印刷電路板之完稿底片製程變異評估

The Evaluation of Artwork Dimensional Variance for PCB Manufacturing Process

俞佳伶(Chia - Ling Yu)

指導教授 ： 江行全 教授

元智大學/工程學院/工業工程與管理學系/碩士(2000年)

摘要

底片尺寸的漲縮變異為影響完稿底片品質的最主要關鍵因素。因為當底片尺寸漲縮誤差過大，將會造成製程中各內層配合度不良，使得許多印刷電路板因而報廢或重工，而造成製造成本的損失。
對此，本研究以完稿底片尺寸漲縮變異為研究主題，分為兩個部份進行探討：第一部份針對所歸納出造成底片尺寸漲縮之關鍵因素︰溫度、相對溼度、機器設備（三次元量床、雷射繪圖機）以及機遇原因，藉由變異成份之統計分析方法，解析並量化關鍵因素之變異程度，以提供給業者作為改善完稿底片製程之依據；之後第二部份則是建構以溫度與相對溼度對於漲縮量的預測模式，以便能及時監控溫度及相對溼度所造成的底片尺寸漲縮。
根據本研究分析結果，就第一部份而言，影響完稿底片漲縮變異的程度，依次為溫度及相對溼度（55.64%）、雷射繪圖機（21.83%）、機遇原因（19.07%）、三次元量床（3.45%）。就第二部份而言，分別針對短邊與長邊的差量，來建構溫度、相對溼度的模式，所得模式的解釋能力分別為短邊（65.7%）、長邊（91.0%）。

關鍵字

印刷電路板 ； 底片尺寸漲縮 ； 變異成份 ； 預測模式

並列摘要

The main factor which would effect artwork’s quality is its dimensional expansion/shrinkage. Because of the error caused by artwork’s dimensional expansion/shrinkage, it would make precise matching of the inner-layers of PCB more difficult. Then manufacturing cost will increase by many nonconforming units.
The focus of thesis study is on the variance analysis of artwork’s dimensional expansion/shrinkage. Part one is to find the factors that affect artwork’s dimensional expansion/shrinkage. The factors are Temperature、Humidity、Machine(Measurement system、Plotter) and other chance causes. Using statistical method of variance component analysis to quantify the variance percentage of these factors. The result of this part offers the PCB industry as a reference to make improvement. Part two is to establish forecasting regression models composed of temperature and humidity. Using these models, one can monitor the artwork’s dimensional expansion/shrinkage caused by temperature and humidity.
According to the results of part one research, the variance percentage of these factors in turn are (1) variance caused by Temperature and Humidity, counted for 55.64% of the total variance、(2) variance caused by Plotter, counted for 21.83% of the total variance、(3)variance caused by other chance causes, counted for 19.07% of the total variance、(4)variance caused by Measurement system, counted for 3.45% of the total variance. And the part two results of research are the two regression models established. The models include temperature and humidity factors for artwork’s short and long side, respectively. The responses of two models are dimensional deviation for the short and long side. The forecasting capability of short side’s model is 65.7%. And the forecasting capability of long side’s model is 91.0%.

並列關鍵字

PCB ； Artwork''s Dimensional Expansion/Shrinkage ； Variance Component ； Forecasting Model

參考文獻

30期，1997年4月，頁72-76。

〔5〕Martin W. Jawitz, Printed Circuit Board Material Handbook,

McGraw-Hill, New York, 1997.

〔6〕Clyde F., Jr. Coombs, Printed Circuits Handbook, 4th

〔11〕Preben Lund, Generation of Precision Artwork for Printed