eXpress provides diagnostic engineers with a mechanism for organizing tests into test programs based on their diagnostic utility. Although intended as a means of optimizing the batches of tests performed on Automatic Test Equipment, this feature can be used to effectively stratify testing for a variety of run-time scenarios.
In eXpress, a test program is a special type of virtual test set that is populated based on attribute settings on tests. Unlike many attribute-related features in eXpress, the special attribute that is used by this feature is automatically defined by the software when you create your first test program.
To create a test program, simply select “New Test Program…” from the Test Programs section of the main Design menu (depicted at right). You will then be prompted for the name to be given to the newly created test program. The first time that you do this, the following prompt will appear:
Select the test program to which you wish to assign tests. If you had filtered the tests displayed on the Test Program Optimization dialog (either by name or usage), you can specify whether you want test assignment to also use these filters.
In the “Cutoffs” section, you can select the total number of tests or cumulative test time that is allowed for that test program. Tests assigned to the selected test program will not exceed (cumulatively) the specified cutoff Test limit(s).
Click on the OK button to perform the assignment.
For this example, a one-minute cutoff was used to assign tests to the test program “TP #1”:
You can examine the test program by selecting the corresponding entry in the “Fill With” dropbox:
Notice that the number of tests assigned to this test program and the cumulative test time are both displayed at the bottom of the dialog. Within a diagnostic study, the candidate lists on the Detection and Isolation panels can now be populated four different ways—all candidates, candidates in test programs, candidates in virtual test sets and candidates in standard test sets.
You can toggle between the different display modes by clicking on the “T” icon in the upper right corner of the candidates list (Note: this icon only appears if the design contains either virtual test sets or test programs that correspond to the panel’s diagnostic usage). To evaluate the diagnostic effectiveness of one or more test programs, generate diagnostics using only those test program(s) as candidates for detection and/or isolation. In the example below, you can see that our one-minute test program can detect only a little more than half of the expected failures. You can toggle between the different display modes by clicking on the “T” icon in the upper right corner of the candidates list (Note: this icon only appears if the design contains either virtual test sets or test programs that correspond to the panel’s diagnostic usage). To evaluate the diagnostic effectiveness of one or more test programs, generate diagnostics using only those test program(s) as candidates for detection and/or isolation. In the example below, you can see that our one-minute test program can detect only a little more than half of the expected failures.
Because no isolation tests were included in this test program, it should not be surprising that the expected isolation to a single component (using only this test program) is poor. In this particular example, the numbers resulted in fault detection being privileged over fault isolation. In situations where certain isolation tests are among the tests most commonly performed during diagnostics, those isolation tests may be included in the initial test program (in place of some of the later, less useful detection tests). In that case, the probability of isolation would be better, at the expense of the probability of detection. Here is where automated test program assignment is particularly useful, since the comparisons are now considerably more complex (and can no longer be performed by simply viewing cumulative Test limits in a spreadsheet). Currently, test programs are handled by the diagnostics in the same manner as a virtual test set. In future releases of the software, however, test programs will take on a larger role—providing one more method for controlling the order in which tests are performed during diagnostics.
