Performance of 50 Completed ATP Projects Status Report - Number 2 NIST SP 950-2 Chapter 4 - Electronics, Computer Hardware & Communications
National Center for Manufacturing Sciences (NCMS) A Collaborative Effort to Address Advanced Technology Needs of the U.S. Printed Wiring Board Industry

COMPOSITE PERFORMANCE SCORE (Based on a four star rating.)

A printed wiring board test vehicle used by the surfaces finishes team, with different solderability test method coupons labeled.

U.S. Producers Loosing in Global Competition
Between 1984 and 1990, the U.S. world share in the manufacture of printed wiring boards declined significantly, from 42 percent to 26 percent, with lost market share captured by lower-cost foreign competitors using increasingly sophisticated technology. In 1991, the profile of the U.S. PWB industry was one of many small firms with little market power or research capability: 725 firms produced PWBs, but most had annual sales of less than $5 million in specific niche markets. These small firms lacked the resources to undertake R&D on a scale that would achieve technology breakthroughs, and could only support R&D that addressed particular and immediate product development goals. Meanwhile, users of PBWs were forecasting that in the near future they would have to procure an increasing share of PWBs from foreign producers whose technical advances were outstripping those of U.S. producers.

ATP Supports Collaborative R&D Initiative for “Leapfrog” Technical Advances
A few leading equipment suppliers with advanced research capabilities, concerned with the decline of the U.S. domestic PWB industry, coalesced to form a research joint venture to address the technology gap faced by domestic PWB producers. These equipment producers, all major consumers of PWBs, utilized their membership in the National Center for Manufacturing Sciences, a nonprofit organization that specializes in brokering collaborative R&D projects, to organize the joint venture. The NCMS helped to organize the collaborative project among four of its member companies.

The joint venture, comprised of the four major companies, with NCMS as the coordinator, applied in the ATP’s 1990 General Competition for a five-year project. The initial participants were AT&T, Digital Equipment Corporation (DEC), Hamilton Standard Interconnect Systems, Inc. (a division of United Technologies, Inc.), and Texas Instruments. Sandia National Laboratories joined the project shortly after it was formed.

ATP awarded the project $13.8 million. Industry participants matched ATP’s funding with $14.7 million, with a total project budget of $28.5 million. Based on the proposal submitted by NCMS, ATP awarded NCMS $13.8 million toward the five-year project, scheduled to break ground in April 1991. Industry participants matched ATP’s funding with $14.7 million, with a total project budget of $28.5 million. During the project, Sandia, funded by the Department of Energy, contributed an additional $5.2 million to the project.

Over the life of the project, membership in the joint venture changed. Eighteen months into the project, DEC decided to withdraw. During the next three years, Allied Signal, Hughes Electronics, and IBM joined, fulfilling the research agenda originally assigned to DEC.

Successful Collaboration in a Horizontally Structured Joint Venture
Studies have shown that horizontally structured joint ventures of direct competitors may experience difficulties in establishing trust and sharing information.⁽³⁾ And most of the participating companies produced PWBs. This largely horizontally structured joint venture, however, appears to have been unusually successful in achieving active collaboration among participants, substantive teaming to carry out major project tasks, and extensive sharing of information. There are several factors that seem to have contributed to its collaborative success.

One factor is that the companies, though producers of PWBs, were also consumers of PWBs and rarely direct competitors with each other. They utilized PWBs for significantly different products.

Another factor that appears to have contributed to the successful collaboration is the administrative and management arrangement. The NCMS, as project coordinator, handled the administrative tasks, including accounting, contract details, and legal and intellectual property issues. A steering committee, staffed by technical personnel from each joint venture participant, managed the technical aspects of the project.

Project Goals
The NCMS project sought progress in PWB technology and manufacturing in four areas: materials, soldering, imaging, and chemical processes. Research teams — staffed by technical personnel from each joint venture participant — were formed for each area.

Although there are two types of PWBs, rigid and flexible, rigid PWBs account for approximately 90 percent of global PWB production. Rigid PWBs are constructed from a fiberglass base, made by taking woven fibers of glass, filling the weave with epoxy resin, and applying pressure to form the sheets that harden and become rigid. The materials team focused on the construction of rigid PWBs, usually made from sheets of double-ply fiberglass, in turn made from two layers of woven glass. The materials team sought to develop cost-saving, single-ply boards. They had to overcome a commonplace obstacle presented by single-ply sheets: these tended to have surface irregularities that led surface metal circuits to meet and short out when the surface warped under heat or pressure. Thus, the materials team had to deliver a regular surface on a single-ply, rigid PWB.

The components on a PWB are connected by soldering component leads to the surface of the PWB. Soldering defects are a significant problem in the production of printed wiring assemblies, basically PWBs that are populated with components, such as integrated circuits. The soldering team aimed to develop ways to automate testing, to reduce the number of defects by identifying and using different materials in the soldering process, and to produce alternative surface finishes. The testing and repair of soldering defects provided an area of potential cost saving: most testing heretofore was performed manually.

