YIELD MODELS
 
Common model to calculate Yield

Yield for each step is presenting percentage of boards which are not identified faulty in that certain step. The defects are not divided equally to each board but they are following Poisson distribution. Yield is calculated from detected defects usind following formula:

  Y = e-DPU

Clustered model

In modern high density boards defects have tendency to be clustered. Poisson distribution is not always the best model to calculate Yield from DPU. Normally clustering can be seen in the first test step, most of the clustered defects are removed and Poisson model is valid for next steps. Yield model for clustered defects is based on negative binomial distribution.
If you want to use clusterd model with first test step, select "Clustered model".

See user guide to read more of yield models.