Issue link: https://iconnect007.uberflip.com/i/1288481
120 PCB007 MAGAZINE I SEPTEMBER 2020 land size to land size; pad to pad size is meaningless. Instead, the edge of the laser drill to the edge of the mechanical drill, as they get closer to each other, start to fail at a higher percentage; the farther they go away, the saf- er they become. Again, you can use production house technology and generate your own cou- pons, and I highly recommend that the OEM checks the coupons that are generated by the board fabricator to make sure everything is cor- rect. Again, the D-coupon matches the struc- ture of your PCB. You are going to evaluate your structures in your process and the coupon that matches your product. If the coupon pass- es, your boards will be good in assembly. The coupons go into a chamber. The OM tests are currently a 24 position test. You have coupon positions and can define different strat- egies of interconnects. In some cases, you may need two or three different coupons to do all of the different drill spans you have designed into your board. Each coupon can only have two different nets. If you have propagated vias of various configurations, you may need multiple coupons on the panel that goes into the cham- ber and is tested. A real-life test is shown in Figures 13 and 14. It is kind of embarrassing, but this really hap- Figure 13: Example of test results. Figure 14: OM test reflow statistics.