Reading time ( words)
The introduction of 01005 chip components and 0.3-mm-pitch CSP devices is pushing surface mount technology (SMT) to its process limits. Miniaturization is driving the smallest and tightest-pitch components, but how far can this go before solder paste is no longer viable? How small a feature can be printed with solder paste, and can this be implemented into a production environment? George Babka, Assembléon America; David Sbiroli and Chris Anglin, Indium Corporation; and Richard Brooks, Kyzen define an experiment and guidelines.
Most factors and critical parameters in ultra-fine pitch printing have been well understood and documented for over 20 years. Some of these are squeegee speed, squeegee pressure, stencil design (technology, thickness, and area ratio), and solder paste. However, as pitch and aperture sizes shrink, additional factors start to affect solder paste deposition (transfer efficiency). What are these factors and can we control them to obtain acceptable results for transfer efficiency and minimized variability?
There are several challenges in implementing an ultra-fine pitch surface mount process. This article covers the solder paste printing portion. Six major variables contribute to transfer efficiency (see Figure 1 and Definitions sidebar): environment; solder paste; tooling; PCB design; print process; and inspection.
Figure 1. Cause and effect diagram of the solder paste printing process.
To simplify the experiments, we have eliminated several of these variables, such as solder paste, environment, and paste inspection. This allowed us to focus on the PCB design, tooling, and print parameters (such as separation speed and blade angle).
Several experiments at Indium Corporation’s Process Simulation Lab in Utica, NY tested a solder paste printer that uses a servo-driven squeegee motor to change blade contact angle. This greatly widens the process window for printing fine features. It can also vary the squeegee angle for different process conditions (after an automatic wipe, after a pause in the process, first board printed in a batch, etc.). The effect is to reduce the overall board-to-board process variation.
The solder paste was a no-clean, lead-free, halide-free material with type 4 powder.**
Tooling was optimized to provide a solid support underneath the stencil and along the entire length of the squeegee blade. Under-side support beyond the length of the board ensured that the force of the blade was distributed evenly along the entire length of the squeegee.
Print speed, separation speed, and time between prints were varied. Additionally, the contact angle was varied from 45° to 60° using the printer’s single swing squeegee head.
Volumetric paste print data was taken on a semi-automatic solder paste inspection system.*** The data was normalized and analyzed for transfer efficiency (TE) on the 0.45-mm-pitch down to the 0.3-mm-pitch devices.
Table 1. Board/stencil patterns of interest.PitchAperture/pad size (mm)Aperture/pad space (mm)0.4220.127.116.110.25; 0.200.15; 0.200.350.25; 0.200.10; 0.150.300.20; 0.150.10; 0.15
A test board incorporated 10 × 10 micro-BGA-style pad matrices of varying size (0.05–0.5 mm) and varying pad spacing (0.05–0.5 mm). A 3.5-mil (89-µm) thick, electroformed stencil with 1:1 square apertures was used. Table 1 shows the board pads and pitches.
Another variable was the solder mask around the solder pads. For very small spaces (<0.2 mm), the solder mask was completely removed between the solder pads. At larger spacing, the solder mask surrounds every pad.
In a previous experiment at Indium Corporation, two different board set ups were compared to see how tooling and rail configurations affect the transfer efficiency of the solder paste deposition as area ratios are reduced below 0.66. An optimized configuration improved the board-to-stencil gasketing. This optimized set up used vacuum tooling, eliminated edge board clamps, and had a slight drive of the board into the stencil.
We observed a much tighter TE distribution for the optimized set up of the 0.625 area ratio openings. We concluded that, as the fine pitch features are reduced below the standard 0.66 area ratio limit, the board set up and tooling process become more critical to obtaining consistent and repeatable results. This dramatic improvement motivated the experimenters to optimize the board set up before performing additional print studies.
Separation Speed and Blade Angle
Once the board setup and fixturing were optimized, the next experiment observed the effects of separation speed and squeegee angle on the transfer efficiency. The squeegee angle was adjusted in 5° increments between 45° and 60°.
