大尺寸TFT-LCD 中線缺陷漏檢改善

Abstract

本文主要在探討液晶顯示器面板模組製程中，大尺寸且高解析度機種因製程難度高容易造成的漏檢問題，如何將不良缺陷攔檢於源頭是任何產業皆需要面對的問題，因為如此除了可以解決客戶抱怨也可加速製程上的改善提升產業競爭力。 在大尺寸且高解析度的面板製程過程中，因解析度提高造成製程難度提升，製程難度提升所造成的不良缺陷機率升高，也因此漏檢的機率也相對提升，其中線缺陷不良長期以來為造成液晶顯示器面板漏檢主要問題，線缺陷時常存在可靠度或不良現象不明顯等問題，若可有效攔檢線缺陷不良對於整體液晶顯示器面板缺陷的檢測將可大幅提升。

在傳統液晶顯示器面板模組檢測主要是使用固定畫面(例如：紅色、綠色、藍色、白色、特定灰階等畫面)做為檢測標準，如此一來針對存在可靠度問題或是人眼不易辨識的淡線將不容易被檢測出，有可能該類問題在終端市場消費者將液晶顯示器產品買回家1~2年後才被發現，此時商譽已經受損、甚至使公司因為單一事件造成嚴重虧損。在第四章中介紹本實驗材料選擇使用對角線長65吋、玻璃厚度0.5mm的面板半成品，使用各種條件做點燈測試，蒐集各類線缺陷不良樣本，針對各種的線缺陷特性做攔檢方法開發，利用數位影像瑕疵分析及以恰好可辨識強度差 (Just Noticeable Distortion, JND) 數值確認其效果。 以本實驗最終結果顯示，針對各種線缺陷不良調整出最佳測試條件，再利用最佳條件做大量的生產驗證確認最後實際量產狀態下發現線缺陷攔檢率大幅提升，檢測穩定性亦可使問題達到有效的控管。

This article focuses on the topics of improving the issue of negligence on inspection of large size and high resolution LCD product. The large size and high resolution LCD product during the difficult production process may easy to occur some display defects and hard to detect. Therefore, control and prevent the defective is always the challenge that needs to face for the industry. What is more, it can also reduce the production cycle time and the key to increase the customer satisfaction to gain higher industrial competitiveness.

High resolution LCD has its limitation to achieve high yield rate due to its difficulty to produce which increase the defective percentage and cause the negligence on inspection to be high. There are several possible disadvantages during the production; but the line defect quality has found to be the main issue for detection failure due to its symptom not obvious. Therefore, control and improve the inspection of line defect can be the main point to prevent the defective and increase yield rate.

Nevertheless, traditional LCD panel inspection process has set to be apply fixed screen with single color (ie., Red, Green, Blue, White, Grey) for inspection standard procedure. This process has low on reliable and high potential to be missing the inspection by visual. The defect product may only be discovering after it launched to the market and cause serious damage to the business goodwill.

In Chapter 4, the 65 inch diagonal and 0.55mm thickness glass panel has been select to be the study material to apply several condition for lighting test. After collecting different data samples and classify the issue for line defect. The study will focus on creating the suitable line defective inspection method and verify it by using auto optic inspection(AOI) of digital images of defects to deal with the different line defects and apply the JND (Just Noticeable Distortion) to ensure the result of new inspection method. The final result aims to achieve and adjust the suitable line defect inspection method and implement it in mass production. Through the study, we can find that line defective inspection yield rate had been highly increased and have good stability and control in inspection.