Automated stencil inspection: miniaturization changes the rules for stencil manufacture.

The days of setting up a stencil production facility with mesh, a roll of stainless steel and a laser are over. When the smallest components were 0402s, the average aperture count was 1200, and the average pitch was 0.65 mm. A basic stencil operation with typical lasers and a manual check was usually satisfactory. But as we move into the era of mainstream 0201 production and the early stages of 01005s, manufacturing a stencil with such small apertures and tight tolerances (approximately 5 [micro]m) will require not only the most sophisticated production equipment, but an advanced set of checks and balances as well.

[TEXT NOT REPRODUCIBLE IN ASCII]

Traditionally, stencils are cut from Gerber data supplied by the circuit board manufacturer; from this data, stencil apertures are laser-cut in the stencil material, be it stainless steel or nickel. Electroformed stencils are produced by means of an additive process, but still rely on the same Gerber data for manufacture. Provided that all the Gerber data are correct and the proper tolerances defined, this process is fairly straightforward for standard SMT stencils. Once the stencil is produced, the product usually goes through a manual five-point inspection to check aperture presence and dimensions. During this process, the technician identifies five points of particular interest on the stencil, visually inspects to ensure the presence of the aperture, and then evaluates the dimensions of each to confirm the integrity of the data and laser. In most cases, this level of process control is adequate for basic SMT production.

As I have said before, with nearly 60% of manufacturing defects directly related to errors in printing, ensuring precision in the print operation is essential for maximum line yield. Obviously, a large part of that optimization is accurate stencil production, precise board-to-stencil alignment and error-free paste on pad deposition--especially in this age of miniaturization. It is safe to say that a simple, manual five-point visual inspection will not be sufficient to assess the integrity of more than 4,000 aperture stencils with dimensions as small as 150 x 200 [micro]m with 0.4 mm pitches (and 0.3 mm pitches on the horizon)! The production and verification processes must be much more robust.

[ILLUSTRATION OMITTED]

So, what exactly does that mean? When manufacturing stencils with such fine features, there is absolutely no substitute for good Gerber data. But it is also possible that the laser being used to cut the stencil is out of calibration, and therefore could cut an aperture the wrong size or in the incorrect position. Because of this condition, when dealing with next-generation stencils, automated inspection will now be tantamount to ensuring the integrity of the fabrication process. With such small apertures at high-density pitch, verification of aperture position is equally as important as dimensional checks. As you can imagine, on a stencil with 4,000 apertures with 150 x 200 [micro]m dimensions, verification--particularly positional verification--is not something that can be done well with a manual five-point check. High-powered stencil inspection systems that can quickly assess aperture size, position and absence/presence will be an absolute necessity.

The other critical aspect of stencil manufacture is Pb-free production. Not only will stencil material selection play a major role in efficient paste release, but alignment and print accuracy must be absolute because of the non-self-aligning characteristics of current Pb-free solder paste formulations. Stencil-to-board alignment must be precise, with little room for error.

The final, notable consideration is the possibility that, even though the board is manufactured to spec, is within its tolerances, and the stencil is manufactured according to the Gerber data supplied, board-to-stencil misalignment may still occur. With such fine features, the board could be at the lower end of its tolerance and the stencil at the upper end (or vice versa) and could result in misalignment and thus manufacturing defects. Though not yet a common problem, in this situation it may be an option to use scanner tools to create data based on the manufactured PCB, in essence, generating the stencil from the physical board instead of the Gerber data to account for any slight misalignment that may cause production errors.

While stencil manufacture has often been viewed--often unfairly--as a rather rudimentary and less sophisticated part of electronics assembly, that mindset will quickly change as the era of miniaturization takes hold. High-level manufacturing and sophisticated inspection capabilities will be a must for next-generation stencil production.

Clive Ashmore is global applied process engineering manager at DEK (dek.com); [e-mail]. His column appears bimonthly.