The soldering team also sought to develop an alternative to solder to prevent the oxidation of copper surfaces on PWBs in storage. Oxidation of the copper on the stored boards renders the contacts—the copper surfaces to which component leads are attached—unsolderable. In the past, contact surfaces have been preserved by solder. The use of solder, however, is messy and imprecise and limits the application of PWBs in circuit designs requiring increasingly fine connections. The goal of the soldering team was to develop a surface finish that would prevent oxidation, but at the same time would leave the copper surface free of obstacles to the etching of fine circuit patterns.

The imaging team sought to investigate and stretch the bounds of the imaging process to improve resolution, conductor yield, and dimensional regularity. It aimed to improve the projection of patterns along which circuits are placed to reduce the number of defects on boards.

The fourth team, originally known as the chemical processes team, later the product team, sought to advance the understanding of the chemical processes involved in producing the copper plating used in the PWBs to interconnect the components. The chemical processes team aimed to develop a thinner copper plating that would adhere to the PWB fiberglass base. This approach was seen as a way to save money by using less copper and reduce processing time. In cooperation with Polyclad, a copper plating supplier, they tested the performance of thinner copper plating. When the chemical processes team ran into difficulty securing financial support from joint venture participants for other copper research not considered high priority, the joint venture steering committee formed another team. The new product team had the more general goal of developing high-density interconnect structures, considered a high priority goal. In making this change, NCMS worked with ATP to redefine the research effort.

A close-up of an individual chip attached to a pinted wiring board.

Research Achievements
The efforts of the various teams paid off. The materials team achieved the noteworthy goal of developing and demonstrating the feasibility of single-ply fiberglass for use in PWBs. They demonstrated that it could be produced with a sufficiently regular surface to allow for reliable electrical performance at a substantial cost savings. It has become the industry standard and the military has adopted it as a reliable product.

The materials team also developed a plasma-monitoring tool, which has a large potential for cost avoidance. Originally developed to monitor printed circuit boards, Sandia National Laboratories, in the post-project period, extended the application from PWBs to detect defects in microchips. If a defect in microchip manufacturing goes unnoticed after the wafer has undergone multiple processes — for instance, if circuits are etched to the wrong depth — it can translate into millions of dollars in expenses or lost sales after the ATP project was over, as part of a Sematech project. The effort resulted in a spin-off company formed by Sandia scientists, Peak Sensor Systems, which has filed 19 patents for related technologies, and has had 3 granted. Now in its third year, the company has three models of the Peak Propak plasma-monitoring device, and has achieved $1 million in sales. The research path is sketched out past the ATP project, in this case, to illustrate the complementarities that often exist across research efforts.

The materials team also developed PWB applications for a block copolymer that facilitates lower copper profiles and thinner materials. Scientists from Sandia, concurrent with the ATP project, developed a block copolymer technology under internal Sandia funding, which they were able to patent. The Sandia scientists as part of the materials team developed PWB applications for the block copolymer technology. Sandia also has a patent pending for prepeg bonding copper.

The soldering team successfully developed better methods of testing solder and produced a surface finish that adequately protects the board in multiple soldering applications. The team also explored applications of the imidazole solution—originally developed by Bell Labs⁽⁴⁾ — to prevent oxidation of copper surfaces in a solder-free manner, and to demonstrate its applicability for a wide range of uses. Joint venture participant AT&T subsequently licensed the technology to LeaRonal, Inc., which commercialized the treatment under the brand name Ronacoat OSP. The company describes its product as a “production-proven, low-cost, environmentally benign alternative” to substitute products.⁽⁵⁾

The imaging team was able to achieve PWB production process improvements. The imaging team developed methods to increase the yield of PWBs without flaws, and in fact, the yield on 3-mil board increased from 30 to 50 percent, and the yield for 2 mil boards from 10 to 50 percent. Another way the imaging team improved quality control was to introduce a test procedure for evaluating the underlying process used to produce PWBs, based on the analysis of a sample board. This development stimulated the creation of a spin-off company—Conductor Analysis Technologies. The company sells this testing service to equipment manufacturers and PWB shops.

The imaging team was also successful in demonstrating the feasibility of a new photolithography tool called Magnified Image Projection Printing. This tool has the potential to provide a contact-free way of printing PWBs, which could in turn eliminate problems caused by the current imaging-etching process. If successful, this technique could significantly improve manufacturing of the boards by allowing for finer, more tightly interconnected lines as little as three mils apart. The NCMS decided to pursue the development of the prototype tool as a separate project, without ATP support, and the final report of this research effort is now available from NCMS.⁽⁶⁾

Before it was regrouped as the product team, the chemical processes team developed thinner copper plating and tested its performance in cooperation with Polyclad. The chemical processes team was able to demonstrate the effectiveness of thin copper plating in adhering to fiberglass. The approach saves money because it uses less copper. It also reduces processing time, because less copper has to be etched away to make connections on the board.

Research efforts by the product team on high-density interconnect structures produced a novel interconnect structure—the Multilayer Organic Interconnect Technology (MOIT)—that has the potential to revolutionize the fabrication of PWBs. By radically increasing the number of connections to and from a device on the board by taking advantage of making connections from the bottom of the board, in addition to the standard surface connections, the MOIT offers the possibility of achieving much higher wiring density than current PWB technology. IBM Endicott is pursuing MOIT.

