Back in Person

The Covid-19 vaccine rollout has begun and we can’t wait to get back to seeing old and new friends in person.

To that end, I want to call your attention to the return of PCB East to the Boston area in June.

We will head to Marlboro, MA, for some 55 hours of training across three days (June 15-17) of printed circuit board engineering training. There, SI expert Lee Ritchey will have a couple of tutorials: Printed Circuit Board Stackup Design for High Performance Products, and also Power Delivery System Design.

We also will offer two full days of Rick Hartley, including a brand new talk titled, “PC Board Design for Optimum Fabrication and Assembly.” As Rick notes, Happy Holden has presented at PCB West a few times where he’s explained how fabricators determine pricing for bare boards and how EMS suppliers determine pricing for PCB assemblies. Happy shares what he calls a “Fab and Assembly Report Card,” which is how manufacturers assess and weight the variables that drive cost.

So, for instance, as most readers know, board size is a major cost driver. But, as Rick explains, what most designers don’t know is that aspect ratio of length to width also has a major impact. Two boards with the same number of layers and same number of sq. inches but with a difference in their respective aspect ratios – say one is much longer than wide – will push up the bare board cost. Same with assembly, which has even more cost drivers than does fab. Rick is going to do is discuss these major cost drivers.

Rick also told me that at PCB West he had discussions during the chat sessions with some of the bare board fabricators in attendance. One of them said (I’m paraphrasing here), “At any point in time as many as 90% of our jobs are on-hold, waiting for correction or clarity from the customer, so we can proceed.” In Rick’s opinion, designers are flying blind when it comes to many the cost drivers and what suppliers need at both the bare board and assembly level, hence the reason for so many delayed PCBs. These delays also add cost.

What Rick wants to do is to highlight and talk about the factors that Really drive up cost, like board size and aspect ratio, layer count, Z-axis uniformity, copper balance, etc.

And Susy Webb will have brand new, two-day tutorial for design engineers, “A Comprehensive Guide to PCB Design Necessities.” Her class will feature an overview of the entire process of board design, from start to finish, addressing the EE designing their own boards or the new designer who needs to thoroughly understand all the steps and processes. She’ll cover everything from the electronics and physics involved, how the rise time and controlling the energy fields impact the signals on the board, choosing parts types, schematics and signal and constraint issues, mechanical issues, and so on. Susy is also doing an all-day webinar.

We are looking forward to these any other presentations, and also to the exhibits on Jun. 16. Registration is now open, so visit pcbeast.com for details.

Living La Via Loca

I got an email a few weeks back from one of our customers who is with a semiconductor manufacturer for whom we make evaluation boards.  He asked some good questions that I answered, of course, and I thought this information may be useful to others, too.  Here is what his email said:

Hi Judy, We received the boards today… and they look gorgeous!  And of course it leads to a question….. not a negative, not an “inquiry” style question….

One thing that caught my eye was how “flush” the board surfaces were.  I wasn’t able to identify a single via on the front or back surfaces of the board…. and (education time) I had an Apps Engr point out that I could see the via plugs by looking at the edge of the EVB.  So I want to understand more about the via fill process, and whether there are different options that influence cost and lead time.

I will share with you my brief answers, which are not overly technical. However, I will follow my comments with a link to a very good in-depth paper from Michael Carano of the Electronics Division of the OM Group on this subject. Unfortunately once you start talking about via filling and plugging it is a bit like opening a big can of worms!

First of all, a semantics lesson:  There are two methods of filling via holes—one is via “filling” and the other is via “plugging.” Via hole filling refers to the non-planer filling of plated through-holes. Via hole plugging refers to the planarization of blind and buried vias, as well as through holes. Via hole plugging pertains to HDI and microvia designs.

(In the case of the above customer they requested the through-hole vias to be plugged with conductive  ink that would facilitate a connection from the front to back sides).

Once the conductive ink has been applied to the holes, the conductive ink must be cured. After curing some of the cured material will protrude from the holes leaving a small bump or “nail-head.” At this point the board must be “brushed” to flatten this protruding bump, with micro precision, to remove the excess material without removing any material from within the via hole. The brushing process must not damage the knee of any unplugged holes. This requires the use of a very flat working surface and special brushes. In some cases there may need to be an additional final cure for hardness. (UV)

Getting back to our customer, we took all these steps to properly fill, cure and planarize the plugged vias. Afterwards, these boards went through final outer layer pattern plating, which further erased any evidence that the plugged vias were present.  This made it impossible for our customer to detect any sign of the plugged vias (unless they peered through the side of the board into the substrate where they could see the front to back plugs). I think they were a little worried we forget to plug them at first glance!

