Evaluation of Solder Paste Printability with Solder Paste Inspection Equipment After Extended Stencil Cleaning Cycle Intervals

Extending stencil cleaning intervals to 50 prints maintained consistent deposition while improving throughput and reducing downtime.
by Takeshi Yahagi, Shantanu Joshi, Jasbir Bath, Ray Welch, Joel Scutchfield and Brent Fischthal

Solder paste printing is an essential part of the electronics manufacturing process. Having a solder paste that can print high solder paste volumes with high repeatability of paste volume will improve process yields. Solder paste inspection is used to validate paste printing volumes and consistency of printed paste deposition. During paste printing, under stencil cleaning is employed to ensure high printed paste volumes are maintained between board prints to reduce printing defects. The frequency of under stencil cleaning varies based on the solder paste used, the product board type, the stencil thickness and stencil aperture openings used during paste printing.

The industry needs a solder paste that has good paste printing volume and requires less stencil cleaning during paste printing. Extending the interval between stencil cleaning cycles will help improve manufacturing throughput and reduce downtime.

The following sections will discuss an evaluation of solder pastes during paste printing with extended stencil cleaning intervals between board prints. Solder paste inspection equipment was used to assess the volumes of solder paste printed. It will help to show how robust statistical analysis and inspection technologies for printed solder paste can offer the industry a method for improving process efficiency, irrespective of the specific brand or product used.

Experimental

The company test vehicle used in the evaluation is shown in Figure 1. The whole test board was printed with solder paste. Solder paste inspection (SPI) was conducted at the locations indicated on the test board. The board pad locations inspected by SPI equipment were the QFP pads (X=0.25mm Y=1.5mm gap 0.15mm), 0.225mmφ CSP pads, 0.250mmφ CSP pads. 0.275mmφ CSP pads, 0.30mmφ CSP pads, (X=0.225mm Y=1.5mm gap 0.175mm) pads, (X=0.20mm Y=1.5mm gap 0.20mm) pads and (X=0.25mm, Y=1.7mm gap 0.25mm) pads. The board size was 100mm x 100mm with a board thickness of 1.6mm.

The board was FR-4. A production paste printer was used with production SPI equipment. Two solder pastes were used in the evaluation. These were Company Paste A and Industry Paste B. was Type 4 no-clean Sn3Ag0.5Cu lead-free halogen-free paste. It would be typical to conduct dry under-stencil cleaning after 15 paste prints with this paste. For Industry Paste B, it was again a no-clean Sn3Ag0.5Cu lead-free halogen-free Type 4 paste. For Paste B, for paste printing, a stencil release rate of > 5mm/sec was preferred with a typical blade pressure of 1 to 1.6lb/linear inch of squeegee blade length. No indication of under-stencil cleaning frequency during paste printing was provided on the Paste B datasheet. The Paste B datasheet indicated that the stencil life at room temperature was 8 hours to maintain consistent paste-printing transfer efficiency (TE).

Continuous paste printing was conducted without stencil cleaning until paste-print issues appeared on the board. Prior to continuous paste printing, the stencil was cleaned using a vacuum-dry stencil cleaning process. The paste print geometry and visual uniformity were closely examined to confirm that each paste print was not only functionally consistent but also visually identical to the prior paste prints for Pastes A and B. Solder paste inspection was conducted after each board print. The evaluation looked to extend the interval between stencil cleaning cycles to approximately 50 consecutive prints for Pastes A and B, with the aim of improving manufacturing throughput and reducing downtime with the solder pastes used.

The paste printer settings used for Paste A and Paste B are shown in Table 1. The metal squeegee blade used had a 55-degree angle during paste printing, and the squeegee blade length was 250mm.

The board pad dimensions for the component pads on the board, which were measured by SPI for solder paste volume, are shown along with the stencil area ratios for these board pads in Table 2. The stencil aperture openings were 1:1 with the board pads. The stencil thickness was 0.12mm, and the stencil was laser cut with electrolytic polishing, with no coating applied to the stencil. The stencil foil was stainless steel (SUS304), and the stencil was manufactured in 2023.

From Table 2, the stencil aperture area ratios ranged from 0.5 to 0.87 on the test board. A stencil aperture area ratio greater than 0.6 should provide good paste transfer efficiency during paste printing, according to the IPC 7525 standard^[1]. There were 404 board pads measured on the test board by SPI, which were broken down by QFP pads, CSP pads, and the other pads on the boards. There were 64 QFP pads on each board. The results for SPI and visual inspection of the printed pads for each board, up to 50 consecutive prints, are shown in the following sections for Paste A and B.

Paste A and Paste B printing results (QFP pads). Figure 2 shows the paste printing results for the QFP board pads for Paste A and B.

Based on SPI and visual inspection results of the paste printed boards for the QFP board pads, Paste A has good paste printing up to 50 paste prints without under stencil cleaning, and Paste B has good paste printing only up to 20 paste prints without under stencil cleaning. For Paste A, there was evidence of consistent paste printed shape and no signs of paste smearing/bleeding during the 50 paste prints. The printed paste deposit also retained sharp printed-paste contours, with no bridging between adjacent print deposits. For Paste B, there was evidence of solder paste print smearing after 10 paste prints and significant shorts after 20 paste prints.

