- Testing [4]
Assembling a reasonable test package can be challenging and tedious, even for experienced engineers.
The successful launch of new electrical products relies on successful design and canny production techniques. It also depends on the test engineer’s ability to write a successful and succinct test plan for the prototype and for the actual production run. The value of product testing in the design and production phases is well known, but it’s also important to realize that there’s an art to determining tests that will economically catch production and design flaws.
In the design phase, the test engineer is trying to find weaknesses in the product. Most times, mechanical and electrical problems can be systematically rooted out by following accepted test regimes found in the various standards published by many bodies worldwide. When the product can be categorized and standards can be found, many problems will be found by simply putting the package through a “standard test package” covering mechanical, electrical, and performance criteria. While this process of gleaning a reasonable test package can be challenging and tedious to perform, the mechanics of doing so are fairly straightforward, and most test equipment and setups can be fabricated from what the engineer can find at hand, or perhaps from a rental source.
However, in some cases, this plan doesn’t work because the product is breaking new ground and the published standards don’t provide a test program that fits. In other cases, standard test equipment will not meet unusual requirements called out by the appropriate Standard. Now the test engineer is faced with a dilemma—does he substitute the test and conduct a similar test with what is at hand, or does he procure custom gear?
Substitution of Like Tests
In many cases, the differences between what a standard calls for and what a test engineer can provide are very small, and the engineer can provide a slightly tougher test. For example, some impulse tests conducted on IT gear specify voltages available on the internal charged capacitor, but readily available surge testers read output voltage. In this case, it’s pretty clear that if a test voltage is set on the output, the resulting test will be more difficult to pass than if the same voltage was set internally to the surge tester, because output voltage will be lower than input voltage. The problem is that if the substituted test is less stressful than the specified test, the product may not pass the real test later. Conversely, if the substituted test is significantly more stressful, the product may be needlessly overdesigned.
Procure Custom Gear
If standards for the new product exist, the advantages and disadvantages of this path are clear: the test engineer tests to the exact requirement, at higher cost and most likely longer timeline because the test equipment must be designed and built either by the test engineer himself or by others. If the product is so new that no standards covering a particular performance or other criterion exist, then the path becomes clearer: the test engineer must determine a test program and procure equipment. The test plan will require consultation with the design team to determine test parameters, but in most cases the design team is too busy to help with this task. The test engineer should try to find a reasonably similar test standard and try to change parameters to fit the new package.
What is the new product similar to? How does it differ? Is the power output higher or lower? Where will the product be installed? How is it powered? What particular hazards does it pose to the user and to the operator? From this list and comparison to a matrix of standards which partially describe the product, the test engineer may be able to piece together a test program. There is no assurance that this test program will be accepted later by outside agencies, but if the product is truly new and truly no standard exists to describe a particular facet, the test engineer is forced to act. When identifying a testing company partner, it’s crucial to find a company that truly has the engineering expertise to create an exceptionally effective customized tester.
Production-line testing has a different goal—to make sure the product is consistent. While hipot testing can ensure that the product primary circuits are correct, further testing is mandatory to make sure the rest of the product is functioning correctly and safely. Although the test itself might be simple, specifying the test equipment to perform the test must take into account simplicity of use, elegance in the determination of a pass/fail point, and the ability of the test equipment to perform for the thousands or millions of cycles needed between annual calibration cycles.
The test equipment itself must be tested to make sure it is functioning correctly before it is put into service, as it is imperative that line technicians have confidence in it from the very beginning, so test results are trusted and acted upon. If the test equipment is prone to breakdown or tripping on false failures, a real problem may be overlooked. Of course, reliability is only one of the parameters; the test equipment must also perform the test function correctly. In many cases, test equipment must be customized to the task, as many of the test parameters will be specific to the manufacturer, line, and devices being tested.
When custom test equipment is specified, it’s important to work closely with the provider and settle on a complete test spec, including test parameters, test speed, failure mode(s), failure indication, data acquisition, and number of cycles between maintenance/calibration intervals. In addition, the ease of program modifications should be explored. As experience is gained with the test program, sequences may be modified, and the difficulty of making these changes should be assessed. Finally, it’s important to ensure that your customized tester includes a manual and complete documentation.
In both the design and production phases of a product’s life, testing is essential to ensure that the design is robust and the product continues to be manufactured with good results. In many cases, custom test equipment is the best fit for these tests. In design, specialized test requirements may require procurement of a custom test solution to avoid under- or over-building the equipment. In production, a custom solution may be the only way to perform all needed tests in an efficient manner.
Jeffrey Lind, president of Compliance West (Del Mar, CA), has contributed his 33 years of extensive electrical engineering expertise to help further the advancement and expansion of the Compliance West, USA brand. Lind approves and implements all business management decisions, runs all financials and budgeting, and oversees the engineering, customer service, and marketing departments. He received his BSEE degree from Cal Poly San Luis Obispo in 1976.