As far as how filling verses plugging vias influences cost and lead time—the HDI via filling does take some extra time and TLC and thus raises the cost, and may or may not affect the lead time. Usually you must allow for one full day to complete the via plugging process.  Obviously if you need a quick turn, the lead time would be affected. However, if the lead time was standard, it probably wouldn’t make much of a difference. It stands to reason that the quantity also comes into play in regards to time and cost.

Thank you, Bill for asking these questions—they were good ones! We appreciate your kind words and commitment to our partnership. We are thrilled you think our boards are “gorgeous!”

I would also like to thank Michael Carano for his expertise and outstanding paper on the subject of filling and plugging vias that can be found here: http://www.electrochemicals.com/viafill07.pdf

Early Thoughts on the Mitchell Administration

The hiring of John Mitchell as IPC’s new president comes as a breath of fresh air to those of us who had long tired of the antics of the previous regime.

Mitchell, the fourth person to run the 55-year-old organization, is the first with electronics industry experience, having spent a combined 16 years at Bose and Alpine Electronics.

As Bill Bader and Jim McElroy at iNEMI have proved, when it comes to running a not-for-profit volunteer organization in this industry, experience counts. The supply chain and regional differences are far too complex and the technology too intense for a greenhorn, especially one who isn’t willing to do their homework.

We envision — and hope — for a return to the days when member input is sought and valued. More than that, however, we are eager to see the occupant of that important position have a vision and tenacity that goes beyond avenging imagined personal wrongs.

Mitchell has his work cut out for himself. The industry is fractured, physically and emotionally. He will have to learn to lead without alienating, something his predecessor never accomplished. He will have to mend fences with the North American board fabricators, on whose shoulders IPC was built but were later ignored or cast out as the organization moved into the more lucrative assembly market. He will need to understand that the suppliers are generally looking to protect declining margins, and yet much of the technological know-how has migrated to that side of the industry, so he will need to convince them it is in their best interest to continue to support IPC’s technical programs, not just the exhibitions. He will need to navigate the treacherous China-US relations, in which the occasionally nasty spells of provincialism and finger-pointing from both sides mask an underlying dual-relationship that neither party can live without. He will have to right an internal culture that has grown distant from its membership. And he will have to do so while determining whether the four (!) vice presidents who applied for the job — two of whom have now been rejected multiple times — are up to the task of working with the man whom the IPC board considered a superior leader.

Based on Mitchell’s resume and conversations with IPC board members, he is the right person for the job. He is first and foremost an engineer. He has a deep business background that belies his age (he graduated college in 1991). He has worked at a high level for a major supplier of consumer electronics, giving him insight into branding and the supply chain intricacies that his predecessors either never had to deal with or were unable to master. We look forward to the next chapter in the continuing story of IPC.

Getting the Right Finish, Before You Start

If you are a fan of John Wooden, the celebrated UCLA coach, you will know that he had more than his share of clever sayings. My husband, who played basketball all his life, and who went to college on a full-ride scholarship—nearly worships Coach Wooden. For this reason, I have heard many of these wise and witty Wooden-isms over the years. One that comes to mind is: “Be quick, but don’t hurry.” With the speed of advancing technologies, we all have to be quick, but we cannot afford to hurry—or be hasty—putting revenue and market share opportunities at stake.

In regards to RF and microwave printed circuit boards, there seems to be some confusion about PCB finishes and their affect on the high performance requirements of these applications. Too often, when considering the available finishes and the potential impact they have on performance, many engineers become both quick and (unwittingly) hasty when they make finish choices based on information, which is conflicted, at best.

I think it’s time to clear up some of these issues, so I am going to spend the next few blog posts talking about these issues. Hopefully, by the time I’m through you will have a much clearer understanding of finishes and which to choose for your product, before you start! I will be drawing from our real-world experiences, as well as looking to experts in the substrate and RF design industries.

