Issue link: https://iconnect007.uberflip.com/i/1307491
NOVEMBER 2020 I DESIGN007 MAGAZINE 43 tion will begin. And we have talked about this before, but the equivalent of thousands of dollars has already been invested in this thing. What is the cost to completion? Phase one would be rated at about $140,000 to just build test vehicles. Can you read that we estimate very conservatively? Time to completion is defined as this point of a phase one report out at a public venue is nine months after the funding is secured. You may have noticed there has been some disruption of government finances in the past five months. There is some uncertainty. We have discussions ongoing about getting government entities to partner on funding through the Cor- nerstone Group. And for those of you who have tried to deal with government funding in the past, I can only recommend working through the Cornerstone Group as opposed to other funding venues that you tried to work with in the Defense Department. It is a far cleaner and more straightforward process, and the Corner- stone Group has been immensely helpful. The future is obvious at this point for this group. We must secure the funding to complete phase one. We need to start getting some real data. Then, we can share that with others. One thing that was not immediately obvious in this from the start is that there was an area reserved on each of the panels that contains those test articles that each of the shops will retain for their own testing. They can perform their own internal and proprietary testing. That is fine. Everybody got something out of this discussion when we started this project. DESIGN007 References 1. IPC, "IPC Issues Electronics Industry Warning on Printed Board Microvia Reliability for High-Performance Products," March 6, 2019. 2. Bill Birch, "Reliability Testing for Microvias in Printed Wire Boards." 3. Yan Ning, "Reliability of Copper-Filled Stacked Micro- vias in High-Density Interconnect Circuit Boards," Doc- toral Dissertation, UMD-CALCE, 2017. This article first appeared in the October 2020 issue of PCB007 Magazine. the other uses a UV/CO 2 /UV laser forma- tion technique. For metallization, both have the same electroless copper bath and the same direct metallization process offered by the same manufacturer. Half of the coupons or panels at each shop will be done with direct metallization, and half will be done with electroless copper. And I must note that comparing different laminate material was not considered in phase one. We'd love to have the luxury of looking at dif- ferent materials, but that's going to have to wait for phase two. We expected to see some failures. I am count- ing on failures. The primary measurements for our statistics are going to be the survival number of cycles and the statistical likelihood of failure. We're pretty confident we'll be able to do some good statistics on the relative signifi- cance of each individual variable. There has been a lot of work with each contributor to this problem. I'm going to be counting heavily on our friends at MTU, working on statistical analysis to see how much statistical reliability we can get from what really is—even though it is several hundreds of D coupons—still a small statistical sample. We need to see if we can determine the statistical significance of inter- actions. One of the goals here is the final report out at a public venue, and the results will be the guide. Phase two and phase three structures are material and methodology. Status Build structures and data packages are ready to release to production, and materi- als have been donated by the material sup- plier. In return for access to the data, the raw material and technical support offered by both the material and the chemical suppli- ers have been very gratifying. The simulation will commence simultaneously with the start of the build. At this point, we have taken this project as far as we can out of pocket. We're waiting for funding. Simulation for the sake of simulation is sort of a sterile exercise. Once we can commence the build of the test articles, then the simula-