SMT007 Magazine


Issue link:

Contents of this Issue


Page 17 of 109

18 SMT007 MAGAZINE I JULY 2021 A lot of customers have started the qualifica- tion testing, and they'll come right back with live product from a first run and say, "How does this compare to what we did on the test board?" ere will be a mindset to be more thorough with your dataset for post-qualifica- tion samples to look at actual live product. Holden: Eric, with the advent of more bottom side termination, the standoff distance is min- iscule. As they go to more wafer-level pack- aging and this miniaturization, especially the standoff, is this potentially causing more issues with contamination or cleanliness? In the past, all the parts were open and easily accessed, but miniaturization in the Z-direction is really tak- ing hold, especially with these bottom termi- nations where a lot of contract manufacturers talk about voids. For one thing, you can only use an X-ray for that, but the other thing is, what else can be trapped underneath between all those terminations? Camden: Right. I always say QFN sent my first kid to college, and at this rate, it will probably send my second one too. I mean, there's not another component in the industry that has shown more failures than the QFN, and abso- lutely for the reasons you mentioned. You can have a four-mil stencil, you can print a QFN, and by the time that large thermal ground pad has soldered down, it's sometimes as low as half a mil standoff high. We've done cross-sections that show you're at a half mil standoff height. Now, if you're using a no-clean process, how are you supposed to vent out the gases? And they just don't outgas reliably 100% of the time the way they're supposed to. ey become intermixed there with your rosin or resin component of your no-clean flux, and they just stay active at the perimeter of the pad between the leads and the ground pad. It's that same standoff issue that we see caus- ing problems with an aqueous process, where you're taking that same QFN and now you're trying to wash out all the residues from under- neath there, and that same standoff height basi- cally just creates a dam around the edge of that component. And if you're lucky, you'll clean 15–20% out between the outer edge of the perimeter of the QFN going to the middle, but that same luck that lets you clean that much also exposes any of the active parts of your flux that were supposed to be outgassed at the time. Effectively you're removing a bandage, and we see a lot of issues around here with partially cleaned no-clean flux residues. It basically exposes all the active contaminants that were meant to be bound within the outer shell. It's so difficult to clean or to properly outgas no- clean flux residues from under QFNs that we see failures to this day. I don't believe it slowed down when it comes to the percentage of fail- ures that we see that are attributable straight to QFNs. QFNs and micro BGAs, any of the bottom terminated components as you men- tioned, are all showing potential for failure just because they're so difficult to process. is is one of the things that I was saying earlier about having the experience to set up a good wash process. It's knowing, "is is a very difficult- to-clean component. I have to be very mindful of where it goes into my cleaner, what my belt speeds are, because I'm going to need extra time, extra exposure to saponify and deionize water to make sure that this thing gets com- pletely cleaned between part and pad." QFNs are still a huge pain in the butt in this industry, and I don't know what the answer is. We've seen improvement with some standoff ideas where you actually put some thermal vias I always say QFN sent my first kid to college, and at this rate, it will probably send my second one too.

Articles in this issue

Archives of this issue

view archives of SMT007 Magazine - SMT007-July2021