Site: Filtronic plc.
Salts Mill Road
West Yorkshire BD18 3TT
Date Visited: Friday, April 30, 1999
WTEC Attendees: D. Friday (report author), M. Iskander, T. Itoh, R. Rao, J. Winters
Hosts: Professor J. David Rhodes, Chairman, Filtronic
Professor Christopher M. Snowden, Director of Technology, Filtronic plc
Professor Peter Clarricoats, Chairman Technology Advisory Board, Filtronic plc and, Chairman,
Defense Scientific Advisory Council, Ministry of Defense, UK
Eric G. Hawthorn, Engineering Director, Filtronic Comtek
Dr. Richard G. Ranson, Director, Subsystems Engineering, Filtronic Comtek
Alan Needle, Managing Director, Filtronic Comtek
Additional unidentified participants came and went throughout the day
Filtronic was founded in 1977 and is a public limited company (plc) incorporated in England and Wales. Professor Rhodes has been Executive Chairman and Chief Executive Officer since its founding. The company is multinational, with several subsidiaries in Europe, Australia, and North America, and it is a world leader in the design and manufacture of a broad range of products for communication across the radio-frequency (RF) spectrum. These products include microwave and millimeter-wave devices, components and subsystems for wireless communications, cellular handsets, components for hardwired networks such as cable TV and high-speed Internet access, and electronic warfare systems. The company claims to be able to supply products at any frequency, for every transmission standard and every modulation system in the world. Its products primarily receive, transmit, filter, and amplify RF signals. Markets are categorized in four broad areas. These markets, and the percentages of Filtronic's business in each of these areas in 1998, are as follows: Wireless Infrastructure (55.8%), Cellular Handset (26.4%), Electronic Warfare (17.1%), and Cable (0.7%). In 1998, the company had sales of £181.1 million ($295.3 million), almost all its wireless business being mobile cellular. The focus in this report will be on the commercial wireless communications aspects of the business, the first two categories and 82.2% of Filtronic's business.
In 1989, Filtronic bought Philcom Microwave, the first of its U.S. interests. This gave Filtronic a technological edge with which to move, more rapidly, into the base station and mobile communications market. In 1994, the company went public on the London stock exchange. The commercial communications market is Filtronic's primary customer base, and national-defense agencies form a second customer base. Filtronic supplies the latter with electronic warfare equipment and systems. The electronic warfare (EW) part of the business remains privately owned and provides access to the U.S. defense market. The advanced defense technology, especially in communications, also provides Filtronic with a critical edge for commercial applications. In the United Kingdom, Filtronic has facilities in Shipley (the Corporate Headquarters and WTEC visit site), Stewarton, East Kilbride, Wolverhampton, and Milton Keynes. The U.K. activities include the design and manufacture of wireless and cable products, EW products, and ceramic and ferrite wireless components. Filtronic has facilities (LK Products) in Kempele and Oulu, Finland, for the design and manufacture of access and subscriber products. In its Brisbane, Australia, facilities Filtronic designs and manufactures wireless and EW products. In the United States, the company has facilities in Salisbury Md., Merrimack, N.H., Santa Clara, Calif., and Natick, Mass. The more recent of these U.S. acquisitions include Litton Solid State, Sage Laboratories, and subsidiaries.
Primary commercial customers are the leading international original equipment manufacturers (OEMs). Filtronic plc employs approximately 3000 employees worldwide, including 250 engineers, of whom 10 % have Ph.D.s. Filtronic refers to its worldwide holdings and resources as the "Filtronic Group." The Filtronic strategic goal is to become the pre-eminent supplier of RF, microwave, and millimetric products.
