Academic Company Events NI Developer Zone Support Solutions Products & Services Contact NI MyNI
1 ratings:
 4 out of 5     Rate this Document

Using Loop on FAIL in CVI Test Executive

Primary Software: LabWindows/CVI Add-ons>>Test Executive
Primary Software Version: 2.0
Primary Software Fixed Version: N/A
Secondary Software:

Problem: I am using Test Executive 2.0 with LabWindows/CVI 5.01, and Loop on FAIL does not seem to work correctly. To reproduce the problem, simply modify the computer.c file:
  • Load the computer.squ file (or any other test sequence).

  • Choose Sequence, Edit from the menu and go to the sequence editor.

  • Select a test sequence and choose Run Options.

  • In Run Options, choose Fail Action = "Loop" and Number of loops = 10, for example.

  • Save the sequence and then run it using Test UUT.

  • Check one of the tests to fail.
The failed test loops 10 times, and then the entire sequence fails. This is correct behavior; however, if the test that initially FAILS now PASSES during subsequent passes (since I am now looping 10 times on that test) then the sequence still shows that the "Test Failed", even though it should say that the test passed. It was my understanding that the idea behind having a loop control is for precisely such a reason, so that a test that may have erroneously failed the first time around, passes and does not cause the entire sequence to fail.

Solution: The intended default behavior is that any test failure causes the sequence to fail. You can uncheck the "Test fail causes sequence fail" checkbox (available in the pop-up window that appears when you press the Set Limit Specification button in the Sequence Editor) to override the default behavior for a particular test. Since the test executive can't determine that a test "erroneously failed", it treats any error on the side of caution.

The idea behind the loop control is in fact the opposite of what you understood. It is to "stress test" a component which may pass the first time but which may fail on subsequent tests, not to give a component that fails several more chances to pass.

As for the behavior that you suggest above, we recommended that you move the loop from the sequence to the test itself and let the test determine the maximum number of fails that still allows the test to pass.

Related Links:

Attachments:





Report Date: 03/09/1999
Last Updated: 02/14/2002
Document ID: 1J8HG2PR

Your Feedback! poor Poor  |  Excellent excellent   Yes No
 Document Quality? 
 Answered Your Question? 
  1 2 3 4 5
Please Contact NI for all product and support inquiries.submit