|
|
|
|
|
|
|
|
Evaluation of 13 Solder Pastes for High Rel
Analysis Lab |
|
Authored By:Scott J. Anson Ph.D., PE LeTourneau University, Longview, TX, USA Michael McLaughlin and Abner Argueta IEC Electronics Corporation, Newark NY, USA Narration and Analysis:
A Lean Six-Sigma Master Blackbelt, Jim has a wealth of knowledge in soldering, thermal technology, equipment and process basics. He is a pioneer in the science of reflow. TranscriptThe authors undertook this project to objectively choose the most robust solder paste available to support the manufacture of increasingly complex high reliability products, assembled at a wide variety of global locations. This paper presents a multivariable test and analysis methodology for solder paste evaluation. The following performance characteristics were tested:
Test boards incorporated backward compatible SAC BGA's, bottom terminated components (QFN's) and other packages. Formal statistical analysis tools were utilized extensively in the evaluation of test data. A weighted matrix was used to integrate all test data into a single ranking value. Results show significant performance variations between paste formulations (not all pastes are created equal). They also indicated that the development of flux chemistry for tin-lead, water soluble pastes typically used for high reliability products has lagged behind lead-free and no-clean pastes. This was demonstrated by lower performance in many individual tests and overall rankings. I feel that this last finding sheds significant light on the growing interest in the seemingly incongruous practice of cleaning no-clean pastes for high reliability applications. SummarySolder paste is the most important material in Surface Mount Technology (SMT) assembly, and contract manufacturers of high reliability electronics face the simultaneous challenges of multisite operations combined with mission critical customer applications. It is essential to select a solder paste that performs at a high level under optimum conditions and also gives stable performance under suboptimal conditions, such as common environmental and manufacturing time fluctuations. This study examined 63Sn/37Pb solder paste alloy with both 63Sn/37Pb and 96.5/3.0Ag/0.5Sn (SAC305) lead free components in mixed alloy reflow. Both water soluble and no-clean chemistries were evaluated. A design of experiments (DOE) approach was employed to examine solder balling and slump resistance. Additional experiments examined print quality utilizing automated Solder Paste Inspection (SPI) in terms of bridge resistance, CSP and 0201 insufficient resistance, and response to a 45 minute print pause. Assemblies were also built to check for defects, joint quality, and voiding. The result is a holistic approach to solder paste testing that includes weighting of numerous factors to assess solder paste performance and robustness to external variation such as humidity and exposure time. ConclusionsEvaluation of solder pastes is a complex multivariable problem. It is not sufficient for a solder paste to excel in many areas but be inadequate in one or more critical areas. Rather, an assessment of best overall must be made to determine a most effective solder paste. It is quite possible that the best solder paste overall not be the best in any one test criteria. The variation of humidity and exposure time over each paste allowed for selection of a robust product that will run in a quantified range of operating environments. The use of the modified FMEA with weighting factors, allowed numerous criteria to be analyzed at one time. This combined with ANOVA in designed experiments and simpler low sample size experiments, provides a practical and balanced approach to selecting an optimum paste when multiple performance criteria are at play. As a group the seven no-clean solder pastes greatly outperformed the six water soluble solder pastes in every category, but especially in slump, solder balling and printing. Discussion with the solder pasted vendors confirmed that most new flux effort goes into lead free chemistries with no-clean first, followed by water soluble. Then remaining effort goes into 63Sn/37Pb chemistries with most effort in no-cleans and then water soluble chemistries. Some vendors disclosed that any 63Sn/37Pb flux development is simply a testing of lead free chemistries in 63Sn/37Pb alloys. It became evident that major solder paste vendors have limited experience blending 63Sn/37Pb water soluble solder paste in type 4 particle size. In order to prevent any supply chain delays, a safety stock of one standard shipment was always held in reserve at the paste manufacturer such that when a shipment is ordered and a batch does not pass stringent quality controls, the safety stock batch can still be shipped. The paste vendor's limited development effort 63Sn/37Pb water soluble sold pastes and limited experience in blending type 4 powders in this metallurgical and chemistry combination, necessitate detailed quantitative paste evaluation and acceptance criteria. This is an result of 63Sn/37Pb water soluble type 4 being the unique domain of high reliability manufactures that produce mission critical electronics where weight limitations result in smaller components and finer paste particle sizes. High reliability electronics manufacturers find themselves in the domain of mission critical products that are built with 63Sn/37 Pb for maximum reliability, and yet these same solder pastes are not the research or product focus of majorsolder paste manufactures. There seems to be a vacant niche for a solder paste manufacturer to differentiate themselves with a focus on 63Sn/37Pb water soluble solder pastes in type 4 for the high reliability market sector. Initially Published in the SMTA Proceedings |
|
Comments
|
|
|
|
|
|
|
|
|