Issue link: https://iconnect007.uberflip.com/i/1389320
JULY 2021 I SMT007 MAGAZINE 13 Dan Feinberg: Eric, have you seen any recent difference in field failure rates, electrical vs. mechanical? It seems there are more portable circuit devices, which are more susceptible to mechanical problems such as drop shock. Are you seeing any trends? Camden: Primarily, we see more electrically driven failures, especially when talking about electrical leakage paths. When we talk about failures in long-term and various end-use envi- ronments, it's primarily electric. CMs are try- ing to make their products more robust if they know the device will be portable. ey're changing the way they clean their boards. We don't see many of the mass-produced products around here because it's cheaper just to replace them than troubleshoot them. I would say we still see a fair amount of elec- trical circuit failures on mobile portable type devices, but I haven't seen an increase in mechanical. Johnson: Eric, where are you seeing upticks in business? Camden: We're seeing more failure analysis that may not have been done because of the methods required. Earlier, I was discussing potted samples, and recently we've been see- ing a lot of potted assemblies. I think they're being tested instead of just treated as "no trou- ble found," because of the lack of components. I think there are a lot of issues and people are afraid that they're not going to be able to get BGAs or they are trimming back on anything that could be considered as a failure. Depotting these parts is a long chemical exposure, very time-consuming, and fairly expensive. Because of the difficulty in get- ting components right now, people are doing more failure analysis than before to stem any failures on a larger scale. As component scar- city becomes a real issue, they will do some FA (failure analysis) on parts that may have been scrapped before. ere was a paper from the mid-'90s that showed ROSE testers serial numbers one, two, three, would come back with different results on three different boards. It really becomes more important for each individual CM to use their equipment on their boards and say, "We're getting repeatable numbers." It doesn't matter if it's one microgram per square inch or if it's a hundred micrograms per square inch, it's what their equipment is measuring in their environment. e SIR drives the dataset that says, "ese materials in this process should work for our product." Take that, along with either some additional ROSE testing or the more advanced ion chro- matography testing, to determine how clean your boards are. at creates a benchmark for process monitoring. You can still use your ROSE tester to do cleanliness monitoring. ey're not telling you that you can't use it, they're just saying you need a better dataset to determine your acceptability criteria as it's no longer 1.56. You still get to do the same test- ing for process monitoring, but now you need a better objective evidence dataset before you start doing the test. Johnson: e ultimate test for reliability is the field failure rate. I would assume the assembler wants to be successful at that metric because that reflects on the assembler. Camden: Oh, for sure. And in many ways: repu- tation and bottom line. Johnson: It puts more of the decision-mak- ing power in the hands of the assembler. Does the typical assembler possess the expertise for that? Camden: e tier ones have that expertise. But from my experience the vast majority of the CMs that we work with are aware of this drive toward the new numbers and are determining it for themselves. ey've been proactive to this new objective evidence way of thinking.