I read with interest in September Circuits Assembly magazine an article entitled "Reducing Variation Through 'Intelligent' Stencils. In this the authors propose that by REDUCING stencil thickness from 0.005" to 0.004", the paste deposit volume will INCREASE by 76%. The test aperture was a 0.010" circle. Their argument,(which I don't doubt), is that by reducing stencil thickness the transfer efficiency is increased by increasing the aspect ratio. In a nutshell - the thicker the stencil the more paste stays in the aperture, the thinner the stencil the more paste goes on the board. This would seem to make sense to me, particularly when you get down to 0.005 - 4 stencils. I very rarely use circular apertures and would be interested in any feedback from "mounties" in production facilities. Anyone tested it? What were the results etc. Cheers, Darby.
I have used both 4 and 5 mil stencils for 10 mil round apertures. We don't have volume measurement capability but we measure height and the 4 mil appeared to have more paste and better uniformity from a visual standpoint than the 5 mil. The unfortunate thing about the 4 mil is that it left very small joints on the larger components like Tant D, coils, etc... So a step stencil may be required for the 4 mil depending upon application. We went back to the 5 mil stencil and increased cleaning frequency and never had any problems. This component was a BGA with eutectic solder balls so we were not excessively concerned with volume as we were with presence. "intelligent stencil design" must also be used with "intellegent board design" for example if the pads were much larger for the coils and tant parts they may not have been insufficient. If you design and build your own product this can be accomplished rather easily but if you are a contract mfg. this can get difficult to achieve.