Filtronic management, early on, saw materials as critical to competitiveness in wireless technology and acquired a U.K. ceramic and ferrite manufacturing company. The main materials-technology roadmap is focused on requirements for ceramic resonators. The fundamental challenge is price vs. performance for the components produced with these materials. Good performance means lower-losses (higher Q), higher dielectric constants (size reduction), and invariance of the material properties under changes in temperature. Present low-cost ceramic resonators have Q-values ranging from 15,000 to 20,000, and are based on ceramics with dielectric constants ranging from 35 to 45. Current, high-end ceramics have dielectric constants from 45 to 55 and achieve Q-values of 30,000. Filtronic representatives stated that one of the targets was a resonator, at 1.96 GHz, with a loaded Q of 50,000. Their roadmap predicted high-Q resonators with dielectric constants ranging from 70 to 200 in 5 years.
Base station technology is another key element in Filtronic's corporate planning and roadmapping efforts, with amplifier and filter performance the key issues. Dual-mode conductor loaded ceramic and high-Q combline filters are mid-priced and also mid performance ( unloaded Qs of 5,000 to 8,000). TE (0,1,d) and dual-mode TE (0,1,d) filters achieve unloaded Qs of 50,000 but at twice the price. There is a need to drop the price of these components. Filtronic is leveraging the Sage acquisition for base station coupler technology. The wireline technology is low cost and yields high performance among competing technologies. Ferrite components and their manufacturing processes for planar circulators is another base station technology the company is roadmapping.
Filtronic sees three viable options for base station filter technology for 3G: GPRS, EDGE, and W-CDMA. One option is modular systems including integrated systems, micro-controllers, and highly linear power amplifiers. The second option is TE (0,1,d) ceramics technology including multimode, helical, and other designs. The third option comprises improved versions of more traditional machined combline filters. Filtronic's dual-mode conductor-loaded ceramic technology has enabled a one-half size reduction. The insertion loss is 0.5 to 1 dB for a typical base station. These devices have a very good Q (not given) and are the result of an 18-month development period combining both ceramics technology and metalization technology.
Filtronic does not have any projects for using high temperature superconductors (HTS) for base station filters. They cited the nonlinearities and reliability as factors that make HTS technology less attractive as a product line. Filtronic also believes that advances in conventional technology have marginalized the performance gains of HTS hardware.
3G hardware will require improved linear amplifier technology and better efficiency, particularly for wideband CDMA. Software radio is doable should be no problems at the lower frequencies. Transceivers need to be cost effective. There is a need for distributed switching systems in base stations. One objective is to make a significant advancement (details unavailable) in Class A amplifier performance by September 2000. Filtronic is taking an integrated approach, without close project management, and researchers believe they have a key advantage expertise and proprietary software for integration of complex filters.
Prof. Rhodes discussed the issue of the higher power (approximately 20 dB) needed for base stations to have the same coverage for high data rate systems (such as EDGE) as for current mobile radio systems. He discussed the use of adaptive antennas, in particular, pencil beam phased arrays on the uplink and downlink at base stations to provide a higher gain. Key research issues were seen to be multiband phased array antennas for base stations, along with filters and efficient and linear power amplifiers. Tower top electronics were also seen as an issue. The use of terminal adaptive arrays was not felt to be practical (because of the expense), as was the use of higher frequencies (because of diversity issues with point-to-point links). DSP power was not felt to be a concern. Other Filtronic comments regarding base-station antennas were "we can't get away from pencil beams from the base stations," and "multibeam RF is absolutely necessary and difficult."
Comments concerning LMDS implied Filtronic does not see a wide market for this. Researchers said that LMDS in the United Kingdom was licensed in the frequency band 40.5 to 43.5 GHz. Experiments performed led them to the conclusion that propagation reliability will be poor with the moist and rainy weather typical of the United Kingdom. They said that point-to-multipoint architectures were not likely to succeed and that perhaps multipoint to multipoint would be a better architecture a dynamic network in which every user would be a possible relay-node. They didn't see such an architecture being acceptable in the more privacy and security conscious United States. They also feel that most demands can be met with existing bands, and envision 1% of the market for LMDS, mostly wireless local loops. They also said one of their subsidiaries (Radiant Networks plc ) had a 25 Mbps wireless technology and that LMDS is not right from either standpoint.
