Stencil Technology and Design Guidelines for Print Performance.
As the squeegee blade travels across the stencil during the print cycle, solder paste fills the stencil apertures. The paste then releases to the pads on the board during the board/stencil separation cycle. Ideally, 100 percent of the paste that filled the aperture during the print process releases from the aperture walls and attaches to the pads on the board, forming a complete solder brick. The ability of the paste to release from the inner aperture walls depends primarily on three major factors:
* the area ratio/aspect ratio for stencil design
* the aperture side wall geometry
* the aperture wall smoothness.
The first factor is aperture design-related while the other two factors are stencil technology-related. The area ratio is the area beneath the aperture opening divided by the area of the inside aperture wall; area ratio = [(LXW)/(2(L+W)T)]. Historically, the aspect ratio is the width of the aperture divided by the thickness of the stencil; aspect ratio = W/T. The generally accepted design guideline for acceptable paste release is [greater than]0.66 for the area ratio and [greater than]1.5 for the aspect ratio.
The aspect ratio is really a one-dimensional simplification of the area ratio. When the length (L) is much larger than the width (W), the area ratio is the same as the aspect ratio. When the stencil separates from the substrate, paste release encounters a competing process: Will it transfer to the pad on the substrate or will it stick to the side aperture walls? When the area of the pad is greater than two-thirds of the area of the inside aperture wall, the paste will probably achieve 80 percent or better paste release.
A laser-cut stencil that is electropolished definitely has smoother inside aperture walls than a non-electropolished laser-cut stencil. Therefore, the former will release a higher percentage of paste than the latter at a given area ratio. Likewise, an electroformed stencil with mirror-type aperture wall finish will release even a higher percentage of paste at the same area ratio. For aspect ratios that approach 1.5 and area ratios that approach 0.66, some stencil technologies are better suited than others to achieve higher percentages of paste release.
The aspect ratio and the area ratio are important considerations when designing stencil apertures. For example, a 20-mil pitch quad flat pack (QFP) with an aperture design of 10 mil X 60 mil in a 5-mil stencil has an aspect ratio of 2.0 and an area ratio of 0.86. Good print performance can be expected with this design using a good quality laser stencil.
However, consider a 20-mil micro ball grid array (microBGA) with a 10-mil aperture in a 5-mil-thick stencil. Because the aperture is round or is a square with rounded corners, the area ratio is the deciding factor. In this case, the area ratio is 0.5, which is well below the recommended value of 0.66. The aperture design can be changed by reducing the stencil thickness or increasing the aperture size, or a stencil technology can be chosen that gives better paste release at this area ratio.
Basically, five stencil technologies are being used in the industry: laser-cut, electroformed, chemical etched, plastic and hybrid. Hybrid is a combination of chemetch and laser-cut. Chem-etch is very useful for step stencils and hybrid stencils.
Laser-cut is a subtractive process. The Gerber data is translated into a CNC-type language that the laser understands. The aperture is cut out by moving the laser head only, moving the table holding the stencil only or a combination of each. The laser beam enters inside the aperture boundary and traverses to the perimeter where it completely cuts the aperture out of the metal, one aperture at a time. The smoothness of cut depends on many parameters, including cut speed, beam spot size, laser power and beam focus. The typical beam spot size is about 1.25 mils. The laser can cut very accurate aperture sizes over a wide range of size and shape requirements.
As with chem-etch, the laser-cut aperture size must be adjusted to the post-processing treatment employed because aperture size change will occur during this process. Figure 1 shows scanning electron microscope (SEM) pictures of laser-cut apertures with no electropolish, with electropolish and with electropolish followed with nickel plating.
Electroformed stencils are made by an additive process as opposed to the subtractive process used for chem-etch and laser-cut. A nickel bath containing nickel ions and a nickel hardening additive is used to electroplate onto a substrate called a mandrel. However, photoresist is first applied to the mandrel. The resist is exposed and developed, forming photoresist pillars anywhere an aperture must be in the stencil. Nickel is electroplated out of the bath one ion at a time until the desired foil thickness is achieved. Then, the nickel foil is removed from the mandrel, creating the completed stencil.
Smooth sidewalls are inherent in the electroform process because the walls take the form of the photoresist pillars exposed on the mandrel. Small and large apertures can be achieved with precise size control. Because the stencil already has smooth aperture walls, additional post-processing steps like electropolishing to make the walls smooth are not necessary. Thus, post-processing aperture size compensations are not needed. Figure 2 is a SEM picture of the sidewall of an electroformed stencil at 250X magnification.