The paste volume information from the SPI equipment was exported for offline statistical analysis using production statistical software to perform the process capability study. Analysis of the SPI data for each printed paste board was conducted for Pastes A and B on the QFP board pads, as indicated in Figure 3 to Figure 6.

An LSL (lower specification limit) of 60% and a USL (upper specification limit) of 160% was used for SPI data analysis. Up to 50 paste prints with Paste A, all the QFP pads have much the same print response, averaging 86% TE, with a CV (coefficient of variation) of 5% (which is very good), with the printed paste volumes adequate for the QFP board pads. CV is defined as the ratio of the standard deviation to the mean (CV = σ/μ). It shows the extent of variability relative to the mean of the population. If 99% of the data falls within three standard deviations (3 sigma) of the sample mean, the coefficient of variation (CV) would be around 16%, which would be termed a marginal 3 sigma process. A CV goal would be less than 10% to provide a wider process window. A 5 sigma process would have a CV of 10%, and an 8 sigma process would have a CV of 6%.

From Figure 3 and Figure 4, with up to 50 paste prints with Paste A, all the QFP pads have much the same print response, averaging an 86% TE, with a CV of 5%. Based on Figure 5 and Figure 6, with up to 20 paste prints using Paste B, all the QFP pads show similar print responses, averaging 90% TE, with a CV of 5%, and the printed paste volumes are adequate for the QFP board pads. Both pastes performed well, with Paste A showing good printability up to 50 paste prints without under stencil cleaning, compared with up to 20 paste prints with Paste B.

Paste A and Paste B printing results (0.3mm diameter CSP pads). Analysis of the SPI data for each paste printed board was conducted for Pastes A and B for the 0.3mm diameter CSP board pads (area ratio of 0.6) as indicated in Figure 7 to Figure 10.

From Figure 7 and Figure 8, with up to 50 paste prints with Paste A, all the 0.3mm diameter CSP pads have much the same print response, averaging 90% TE, with a CV of 4%. Based on Figure 9 and 10, with up to 20 paste prints using Paste B, all 0.3mm diameter CSP pads show similar print responses, averaging 91% TE, with a CV of 4%, and the printed paste volumes are adequate for the QFP board pads. Both pastes showed good printability.

Paste A and Paste B printing results (QFP + CSP + Other pads). The SPI data for 50 board prints with Paste A for all the board pads, which included QFP, CSP and other pads, and 20 board prints with Paste B for all the board pads, which included QFP, CSP and other pads, are shown in Figure 11 to Figure 14.

For Paste A SPI data and analysis as shown in Figure 11 and Figure 12, for up to 50 paste prints, all the pads have much the same print response, averaging 87% TE, with a CV of 6% with the printed paste volumes adequate for the board pads. For Paste B SPI data and analysis, as shown in Figure 13 and Figure 14, for up to 20 paste prints, all pads have similar print responses, averaging 89% TE with a CV of 7%. It was noted for Paste B in Figure 13 that there were low volumes of paste printed during the second board print, with a volume of 30%, and during the 20th board print, with a volume of 20%. This may be due to the challenging stencil aperature area ratios used on some parts of the board.

Overall, Paste A showed good paste printability up to 50 paste prints without under stencil cleaning being needed, and Paste B showed reasonable printability up to 20 paste prints without under stencil cleaning being needed. Paste A needed much less stencil cleaning than Paste B, which would translate into higher paste printing production efficiency.

Conclusions

The study showed that Paste A had good printability up to 50 paste prints without under stencil cleaning being needed, and Paste B showed reasonable printability up to 20 paste prints without under stencil cleaning being needed. Paste A needed much less stencil cleaning than Paste B, which would translate into higher paste printing production efficiency.

For Paste B, the smearing and bridging observed visually could not be detected based solely on changes in transfer efficiency (volume %). As a future task, additional SPI parameters and/or image-based indicators should be evaluated to detect such behavior.

The study showed that robust statistical analysis and solder paste inspection technologies can be used to develop a reliable method for assessing and improving process efficiency.End of article content

References
1. IPC-7525, Stencil Design Guidelines.

Ed: This article was first presented at the 2026 Apex Expo Advanced Electronic Packaging Conference, with original work published within conference proceedings, and is republished here with permission of the authors.

Takeshi Yahagi is with Koki Co. (koki-global.com). Shantanu Joshi is SME – Advanced Soldering Technology at Koki Solder America (koki-global.com). Jasbir Bath is support advisory engineer with Koki Solder (koki-global.com). Ray Welch is customer program manager at Koh Young Technology (kohyoung.com). Joel Scutchfield is general manager of SMT and advanced packaging at Koh Young Technology (kohyoung.com). Brent Fischthal is head of global marketing at Koh Young Technology (kohyoung.com); brent.fischthal@kohyoung.com.