Using SPI to identify root causes: early inspection can resolve end-of-line defects tied to printing errors.

Paste printing plays a key role in the quality of Pb-free manufacturing as the print process provides a vital foundation for the remainder of the SMT assembly process. To deal with post-reflow joint-related issues, it is crucial hat manufacturers understand how Pb-free will impact end-of-line quality based on specific root causes found at paste deposition. These root causes can first be identified through the structured implementation of 3-D solder paste inspection (SPI), and further mitigated with the use of 3-D SPI for better process control and identification of variations. In addition, careful examination of the SPI data after board assembly can point to the root cause of issues. This intelligence can be fed into the inspection criteria to prevent issues by establishing more effective thresholds for error and variation.

To conduct effective root cause analysis, the manufacturer must first understand how to relate specific issues found at post reflow to the paste process. Several areas of concern have been identified based on experimental studies and initial experience implementing Pb-free processes. The lower wetting forces found with Pb-free solder paste in reflow are leading to more prevalent bridging after reflow. Paste inspection systems could typically be programmed to accept a minimal amount of bridging on normal eutectic due to this reflow phenomenon. Threshold considerations will have to be revisited with the more constrained Pb-free process. In addition, volumetric and area tolerances could traditionally be wider due to the ability of the downstream process to correct for print inconsistencies with more robust reflow characteristics.

One study published last year showed that for small chip devices defects due to print registration issues were double those caused by component offset. (1) The lack of wetting forces on the offset paste contributed to a higher rate of tombstoning. This suggests an advantage to placing more emphasis and visibility around the registration of the paste print. Therefore, root cause can be conducted using registration data gained from the SPI tool. Knowledge gained from this exercise will allow the manufacturer to apply preventative measures more effectively.

To conduct this level of root cause analysis on the paste process, the correct resolution of data is required from the 3-D SPI sensor. Attribute (or pass/fail) data may not effectively identify the issues, but rather point to the fact that there are issues. These calls are subject to the user's ability to establish accurate pass/fail criteria, which can be subjective in nature. Quantitative analysis of paste volume, height and area is vital to understand and promote more effective root cause analysis due to the magnitude of variables in the printing process. By applying the quantitative data, trends and variations can be monitored and identified--which may not manifest into conclusive attribute level failure data. Also, proper characterization of the process is vital if the manufacturer expects to understand its capability and identify when variations occur. Using the trend and variation analysis against the initial process characterization, the user will gain insight to how printing issues could propagate to post-reflow related defects.

Initially, root cause analysis is applied in a more reactive fashion with the industry gaining an education on how the paste process contributes prominently to end-of-line defects. This knowledge can then be translated into more preventative measures, thus enabling a more proactive culture in the manufacturing process through more careful examination of 3-D SPI data

When considering root cause associated with the paste printing process for Pb-free manufacturing, manufacturers will gain from following specific processes and training. An understanding of the typical failure modes of post-reflow defects due to Pb-free solder characteristics is crucial so that the 3-D SPI sensor can be used most effectively to find and quantify those issues. Also, 3-D SPI must be maximized through the use of both attribute and measurement data with a focus on initial characterization of capability.

Reference

1. S. Aravamudhan, et al., "Self-Centering of Chip Components in a Pb-Free Assembly as a Function of Component and Solder Paste Print Offsets," SMTA International Proceedings, September 2005.

Jeff Harrell is automated optical inspection product manager at Agilent Technologies (agilent.com)

[ILLUSTRATION OMITTED]