| Summary: | 碩士 === 義守大學 === 工業工程與管理學系 === 91 === MSA (Measurement System Analysis) of QS-9000 adopts regression approach to evaluate the accuracy. Intel judges both statistical significance and technical significance to verify the accuracy. However, both approaches do not provide any quantitative criterion as a rule of acceptance. In order to improve them, Philips uses “Range Method” as statistical way and adopts “Tolerance Percentage” as acceptance criterion. We prefer Philips’ way due to the suitability for industrial application.
Review back published articles, and no one was found to tackle ever with two-dimension measurement issues. This study transfers the coordinates to norm distance, and then the traditional MSA can be applied to the two-dimension measurement issues.
The measurement errors of Philips-CMS are assumed coming from primarily the prober system itself and the background vibration. On the other hand, the floor is the potential source of background vibration. This paper adopts MSA of QS-9000 to evaluate both its measurement precision and accuracy. System repeatability is defined to the extent of precision incurred within the Prober System only. Floor reproducibility is defined as the degree of vibration incurred by the floor vibration.
Due to good analysis, we can identify microscope out as the main cause. Thus, after the microscope is fixed properly, the Prober 4090u-57 is found precisely enough.
|
|---|