A recent study suggests that area ratios lower than 0.65 can provide acceptable print quality.  The print performance of five different stencils with area ratios ranging from 0.4 to 1.3 was studied. Four laser stencils were used: laser with electropolish (L-EP), laser with electropolish and nickel plate (L-EP/NP), and two lasers (Laser A and Laser B) with other post-processing. One electroformed stencil (E-FAB) was used. All stencils were 5 mils thick.
Figure 3 shows the relative paste volume (actual/theoretical times 100) for a 20-mil pitch QFP with 9 X 80 mil with an area ratio of 0.81. Figure 4 depicts the solder bricks for this same QFP. The EFAB gives the highest relative solder volume, but all five stencils give acceptable relative solder volume. The solder bricks appear well defined with no sign of insufficient pad coverage.
Figure 5 shows the relative paste volume for a 20-mil pitch microBGA with 10-mil square apertures with 2.5-mil radius corners. The area ratio for this aperture design is 0.5. The E-FAB gives good solder volume, but the others provide insufficient volume. Figure 6 depicts the solder bricks, which show good pad coverage for the E-FAB, but incomplete and inconsistent coverage for the other stencils.
These data demonstrate the roles that both stencil aperture design and stencil technology play in print performance. Specifically, when the stencil printing application dictates a very tight area ratio, select an electroformed stencil. Also, try to design stencil apertures with large area ratios. When the area ratio is at 0.8 or above, laser and E-FAB give acceptable print results. However, with an area ratio of 0.5, E-FAB is the only technology in this study to give acceptable print results.
Table 1 is a useful stencil aperture design guide that relates aperture size, area ratio and stencil technology.  Although the 20-mil pitch microBGA has good print performance with a 0.5 area ratio, Table 1 recommends area ratios of 0.55 or higher for E-FAB. The print setup for the study was optimized and may not be achievable for everyday printing. Table 1 is for a 5-mil-thick stencil.
Depending on the application, steps may be required in a stencil. Some of these applications are listed below.
Step-down area for fine-pitch components
One example is an 8-mil-thick stencil for printing solder paste in and around a through-hole connector with all other surface-mount components stepped down to 6 mils.
Relief step on the board side of the stencil
Relief steps are desirable when a protrusion or high spot on the board prevents the stencil from gasketing during printing. Examples are barcodes, test vias and additive trace lines. Relief step pockets are also used for two-print stencils.  Two-print stencils are used mainly with mixed technology requirements such as through-hole/surface mount or surface mount/flip chip.
In the through-hole/surface-mount case, the first stencil prints all of the surface-mount solder paste with a normal 6-mil-thick stencil. The second stencil prints paste for all of the through hole components. This stencil is normally 15 to 25 mils thick to provide sufficient solder paste for the through holes. A relief step (typically 10-mil deep) is on the board side of this second print stencil at all locations where surface-mount solder paste was printed during the first print. This relief step prevents smearing of the surface-mount solder paste during the through-hole printing.
An example of a step-up stencil is one that is 6-mil thick in all locations except in the area of a ceramic BGA where the stencil is 8-mil thick. Another example is a stencil that is 6-mil thick except in the area of a through-hole connector where the thickness is 10 mil. In this case, the 6-mil-thick area should be at least as wide as the squeegee blade.
When designing apertures, the type of stencil and the aspect ratio/area ratio must be considered. When the area ratio is 0.8 or above, a wide range of stencil technologies can provide acceptable print quality. However, when the stencil design requirement dictates an area ratio between 0.55 and 0.80, electroformed stencils should be considered.
Step stencils still play an important role in stencil design. Applications where step stencils are very convenient include ceramic BGAs, raised via pads on the substrate, additive traces on the substrate, surface-mount/flip chip mixed technology and through-hole/surface-mount mixed technology.
The author would like to thank Jay Hinerman, Jerry Hingtgen and Jeff Schake of DEK for performing the printing and providing the solder volume data for the study.
William E. Coleman is vice president of technology with Photo Stencil, Colorado Springs, CO.
(1.) Coleman, W. (2001). Stencil technology: Paste release versus print area ratio. Proceedings of APEX 2001.
