Issue link: https://iconnect007.uberflip.com/i/1305670
NOVEMBER 2020 I SMT007 MAGAZINE 39 Horner: When you're measuring the results, it is based on the metric. A lot of times, the quality reports that come off a test or inspection system are not based on a met- ric. What can be accom- plished with TestWay is a comparison of theoret- ical coverage versus measured coverage. When you're looking at test coverage, somebody says, "I have 90% coverage, and that's all very sub- jective. But if you base it on a metric that says, "I have a PPVS metric," then we have some- thing that's based on presence, polarity, value, and solder. The PPVS is a metric that is a subset of PCOLA/SOQ, an industry standard that is loosely adopted. I think Keysight and other ATE suppliers are embracing this type of metric, but not everyone. But if you have a baseline that you could take to management and say, "I have 90% coverage," then you would have a better idea of what that 90% coverage entails. Webb: It relays it back to a metric. The problem that you have is if you start looking at coverage from different types of machines, you have a particular AOI machine and a particular flying probe. Those all produce a coverage report like Bert says, but they may use different metrics to evaluate the coverage. The good thing about bringing in a tool like TestWay is that it resolves everything back to the PPVS or PCOLA/SOQ and easily creates a combined coverage report for you from any of the tests that you're going through—something that would be very diffi- cult to do manually with a spreadsheet. With TestWay, it is done instantly by resolving all that back to the metric that Bert says. Johnson: You could use multiple test styles, fly- ing probe, optical, and bring all of that infor- mation together into one place for one cohe- sive set of results. Webb: Exactly. There's no one strategy in our industry that would cover all your potential defects. You need to look at multiple strategies to come up with decent test coverage. With ICT, certain faults can be caught with AOI, and others can be caught with X-ray. You need to be able to look at what each one of those is bringing to the table and make sure that the sum total of all your strategies is giving you good coverage. When you do that, you're going to wind up with the overlap, as we said. Not only will each strategy be testing something uniquely, but you'll wind up with overlap. The tool will identify that for you. Then, the user can make the decision about whether they want to reduce that overlap and have a leaner test and more cost savings in your test department. The tool will help you find the right amount of tests and keep the cost low. Johnson: That's great. For somebody like Bert, in your facility, there's a lot of tooling, setup, and optimizing. You are living and dying prof- itability-wise in the effectiveness of your test coverage tool. Of course, there are other pro- duction facilities with fewer products but lon- ger lead times. It would seem to me that the value in your product to a more production- oriented assembly house would still be there, but it would be a different set of returns on investment. Webb: It's very interesting that you asked that question. We deal with both OEM, CM, and EMS companies. They can both be customers of our tool, and they are. As you say, the ROI calculation can be a little bit different. An OEM can look at what the DFT can do for them. In fact, coming from my previous employment, we were an OEM using the tool. With DFT, we could reduce our board spin. And we had some costs associated with every time a board spin happened, and it doesn't take too many board spins to pay for the tool since you have already paid for it. Coming from the CM side, they're not always engaged early enough to have a real impact in a timely matter on the DFT side. The OEM might have already completed its design, gone through layout, and thrown it over the wall to Bert Horner