About Mike

Mike Buetow is president of the Printed Circuit Engineering Association (pcea.net). He previously was editor-in-chief of Circuits Assembly magazine, the leading publication for electronics manufacturing, and PCD&F, the leading publication for printed circuit design and fabrication. He spent 21 years as vice president and editorial director of UP Media Group, for which he oversaw all editorial and production aspects. He has more than 30 years' experience in the electronics industry, including six years at IPC, an electronics trade association, at which he was a technical projects manager and communications director. He has also held editorial positions at SMT Magazine, community newspapers and in book publishing. He is a graduate of the University of Illinois. Follow Mike on Twitter: @mikebuetow

AR/VR: What’s New?

CES 2020 offers the latest augmented and virtual reality technology, creating immersive experiences across industries from health care to agriculture and manufacturing to retail.

The latest in AR/VR technology from CES 2020.

What a Waste

I like articles like this one from Forbes — even if they are out of date, because they cast attention on the intrinsic complexities of manufacturing.

But, while Forbes may think manufacturing inefficiencies cost the industry $8 trillion a year (no idea where that figure comes from, by the way), it points to the need for tools like the digital twin.

The Disruptors

In his most recent monthly column my friend Gene Weiner pointed out the irony of calling a printed circuit board a commodity. “When did printed circuits become mass-produced unspecialized products?” he asks.

As I told Gene, it’s true PCBs are custom, but in some (many?) instances they are highly mass-produced. Think residential LED lighting, or game consoles or the most popular phones, etc. My sense is that calling them “commodity” or “custom” is a bit of a red herring. Large OEMs call PCBs commodities even though they know each is specific to a particular program. I think to most its a catchall word for “volume procurement.” 

There is real tension on costs between customers and suppliers, and without artificial price bottoms I can see no reason that won’t continue. I think it’s highly unlikely sellers will be able to change the pricing model as long as the process technology remains essentially the same. It’s too entrenched, and often the buyers (which more and more often are EMS companies) have lower margins than the fabricators, so there’s no incentive for them to switch to a pricing model that will cost them more. A move to a different process technology (3-D/additive manufacturing) might well change the pricing curve.  

It’s not that simple, of course. As double-Ph.D. and former CTO of Multek Craig Davidson once said, “Never underestimate the tenacity of incumbent technology.” It’s perhaps the best quote on technology adoption I’ve ever heard.

I used to think that with the many brilliant minds in our industry, we would successfully disrupt ourselves — in other words, the ideas that obsolete our current ways of doing things would be conceived and implemented by one of our own — maybe some genius in the basement of Apple or Intel or IBM. I’m less confident of that now.

I now lean toward the notion that the true disruptive change will come from someone outside our industry who has knowledge of both hardware and software and who uses a truly novel process to develop a widget that the public embraces. The mass appeal of the product (not the process, to which the general public is ambivalent) will be the proof of concept. And the investment money will follow.

Selecting Reflow Oven Length

Folks,

You are putting in a new assembly line to assemble some large boards for which your company just received a three-year contract. The boards are 45cm long and you expect the cycle time from the component placement machines to be 40 seconds per board. Your boss is pressuring you to get another 5-zone oven, as they are cheaper and take up much less space than a 7- or 10-zone oven. But, you are concerned that a 5-zone oven may not have the capacity that is needed to keep up with the component placement machines. Let’s make some calculations and see if your concerns are justified.

Table 1 shows some typical reflow oven metrics:

Let’s assume that you will be using a typical modern SAC solder paste. By studying the reflow profile above, we see that the amount of time needed in the heated zone is about 4.5 min. or 270 sec.

So if we choose the 5-zone oven the belt speed will be:

Belt Speed = BS= Heat Tunnel Length/Time in Heated Tunnel = HTL/Time = 180 cm/270 sec. or 0.66 cm/sec

The component placers will be presenting a 45cm board every 40 sec., so the belt speed needs to be:

BS = Board Length/Cycle Time = BL/CT = 45cm/40 sec = 1.125cm/sec

So clearly a 5-zone oven won’t work. What about a 7-zone oven? Let’s calculate the belt speed for this oven.

BS = HTL/Time = 250cm/270 sec. or 0.926cm/sec

Now we can see that the 7-zone oven won’t do the job either.

How about the 10-zone oven? Let’s see if the belt speed is greater than the 1.125 cm/sec needed.

BS = HTL/Time = 360cm/270 sec. or 1.33cm/sec

Success! Since 1.33cm/sec is greater than 1.1125cm/sec, this 10-zone oven will work. The extra belt speed will permit a small amount of spacing between the boards. Let’s calculate what it will be:

BS = (BL + Spacing)/CT = 1.33cm/sec => BL+ Spacing = BS x CT => Spacing = BS x CT – BL

Spacing = 1.33cm/sec. x 40 sec – 45cm = 53.2cm – 45cm = 8.2cm

To summarize: For our 45cm board that has a cycle time of 40 sec., we need a 10-zone oven with a heated tunnel length of 360cm. There will be an 8.32cm spacing between the boards in the oven.

If you would like an Excel spreadsheet to make these calculations send me an email at [email protected].

Cheers,

Dr. Ron

Not with a Bang …

Cemtrex’s run in EMS ended today with a decided whimper. The New York-based company sold its remaining contract assembly assets to a German private equity group for about $7 million.

It brings to a close one of the stranger stories in recent EMS memory. Cemtrex grew through acquisition, buying up a German EMS company, and later another, and green-fielded plants in Romania and India.

