Everyone’s product is “mission critical,” but the specific mission must be understood.
Reliability, like quality, its kissing cousin, is a given, and anyone in the design and manufacturing world knows that whatever is worked on is assumed will be reliable and of the highest quality. That said, what customers need in the way of a reliable product does not always square with what they ask for in a reliable product.
Nowhere is this more evident than when core technologies and processes change, and customers ask for the “old” performance from solutions designed for a “new” application.
This was especially the case when RoHS emerged on the scene a decade ago. The surface finish of choice for decades was tin-lead. Under RoHS, lead was banned, and it was assumed there would be a simple plug-and-play replacement the chemical companies had ready in the wings, available for an eagerly awaiting industry to embrace and switch to. Instead there was one expensive option, gold, and a host of less expensive options, all in the not-ready-for-prime-time beta stage. While all offered a quality option, reliability was far from assured. The military folks, used to producing product that had a very long lifespan operating in harsh, environmentally challenging environments, did not like the idea that some of the new options could create “whiskers” that could ultimately make a device fail. The folks who produced high-volume consumer products with long supply pipelines did not like the tarnishing that product left on the shelf may display with other options. And nobody liked the cost adder to use gold as a surface finish, despite its track record.
Reliability, previously defined as being long-lasting, bulletproof and cheap, was going through a paradigm shift.
Over the past few years I have been amazed to hear some of my military friends, the ones who once said they would never give up tin-lead HASL as a finish, enthusiastically embrace what they had considered a “dangerously” unreliable surface finish, silver.
Why the change of heart? It was not because the surface finish improved, but rather the customer took stock of what the end application was and decided that for some electronic devices that have a relatively short-use life or technological lifecycle, the probability those items would be in service long enough for any whiskering to take place was minimal. For the same reasons many commercial customers have approved tin as a finish on short lifecycle applications. Military folks, as well as commercial customers, are rethinking what reliability means to them application by application.
The same is the case with laminate. Only a few short years ago FR-4 was FR-4 – all was virtually the same, and the decision as to which to use was based on price and availability. As changes have taken place in assembly, however, where circuit boards are more often exposed to higher temperature thermal shocks, combined with the characteristics of those newer environmentally friendly chemistries and surface finishes utilized in the fabrication process, FR-4 laminate is changing. Different and higher Tg requirements are redefining the notion of producing a reliable product. More questions need to be – and are – asked about the application and processing environment before materials are specified. The old paradigm of any FR-4 (or the cheapest) has transitioned to a more application-focused selection process to ensure reliability.
But still reliability measurements are often focused on expectations of the past rather than needs of the future. Testing for failures that no longer pose the risk they once did can cost the supply chain significantly, while not contributing to reliability – or addressing newer risk issues.
I am always amazed when a customer conducts an audit and drills down on some aspect of the process that 20 years ago caused that customer heartburn. What almost always results is their realizing, often embarrassingly so, that that process of concern is no longer used. Equally, those same customers never consider the chemical system a given fabricator is using compared to others. Dow, MacDermid, OMG, etc. all address process chemistries for PCBs differently, but customers do not seem to understand if and how those differences may or may not impact reliability. Ditto for individual companies’ detail to managing process and procedures. Often during an audit customers expect to see some process, procedure or material and supply they were familiar with decades earlier, rather than being open-minded enough to understand what and why a different process, material or supply is being used by their supplier today.
Most important, as reliability requirements evolve, so do the materials and processes required to deliver that reliability. It is increasingly more important that customers and suppliers fully understand the end-application and any idiosyncrasies that should be taken into consideration when designing and building the product, rather than just having bandied about the standard quality statement such as “all of our products are mission critical.” Everyone’s product is mission critical, but unless you understand and build for the specific mission, reliability cannot be assured.
And there is a cost associated with reliability. Some applications can be delivered for far less than others, but they all carry a cost. If over-specified – or underspecified – the cost can spiral out of control.
Which comes back to the danger of assuming that what was appropriate in the past is really up to the challenges of the present and future. This places the burden of responsibility on all parties. The customer needs to better understand what material changes have taken place that could impact the finished product, while the manufacturer needs to understand what environment the product will see during the rest of the manufacturing process and in the field.
We all want reliability. We all expect that when we order a product, whether it’s a completed device or any of its many components, we will get that reliability. But no one should assume the reliability standards we have historically expected are still valid for product that is expected to perform in more demanding environments and is produced with next-generation materials and processes. All those little changes that take place throughout the manufacturing process add up and often require everyone to rethink what level of reliability is needed for which situation.