(2.) Coleman, W. (2000, July) Stencil design for mixed-technology placement and reflow. SMT.
The Impact of Stencil Cleaning on Surface-Mount Print Yields
According to one study, 51 percent to 72 percent of all solder defects are the result of screen-printing problems.  Particularly when smaller components are required, an unclean stencil or one that has been altered by the cleaning process will increase misprints while decreasing throughput and profitability.
A primary way to assure consistent solder paste deposition is to guarantee stencil cleanliness. Stencils must be cleaned to facilitate optimal prints. Solder paste residue on a stencil can interfere with good gasketing of the board to the stencil or transfer paste from the bottom of the stencil onto subsequent board prints. Solder paste left in stencil apertures can impede paste transfer, resulting in open solder connections or insufficient solder joints.
However, selecting the proper stencil cleaning process is not as easy as when use of a CFC solvent and a vapor degreaser were permissible. Today's assembler must be concerned with stencil cleanliness and also aware of user safety, environmental concerns and profitability.
Stencil cleaning has been identified as the most hazardous process with the greatest potential environmental impact associated with printed circuit board (PCB) assembly. The lead contained in the solder paste is poisonous. The chemicals used for cleaning pose concerns for fire, explosion, high alkalinity, user health, fetal disorders, scalding water, CFC and VOC emissions and heavy metal wastewater discharge.
Because few stencil cleaner manufacturers will guarantee the performance or environmental soundness of their systems, PCB assemblers must evaluate stencil cleaners for efficacy, safety and health concerns, while balancing environmental restrictions with cost and overall processing speed. In addition, because no industry standards for stencil cleanliness exist, suppliers can claim that they have a safe and effective method of stencil cleaning without having the hard data to back these claims.
However, the U.S. Environmental Protection Agency (EPA) and some state EPAs now offer programs to help set standards and certify stencil-cleaning processes for environmental and user safety and verify the manufacturer's claims of efficacy. Processes that do not have EPA certification should be scrutinized closely to determine the reasons why.
In some cases, cleaning the surface of the stencil can be accomplished without any chemical; a roll of lint free paper simply wipes the stencil. However, thorough cleaning of the surface and apertures requires the use of a chemical, and herein lies the problem.
When selecting a stencil cleaning process, the chemical should be evaluated first. The cleaning machine is secondary and will only determine the initial capital cost, how the chemical is applied, footprint, user interface and maintenance access. The chemical will dictate the wetting or solubility of the contaminant, user exposure to health and safety hazards, operational cost, environmental impact, odors or hot water vapors, cycle times, ability to clean different solder pastes, energy use for those chemistries requiring elevated temperatures, maintenance schedules, storage and transportation requirements, waste management procedures, and exhaust or other special installation requirements, especially if the chemical is a low flash point solvent such as alcohol to isolate potential fire and/or explosion hazards.
Once the chemical is identified, the method of applying it can be determined. Manual application is by far the most common methodology worldwide. However, manual cleaning usually yields the least desired results: blocked apertures, stencil damage and inconsistent cleaning.
Mechanical application of the chemical can be achieved by one of three available technologies: spray in air, spray under immersion or ultrasonic cavitation. High-pressure sprays are known to bend delicate land mass areas between fine-pitch apertures. However, without high-pressure, the chemical may not penetrate into apertures smaller than 20-mil pitch, which results in residual contamination.
Assuming the chemical is compatible, ultrasonic cavitation has become the technology of choice to apply the chemical. Ultrasonic cavitation can deliver the chemical safely into ultra-fine-pitch apertures, thereby achieving optimal cleaning results. However, without the proper chemical, ultrasonic cavitation will have little effect.
Whether using sprays or ultrasonics, without the proper chemical to wet or dissolve the contaminant, the results will be similar to washing soiled hands without the proper soap--less than 100 percent effective.
1. Surface Mount International Technical Program. (1993.) Vol. 1, pp. 157-166.
Bill Schreiber, President
Smart Sonic Corporation
Relative paste volume of 20-mil pitch QFP. Stencil Type Relative Volume L-EP/NP 82.7 L-EP 81.4 L-A 88.9 E-FAB 93.6 L-B 87.0 Relative paste volume of 20-mil pitch BGA. Stencil Type Relative Volume L-EP/NP 64.71 L-EP 55.53 L-A 52.30 E-FAB 91.82 L-B 50.50