News

Optimizing a process with limited access to process factors.

If all the individual manufacturing steps required to assemble a "typical" circuit board are evaluated, one step is truly a process: material printing. No other step requires the integration and optimization of as many individual factors to produce optimum results.

Factors that influence the material printing process include but are not limited to:

• PCB design, finish, flatness and thickness.

• Soldermask (type and registration).

• PCB fabrication (stretch, actual versus specified pad sizes, etc.).

• PCB markings (ink) and labels (applied before printing process).

• Material to be printed (solder paste, adhesive, etc.).

• Stencil design and fabrication (laser cut, chemically etched, etc.).

• X-, y- and z- axis support.

• Environmental conditions (temperature, humidity, air flow, etc.).

• Printing equipment.

• Operating parameters (squeegee speed, squeegee pressure, etc.).

• Squeegee blades.

• Stencil wiping frequency, chemistry and effectiveness.

• Operator training.

• Technical support.

• Supplier support (material, equipment, tooling, etc.).

This is not a complete list, but it highlights some of the many factors to consider, understand, evaluate and optimize to achieve optimum results. Yet many assembly operations control only a portion of these factors. For example, many EMS companies may have little or no control of board design, finish or stencil design. Others may have no control over the solder paste used or one or more of the other factors named.

The question then becomes, How does one optimize a printing process for any particular product when they do not control several of the factors that influence its performance? And how does one do this while the customer mandates better quality, faster cycle time, shorter setup time, and so on?

Focus on what can be changed. Here are seven steps for how to change those "changeable" factors to overcome deficiencies of factors that cannot be altered.

1. Take inventory of the factors for which that particular customer will or will not permit changes for that particular product.

2. Identify which of the changeable factors will provide the most process improvement based on the types of defects experienced. As always, viable data are key to making correct decisions. If you lack the required data, start collecting them immediately. Data collection does not have to be delayed for the implementation of a sophisticated SPC system; it can start with a pencil and paper. Automating data collection should be a secondary concern. Train the operators to collect the data, coach them (to reinforce the training) and monitor that it is being done at all times on all shifts.

3. Brainstorm possible process improvements that will achieve the identified improvements based on your work in Step 2. Consider short-, medium- and long-term improvements. For example, a short-term solution would be changing a stencil design aperture to accommodate a less-than-optimum sized pad. A relatively minor investment in a new stencil will provide significant benefits in reduced defects. A medium-term solution would be to redesign the board with the correct component pads (if this is a changeable factor). A long-term solution would be new printing equipment that is more accurate and perhaps includes post-print inspection capability.

4. Experiment with the proposed change. It is vital that these experiments be formally done (design of experiments) so data collected are meaningful. Building a few boards to prove out a solution does not validate or prove the change.

5. Once you have experimented with the change, implement it in a controlled run of products. Collect the data and verify the effectiveness of the change.

6. Implement the change. Make sure all required tooling, training, documentation, etc., are available and understood by all involved.

7. Move to the next process optimization opportunity and repeat the process. Continuous improvement has no conclusion.

Some steps are ongoing. Always provide your customer design for manufacturability suggestions and training. Show your customer how a particular change will reduce cost, improve quality, etc and support it with data that will quantify the proposed design change.

"If it is not broken, do not fix it" is not a continuous improvement philosophy. "If it is not broken, how do we make it better" with or without control of all the process factors should be the primary goal. That said, do not give up the effort to change factors that you currently do not control. If a compelling case, supported by statistically valid data, can be made for change, then perhaps the customer will permit it to be implemented.

Joe Belmonte is project manager, advanced process development, at Speedline Technologies (speedlinetech.com); jbelmonte@speedlinetech.com. His column appears semimonthly.

• Prev
• Next