Public Recognition
Industry recognition of the successful research efforts of this project includes best paper awards to two of the many research papers that came out of the project. One of these awards was presented at the fall 1994 meeting of the Institute for Interconnecting and Packaging Electronic Circuits (IPC⁽⁷⁾).⁽⁸⁾

The president of NCMS, John DeCaire, also bes-towed public praise on the project at a press conference in 1997. He stated that the ATP project had “quite literally saved the [then] $ 7 billion U.S. PWB industry—a key segment of the $20 billion domestic electronic interconnection industry that employs over 200,000 people.”⁽⁹⁾

Technology Diffusion and Industry Impacts
The NCMS project achieved significant technical successes in each of the four areas targeted by the joint venture participants. Three of the efforts—single-ply boards, thin copper plating, and conductor analysis technology—delivered commercial advantages to the PWB industry and economic benefits to the equipment manufacturers who are their customers.

The project’s successes and the acclaim received have furthered PWB technology diffusion within the industry. Technology diffusion among industry suppliers was aided by the decision of project leaders to include suppliers to the PWB industry in the process of developing the technologies. In fact, over the life of the project, almost every team meeting involved working with suppliers across various industries such as glass manufacturers, weavers, laminators, resin manufacturers, equipment manufacturers, printers, and chemical suppliers.

Producers of PWBs report gains from project related advances that have been implemented into production. For example, 85 percent of the PWBs used by AT&T now incorporate single-ply technology, and this conversion to single-ply technology has saved AT&T at least $3 million per year. Another company reported a 50 percent decrease in solder defects, due to the soldering process improvements developed during the ATP project.⁽¹⁰⁾

The ATP catalyzed a small group of companies with advanced research capabilities to undertake a wide-reaching project that would benefit the entire industry. The resulting NCMS-coordinated project conducted a research effort that would otherwise not have been pursued by the individual equipment manufacturers and PWB producers without the ATP, or only at a slower pace and in a far more costly way. The project encouraged the research teams to pursue two technologies—single-ply boards and thin copper plating—that challenged conventional wisdom, which held that these options were unfeasible.

An independent study undertaken on behalf of the ATP concluded that the ATP's presence supported and sustained research outcomes. The study's survey of joint venture participants found that "of the 62 research tasks completed by the PWB joint venture, about one-half would not have been undertaken at all in the absence of ATP funding, and the remaining one-half would have been belayed by at least one year without the ATP, in an industry where timing is critical."⁽¹¹⁾ In addition, the study found that by collaborating on the research, the companies saved an estimated $35.5 million on those tasks they would have undertaken anyway, although at a slower pace. Savings resulted mainly from the labor efficiences achieved and the duplicative test equipment avoided. (The study did not assess in a systematic way the benefits from the part of the research that would not have been done at all without the ATP.)

____________________
1. The Sandia scientists received a patent for the concurrent technology supported by internal Sandia funding —“Block Copolymer Adhesion Promoters via Ring-Opening metathesis Polymerization,” patent number 5,603,985.

2. IPC, “Management Briefing on the Global Outlook for 1999,” available by request from the IPC director of market research via e-mail <CarlaWehrspann@ipc.org>.

3. See, for example, Dyer, Perspectives on the Determinants of Success in ATP-Sponsored R&D Joint Ventures: The Views of Participants, NIST Contractor Report (In Press 2000).

4. Bell Labs patented the formula in 1983, patent number 4373656, and AT&T licensed the technology to LeaRonal, Inc. Viasystems is the successor owner of the patent.

5. See LeaRonal’s, website at <www.learonal.com> for a listing of products and their applications.

6. See the NCMS website at <www.ncms.org> for a description of the report, Magnified Image Projection Printing Project Final Report. The MIPP project was dissolved prior to the completion of the illumination system design and building of the prototype due to the withdrawal of the system integrator, and no viable commercial integrator showed interest in participating in the project. The intent of the report is to preserve the knowledge generated during the project if any future efforts are undertaken.

7. IPC is a U.S. based trade association of nearly 2,500 member companies representing all facets of the electronic interconnection industry, including design, PWB manufacturing, and electronics assembly.

8. Albert Link, Early Stage Impacts of the Printed Wiring Board Research Joint Venture, Assessed at Project End (1997), p. 28.

9. May 13, 1997, NCMS News Release, “Landmark Collaborative Research Program Credited with Saving Domestic Printed Wiring Board Industry.”

10. Link, Early Stage Impacts p. 26.

11. Link, Early Stage Impacts, p. iv.

Return to Table of Contents or go to next section.

Date created: April 2002
Last updated: April 12, 2005

ATP website comments: webmaster-atp@nist.gov  / Technical ATP inquiries: InfoCoord.ATP@nist.gov

NIST is an agency of the U.S. Commerce Department
Privacy policy / Security Notice / Accessibility Statement / Disclaimer / Freedom of Information Act (FOIA) /
No Fear Act Policy / NIST Information Quallity Standards / ExpectMore.gov (performance of federal programs)