Decoupling Performance Class and Quality Assurance testing—YES, and add Reliability

‘Functional Performance’ Classes cannot be equated with ‘Reliability’ classes–they are not the same. I have tried unsuccessfully in the past to put a disclaimer into the scope of these documents, to alert the readers that only an appropriate ‘Design-for-Reliability’ would assure reliability.
While good quality—and
IPC-6012 and IPC-6013 assure quality, is a requirement for reliability, it is far from sufficient. That is why we have IPC-D-279 and all the IPC-970x’s. Neither IPC-6012 nor IPC-6013 were written with reliability being in any way addressed. There is at least one documented situation—chip capacitors, where Class 3 product can fail before Class 2 product. If I can get Class 3 without cost penalty, sure, why not, but do I have to have it for reliable product—NO. In most cases, Class 3 is specified for contractual, not technical reasons, and because there is a more complete paper trail [which is not a bad thing].

During APEX/EXPO there was talk about differentiating electronic packaging by applications. It also has become quite apparent, that because of the much higher soldering temperatures required for most RoHS-compliant Pb-free solders, the processing windows for many PCBs and components have significantly narrowed, jeopardizing the very survival during the assembly processes.

The consequence of this, it seems, it that survival, quality and long-term reliability [that is actually the title of a workshop I gave at APEX/EXPO] issues need to be addressed across the traditional boundaries demarcated by the various committees. This needs to start at the design and specification stage; the choice of base materials and their pertinent properties like Tg, Td and Alpha-2; the processing with such diverse considerations as copper plating thickness, peak soldering temperatures and the number of solder temperature excursions; when, where and how to bake out the moisture; and, of course, the severity of the operating environment, the expected life time and the acceptable failure probability.

With SnPb-soldering, much of this worked independently and often by pure serendipity—but all that came to an end June 1, 2006.

I do not know how to bring this interdependence to bear on the IPC committee structure so it is effective and avoids ‘turf wars,’ but it has to be given serious consideration.

Post a Comment

Required fields are marked *


%d bloggers like this: