Issue link: https://iconnect007.uberflip.com/i/1345210
MARCH 2021 I SMT007 MAGAZINE 27 PCQR 2 e Printed Board Process Capability, Qual- ity, and Relative Reliability Benchmark Test Standard and Database, IPC-9151D, defines available test coupons and the tests to be performed on them. ese coupons can be arranged onto a variety of test panels, currently 11 different options primarily differing in layer count, via structures, and trace geometries. Ideally, vendors would each produce the panel that best showed the transition from successful to unsuccessful production, establishing the boundaries of their production capabilities. To facilitate comparison of results between shops, we chose to focus on three panel types—10R, 18R, and 24VH—roughly corresponding with low, medium, and high technical capabilities. Vendors self-select the panel types they wish to attempt, though vendors who easily accom- plish their chosen panel are encouraged to attempt a more difficult one on a future test cycle. Figure 1 shows the dielectric and copper specifications, and through and blind via struc- tures of the 10R-E panel intended as an entry point for the company's lowest-technology, lowest-cost suppliers. If a low technology ven- dor does not have equipment for microvia pro- cessing, they can choose not to build any or all of the V2 structures without impacting test of the rest of the panel. As a supplier's technology increases, they can advance to a medium complexity panel, such as the 18R-E as shown in Figure 2. is board focuses on a higher level of technology with a reduced via size, highlighting a suppli- er's copper plating reliability and registration capability. A more complex panel, such as the 24VH-E panel illustrated in Figure 3, allows higher- technology vendors to demonstrate more advanced via structures such as buried vias, skip microvias, and backdrilling, as well as the complex registration requirements of a higher layer count board. To better evaluate suppliers' ability to main- tain process controls over time, a test sub- mission is built in three groups of three pan- els each, with the jobs spread out over several weeks. All panels are sent, untested, to a third party for evaluation of parameters such as con- ductor and space yield, via registration and Figure 1: 10R-E panel stackup and via/trace structures.