How to Create Successful Automated Test Equipment Self-Tests
Automated self-tests are crucial for discovering otherwise unforeseeable faults in your automated test equipment (ATE). Here are the five tried-and-true best practices for creating successful ATE self-tests that will provide reliable and easy-to-use test data, while eliminating user confusion.
1. Have a Clear, Easy-to-Read Report
Knowing exactly what went wrong during a test is critical. Report results should include the measured value, the expected value and the limits used to determine that a failure occurred. Nothing frustrates a diagnostic effort more quickly than a vague description of what went wrong. Imagine a test report that says, “Voltage Test: 5.0126 V (FAILED).” Is this quantity close? Is it supposed to be fewer than five? Is it supposed to be zero or 10? Having a clear report will terminate all of this confusion.
2. Show Your User What You Actually Tested
Letting your user know exactly what’s happening in the self-test can help them fix the ATE if something is wrong. Suppose you are testing relays. Are you asking the relay card if it is OK? Are you closing the relay and measuring the resistance with a digital multimeter (DMM)? Are you driving a current through the relay and measuring the voltage drop? These are all important questions that will impact how the user diagnoses a problem. If you’re using a voltage and a current to calculate a resistance, it is important to provide the user with both the voltage and the current values, not just the result. If the calculated resistance fails the test, it could be due to poor resistance or a problem with the application of the current or voltage.
3. Be Thorough
Leave no stone unturned. If your self-test has a blind spot, that blind spot becomes even greater in the eyes of your faithful user. Nothing is more frustrating than thinking everything is tested, and then later discovering that a considerable portion is missing. Leaving a gap in the testing is where the elusive problem originates. It would be highly beneficial to randomly select a few things that could go wrong in the tester and examine your test logic. Would your test detect if a particular wire was disconnected? What if it was shorted to its neighbor? What happens if something isn’t powered on when the test begins? If you’re testing many inputs in parallel, what if two inputs are swapped? These are the essential questions you must ask yourself to unveil the blind spots of your self-test.
4. Test Consistently With the Manner the System is Used
Whenever possible, the wrap-around self-test should conduct tests in a similar manner to how the test system will be used when deployed. If a DUT is going to provide a 30V signal, the self-test should also test those channels with a 30V signal. In addition, if a relay is expected to convey 2A, the relay should be tested with 2A. This consistency ensures that the self-test is subject to parameters that are as close as possible to the real-life test scenarios.
5. List All Involved Components
When a failure occurs, it is important to detail all of the components that we have used in the test. Include components such as the cable numbers of all cables carrying signals for the test, input pin, output pin and data acquisition card name and position. Your goal here is to provide a checklist of every area that the user might need to inspect if something is wrong.
ATE self-tests with unclear, inconsistent and lacking information can cause as many problems as they solve. By putting in careful consideration around these five areas, you can easily create valuable self-tests that generate clear data and build user trust.
Take a look at part one of our self-test blog series to learn more about the top five benefits of ATE self-testing.
If you are developing your own system or planning on outsourcing your test system design to a company like G Systems, check out our blog about communicating your test requirements.