The purchase of LK Products has extended Filtronic's expertise and its market to include handsets. Customers now include Nokia, Ericsson, Siemens, Bosch, AT&T, Hyundai, etc. Clearly, the only directions here are to make the handsets smaller, less costly, and more multifunctional and to improve their performance. Since handsets are simply housings for the real technology, the only solutions are in advances in the components discussed elsewhere. Filters are a key component and in addition to the ceramic technology discussed under "Materials," surface acoustic wave (SAW) filters are showing promise. Helical filters, traditionally used in handsets, will likely drop out of use as ceramics and SAW improve. The one exception is the handset antenna and integration of the antenna with its housing as well as the RF front end. Improved integration was seen as a key issue here. The cellular air interface also needs attention, particularly the antenna technology, the effects of hand placement, and the user-body effects on antenna performance. Additional comments were "3G specifications, such as 2 Mbps are political issues," "GSM will be driven by power requirements," and "multi-antenna handsets will be essential."
Filtronic does not use HEMTs but developed a capability (by acquisition) for fabricating pHEMTs. Their pHEMTs have very high linearity, TOI up to 20 dB above P-sub-I. Researchers developed the capability for producing MMICs up through mm-waves. Filtronic now has MBE and E-Beam capability, several class 1000 and Class 100 Clean Rooms, and GaAs, InP, and YIG fabrication technology. Filtronic sees semiconductors as one of the most critical technologies, where it needs to maintain cutting-edge capabilities in design and manufacture, in order to remain competitive. The hosts had an interesting observation on competing semiconductor technologies particularly SiGe: They see the frequencies for Si technology being pushed much higher and the costs for GaAs being pushed down. The end result is that SiGe technology will lose its perceived advantage and drop out as an alternative for wireless. The company has no SiGe fabrication capabilities.
Everyone will go to the limits to use these frequencies as much as possible. Some applications assumed to be the domain of higher frequencies will be done in the future at lower frequencies, and industry will develop much more competence at exploiting these frequencies.
Receiving much attention at present, but the ultimate applications and implementations will have to sort themselves out. Many technical (hardware) challenges remain before this technology is competitive. Most demands can be met in the existing band.
Chairman Rhodes stated in his overview that the company's technological niche requires that it operate in one of the world's most highly competitive markets. Very rapid advances in technology characterize this market and its industries. Filtronic, he said, competes on the basis of cost, quality, performance, and manufacturing and delivery deadlines. All R&D and corporate acquisitions are directed toward advancing the technology, the expertise, and the production capability necessary to stay even with or ahead of the competition for the next major procurement contracts. If Filtronic falls behind, it will be adversely affected, and even its OEM customers may become competitors if they can manufacture equivalent products at less cost. Filtronic has a long-term planning process and showed us for perusal (but did not provide) a proprietary copy of their strategic plans, the Filtronic corporate roadmap. Our perspective however, was that the technology requirements forecasting process was critically driven by the need to make the next logical improvement in existing devices, components, and subsystems. The rationale was clear. If the company's product is not competitive in the next few rounds of procurement competition, the long-term future will be irrelevant. Although this essential aspect of the planning process was in the forefront, it was balanced by a longer term vision that encompassed new technologies, experimentation, and prediction of customers' needs through the year 2005. Director of Technology Snowden stated that the horizon was only 5 years out, due to the rapid evolution of markets and technology. William Smith, who spoke regarding the manufacturing operations, said the basic mode of operation is "get the call today, deliver the product tomorrow." A typical product may go from specifications to a volume-manufactured product in less than eight weeks. This is clearly a very difficult environment in which to do long-term technology forecasting. However, Filtronic did develop and provide insights into the next steps in wireless hardware improvements.
The Filtronic management left the WTEC study team with its vision of the five key research topics that, over the next five years, will have an important impact on future wireless communications technology products. The key topics are the following:
Filtronic hosts stated that performance is the bottom line and given the success and rapid growth of Filtronic plc since its inception, this vision may not be too far off.