-
- News
- Books
Featured Books
- smt007 Magazine
Latest Issues
Current IssueBox Build
One trend is to add box build and final assembly to your product offering. In this issue, we explore the opportunities and risks of adding system assembly to your service portfolio.
IPC APEX EXPO 2024 Pre-show
This month’s issue devotes its pages to a comprehensive preview of the IPC APEX EXPO 2024 event. Whether your role is technical or business, if you're new-to-the-industry or seasoned veteran, you'll find value throughout this program.
Boost Your Sales
Every part of your business can be evaluated as a process, including your sales funnel. Optimizing your selling process requires a coordinated effort between marketing and sales. In this issue, industry experts in marketing and sales offer their best advice on how to boost your sales efforts.
- Articles
- Columns
Search Console
- Links
- Events
||| MENU - smt007 Magazine
Statistical Tools Promote Bottom-line Improvements
December 31, 1969 |Estimated reading time: 7 minutes
By reducing process variation, SPC helps maximize yields, minimize rework, increase profits, satisfy customers and meet vendor-certification standards.
By Jeffery L. Cawley
Every process has natural variation. The key to achieving bottom-line benefits is to identify which variation is within the statistically predictable range and which variation is outside. If a process is out of control (i.e., showing unusual variation), work to eliminate the out-of-control conditions, then perform a capability study to determine if the process meets specifications predictably and whether tighter specifications can be met.
Figure 1. The X-bar chart (top) shows a process that is varying greatly in its average solder height, while the Range chart shows an in-control process.
Many electronics assembly operations initially use process capability analysis to select, optimize and certify production equipment. Once the equipment is up and running in a process, variable and defect tracking can be performed with statistical process control (SPC) charts to monitor daily production. If control charting reveals serious problems, an SPC engineering study can reveal the problem sources.
How a Manufacturer Might Use SPC
Keeping solder height within control limits is critical for board manufacturers - it determines how well components will sit on the board and whether copper in the components will migrate if boards are stacked. Inadequate solder height can indicate many process problems, including solder stencil wear-and-tear, a calibration problem with the machine applying the solder, inadequate stencil cleaning or a raw material problem. These problems can lead to higher production costs because of waste or rework. The following process example illustrates how to use SPC to bring an out-of-control process back in line.
Figure 2. The process is in control, but the Cpk (right) indicates excessive variation.
In this example, the specification for solder height is 0.25 to 0.55 mm. Periodically, samples of five units made at the same time are measured for solder height. The data is analyzed using X-bar/Range charting to determine whether the height variation indicates an out-of-control situation.
Data from a full day's run are charted in an X-bar/Range chart (Figure 1). This chart shows that the process is out of control on the X-bar chart, but in control on the Range chart. This indicates that the process average is being influenced externally, causing inadequate solder height. The problem might be with the raw material, the machine applying the solder or the measuring device itself. Investigation reveals that the inadequate solder heights all come from lots using raw materials from a single vendor. Further analysis of lots from that vendor reveals the vendor's inability to produce consistent material. The problem is solved by dropping the vendor as a supplier.
With the out-of-control condition removed, the process is run again and charted, using raw materials from reliable vendors only. The process is now in control (Figure 2), but suffers from excessive variation and is not capable of producing output within specifications (note the Cpk of 1.05). This situation indicates significant opportunities for process improvement because the in-control condition demonstrates that the variation source is the process itself.
Figure 3. The process is now in control and capable of producing acceptable output.
It is discovered that the machine needs maintenance and recalibration. This is done and the results for the next day's production (Figure 3) indicate a process in control and capable of producing output well within specifications (note the Cpk of 1.438).
Defect Tracking with Pareto Diagrams
After board assembly, Pareto analysis is used to examine the relative contribution of different solder-related attributes that lead to printed circuit board (PCB) rejects. Pareto diagrams commonly are used to rank the relative frequency of different defect categories. With these diagrams, a quality control staff can assess the relative contribution of different defects and assign priorities for addressing their causes (Figure 4). Defects also can be ranked by categories other than frequency of occurrence, such as cost.
Figure 4. A Pareto chart is an effective way to compare defect causes.
