Updating the Design Standard

IPC-2221A, as most designers know, was released in 2003. Since that time, lead-free has gone from a niche technology to a mainstream one, and its added a generous dose of complexity to the design decision tree.

In November, IPC-2221B was, at long last, released, and it’s a good opportunity to reflect on the process of how standards are developed, and why it took nearly a decade to get the latest rev out.

In that time, for example, the task group responsible for J-STD-001, the industry soldering standard, came out with a “D” and an “E” revision, and work is underway on the “F” revision.  Certainly the changes lead-free brought about affected electronics assembly at least as much as design: it’s hard to pin the problem solely on technical reasons.

My experience working on the J-STD-001 suggests the difference in publishing frequency comes down to how the respective task groups call a halt to the changing technology. Going back to when Jerry Rosser was chair of the J-STD-001, for example, that group has a 15-year-plus history of calling a hard stop to new technical additions after a set period of time. Jerry knew the standards would always be disrupted by new technology, and therefore he initiated a plan — still followed — whereby after a period of months, the spec would be frozen, and the only changes from that point to the revision in progress would be to ameliorate the grammar and, on the rare occasion, fix legitimate technical disputes. If, for example, new chip-scale package requirements weren’t ready by the hard stop, they were tabled for the next rev. The ensuing document was never perfect, but it was far more timely than would have been otherwise possible.  To wit, the task group published four revisions, plus one amendment, in one decade alone.

Based on interviews with the coordinator of the IPC-2221, a different tack was taken, which slowed the process considerably.  It took the task group several years before calling a hard stop. Another difference is the decision to include a fair amount of tutorial in the design standard, whereas the soldering task group stripped all that info out years ago, opting instead to segment it into a separate handbook.

Now, to be sure, there is a design guide, but it was published in 1992 and has never been updated. I understand the philosophical reasons for doing things the way the design task group has done them, but in the interest of faster time-to-market, I think it’s time to reconsider whether there’s a better way. Thoroughness has value. But so does expediency.

The Cost of Misunderstanding Standards

An article in the latest issue of Assembly magazine asserts that use of standards, specifically IPC-A-610 and J-STD-001, raises the cost for US manufacturers and has led to the widespread offshoring of assembly.

The premise of the article, authored by a Dr. James A. Smith of Electronics Manufacturing Services Inc., is that standards drive up costs. This is stunning in that it completely mischaracterizes a core reason standards exist: to ensure widespread uniformity to a predefined level of quality.

Indeed, as someone who has traveled extensively abroad — I have spent more nights in Shanghai than any city other than the ones I have actually lived in — I can unequivocally state that manufacturers in China, Taiwan, Malaysia and so forth use IPC-A-600 and IPC-A-610 almost exclusively. And the reason is, those are the standards that their Western customers demand. Southeast Asia might offer lower labor rates, but that has nothing to do with IPC-A-610. As they used to tell me in stats class, correlation isn’t causation. I’m surprised Dr. Smith’s grad school teachers didn’t drill that conceit into him.

In the article, Dr. Smith asserts that “cost-plus” contracts reward poor manufacturing by ensuring that the assembler gets paid a set margin even if low yields lead to high rework costs. Besides being expensive, rework, of course, can be detrimental to the long-term board quality. Says Dr. Smith:

Some types of heat damage—lifted pads, delaminated circuit boards, and melted component bodies to name a few—are easily recognized. However, soldering iron heat causes serious degradation inside components such as ICs where the damage can’t be seen. The most prominent example of such damage is accelerated growth of the intermetallic (“purple plague”) between the gold wire bond and the aluminum pad on the chip substrate. As the intermetallic grows, electrical resistance inside the connection increases and switching characteristics change; depending on the sensitivity of the circuit, this change alone can be fatal. Even worse, Kirkendall voids develop in place of the pad material and breaks develop around the edges of the pad.

Therein lies the problem: Dr. Smith gets the technical details right and yet extrapolates from them a complete fallacy, writing “touchup and rework are all about deceiving the customer who, unwittingly, receives product with higher probability of premature failure.”

I am a former member of the IPC technical staff responsible for IPC-A-610 and J-STD-001. Having spent many a weekend in J-STD-001 meetings, I can state from experience that many defense contractors pushed to ease certain requirements in order both to save money and improve reliability. In one instance that springs to mind, Boeing provided ample evidence that minimum hole fill could be reduced because, they found, although a higher percentage of hole fill was seen as more reliable, in practice inspectors would have the rework technicians hit nonconforming holes with the solder gun, and the additional temperature excursion *reduced* long-term reliability more so than the greater volume of solder in the hole could increase it. These types of discussions don’t show up in the final boxscore, but you can’t understand the outcome of the game without knowing them. If Dr. Smith hasn’t been able to “unearth the data” behind the standards, it is because he hasn’t attended the meetings.

He also takes aim at the American industry for ignoring the teaching of the great quality gurus. “Instead of the focus on results emphasized by Deming and Juran, industry has embraced paperwork bureaucracy.” In fact, Dr. Deming focused on process, with the idea being a perfect process would net perfect results, and his “knowledge of vriation” concept runs through J-STD-001. At its core, J-STD-001 is a process driven document; a company that doesn’t understand SPC and process deviations has no hope of properly instituting it.

This all indirectly raises a separate point, however, namely: that it is critical for the US to maintain control of the standards. As my old friend Dieter Bergman used to tell me, he who controls the minutes controls the meeting. By authoring the standards (and making sure that the key Western OEMs are active contributors), the US can ensure a place at the international table far superior to the one our depleted manufacturing base would otherwise allow.

No More A-‘void’-ance

Did avoiding the voids just get less important?

A new study by Rockwell Collins, RIM, Indium and Dage shows voids are not a “primary” issue to BGA solder joint integrity.

New BGA voiding requirements to be proposed to J-STD-001 task group based on study’s findings, Rockwell engineer Dave Hillman says.