Focus was not its strong suit. Along the way, it became involved in virtual reality software and proprietary IoT device design and manfuacturing. It set a goal of $500 million in revenue, but its most recent quarterly results, released this week, put it on a run rate of about $90 million annually, which is behind its pace of just a year ago.

Then there was bewildering and lame hostile takeover attempt of Key Tronic, at the time a Top 50 EMS in terms of revenue. In offering a 1:1 stock swap, Cemtrex called out the larger EMS for its seemingly underwhelming profitability. Key Tronic’s response — at once clinical and dismissive — was one for the ages: “Our initial research shows [Cemtrex] reports approximately $45 million of EMS revenue. In our opinion, this does not qualify [Cemtrex] to make any statements as to how it might operate an EMS business like KeyTronic which is over 10 times [its] current size in terms of revenue.” 

Cemtrex never filed the paperwork for a potential acquisition and the proposal quietly vaporized. And today, so did its aspirations of EMS supremacy.

Ralph Morrison, RIP

I don’t have all the details, but the industry icon Ralph Morrison passed away Aug. 2.

Ralph spent more than 50 years in the electronics industry. He was responsible for teaching thousands of engineers all about the fundamentals of how voltage and currents work. Up until about eight years ago, he was a constant presence at the PCB West design conferences.

He wrote several books, including Solving Interference Problems in Electronics, Grounding and Shielding Techniques in Instrumentation, and The Fields of Electronics: Understanding Electronics Using Basic Physics.

I didn’t know Ralph in his prime, but we developed an amiable relationship over the past 18 months as he did some writing for us. I truly enjoyed him — he had an economy with words that I will always strain to achieve.

Last year, he put together a series of articles for PCD&F: Voltage in the GHz World, Limiting Radiation from Logic Circuit Boards, Maxwell and Transmission Lines, and The Layout of Fast Logic on Printed Circuit Boards.

According to longtime friend Dan Beeker, a memorial service will take place in October. I will miss his emails and occasional phone calls, and most of all, his kind words about the role we editors play in helping designers and engineers better understand their craft.

Silicon Valley Not Paved with Gold

Is the bloom off the rose in the Silicon Valley?

For years, manufacturers have insisted on putting factories in the greater San Jose area. The CIRCUITS ASSEMBLY Directory of EMS Companies lists hundreds of entries with Silicon Valley zip codes. Damn the costs — siting near customers — actual or desired — takes precedence!

In the first quarter, the most up-to-date data available, industrial space vacancy rates were 2.7%, near an 18-year low. That’s despite more than 200,000 sq. ft. of new industrial space coming online in the period, on top of about 3 million sq. ft. of new industrial space that came online last year.

Ironically, industrial space rents, while climbing, are a relative bargain. The average rent was $1.27 per sq. ft. in the March period, more than twice that in 2010 ($0.60 per sq. ft.), but well below the national average. That comes to more than $381,000 in rent a year for a modest 25,000 sq. ft. factory. But tack on energy, and labor costs — unemployment rates are not only lower than the national average, but workers earn a small fortune — and it all adds up to a very expensive enterprise.

Today the pendulum is shifting, if only bit by bit. We are seeing furloughs, layoffs and even some big names starting to blink. Jabil, Creation Technologies,
and this week, Benchmark are among those closing factories in Silicon Valley.

Will more follow? In an industry where margin and cash flow often make all the difference, it won’t be a surprise if more players head for lower-cost pastures.

Trade Wars Battlefield Goes Beyond US-China

The US-China trade war shows no sign of abating. And as predicted, it is wreaking havoc on the electronics supply chain. While OEMs like Huawei get most of the headlines, suppliers of semiconductors are feeling the pain.

Even assemblers are getting caught in the crossfire: When the US government initiated a ban on doing business with Huawei, one of its EMS firms, Flex, withheld a reported $100 million in materials and shipments. According to some Chinese blogs, the OEM now plans to sever its business ties with Flex.

Less publicized, but perhaps just as crucial, is the ongoing spat between Japan and South Korea. The Asian nations are locked in a dispute over a group of uninhabited islands located in the straits between the two countries. Relationships are further frayed over issues of war reparations dating to World War II. Japan has struck first, taking steps to put controls on exports of key semiconductor fabrication materials to South Korea for fear they could be used in military applications. South Korea has not publicly countered, but the nation is home to two of the three largest memory device makers.

The tension between Japan and South Korea further complicates an already cloudy memory market picture. DRAM prices are already threatened by a global inventory glut, leading Gartner to forecast a 4% price drop this year. By choking materials supplies, Japan could inadvertently help lower inventories and boost margins. But there’s a point where parts availability could tip the wrong way, ratcheting up lead times and leaving buyers scrambling for sources.

By sales, Samsung and SK Hynix manufacture about 23% of the world’s memory chips. Their combined sales last year topped $110 billion. But there’s no simple answer. Japan is in many cases a sole source of many of these critical materials. How long will South Korea be willing to suffer? And what lengths will it go to to protect its semiconductor dominance?

What’s the Difference?

Here’s the headline:

F-35 Jets: Chinese-Owned Company Making Parts for Top-Secret UK-US Fighters”

Sounds dramatic.

As it turns out, Exception PCB, a UK-based fabricator, is building circuit boards for the next-gen airfighter. And Exception is owned by Shezhen Fastprint, the Chinese fabricator.

All angst aside, however, how is this different than TTM being the largest board supplier to the US Department of Defense?