Testing Errors: Data and Calculations in an IC Manufacturing Process
Richard Hays Williams, R.G. Wagner, Charles F. Hawkins · 1992
Test data on 7,981 MOS integrated circuits (ICs) sampled over a six year time period were analyzed to determine the magnitude of the Type I and Type I1 test errors in a high reliability, low volume production test environment. Type I test errors are those in which the tester fails good product and Type I1 errors are those in which the test process passes defective product to the customer. The data allow either direct calculation or estimation of all test distributions of the manufactured population. Results show that the Type I error constant (a) is simcantly higher than the Type I1 error constant ((3). The test distributions are modeled and compared with results of others. Knowledge of actual error values can be used to monitor and improve product quality.