A large electronics OEM near Chicago, Illinois was concerned about their stencil printing costs. Despite their best efforts, they observed persistent stencil printing problems which led to increased rework, reduced through-put and added costs. To respond to this situation, and as a general effort of “continuous improvement” they developed and deployed a systematic, reliable stencil roll testing protocol which may be useful at other companies.
The company is a leading provider of digital motor controls used in wide variety of industrial applications. This OEM produces hundreds of different PCB designs for a wide variety of controllers. Since almost all of their PCBs are considered “Class II” boards under the IPC standards of design and quality, quality is vital because these products are expected to survive for years or decades. The company’s main goals in this study was to improve their quality, enhance yield, reduce their warranty costs and enhance their reputation as a premier OEM in their field.
The company has four SMT lines using MPM printers. Using an industry-standard stencil roll and wetting the paper with alcohol during cleaning they cleaned every two prints. In general, printing took 15 seconds per board and the cleaning process (a three-step, wet/dry-dry process, with vacuum) required about 60-70 seconds. Total average through-put, including board positioning and alignment, was 57 seconds.
As with most companies, this OEM knew they had certain board designs that were more problematic to print than their other designs. To ensure sufficient statistics for a meaningful analysis, the company focused on these moderately high-volume “troublemakers.”
The engineering team defined a set of visual “process indicators.” For example, one board had a 20-mil pitch QFP that suffered frequent bridges. Another board had a 50-mil PLCC that was often the source of rework. So those defects became the “process indicators” that defined unsatisfactory cleaning.
To begin, the process engineer collected benchmark yield information on these boards using their standard stencil printing and cleaning process. The benchmark indicators re-confirmed their current manufacturing process and results; the process indicator defects began to appear after more than two prints.
“This is very typical of electronics assembly today,” notes J.D. Buller, the manufacturer’s representative who worked with the client OEM. “Almost every PCB has a fine-pitch package that causes stencil printing problems. By collecting baseline statics, these components easily serve as the cleaning benchmarks.”
Testing and results:
The next step was to test a new cleaning process. The first test used “FP” stencil rolls from MicroCare. The FP fabric rolls are a drop-in replacement for older paper rolls, so no cleaning parameters were changed. Using the “process indicators” this one improvement delayed cleaning until after seven printings.
The process team wanted to see if further cost-savings could be achieved. These tests involved using simplified wet-dry cleaning process which saved approximately 20 seconds per cleaning cycle. This proved successful with the FP rolls. Eventually, the testing expanded to evaluate wiping dry (with vacuum) only. On the 20-mil component, the process required cleaning after five printing cycles. On the popular 50-mil component, they were able to jump to cleaning only every 15 prints.
“It was a true ‘win’ for this customer,” Mr. Buller notes. “The client’s consumption of stencil rolls is down about 66%. Their use of flammable alcohol in the stencil printer is completely eliminated. Their throughput has increased because a typical board now is printed in just 21 seconds, plus as yield has improved the cost of rework has been reduced. Their development of reliable process indicators has greatly enhanced the performance of their stencil printing process.”