January - February 1999
eMail the Editor
Jack Fisher on Substrates: An Expert Looks at The Issues
Prior to joining ITRI, Mr. Fisher held several senior management and staff posts at IBM, and is a 31-year veteran of IBM's technology manufacturing and development org-anizations in both Endicott, N.Y. and Austin, Texas. He has also authored a book on the printed circuits industry and has written numerous papers for trade journals and conferences. He was educated at the University of New York, Binghamton. Mr. Fisher also serves as a member of the National Electronics Manufacturing Initiative technical committee and the assembly and packaging committee for the Semiconductor Industry Association. Readers may contact Mr. Fisher at jfisher@itri.org or by phone at 512.833.9930. |
Editor's Note: For this special issue highlighting substrates, we interviewed Jack Fisher, chief technical officer of the Inte-connection Technology Research Institute (ITRI), Austin, Texas. ITRI was founded in 1994 by the internationally-respected trade organization IPC.
Q What are the major influences on printed wiring boards?
A There are two major in§uences on PWB's. First is the continuous drive for smaller, lighter and thinner component technology. Second, area array components like BGAs and CSPs and, in some cases, direct attach §ip-chip are pushing the wiring and connection density of PWBs to new density levels. The OEMs that manufacture handheld and portable consumer products are the force behind both of these trends.
Q What are today's leading methods of making high-density PWBs?
A At last count I believe someone had identified over 30 different manufacturing processes to make HDI PWBs. Many of these are proprietary and can only be used with a license, while others are more universal. It appears likely that two of these technologies: laser-formed vias and photo-formed vias will become the dominate technologies.
Q Why are laser vias and photo-vias becoming the most popular choices?
A The reason is probably the infrastructure. Both of these technologies are being used by more fabricators than the other alternatives and have very competent capital and material suppliers behind them. This is not to say that the other processes will disappear. Some will, but several of the Japanese-developed technologies like B2IT and ALVIH have large, vertically-integrated companies behind them, and will continue to be employed.
Q Why has the price of HDI substrates fallen?
A Unfortunately for fabricators, the price of all PWBs has fallen this year. Charles Lassen of Prismark Partners has prepared a chart that shows price changes over time for HDI boards, high-density conventional boards and commodity PWBs. Over the past five years, HDI boards have typically dropped in price by about 15% per year. In some cases, I would expect to see 17-20% price reductions this year.
Q How will CSPs impact PWBs and vice versa?
A That's a difficult question. However, flip chips and CSPs will definitely affect PWB's by requiring increased density. Alternatively, how PWBs will affect flip chips and CSPs is more difficult to answer. High density, low-cost PWBs are required for large-scale development of the flip chip and CSP markets. PWBs could inhibit both of these component packaging technologies. If the U.S. fabricators maintain their conservative approach to investment in capital and R & D, the U.S. may not be ready with the PWB technology to allow §ip chips and CSPs to fully utilize their capabilities.
Q Will we see major improvements in wiring density and size reduction from HDI boards?
A ITRI, my organization, has seen very significant enhancements in wiring density and size reduction as OEM's have introduced HDI into their products. One U.S. OEM in the communications industry has experienced a 35% reduction in the size and weight of its new products and has not seen an overall product cost increase. Happy Holden, TechLead Corp., has demonstrated at several different seminars that with the proper design rules a 3x-4x increase in wiring density is achievable for the same space.
Q Are existing materials suitable for HDI substrates?
A The materials for HDI are developing rapidly. There are already several major suppliers of photo-imageable dielectrics. There are both liquid and film versions readily available. Development needs are thinner dielectrics and increased peel strength. Development of positive acting dielectrics is also a possibility. Pre-coated foils is another material area that is seeing a lot of activity, and these foils are readily available today. Development needs are thinner copper for ultra fine line etching and new epoxies with higher Tgs and improved electrical performance.
Q What challenges do fabricators in the HDI business face?
A The obvious challenge is capital. All HDI technologies require a capital investment and capital money is difficult to come by right now. Internal plant rearrangement is also a very real possibility. For many HDI applications, clean, contained manufacturing areas are required. Yield of products with microvias is not as good as the yield without microvias. This is an obvious cost factor. Capacity lost due to low yields is less obvious. Some fabricators have said that for every HDI board the put into the line they lose the capacity equivalent of two standard boards.
Q What's happening on the design front for HDI substrates?
A Design tools and design experience is a problem. Dr. Charles Bauer, TechLead Corp., recently drew a parallel between HDI design and the introduction of blind and buried vias into PWBs in the early 70's. In the 70's, the design tools and the designer experience just wasn't there for blind and buried drilled vias. It's the same story today for HDI. The IPC is aggressively working on a design guide and specifications for HDI. Seminars are available for those that can attend them, but more is needed. We need to be very proactive in this area.
Q What factors will enable HDI substrates to succeed?
A The ultimate success factors for any technology are widespread acceptance and infrastructure support. Widespread acceptance has two sides: The fabricators have to accept a technology as viable and capable of high volume manufacturing. They also need to be very sure that their investment will pay off. The other acceptance comes from OEMs. They also have to accept the technology as viable and capable, but equally as important, they have to believe it's reliable, since they are the ones who are going to put their names on it.