在現代半導體產業中，封裝上片製程的品質控制對於確保最終產品的性能和可靠性至關重要。傳統量測方法存在設備內建軟體的不便利性與人工計算誤差的問題，以及生產過程中面臨晶圓上的微粒汙染和取錯晶片等問題，都會導致品質控制不穩定的問題發生，故需要有效的解決方案以確保產品品質和生產效率。為解決上述問題，本研究提出了一個結合MATLAB和C#工具的綜合解決方案。首先，針對銀膠攀爬高度量測問題，開發了一個基於MATLAB的量測程式，該程式操作簡便，能夠自動進行量測和計算，從而減少手動計算誤差並提高量測準確性。其次，針對微粒汙染和取錯晶片問題，開發了一個基於C#的視窗應用程式，該應用程式能夠快速且準確地檢測並處理相關問題，提高生產效率並降低賠償風險。為進一步提升量測速度和準確度，本研究利用MATLAB的神經網路功能進行視覺訓練，使其能夠根據實際情況快速進行量測並自動判別結果。該創新方案不僅加快了銀膠高度的判定，還為未來實現自動化工廠提供了堅實的技術基礎。;In the modern semiconductor industry, quality control in the die bonding process is critical for ensuring the performance and reliability of the final product. Traditional measurement methods face issues such as the inconvenience of built-in software in equipment and manual calculation errors, in addition to challenges like particle contamination on wafers and incorrect chip placement during the production process. These issues can lead to instability in quality control, necessitating effective solutions to ensure product quality and production efficiency. To address these challenges, this study proposes an integrated solution combining MATLAB and C# tools. Firstly, a MATLAB-based measurement program was developed to tackle the issue of die attach paste height measurement. This program is user-friendly and capable of automatic measurement and calculation, thereby reducing manual calculation errors and enhancing measurement accuracy. Secondly, a C#-based Windows application was developed to address the issues of particle contamination and incorrect chip placement. This application can quickly and accurately detect and handle these problems, improving production efficiency and reducing the risk of compensation. Furthermore, to enhance measurement speed and accuracy, this study utilized MATLAB′s neural network capabilities for visual training, enabling rapid measurement and automatic result determination based on actual conditions. This innovative solution not only accelerates the assessment of die attach paste height but also provides a solid technical foundation for the future implementation of automated factories.
