When the Analyst specifies that individual test sets are candidates for fault detection or fault isolation, then all tests that have been defined within the selected test sets may be utilized by the diagnostics. The Analyst should plan to use test set tests (as opposed to using Output Flags or Net Functions as tests) for all final analyses. There are two reasons for this. First of all, test definitions allow a degree of control (selected states, non-detectable coverage, non-monitored stimuli, symmetry, interference, etc.) that is simply not available when merely testing output flags or net functions. Secondly, test Attributes (such as test time or cost) can only be associated with test set tests, thereby enabling the calculation of several isolation metrics.
Tests Set tests and output flags can be utilized together in diagnostics (in fact, this is common practice when performing a test point placement study). Because the Test Candidate Groupings in the predefined Diagnostic Algorithms nearly always use test set tests before they will use output flags as tests, diagnostics generated using both tests and output flags may indicate ways in which the existing set of tests can be improved by the defininition of additional tests.
