Our industry is full of tales describing the work and effort needed to overcome fabrication hurdles to produce a complex design. Manufacturing a custom product with 100+ processing steps can be tricky enough, but when you add in a highly complex set of design attributes, pour engineering resources into defining the process, and verify that it is repeatable and reliable, that comes with a certain sense of pride and satisfaction. Today, I want to share a case study of one of those types of designs. And while this example will be one of those stories that will be discussed for years to come, it is also a case of something going wrong even when everyone is doing things right.
This tale starts approximately three years ago when we reviewed a complex design for a medical application. The product has contact with human skin, requires FDA approval, and was expected to have multiple prototype revisions and both test and pre-production runs before going to full production. The design itself was extremely tricky and required a very specific material set and multiple surface finishes and lamination cycles. Countless hours were spent in engineering discussions, trying to get an understanding of which elements had some flexibility and could be adjusted to make the design more manufacturable and how to overcome the challenges for the elements that had no flexibility. There were three or four revisions and subsequent engineering reviews to make adjustments after each build.
There were several small tests lots of these PCBs run as the final product was going through qualification. I think everyone involved with the fabrication of that design felt a strong sense of pride and satisfaction when we moved to the pre-production phase with a process and product that was performing well and with a good yield given the technological hurdles involved.
Until one day, several months ago, when the company doing the PCB assembly noticed a slight color and texture difference between some of the pieces. Significant time and effort were put into review and analysis of those parts from both a PCB fabrication perspective and from an end-use perspective. The final analysis showed this to be measling acceptable per IPC standards. Good, right? The product should be usable. Unfortunately, while the parts meet the IPC specification they were built to, it was determined that this condition was not going to be acceptable for end use.
Once again, we all had to sit down together to figure out a solution to this latest hurdle. Product that was built is now being sorted at the contract manufacturer, further testing on the final product is being done, and there are discussions about a potential field recall. The engineering groups throughout all segments of the full build cycle are meeting to find the best path forward. In one recent meeting, we talked about how to have a specification written to quantify the texture issue in a way that can be inspected, and parts will now be screened before shipping to assembly. I imagine there will be additional discussions as we move toward the long-term solution.
In these situations, I think there is a natural tendency to want to place blame on something that is frustratingly causing a considerable delay in product launch and comes with considerable cost to all involved. But this example really shows that even when everyone is doing everything right, something unexpected can still go wrong.
It would be satisfying to be able to finish this with a list of lessons learned from this example, but from a technical perspective, I don’t think we will be able to do that. I believe the lesson learned in this case study is that strong communication and a timely response have been critical from the beginning of this project and are key ingredients to working through complex design requirements. Frank discussions during the process development phase were critical to developing a reliable and repeatable fabrication process. And when this latest issue was identified, a swift response, testing, and reporting were crucial as we worked to understand this texture anomaly that was ultimately determined to have a significant impact on the end use.
As far as a final resolution, this is still to be determined. Watch for a future column to help bring an ending to this PCB tale—one that will be used as an example for years to come.