Today, I am just going to cover the major available finishes, and which ones seem to be preferred by those with high speed applications:

  • Tin (Lead free)
  • HASL (Hot Air Solder Leveling) Tin/Lead 63/37
  • ENIG (Electroless Nickel Immersion Gold)
  • ENEPIG (Electroless Nickel/Electroless Palladium/Immersion Gold)
  • Hard Gold
  • Soft Gold
  • Immersion Tin
  • Immersion Silver

In high-speed applications, the prevailing wisdom suggests ENIG, ENEPIG, hard gold, soft gold and immersion silver are the best choices. Gold is a natural choice due to the fact that it does not oxidize and that it is wire-bondable. Immersion silver is gaining some traction due to the excellent conductivity, but it oxidizes and it is not wire-bondable, which keeps many from choosing this option.

Unfortunately, I must leave us barely posed in the starting blocks, in regards to finishes! In two weeks, however, I will sound the starting shot, and we will be off to the races. I will discuss each finish in more detail with the pros and cons of each.

If you have specific questions you would like to submit about this subject, please post them here in the comment section or email me at: [email protected].

I’m looking forward to tackling this complex and critical subject together!

 

–Judy

Summer Doldrums

Is it cyclicality, or … ?

Many reports, anecdotal and evidentiary, point to a general slowing in PCB production and sales over the past quarter.

Yet there are some reasons for optimism:

I am of the mindset that what we are seeing is a return to cyclicality after roughly two years of recession followed by a year-plus of bottled-up demand. Clearly there’s some market turbulence ahead, especially when we take the macro vectors into account. Some of the end-markets need a boost: Now that Windows 7 has taken over, PCs are stagnant, with new tablet demand offset by rather humdrum desktop/laptop interest coupled with some migration to smartphones. Nokia and RIM are skidding, and Apple can’t make up for everyone’s lack of flair. Autos are a big-ticket item and many consumers today need stronger feelings of job security before taking on new debt.

A forecast slowdown in US defense spending (the nation’s fiscal year starts in October) could be partially offset by new deliveries of jumbo passenger jets (Boeing last month announced a record single order and will ship its first Dreamliner next month).

The tea leaves are murky. We hope for the best.

One-Stop Shop

If you are looking for a snapshot of the latest (or historical) market statistics, we’ve begun compiling all the data from a host of sources in one place on the CIRCUITS ASSEMBLY website.

Among the data we are posting include book-to-bills and sales and orders of:

  • Semiconductors
  • Passive components
  • Printed circuit boards
  • Key end-markets such as PCs, servers, mobile devices, etc.
  • Wafer utilization.

Sources include SIA, SEMI, Gartner, IDC, IPC, ZVEI and other research firms and associations. Check it out!

Good Fiction

Ian Fletcher (not to be confused with the scribe behind the James Bond novels) yesterday wrote about the problems with American manufacturing.

Whereas I agree with his larger points — manufacturing output does not necessarily measure manufacturing health — there’s a couple of problems with some of his supporting evidence.

For one, there’s his contention that because Japan supplies “over 70 percent of the world’s nickel-metal hydride batteries and 60-70 percent of the world’s lithium-ion batteries,” it give the country “a key advantage in electric cars.”

Well, maybe. Being a supplier of a critical commodity or technology does not, history shows time and again, necessarily translate to ownership over the end-market. Labor and other ancillary costs have a tremendous affect on the procurement decision tree. China, of course, had no real technological advantage over Japan, Taiwan or the US in printed circuit boards. It just had loads of cheap manpower, and a government so eager to build its tech base that it moved land and sea (sometimes literally) to make it possible.

Which brings us to nit No. 2. Says Fletcher: “The Obama administration shows no awareness of any of this, despite scratching a hole in its head over why job creation has stalled. (Hint: it hasn’t stalled in the nations, from China to Germany, running trade surpluses with us in manufactured goods.)” No, unemployment hasn’t stalled in either country, but for reasons other than what Fletcher asserts. China, of course, has an abundance of cheap labor that no country save India can match. In Germany, on the other hand, the unemployment rate is 6.8%, which is better than in the US, but hardly great relative to classical standards. As recently as 2004, it was 9.7%, after which Germany enacted a series of financial reforms. The nation now stands as the financial bedrock upon which lies much of the rest of the beleaguered European Union. Given that, I would argue that Germany ability to position itself financially, rather than any politically driven manufacturing strategy, is at the core of its current export success and employment stability.

