Issue link: https://iconnect007.uberflip.com/i/1288481
SEPTEMBER 2020 I PCB007 MAGAZINE 115 see anything at 200x, we micro-etch, which is where we can no longer see any separation. But if we go to a higher magnification, we see what it looks like after micro-etching. The top picture of Figure 5 is at 300x. IPC requires that we start at 200x for a microvia, and if there is something that's questionable, go up to 500x. If we can't identify anything, it's deemed accept- able. But if we take those same pictures and go to 3,000x, we can see that separation is taking place. Our standards for what's required from a fabricator may not identify this separation, which leads to problems. They pass all of the cross-section evaluations and get shipped, but nobody did a thermal test of the D-coupon to show that there were weak mi- crovias in the build. Instead, it will be found in assembly when your fingers are pushing down on the component. Figure 6 is an example of the as-polished condition. On the left, you can see the three stack microvia, and there is no sep- aration. Right after this point, when somebody evaluates it, they are going to take a micro- etch and look. The center pic- ture shows a 2X tin-lead reflow, which is about 230°C, and it's a lot less stress- ful on the specimen. You can't really see a sepa- ration, but there might be a separation between the top and the middle of microvia on the tar- get path. On the far-right column, you can see a separation occurring at the target pad on layer 2 to the microvia between one and two. At that point, it should be deemed an unac- ceptable condition and non-conforming. It is so thin and slight that it could be missed. Once they ship, you are not going to see this demar- cation. Figure 5: IPC-6012 requires examination from 200x up to 500x. Figure 6: How older methods fail us.