Reading time ( words)
In order to duplicate the original manufacturing process as closely as possible, the rework of complex devices can be accomplished by paste printing solder paste in a select location. Ever since the advent of widespread usage of surface mount devices such as fine pitch QFNs or BGAs, this practice has continued to grow. Today’s pitches are commonly 0.5–0.4 mm with packages of tiny outline sizes 510 mm square, making the rework of such devices a challenge. Spacing to neighboring components continues to be compressed so the rework techniques are getting more challenging from a practical point of view.
The process of printing consistently has given rise to a more modern technique for printing solder paste in a given location for fine pitch devices without there having to be poor yields or highly skilled technicians to perform the work. A soon-to-be-reported study compares the older miniature metal stencil printing process to the more modern plastic film with adhesive approach. This report quantifies the differences in performance for the first time.
The miniature metal stencil, while matching the original SMT printing stencil, has several shortcomings with respect to the rework printing process. This technique, while perfectly capable for most SMT boards, has fallen out as preferred and capable method for several reasons. With the tighter spacing between components these “mini” metal stencils are too large for modern board layouts as there needs to be an oversized area compared to the print area for the holding and supporting of the stencil. In addition, the finer pitches and pad sizes makes it difficult, even for the most skilled rework technicians, to perfectly print in a single pass the correct volume of solder paste onto the PCB. With ever-thinning stencils due to lower paste volume requirements, these metal stencils are easily bent or damaged during cleaning or squeegeeing, meaning that it is difficult to retain a co-planarity with the PCB. The lack of coplanarity causes solder paste to squirt out from underneath the stencil. These problems make the miniature metal stencils less suitable today for the deposition of solder paste during rework.
Plastic film stencils have taken over where the mini metal stencils have left off. They offer up some distinct advantages based on the user complaints over their mini metal stencil counterparts. Their repositionable adhesive backing allows the user to move the stencil around creating micro-fine alignments on the PCB after the macro adjustments have been made. In addition, the adhesive helps ensure co-planarity with the PCB even if the board is slightly warped. The sticky backing also allows the user to use multiple swipes with the squeegee, thus ensuring the apertures are “filled” completely. The flexible nature of the stencil allows for the stencil to be used in very tight areas, which need to be printed prior to rework. These stencil designs even allow for “flaps” to be built into the stencil design and prevent solder paste from being pushed outside of the rework area requiring further clean-up. For these and other reasons the adhesive-backed plastic film stencil is now the preferred rework stencil.
In a recent study, the results of which will be published at the APEX conference in March of 2016, the printing consistency between these two types of rework paste printing stencils will be revealed. In the meantime, select data from the study as well as the results of that study are included in this discussion.
In this study, a solder paste inspection measuring system measured the solder paste volume and solder paste “brick” height at each of the pad locations. After initial calibration, an ASC AV862 offline semi-automated measuring system was utilized to take the measurements. This system used Gerber data as input information in order to find the location of each of the solder pads. After this set up the camera scans each of the locations that it recognizes as having solder paste and measures the height and volume amongst seven parameters per pad location. These results were then tabulated, graphed and analyzed.
Editor's Note: This article originally appeared in the February 2016 issue of SMT Magazine.
