Weiner’s World


Reading time ( words)

Editor’s note: This blog was originally published in August 2016 at www.weiner-intl.com  and is being reprinted here with special permission from the author.

Solving Problems in High Density Electronic Packaging

This month, I was a guest at the High Density Packaging Users Group (HDPUG) meeting in Nashville, Tennessee. The consortium, composed of more than 50 companies (small and large) in the electronics packaging supply chain, conducts projects to solve real world problems or develop data for product parameters, package/component life, and production processing. All status presentations were short and to the point. HDP announced the impending completion of eight of its projects. Different companies around the world recommend and participate in the projects of import to their businesses. The results are available for use to members only. The rapidity and output of this particular consortia surprised me. It clearly appears to provide an affordable solution for problem solving and data development in the day of vanishing R&D (especially D) funds.

Lincoln International Adds an Innovative and Global Joint Venture & Partnering Practice

Lincoln International, the leading global, mid-market investment bank, announced today that it has added a new global capability in joint venture (JV) and partnering advisory service.  Lincoln International’s JV & Partnering team’s expertise spans the entire JV and partnership lifecycle, from creating sustainably successful new JVs in emerging economies, to resetting, restructuring and exiting existing JVs and partnerships, globally. A particular area of focus for the team will be driving improved value from the diagnosing and resetting of terms and operations of individual (or portfolios of) underperforming JVs, partnerships and minority stakes.

The Reshoring Conversation Continues

Here are comments from the industry icon who has visited more printed circuit factories in the world, than any other industry member— Dr. Hayao Nakahara, President, N. T. Information, Ltd:

I have a comment or two on reshoring.

My boss at Photocircuits used to drive me out of my office by saying, "You cannot do business sitting on your rear!" Seeing is believing. In the early 1990s, I attended one of IPC's activities and commented, "Watch out for China in the future." One person in the audience stood up and said, "Hey Naka, we've got technologies; let Chinese take care of the cheap and dirty stuff." Where is he now?

China made about $450 million worth of PCBs at that time. It produced $29 billion in 2015. The North American PCB output peaked in 2000 at about $12 billion. In 2015, it made only $2.6 billion (actual production plus $500 million in imports for resale by PCB makers). There are now only two fabricators in North America that have revenue (made in North America) of more than $100 million. Jiangsu Suhan, ranked 50th in China, had revenues of $110 million!

Since my native tongue is not English, I don't know exactly the meaning of word "reshoring." Using my maximum imagination, "reshoring" is to shut down the plants built in foreign lands and bring the business (maybe equipment as well) back to your home land. There have been a few new PCB plants built in the United States in the last few years. I vaguely recall one in Texas, one in New York (GE Medical?) and one in New Hampshire. They are all very small. In my interpretation of reshoring, these are not "reshored" plants.

According to sources, GE Medical got fed up with PCBs sourced from China. I heard the rumor that the Chinese supplier offered 11 pieces for the price of 10. This happens when a buyer keeps looking for cheap suppliers.

Now, look at U.S.-headquartered PCB materials and equipment suppliers. More than 80% of the chemicals used for PCBs are now produced in Asia Pacific, mostly in China. We have only one copper foil manufacturer in the U.S.—Oak Mitsui, whose emphasis is more on thin copper foil for car batteries.

Where does DuPont (sorry, now Dow Chemical) manufacture dry film products, which it invented in 1968? How about laminates? U.S. makers dominated the world in the 1960s through part of 1980s. Only Isola, Park/Nelco and Rogers remain on U.S. soil, and the latter is making more and more overseas. The majority of Isola and Park's production is made in Asia.

How about equipment? Excellon, which dominated the world, is still surviving with a spare parts business and a few mechanical and laser drilling machines sold only in the U.S. What happened to multilayer press makers? TMP, Wabash, Pasadena, and so on? Chemcut is still surviving at the corner of Pennsylvania, but I saw only a few etching machines in Asia made by Chemcut. ESI is doing better than the rest of U.S. equipment makers, but most of its laser drilling machines for the PCB industry are made in Singapore. DIS, based on Long Island, is doing exceptionally well in the international market against ADARA.

ESI and DIS are the only U.S.-based equipment makers I see often at PCB fabricators operating in China, whose output accounts for more than 45% of the world output.

What I am driving at is that if some dreamer tries to build a good size PCB plant on U.S. soil, he has to import practically all materials and equipment from Asia Pacific. Despite the labor laws in China regarding overtime, factory workers in China still work 12 hours a day and six days a week. Unfortunately, labor costs have been rising continuously in China, which is propelling factory automation. Robot panel loading is becoming a common scene in China (VCP plating, AOI, router, etc). Many have adopted the spray coating of solder mask inks, which allows total automation for this function.

I recently visited a PCB factory in China. "Lay-Up" panels are unloaded onto a conveyorized de-pinning line, pins are removed, "lay-up" structure is taken apart and each panel is sent automatically to a deburring machine, then carried into pre-cleaning and sent to a loading station where a robot mounts the panel onto a VCP plating line. This maker has an output of about $120,000/employee/year, about two-times the average of Chinese PWB factory output.