Figure 2. Variability chart for volume percentage (TE): component pad 0.20 × 0.20 (area ratio of 0.56).
For the larger pad size (area ratio 0.70) there was no significant difference between the various squeegee angles and separation speed. For smaller paste depositions (area ratios below 0.66), the differences become more pronounced (Figures 2 and 3).
Figure 3. VMR chart for separation speed and blade angle.
For short-term decision making, comparing data sets in a box-plot format can be unreasonable, but fortunately a short-cut method can put entire sets together on a single page. Consider that the (100%) target value of the axis setting on the box plots indicates the transfer efficiency data could have a 1:1 relationship with the variance. This is because the standard deviation is the square root of the variance. Consequently, good print quality will have a transfer efficiency variance value less than 100%. To compress the information from multiple print trials, the variance-to-mean ratio (VMR) can be taken from a set of VMR values plotted on a line chart.1 Points on the line would represent the aperture size and pad design combination. Points that remain below 1.0 indicate that the print quality is good. An entire line below 1.0 visually indicates that the entire combination of apertures and pad designs has good print quality. Several lines on the VMR line chart can represent an alternative attribute in the process; for example, stencil separation speeds.
Entire VMR lines that remain close will show similar print quality. VMR lines or points on a VMR line that diverge can show precise differences in the stencil printing process. At times, the transfer efficiency specification limits may be unknown. The line chart in Figure 3 indicates a pronounced difference for fast separation speed. In a production setting, an adequate volume of paste is present from a transfer efficiency of 65–85% on 200–250 µm (8–10 mil) square apertures. The true target may not be 100% transfer efficiency as implied in the box-plot, but instead 65–85% TE with low variation. The VMR line charts can be used to show qualitative and quantitative differences. The actual volume, SPC specification limits, and other information remain important to clearly characterize a 0.4-mm-pitch process. However, the VMR line chart technique for precision stencil print process combination will show its benefits for characterizing attributes for fine feature assembly designs.
Table 2. Determining optimum squeegee pressure for various print speeds and print angles. Squeegee speed (mm/sec) 255075100125150200Angle (degrees)45
404555657590 5035455050657075 5530404055606570 6025404040555570
The lower squeegee blade angles (45° and 50°) provided better and more consistent results compared to the higher angles (55° and 60°). Additionally, the faster separation speed at low angles provided more consistent print volumes compared to a slow separation speed. The fast drop results were different to those expected for a more consistent paste volume. Previously, it was thought that if the release of the printed board from the stencil was slow, it would allow the paste to release more completely from the walls. The results from this experiment for these apertures appear to contradict this standard belief.
Blade Angle and Print Speed
For the final experiment, the squeegee angle was varied from 45° to 60° and the squeegee speed from 25 to 200 mm/sec. The goal was to determine the effect on transfer efficiency of the various squeegee angles and squeegee speed on different pad sizes and pitch.
First, the optimum squeegee pressure for each squeegee angle was determined. This was achieved by starting with a very low pressure (<0.5 lbs per inch of blade length) and then increasing the pressure until a clean wipe of the solder paste across the stencil was obtained. This was completed for each squeegee angle and speed (Table 2).
The print results of 0.25 mm (AR = 0.7) and 0.20 mm (AR = 0.56) pads with the two spacings (0.2 mm and 0.15 mm) were compared. The results (Figure 4 shows values for Component ID=0.200 x 0.20) showed no significant variation between prints, but that the average transfer efficiency is different. The average print volume for the tighter pitch (0.15 mm) is approximately 100%, and is an average of 15% higher than the 0.2-mm pitch. The increase in transfer efficiency is believed to be caused by the absence of the solder mask between the pads. This suggests that pads without interstitial solder mask may provide a better local gasket of the stencil to the board and thus increased transfer efficiency.
Figure 4. Variability chart for volume (%) Component ID=0.200 x 0.20.
Analysis of print height and print volume data support the idea that local stencil thicknesses differ and enhance volume transfer efficiency (Figure 5). Next, we observed the transfer efficiency results for the 0.25- and 0.20-mm aperture openings with a 0.10-mm aperture spacing. Figure 6 shows a large variation for 45 and 50° squeegee angles compared to the 55 and 60° angles (the graph for 0.20-mm aperture openings is similar).
