In the field of semiconductor manufacturing, chip welding is a key process, and the visual positioning system of the die bonding machine is the core component to ensure the accuracy of welding. Its accuracy not only determines whether the chip can be accurately placed at the target position, but also directly affects the quality of welding, which is related to the performance and reliability of electronic products.
The visual positioning system of the die bonding machine is usually composed of industrial cameras, lenses, light sources, image acquisition cards and image processing software. When working, the industrial camera collects images of the chip and the welding target position with the assistance of the light source, and the image acquisition card converts the optical signal into a digital signal and transmits it to the computer. The image processing software uses edge detection, feature extraction and other algorithms to identify the position, shape and other information of the chip and the welding point, and then calculates the deviation between the chip and the target position, and guides the welding head to make precise adjustments. The role of this system is to overcome the error and instability of manual positioning, achieve high-precision positioning at the micron level or even the nanometer level, and lay the foundation for high-quality welding.
When the accuracy of the visual positioning system is insufficient, it will cause many problems to the welding quality. Positioning deviation may cause the chip and the welding point to be unable to be fully aligned, resulting in welding offset, reducing the welding area, reducing the connection strength, and easily causing cold welding. In addition, low precision may also cause uneven distribution of welding pressure, too deep welding in some areas, damaging the chip or substrate; while welding in some areas is too shallow, forming cold welding, affecting the electrical connection performance. In the long run, these welding defects will cause problems such as unstable signal transmission, performance degradation and even failure of electronic products during use, seriously affecting the reliability and service life of the products.
Hardware upgrade is an important way to improve the accuracy of the visual positioning system. The use of high-resolution and high-frame-rate industrial cameras can capture clearer and more detailed images and reduce positioning errors caused by image blur; with high-precision optical lenses, the image magnification and imaging quality can be improved, and the ability to recognize tiny features can be enhanced. Optimize the light source design, select the appropriate light source type (such as ring light, backlight, etc.) and light intensity according to the material and surface characteristics of the chip and welding point, ensure that the image has good contrast and clarity, and facilitate subsequent image processing. At the same time, regularly calibrate and maintain the hardware equipment to ensure stable equipment performance and reduce the problem of reduced accuracy caused by hardware aging and wear.
Algorithm optimization is also critical to improving the accuracy of the visual positioning system. Improve the image preprocessing algorithm, such as using more advanced filtering, noise reduction, and image enhancement algorithms to improve the quality of the original image and reduce the impact of interference factors on feature extraction. In terms of feature extraction and matching algorithms, introduce deep learning algorithms, such as convolutional neural networks (CNNs), to automatically extract the features of chips and welding points through learning and training a large amount of welding image data, and improve the accuracy and robustness of feature recognition. In addition, optimize the positioning algorithm, use Kalman filtering, particle filtering and other algorithms to dynamically correct and predict the positioning results, compensate for the positioning deviation caused by mechanical vibration, environmental interference and other factors, and achieve more accurate positioning control.
Environmental factors have a significant impact on the accuracy of the visual positioning system. Controlling environmental conditions can ensure the stable operation of the system. Keep the temperature and humidity of the welding workshop stable to avoid thermal expansion and contraction of chips, substrates and equipment components due to changes in temperature and humidity, which affects the positioning accuracy. Reduce electromagnetic interference in the workshop, and perform good grounding and shielding on electrical equipment to prevent electromagnetic interference from affecting the normal operation of industrial cameras, image acquisition cards and other equipment. In addition, control the lighting conditions in the workshop to avoid strong direct light or uneven lighting, ensure that the visual positioning system works in a stable lighting environment, and reduce positioning errors caused by fluctuations in environmental factors.
The visual positioning system of the die bonding machine needs to be precisely coordinated with other subsystems (such as motion control systems, welding head control systems, etc.) to achieve high-precision welding. During the system integration process, accurately calibrate the coordinate relationship and motion parameters between the subsystems, establish an accurate mathematical model, and ensure that the visual positioning results can be accurately converted into motion instructions for the welding head. Regularly calibrate the entire welding system, correct the system's cumulative errors through standard parts testing and error compensation, and improve the overall accuracy and stability of the system. At the same time, establish a system operation status monitoring mechanism to monitor the working status and accuracy indicators of each subsystem in real time, and promptly discover and solve potential problems.
Improving the accuracy of the die bonding machine visual positioning system is a continuous optimization process. Pay attention to the latest technology trends in the industry, and continuously introduce new hardware equipment and algorithms, such as the use of new sensors, quantum dot imaging technology, etc., to promote technological innovation in visual positioning systems. Enterprises are encouraged to cooperate with scientific research institutions to carry out industry-university-research projects, tackle technical problems in actual production, and explore more advanced positioning methods and solutions. In addition, a user feedback mechanism is established to collect problems and improvement suggestions encountered in actual production, and the design and performance of the visual positioning system are continuously optimized based on feedback to meet the growing demand for high-precision welding.
