Identifying and quantifying defects is key to achieving greater yield and higher quality using this systematic approach to problem solving. Today, SMT manufacturers find it challenging to maintain their competitiveness and profitability. The reasons are increasing complexity of SMT processes, higher customers'' demands on quality, and shorter lead time matched by cost reduction. Faced with these challenges, many manufacturers have explored manufacturing solutions to maximize resources, through optimizing processes by adopting an improvement organization culture.

One of the most widely implemented solutions is the Six Sigma program.
Some successful case studies of Six Sigma are companies such as Motorola and GE. Six Sigma is a management philosophy which aims at building world-class quality standards and quality culture. It relies on the understanding of the processes of supplier-manufacturing-customer value chain to derive a systematic methodology to optimize the processes through employment of powerful problem solving tools. Case
study: Six Sigma methodology/tools application in SMT The manufacturer was encountering a high SMT defect rate, resulting in unnecessary rework after reflow. A certain percentage of the defects even escaped into the assembly line and was distributed to customers. Realizing the seriousness of the problem, the organization decided to adopt the Six Sigma methodology as their solution in solving their quality issues. A project team was set up to focus on these issues and the following was what they did.

Define phase : In order to identify the top issues, a production total defects breakdown graph is derived for analysis. Through the diagram (Figure 1), the team realizes that the top issues affecting process quality are component soldering and placement quality. Figure 2: SMT defects breakdown by line (product) Measure phase: Next, a stratified Pareto chart categorized by different production lines (or by components) is created (Figure 2) to identify the defects. Soldering and placement defects are still the "top-common" issues for all the lines, affirming the accuracy of the identified problems. Proceeding on, a model line is chosen for study, in order to retrieve a more detailed analysis.
Analyze & Improve phase: Firstly, the focus is analysis of placement misalignment defects. In this particular example the placement machine is configured with two 12 nozzles revolver heads. To analyze the root cause, a multiple variation study is conducted to analyze head, nozzle and placement angle variation. The multiple variation chart and main effects plot chart (Figure 3) display the measurement results, indicating largest variations coming from different head and nozzle variation. On average, placement head to head varies about 30um and nozzle 3 (for example) varies about 20um.

Figure 3: Main effects analysis for placement After further investigation, the following root causes are identified and counter actions proposed: - Nozzle design: Nozzle design is not suitable for component shape.
Solution: Design a special nozzle for this type of component. - Nozzle
vacuum: Nozzle is worn out and dirty. Solution: Implement a new method to clean nozzle and set up a schedule to check and change nozzle on a regular basis. - Feeder operation: Daily procedures executed on feeders are not suitable, trouble-shooting methods are not correct and maintenance quality is not controlled. Solution: Standardize feeder maintenance items, time-interval, and checklist and operation instructions to control feeder quality. - Head calibration: Head variation is not monitored during maintenance. Solution: Update maintenance checklist and methods to control the quality of maintenance. Based on these recommendations, action tasks are implemented and maintained as placement process control plan for daily operation controls. Next, the printing process is analyzed to rectify soldering bridge and open defects. As solder volume is a critical output in reflecting soldering defects, a DOE (design of experiment) was conducted to optimize process setting in order to obtain desired solder volume specification. The experiment''s analysis (Figure 4) result at the printing process, the outputs are printing volume measurement (which reflects soldering quality) and component pull force measured after reflow (which reflects process reliability). From this overlaid plot, the desired process window for input is obtained; setting printing speed at 0.5 level and pressure at 0 level. These settings enable both solder volume and pull force specification requirements to be met, whilst achieving a robust process design window.

Figure 4: Printing parameters optimization area Control phase: In this phase, the process control plan and parameter settings are finalized based on the experiment''s results and analysis. The checklist is updated, and a standardized process is implemented and monitored by line operators and technicians on a daily basis. About one month after implementing the new process, the Pareto chart of defect type is updated and logged. A comparison of results before and after the project indicates a significant reduction in soldering and placement defects with the new processes.

Conclusion : Six Sigma fosters a solution oriented culture that focuses on improving processes, which eradicates current problems permanently and streamline them to achieve targets through the DMAIC methodology. This practice and related tools application enable project teams in SMT manufacturing to adopt a step by step process to systematically understand process characterization. Through continuously improvement, new processes to make it easier for companies to achieve benefits like higher quality, shorter lead time and cost reduction in these challenging times.