameter collector, coupled to an ultrasonic wire bonding machine, may collect ultrasonic bonding parameters continuously during the ultrasonic wire bonding process, machine learning training may be performed by a convolutional neural network, and a bonding defect may be detected based thereon.