I have the privilege of working for a circuit board supplier that does not own any factories. Some may question how this can be advantageous. I believe the best answer is that we are not locked into the limitations of one facility, and we can offer multiple specialties from different factories. No one PCB factory can truly be a jack of all trades, do everything, and manage to be the best at it.
But don’t get me wrong; finding the right factories is not an easy task. Anyone can take customers’ files and send them to whichever factory is available. But what guarantees does the customer have that the factory used is reliable in producing the design?
One of the most important areas that must be a focus is having an infrastructure in place local to the factories. For example, we have a factory management team that consists of 73 team members in China and is dedicated to finding, approving, and ensuring that the factories used are the best at what they do. It is a time-consuming process and takes a lot of hours, but the end result is worth the effort.
Identifying the factories should be a major focus. This could be triggered by a need. Perhaps you identify a need for another source for rigid-flex or HDI boards. A shortlist should be created from the thousands of factories operating. We then recommend that these be grouped into product types. Information then needs to be gathered on the potential candidates—everything from how financially stable they are to how they have incorporated a sustainability plan.
After the candidates that didn’t meet the requirements are eliminated, we recommend proceeding with benchmarking exercises, followed by on-site visits. If everything is approved, the next best step will be to load sample runs and do laboratory evaluations. Once the samples have been approved, thorough on-site quality and sustainability audits should be conducted.
When the sourcing process begins, you will be looking at possibly thousands of candidates. By the time the list has been narrowed down to the potentials, and they evaluated the factories’ performance on multiple levels, the list should be down to the top one to three candidates. Figure 1 shows an example. While working through the sourcing and auditing process of one factory, we can see their advertised capabilities, but when it came time to produce, the results showed they were lacking in these capabilities.
Figure 1: The image to the left is a factory’s claimed capabilities, but after multiple small batch tests, we discovered that the capabilities were not as reliable when produced this way. Therefore, we had to reduce their capabilities for what we feel is acceptable.
And although it sounds fairly simple when put on paper, the actual process can take up to 60 weeks to complete. In the end, what you have are the best possible candidates available.
Figures 2 and 3 show examples of results from an audited factory. Here, you can see that reliability is not the only factor to consider before approving a factory. We don’t recommend accepting boards that will be subcontracted out due to lack of capabilities within the audited factory, nor considering factories that charge above market price for certain capabilities.
Figure 2: This is another example of where a factory claims to produce a variety of technologies.
Figure 3: Upon inspecting small-batch prototypes from this particular factory, we discovered the technologies claimed to be produced from the factory cannot all be produced reliably or efficiently. Therefore, we can only produce in the areas where they excel.
Of course, this is by no means the end of the process. Once agreements with the factories are approved and signed, then we recommend training. For example, we established our own PCB specification. When a new factory is approved, we train personnel to our needs and requirements. Even though we realize that the factories have done training, we have demands that are above and beyond IPC and industry standards that require specialized training, including for exclusive, dedicated equipment for all produced boards (Figure 4).
Figure 4: Specialized training, inspection, and auditing within the factories are key to producing the most quality product.
Even at this point, the process is not over yet. In fact, it never really ends. We must consistently be evaluating the factory on numerous levels. Once the factory becomes a part of a preferred supplied list (PSL), there should be continual quality and performance monitoring, process audits, quality, and pre-production engineering (PPE) audits on a consistent basis. If something goes wrong, a root-cause analysis should be conducted. We also recommend conducting sustainability audits, which include everything from human rights to environment and business ethics. It’s a good idea to hold factory meetings on a regular basis. It may be difficult to achieve this level of engagement with the factories. For instance, we strive to have a minimum of 15% of our main factories’ revenue or rank within as one of their top five customers and are able to reach these figures by our combined global purchasing power.
If you don’t have the capacity to achieve all these activities and standards, there are many advantages to working with a company in this business model. Recently, a customer came to me and asked me to produce a board they were having some issues with. I suggested the particular factory they had already been using, and the customer was reluctant at first. Sure enough, when the factory produced the board with our particular set of standards set in place at the factory, as well as the specialized training staff, the board was produced with much higher reliability than when they worked directly with the factory.
We have created an infrastructure with the ability to take the time to source, qualify, and constantly evaluate performance, which many companies may not be able to do in-house. Aside from having multiple sources to meet all PCB requirements, customers can have peace of mind, knowing that their partner will be around for the long run.
Ruben Contreras is a program manager and field application engineer at NCAB Group.