The Printer and the Pea
Seemingly small issues can have big print implications.
Most reading this column are probably familiar with "The Princess and the Pea" fairy tale by Hans Christian Andersen, wherein a tiny pea under 20 mattresses keeps a true princess from a good night's sleep. Similarly, in the stencil printing process, seemingly small settings or interferences can prevent a good print outcome – particularly as assemblies continue to move toward higher densities and miniaturization. Our team had such a scenario recently while troubleshooting the cause for incomplete deposits on a complex board assembly.
This project involved printing on ultra-thin substrates – approximately 400µm thick. That's about the thickness of a playing card, so very, very thin, which makes for challenging print stability. Moreover, the panel was on the larger side, about 400mm x 400mm made up of multiple PCBs with no routing, making it more difficult to ensure coplanarity during printing. Tooling stability is always critical and even more challenging with very thin substrates. Today's tooling support solutions number a handful of options: Equipment standard tooling pins, reconfigurable pin systems like Grid-Lok, smart automatic pin placement, and dedicated tooling blocks with and without vacuum. Given the size and thickness of this substrate, our team used dedicated tooling blocks with vacuum to offer the most robust support for the process analysis.
We were printing only one side of the substrate, which contained apertures ranging from 01005s with dimensions of 200µm x 150µm to larger apertures measuring approximately 1.5mm x 2mm, and were using an 80µm thick stencil. The smaller apertures put the area ratio below the generally acceptable 0.6 rule, which also added to the challenge of this print. Finally, underneath the live print area for each PCB – on the opposite side of the substrate – there could be no contact, so any tooling support had to sit between the live print areas. Envision multiple 20mm x 20mm pockets, each surrounded by metal webbing in a grid configuration. Around the perimeter, vacuum channels secured the substrate into the tooling block.
The team honed in on the 01005 apertures due to the area ratio challenges, and the initial prints of the smaller dimensions were excellent. The pressure and speed balance were good and met the print quality criteria. The larger apertures did not fare as well, however. Upon inspection, we discovered that the larger apertures were missing material on the front edge in the direction of the print – or where the blade first passes over the aperture. When the squeegee traversed in the opposite direction, the back side of the aperture experienced the same phenomenon. Our engineers had observed this under other conditions – when using rubber squeegees or when printing too quickly with certain materials, for example. But this was unexpected with a metal blade positioned at the proper angle and with recommended pressures. Beginning with speeds and pressures, adjustments were made that only slightly modified the amount of material missing on the front edge of the larger apertures. Next, the team moved to the squeegee blade angle, going from a 60° angle to a 45° blade, but the problem persisted. Changing the solder paste material to a formula preferred in our lab also had little effect.
Going back to the basics, and as I always recommend in these columns, our team dissected the process parameters from the beginning, from the bottom up. The culprit was revealed by taking out the stencil, leaving the tooling setup in, bringing the board under control, and then using a dial test indicator (DTI) clock to measure the flatness of the substrate in the tooling block with the vacuum on. Because the panel was very thin and the vacuum was set to hold it tight, the vacuum pull was sucking the board into the pockets in the tooling plate. When analyzed, it was clear the substrate was slightly scalloped and wasn't being presented flat to the stencil; a gap was interfering with the filling process. A simple adjustment to the vacuum flow eliminated the filling issue, and the substrates printed perfectly with the original squeegee angle, speeds, pressures and specified solder paste. Problem solved! Little things can, indeed, have a big impact on the result.
is global applied process engineering manager at ASMPT (asmpt.com); clive.ashmore@asmpt.com. His column appears bimonthly.