The same data used to generate Pareto diagrams can be plotted as percent defective control charts (p-charts), as seen in Figure 5. Several attributes - the top four defect categories, for example - can be grouped in one display (Figure 6). In the example, the four most common defects exhibit significantly different characteristics. "Pinhole/void" and "open joint" defects indicate serious early problems that appear to have been corrected (indicated by out-of-control and pattern rule violations). "Cold solder" shows the opposite - defect levels appear to be shifting upward, as the chart indicates that the early low defect rate is unusual. "Insufficient solder," on the other hand, is in perfect control, meaning that while the defect rate may be higher than desired, the process itself is performing as expected. This example illustrates the value of SPC charts - the actions taken to effectively address the four largest defect categories will be quite different.
Spreadsheet vs. SPC Software
Although SPC control charts can be produced manually, collecting and charting data disrupts the production process while putting extra burden on assembly and inspection staff. Significant room for errors exists with this method and opportunities are lost because data and charting are not integrated into vendor and in-house reporting systems. Additionally, data retrieval over any length of time is extremely difficult.
Figure 5. P-charts are created from the same data used to make Pareto charts.
Spreadsheet-based charting improves on manual methods. Electronics manufacturers can use spreadsheets to do SPC if analysis is limited in scope and primarily done by engineers. While spreadsheets offer the advantage of availability, they offer little in terms of usability. Spreadsheets take significant amounts of time to set up for SPC use. Automating analysis involves substantial amounts of macro-writing and programming into which mistakes can be built that require time-consuming debugging and logic validation. When process changes are made or analytical needs change, the spreadsheet has to be reprogrammed. The customized nature of the spreadsheet design means that having non-engineers use it for SPC can be challenging.
Specialized SPC software applications eliminate those disadvantages. They chart and analyze, and are flexible when process analysis needs change. But these software applications also can come with a set of problems. Some are very complex and difficult to use by those without an advanced knowledge of statistics. Others are not compatible with process information databases. And some are only designed for certain types of SPC analysis.
On-line vs. Off-line
On-line SPC software for plant-floor data collection and SPC charting is a valuable tool for tracking a process in real-time. Often integrated with supervisory control and data acquisition (SCADA) systems, on-line SPC software collects data automatically from process instruments and can be configured to generate control charts at the operator's workstation. If the software offers high-quality graphics and automated charting, operators can be trained to recognize when the process is going out of control or developing a bad trend. With this immediate feedback, the process can be fixed quickly before creating large amounts of scrap and rework. This type of charting also can be used to satisfy vendor-certification requirements. On-line software can be combined with on-line problem-solving manuals that provide operators with correct responses to out-of-control situations.
Figure 6. Pareto data is displayed alongside p-charts of the top three defects, illustrating an effective way to use SPC information.
Off-line SPC software is used for looking at long-term trends. Many electronics manufacturers combine this software with databases containing process and test data that have been collected both automatically and manually. Powerful off-line SPC software allow regular reporting functions to be done automatically or by clicking on a desktop icon. Some applications allow the off-line analysis to be shared across multiple facilities via an intranet.
Conclusion
When selecting SPC sofware, keep the following points in mind:
- On-line SPC software will not allow users to see the big picture. To make real bottom-line process improvements, off-line software is needed. The best solution is to have both, so the real-time picture as well as longer trends can be seen.
- Ease-of-use is critical. Some applications can only be used effectively by expert statisticians. For a successful SPC program, people of all abilities - from engineers to operators - need to learn and use the software.
- High-quality graphics are critical. SPC charting gives a clear visual picture of the situation so users can take action. The software needs to provide high-quality charting in which multiple charts can be scaled and compared side-by-side.
- The off-line application should connect to the company database, as well as to databases that will eventually be used. Do not take connectivity for granted. Make it a priority when evaluating demo packages.
- It should be easy to share analysis and charting functions over an intranet.
With supply-chain issues tightly bound with SPC, Web-based communication is inevitable among suppliers and other facilities.
JEFFERY L. CAWLEY is vice president of Northwest Analytical Inc., 519 SW Park Ave., Portland, OR 97205-3207; (503) 224-7727; Fax: (503) 224-5236; E-mail: jcawley@nwasoft.com.