薄膜式單層多點電容觸控感應器製程研究

Abstract

摘 要 各類型的觸控感應器因手機產業的激烈變化，演變成少量試產、導入量產、結束生產的短週期生產製程模式。應客戶需求設計線路及指定材料，工廠端需快速反應；甚至當客戶要求本廠作為「第二供應商」時，也會提供他廠「第一供應商」的材料表進行生產，因此，部分材料有時並未經過廠內驗證，旋即快速導入並進行試產、量產。而產品會依市場反應及銷售狀況，來決定生產週期，若是銷量不佳，就會被砍單並停止量產作業，所以迅速應答的製程流程及良率提昇手法變成首重課題。 電容式單層多點觸控感應器是一種新式設計，製程設計及材料都不同以往，生產時只有以有經驗的工程師進行製程參數的設定，抓取各參數的容許範圍，利用 8D 及 PCDA 的手法來調整改善並判斷生產狀況及結果；第二階段投入時，使用變異數分析各生產參數，得到各參數對製程的影響程度，及其交互作用結果。單層多點感應器的製程參數包括了：外型尺寸及老化後電阻值結果、黃光蝕刻 ITO 圖案化後關鍵尺寸大小、網板印刷絕緣膠材厚度、外型尺寸及網板印刷銀漿線路位置尺寸精度、厚度及銀線關鍵尺寸大小等等。為了得到快速正確的生產參數，各製程工序必須系統化及數據化，而利用實驗設計變異數分析方法計算最佳製程參數組合，可減少傳統試誤法的時間浪費及穩定度不佳風險。故專題中利用變異數分析而得到的最佳化參數為：老化（ITO 烘烤時間45 min 、溫度 160 ℃）、乾膜曝光（能量 32 mj）、顯影（時間 45 min、溫度 160 ℃）、銀漿網印（刮刀壓力 25 kgf、刮刀速度 100 mm/s、網板距平台高度 3.0 mm）、絕緣膠網印（刮刀壓力 35 kgf、刮刀速度 80 mm/s、網板距平台高度 3.0 mm）。 利用數據分析得到生產參數，可以有效減少測試時間，大幅減少因試誤法所產生的不良因子，而變異數分析所建立的分析模型，可以代入後續其他生產的 ITO Film 、絕緣膠及銀漿等材料，進而減少生產材料浪費，並提升良率且穩定品質，加速產品生產。

Abstract Various types of touch sensors have evolved to be manufactured in a production mode starting from a small trail production to the mass production due to swift changes in the competitive cell phone industry as it can shorten the lead time of production. The applied materials are assigned according to customers’ requirement for circuits thus a factory needs to have a quick response to customers’ needs. Even when the customers ask the factory to become the second supplier, the factory needs to use the “first supplier’s “material to undertake production. Hence, some materials are not certified onsite and the trail production starts off quickly and goes on with mass production. The production schedule will be decided according to the market feedback and sales performance of products. If the sales performance is not satisfactory, the orders would be cut off and the mass production would be forced to cease. Therefore a timely response to production procedure and better yield rate are becoming the key issues to tackle.

Surface Capacitive single layer multi-touch sensor is a new design with different production paths and materials as to distinguish out from the old style. The experienced engineer would set up the production parameters during production to retrieve the tolerance range of each parameter. 8D and PCDA approaches are applied to adjust and improve the production situation and results. When the second stage production starts, the production parameters are generated using Analysis of variance (referred to as ANOVA hereunder) to attain the result of each parameter’s influence on the production and interaction effects. The production parameters of single layer multi-touch sensor include: dimension of models, resistance value result after anealing, critical dimension after photo etching on the ITO pattern, the thickness of isolation rubber for screen printing, accuracy of models dimension, alignment accuracy of screen printing silver circuit location and thickness, and critical dimension of silver circuits. In order to obtain the immediate and correct production parameters, every production process needs to be systematized and digitized. The variance data obtained by the experiment design can compute the optimal parameter combination of production in order to avoid the risk of instability and inaccuracy caused by the traditional calculation. The most optimal production parameters obtained by variation analyses in this project are as follows: Aging (ITO baking time 45 min, temperature of 160 ℃), dry film exposure (energy32 mj), development (time 45 min, temperature of 160 ℃), silver screen printing (squeegee pressure: 25 kgf, squeegee speed :100 mm / s ,height from screen to work stage: 3.0 mm), insulating rubber screen printing (squeegee pressure 35 kgf, squeegee speed of 80 mm / s, height from screen to work stage: 3.0 mm).

The production parameters obtained by data analysis can be applied to shorten the testing time and to reduce defect factors largely caused by trial and error methods. The analysis models established by ANOVA can be used for material production of ITO film, insulation rubber and silver paste etc. as to avoid failure of material production, improve defect-free rate, stabilize the product quality and speed up the production schedule.