Keeping overseas production moving
An exclusive interview with a test engineer
-- Test & Measurement World, 11/1/2009 1:00:00 AM
Jonathan Davis spent several years in Asia working for a fabless semiconductor company that makes discrete and integrated power-management devices. He was responsible for product engineering, test engineering, failure analysis, product reliability, and new product introduction. He worked with several engineering groups within his company to oversee production, test, and quality at various contract manufacturers.
Q: Who tested your production parts?
A: Most production test was done by contract manufacturers. Some manufacturers were not set up for test, so we would use a third-party test house, but we preferred to have the manufacturers run the tests as well.
Q: What was a typical test setup?
A: A typical tester consisted of a handler and an ATE (automated test equipment) system that could produce and measure voltage and current. The handler needed a test fixture with fittings and attachments that could have 60 to 70 contact points. Parts were vibration fed into the handler, and they needed to make contact with the test fixture for measurements.
Q: Did you develop the testers for the manufacturers?
A: We configured and programmed the test systems and delivered them to the manufacturers. We purchased many test fixtures, because the parts came in numerous package sizes and the pin functions differed. When discrete parts used the same package, we could use the same test fixture and just change the ATE programming for each part. A third-party vendor constructed the test fixtures for us. I had used that vendor at a previous job so I knew that the vendor could fabricate high-quality fixtures at reasonable prices. I also purchased handlers and testers to consign to the manufacturers when needed.
Q: Who did the test-program development?
A: For new designs, a separate group in the US developed the test software. Those engineers worked closely with designers. We were responsible for transferring their test programs and hardware to the vendor. In those cases, my group would modify existing test parameters and limits to debug the test application on the mass-production test station. We had a separate product line that, while similar to discrete components, had different packages and pin functions. Thus, we designed a universal load board for these parts so we wouldn’t need a new load board for each new device.
| Read past Test Voices columns at www.tmworld.com/testvoices. |
Q: What other technical problems did you overcome?
A: After we built prototype parts, we’d ship them to the US for engineering evaluation. But we didn’t want an expensive handler underutilized in the design center. We designed two versions of the load board—one for production with the contactor and test fixture, another with a hand socket for engineering. Engineers in the US would use their fixture to develop and debug the ATE software. But we usually had to perform additional software debug in Asia where we used the handlers.
Q: What issues did you encounter working in Asia?
A: The time difference was a big factor. We often had to work late to speak with engineers in the US. Language was also a barrier to productivity. I could speak some of the local language, but not fluently. Cultural differences also come into play. Here, if there’s a technical problem, we assume that either the employee or the manager is at fault. That is, the manager may not have provided clear instructions. Over there, problems tend to fall on employees. Because of the language difference, I knew that instructions may not be clear and I tried to be flexible. For example, an engineer might take data incorrectly or not correlate the data to the parts. I learned that I needed to be clear on technical points like those.
No related content found.
- 0 rated items found.
Datasheets.com Electronic Parts & Inventory Search
185 million searchable parts
- Part Number
- Description
- Inventory
- Products
- Manufacturers
View All Webcasts
