Optimal Tool Replacement For Processes With Multiple Characteristics Based On Capability Index

Abstract

本篇論文研究結合工具耗損之修正製程能力指標和無母數信賴區間方法。我們應用了percentile bootstrap (PB) 觀察全製程良率C_PL^(T^' )之信賴區間。此研究方法應用於太陽能電池效能之量測，然而影響太陽能電池效能有三個主要特性，分別為理想電流 (Isc) , 開路電壓 (Voc) 和阻值 (Rsh) 。此三個特性影響I-V曲線表現，而I-V曲線為太陽能電池轉換效能重要的特性因子。此三個特性的製程量測為依據量測機台的探針，在這個案例中當探針呈現不可靠的狀態下間接影響太陽能電池轉換效能的量測。也就是說當量測太陽能電池轉換效能為逐漸下降的情形時，並無法得知是探針耗損需要被更換亦或太陽能電池本身轉換效能不符規格。因此我們提出透過C_PL^(T^' (LB) )製程指標特性下去判斷真實製程狀態。在量測製程上，我們提供更可靠的製程指標特性C_PL^(T^' (LB) )以判斷探針的更換準則。

This paper considers the modified process capability index with the tool wear process with the confidence bound of the bootstrap estimates. This paper applies the percentile bootstrap (PB) method to the overall process yield measure C_PL^(T^' ) to obtain the confidence bounds. This paper applies this modified process capability index to the measurement of efficiency for silicon solar cell. It is noted that the ideal current source (Isc), open-circuit voltage (Voc) and the shunt resistance (Rsh) are the three important factors for the I-V characteristic curve. The probe measures these characteristics in the producing process. In this case, if the probe is not replaced, the efficiency of the silicon solar cell is unreliable and the value of silicon solar cell is drop down. Based on the value of C_PL^(T^' (LB) ), it can be judged that the process is “emergency” for characteristic measuring of efficiency. In the actual measuring process, we propose the more reliable index of the lower confidence bounds C_PL^(T^' (LB) ) to judge the replacement of the probe.