Issue link: https://iconnect007.uberflip.com/i/253291
14 SMT Magazine • February 2014 technology for the production of a high-densi- ty, highly miniaturized PCB assembly 1 . The test vehicle used in that study is shown in Figure 1. The PCB design packed nearly 15,000 paste deposits in a 3x7 area; 8,500 of those were 0.5 mm microBGAs. The study used print yields, transfer efficiencies, and print volume consis- tency as metrics to evaluate a number of stencil technologies, including electroformed nickel stencils, electroformed nickel that had been laser cut, and two different types of laser-cut stress relieved stainless steel (SS). The study con- cluded that the best print quality was produced with laser-cut fine grain (FG) SS foils with two- part Self-Assembling Monolayer Phosphonate (SAMP) nanocoating applied. With the key materials identified, a new study was launched to optimize the laser cut- ting parameters on the FG SS. It tested three experimental parameter sets against the process of record (POR). Prior to the outset of the tests, a new two-part SAMP nanocoating was intro- duced to the market, so additional tests were planned to benchmark the new generation of nanocoating against the original one. In response to recent reports that cite square apertures as superior to circular ones on fine features 2 , a leg was added to the DOE that di- rectly compared the two. As the time to execute the experiments ap- proached, new experimental SS foil materials were introduced, as was a new electroforming process, so another leg was added to analyze their performance. During the execution of the tests, two additional runs were added to begin understanding the relationship between nano- coating and stencil under wipe frequency. Upon review of the results, a final run was added to benchmark the performance of a dif- ferent laser stencil cutting process. All the tests were executed in a similar fash- ion, using the same ten-print test and the same metrics for analysis. Detailed information on the derivation of the area ratio, transfer effi- ciency and coefficient of variation metrics used in this study is provided in the original report, cites as reference #1. experimental Setup Test Vehicle The original test vehicle shown in Figure 1 was used for a multitude of comparative tests over a two-year span. It is a production PCB that offers vast amounts of comparative data. The design was recently revised; the new test vehicle used in this evaluation is shown in Fig- ure 2. This new design replaced some of the mi- croBGAs with FETs, reducing the number of BGA I/Os from 8500 to 2176 per board. The number of 0201s increased on this design, from 1900 deposits per print to 3712. A 10-print test using the new TV now produces 21,760 BGA data points and 37,120 0201 data points. Test Methods For each stencil, 10 prints were produced sequentially on a well maintained and cali- brated 2009 DEK horizon stencil printer using, FINe-TuNING THe STeNCIL MaNuFaCTurING PrOCeSS continues FEATUrE Figure 1: Test vehicle used in previous tests (non- bgA circuitry on closeup is intentionally blurred). Figure 2: New test vehicle.