Premium
Utilizing design of experiments, Monte Carlo simulations and partial least squares in snapback elimination
Author(s) -
Church Mike,
Lynch Richard O.
Publication year - 1998
Publication title -
quality and reliability engineering international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.913
H-Index - 62
eISSN - 1099-1638
pISSN - 0748-8017
DOI - 10.1002/(sici)1099-1638(199807/08)14:4<227::aid-qre189>3.0.co;2-u
Subject(s) - snapback , monte carlo method , process (computing) , reliability engineering , engineering , computer science , mathematics , statistics , electrostatic discharge , electrical engineering , voltage , operating system
The first test structures built on a newly developed semiconductor process revealed that product could be susceptible to an operational fault called the snapback condition. The process architect identified five factors that might be adjusted to greatly reduce the occurrence of snapback. A response surface type of experiment was run on the process simulator so that the ideal combination of settings for these five factors could be identified. Monte Carlo simulations were then run at the new settings for those process factors. The data from the Monte Carlo simulations were analysed using partial least squares to identify the process variables that would be most critical to control in maintaining a snapback‐resistant process. The new settings were confirmed on actual product. © 1998 John Wiley & Sons, Ltd.