Stencil technology DoE
Process surveys to investigate recent advances in materials and processing techniques should be undertaken regularly by process and manufacturing engineers to make current processes more capable. The capability of current processes should be the ultimate arbitrator on deciding what processes to investigate first. The aim of these investigations is to reduce the variability of the current processes by investigating new materials, equipment, and processing parameters. The investigation should be universal in nature, affecting as many existing and new products as possible.
An example is an investigation into the technology of solder deposition using stencils for SMT PCBs. The DoE should examine alternative technologies from different suppliers, and rate the soldering quality produced by the stencil types. The following is a discussion of the issues encountered and decisions to resolve them in the DOE. These issues could be useful when conducting similar survey DoE's:
• The quality characteristic was the height of the solder “bricks” formed by the solder deposition operation through the stencil. Minimum variability of the solder brick height was shown to result in good soldering with reduced defects. The height proved to be difficult to measure because the solder brick top surface was not uniform, and individual readings of the solder brick height varied according to the presence of solder spheres in the paste. The volume of the bricks proved easier to measure by a laser detection system, and it was decided to measure the volume of the solder bricks as well. A combination of the two, the area of solder, was chosen as the quality characteristic; it is equal to the volume/height. The measurement of the variability of the solder areas was repeated several times and transformed as a S/N value. The statistical analysis was performed on the single number representing the variance of solder areas for each experiment line.
• The factors chosen were the stencil technologies available. They differed in the creation of the holes for depositing the solder paste on the surface of the PCB^. The technologies included band, chemical etching, laser drilling, and electroforming. Several suppliers for each technology were included for a total of eight levels of stencils. Other factors included the use of paste with or without aqueous cleaning (C or NC) after soldering, snap-off distance (5 or 0 mils), squeegee pressure (35 or 30 lbs), lead orientation angle of the components (90 or 0 degrees from the squeegee travel), and lift-off pressure (75 or 60 lbs).
• A specially made test PCB used was used, with 208 components, 19.37 mils lead pitch on each PCB.
• Other known factors affecting solder deposition were fixed for this DoE. They included the stencil thickness of 0.006^ (6 mils), stencils with aperture sizes of 10 x 55 mils, and using the same stainless steel squeegee for all experiments. One squeegee pass at the same speed was used for all experiments. These factors were determined to be not significant in earlier experiments.
• The stencils were used to deposit solder paste on bare copper substrate for all experiments.
An L16 orthogonal array was used. The factor assignments are shown in Table 8.7. The selected factors were assigned to specific columns so that the interactions of interest were isolated. The eight levels of stencil technology were used in a multilevel combination column, consisting of the columns 2, 4, and 8. The interaction of the columns forming the stencil levels were confounded with some of the other factors, as shown in Table 8.7 • A total of 480 points were measured for the experiments, consisting of 10 replications of the 16 experiments in the L16 design. Each replication consisted of taking 10 points measured three times on the deposited solder pattern on bare copper substance.
The L16 design is presented in Table 8.8, with the interaction columns not shown. The ANOVA analysis is shown in Table 8.9, with the percentage contributions only. The total degrees of freedom (DOF) is equal to number of experiment lines minus 1. The DOF of the stencil factor is equal to three, since three columns with two levels each were used. The interaction of the paste and stencil was also equal to three, since three stencil columns were used.
The analysis clearly indicates the importance of the stencil technology quality of the solder deposition. Of the factors selected, snap-off, lead orientation, and squeegee pressure were significant, whereas paste selection and lift-off pressure were not significant. Only one interaction, paste selection and lift-off pressure, proved significant, even if paste was not considered significant.