Figure 5. Local thickness of the stencil would affect the overall area ratio and, as a result, the transfer efficiency. Analysis of the paste height indeed indicates that local thickness differs.
Figure 6. Variability chart for volume (%) Component ID=0.250 x 0.10
A previous experiment at Yamaha showed that as the squeegee angle increases (45–65°), the actual paste filling pressure is almost halved (from 13 to 7.6 KPa). One possible reason for the increase in print variation for the lower squeegee angles is too high a paste pressure, causing the solder paste to be squeezed out between the pads. This data suggests that that an optimum pressure for printing these apertures might be achieved at angles of 55° and above. Figure 7 shows larger volume variations for 0.250-mm apertures.
Figure 7. Analysis of print volume data compares volume variation among blade angle.Conclusion
Throughout the years of solder paste printing, there are well-documented reports on the critical parameters of squeegee speed, squeegee pressure, and stencil design (technology, thickness, and aperture size).
Our results have also shown that:1. The board set-up is critical in printing fine features. Specifically, the board gasketing is improved dramatically by holding the board flat without any topside clamps during printing. Additionally, supporting the stencil along the entire length of the squeegee blade improves print definition and transfer efficiency further.2. A fast separation speed can significantly reduce transfer efficiency variation. 3. As the pad size and spacing are decreased, the squeegee angle has a greater effect on transfer efficiency and the consistency of the print deposition.4. Solder mask in between ultra-fine pitch apertures can be a significant source of variation in overall transfer efficiency. 5. Changing the squeegee angle during a print run can offer significantly greater consistency after stencil wipe, changeover, or pauses in the print cycle.
These experiments revealed several new print parameters that can contribute to the process variation of ultra-fine-feature printing. As the pitch and stencil aperture openings decrease, the variables of squeegee blade angle and squeegee speed become an important factor in minimizing insufficient prints and maintaining low print variation.
For ultra-fine-feature assembly, insufficient print deposits can led to the undesirable task of rework. Even the smallest process variations have significant consequences here, so the entire printing process must be characterized prior to implementation.
* Yamaha YGP solder paste printer, sold and supported by Assembléon.** Indium 8.9HF no-clean, lead-free, halide-free solder.*** Koh Young 3020T 3D SPI system was used for inspection.
DefinitionsTransfer efficiency (TE): The amount of material deposited on the stencil in relation to the theoretical maximum volume of material that could be deposited. This is typically expressed as a percentage of the maximum volume.Print area ratio or area ratio (AR): The area of the aperture opening divided by the surface area of the inside aperture wall. For a rectangular stencil aperture: Area Ratio = (L × W)/(2 × [L + W] × T)Contact angle: A function of the squeegee blade holder; the angle formed between the stencil and squeegee blade as they first make contact with no force between them.Attack angle: A function of the contact angle, blade compliancy, print speed, and paste rheology. The attack angle is the angle formed by the sum of the static and dynamic forces acting on the squeegee blade during the print stroke. REFERENCES:
1. Anglin, C., “Establishing a Precision Stencil Printing Process for Miniaturized Electronics Assembly,” IPC Technical Conference, March 29–April 2, 2009.2. George Babka, " Moving Towards a Stable Process: Minimizing Variation in Solder Paste Printing,” Originally distributed at the International Conference on Soldering and Reliability,” Toronto, Ontario, Canada; May 20–22, 2009.
George Babka, Assembléon America Inc., Alpharetta, GA, may be contacted at firstname.lastname@example.org. David Sbiroli, Indium Corporation, Clinton, NY, can be reached at email@example.com. Contact Richard Brooks, Kyzen Corporation, Kyle, TX, at firstname.lastname@example.org. Chris Anglin, Indium Corporation, Clinton, NY, can be reached at email@example.com. Click here to send a group email to all of the authors. Originally published in the Proceedings of the SMTA International Conference, San Diego, California, October 4-8, 2009.
SMT, February 2010