Finally, Fletcher ignores a far more significant data point: Japan. Japan’s economy is some 60% larger than Germany’s, and its internal manufacturing supply chains are legendary. Its unemployment rate was 4.9% in December, which appears stellar, until one realizes that figure is 188% higher than the nation’s average from 1953 through 2010. Its GDP is on a roller coaster, having contracted in the December quarter. Yet it runs a $45 billion trade surplus with the US. Here, Fletcher’s contention that trade surpluses and manufacturing supply chains go hand in hand with job creation falls on its face.

Fletcher is on the right track, but some of his supporting details are closer to the James Bond series: good fiction.

Good Fiction

Ian Fletcher (not to be confused with the scribe behind the James Bond novels) yesterday wrote about the problems with American manufacturing.

Whereas I agree with his larger points — manufacturing output does not necessarily measure manufacturing health — there’s a couple of problems with some of his supporting evidence.

For one, there’s his contention that because Japan supplies “over 70 percent of the world’s nickel-metal hydride batteries and 60-70 percent of the world’s lithium-ion batteries,” it give the country “a key advantage in electric cars.”

Well, maybe. Being a supplier of a critical commodity or technology does not, history shows time and again, necessarily translate to ownership over the end-market. Labor and other ancillary costs have a tremendous affect on the procurement decision tree. China, of course, had no real technological advantage over Japan, Taiwan or the US in printed circuit boards. It just had loads of cheap manpower, and a government so eager to build its tech base that it moved land and sea (sometimes literally) to make it possible.

Which brings us to nit No. 2. Says Fletcher: “The Obama administration shows no awareness of any of this, despite scratching a hole in its head over why job creation has stalled. (Hint: it hasn’t stalled in the nations, from China to Germany, running trade surpluses with us in manufactured goods.)” No, unemployment hasn’t stalled in either country, but for reasons other than what Fletcher asserts. China, of course, has an abundance of cheap labor that no country save India can match. In Germany, on the other hand, the unemployment rate is 6.8%, which is better than in the US, but hardly great relative to classical standards. As recently as 2004, it was 9.7%, after which Germany enacted a series of financial reforms. The nation now stands as the financial bedrock upon which lies much of the rest of the beleaguered European Union. I would argue that Germany ability to position itself financially, rather than any politically driven manufacturing strategy, is at the core of its current export success and employment stability.

Finally, Fletcher ignores a far more significant data point: Japan. Japan’s economy is some 60% larger than Germany’s, and its internal manufacturing supply chains are legendary. Its unemployment rate was 4.9% in December, which appears stellar, until one realizes that figure is 188% higher than the nation’s average from 1953 through 2010. Its GDP is on a roller coaster, having contracted in the December quarter. Yet it runs a $45 billion trade surplus with the US. Here, Fletcher’s contention that trade surpluses and manufacturing supply chains go hand in hand with job creation falls on its face.

Fletcher is on the right track, but some of his supporting details are closer to the James Bond series: good fiction.

Seeing Big Possibilities

A comment in old friend Dominique Numakura’s weekly newsletter prompted this audacious thought.

First, though, the background. Dominique has been detailing this year’s JPCA 2010 Show. As he points out in his June 13 issue, there was a new exhibition focused on large-scale electronics. In it, it was revealed how market demand for massive flat panel TVs, digital signage, photovoltaic cells and surface light sources creates concurrent demand for large PCBs. And he notes how, while suppliers like the trend — more volume sold — PCB makers aren’t necessarily embracing it because of the bigger boards require tooling changes and capital equipment investments.

But (!) — and this is where Dominique leaves off and audacious Mike comes in — big boards are what North America traditionally has done well. Could it be that this trend, primarily for industrial and certain consumer electronics — might actually spark a modest return to prominence for domestic fabricators?

Seeing Big Possibilities

A comment in old friend Dominique Numakura’s weekly newsletter prompted this audacious thought.

First, though, the background. Dominique has been detailing this year’s JPCA 2010 Show.  As he points out in his June 13 issue, there was a new exhibition focused on large-scale electronics. In it, it was revealed how market demand for massive flat panel TVs, digital signage, photovoltaic cells and surface light sources creates
concurrent demand for large PCBs. And he notes how, while suppliers like the trend — more volume sold — PCB makers aren’t necessarily embracing it because of the bigger boards require tooling changes and capital equipment investments. 

But (!) — and this is where Dominique leaves off and audacious Mike comes in — big boards are what North America traditionally has done well. Could it be that this trend, primarily for industrial and certain consumer electronics — might actually spark a modest return to prominence for domestic fabricators?