But this takes a lot of costly "internal" development and requires strong financing. Is there any U.S. PCB maker which can afford to or is willing to build a system like this today?

There are several U.S.-based PCB makers operating overseas, mostly in China. TTM Technologies, Multek, Sanmina, 3M, Amphenol PCB, and MFlex (mistake, this was recently sold to a Chinese metal stamping company). These are the ones that I know. If these makers shut down any one of their plants and bring back the business to U.S. soil, this would be true re-shoring. Do they have plans to do this? Hell no!


On top of this, many U.S.-based PCB makers have partners in China (or should I say subcontractors?). PCBs bought by these U.S.-based makers and sold in the U.S. amounts to no less than $500 million a year.

There will always be a need for PCBs made locally for logistical reasons. Design/develop stage, pre-production, design change iterations, small quantities, quick delivery, etc. are some of the reasons for locally made PCBs. But, once the volume begins to build up, ‘price’ becomes everything.


Price comparison of PCBs made in China and the U.S. is not easy because the quantity is normally different. However, hypothetically, for larger volumes, say more than 1,000 panels per order, the Chinese price is invariably 30% cheaper. Except for some very special products, every PCB made in the U.S. can be made in China if there is a need. I have seen 116-layer probe cards in China! I have seen 84-layer backplanes! 34-layer rigid-flex circuits!

Well, let's forget about these endless bashings and stop complaining. The fact is, the U.S. PCB industry is still reasonably healthy—but frail. Many small shops are profitable. The bottom line is profit. Let's keep it that way. Many Taiwanese makers make profits of more than $200–300 million after tax. If they want to, they could use a large chunk of these profits for investing further—mostly in China. There is only one Taiwan PCB maker that has shut down its China plant—Boardtek, because its Chinese workers could not make the types of boards it fabricates at a high enough yield. Unitech chose YiLang, Taiwan, for a new plant several years ago. It has recently broken ground (on 150,000 square meters of land) in Nantong for a new plant. Dynamic broke ground in Huangxi, Hubei Province, in June to build a brand new facility with a final capacity of 5.5 million square feet per month. Unimicron and Wus built plants there, too. AT&S's Chongqing plant is gigantic. Even Taiwan and some European makers are not at all considering reshoring.

My final remark: PCB manufacturing is a volume business if a manufacturer wants to grow sale revenue. Our current production capacity has been cut down to 1/4 of its peak time in 2000, during which we had about 96,000 people working in our PCB industry. Today, it is somewhere around 23,000, plus or minus a few, by adjusting with part-timers. IBM Endicott used to spend $100 million each year in the 1980s for PCB technology development with 600 R&D people. Today, the surviving entity is called i3 and has only $30+ million in revenue. Hewlett-Packard once had 11 plants worldwide. The last surviving plant (Multek Germany) was shut down three years ago because it kept losing money despite its high-technology profile. Viasystems shut down all the plants in North America and Europe. Now it is a part of TTM Technologies.

If and when the management of TTM, Multek, and/or Sanmina decide to shut down their overseas (China) plants and bring back the business to the U.S., I will remove my hat and bow deeply, saying ‘I was wrong.’

PCB West's attendance increased for the seventh year in a row to nearly 2,000—the show's highest turnout in 14 years. Technical conference registration rose more than 20% year-over-year. Attendees gravitated toward sessions on resolving fundamentals and practical solutions to engineering and design problems. More than 24 designers underwent IPC certification during the conference as well. The sold-out show floor featured more than 100 companies occupying 110 booths.

Share

Print


Suggested Items

Calumet Electronics and Averatek Team Up on A-SAP

12/02/2019 | Nolan Johnson, PCB007
Nolan Johnson talks with Brian Hess of Calumet Electronics and Mike Vinson of Averatek about the new, insertable additive processes that the companies are working on together to help factories running primarily subtractive processes to quickly convert to very high-density interconnect (HDI) features, including trace and space from 2.5-mil line and space to 1-mil line and space and below.

Decreasing Bend Radius and Improving Reliability- Part II

11/22/2019 | Kelsey Smith, All Flex
Many of the issues that arise when using a flex circuit come from a lack of knowledge about how to properly design one, especially when the circuit is required to bend. Many novices will design a circuit that calls for bending the flex in too tight of a bend radius, which can cause damage to the circuit and lower the reliability of the end product. This series of articles will focus on the seven key aspects to consider when designing for maximum durability and maximum “flexibility”.

Flex Standards Update With Nick Koop

11/21/2019 | Andy Shaughnessy, Design007 Magazine
This month, I interviewed Nick Koop—director of flex technology at TTM Technologies, a veteran “flex guy” and instructor, and a leader of several IPC flex standards committees. Nick provides an update for the committees he’s involved with and discusses some of the challenges that he sees as more designers enter the world of flex.



Copyright © 2019 I-Connect007. All rights reserved.