HTS: Poised for Growth

High-temperature superconductor (HTS) technology is not a new, innovative concept. Vendors have been developing the technology since the late 1980s. But it's only been recently that companies actually have started deploying these filters.

Illinois Superconductor was first to market with a commercial order in the summer of 1996 with Southwestern Bell Mobile Systems. Conductus tested its first units with CellCom in 1996 and announced its first product last year. Superconductor Technologies conducted field trials in 1998 with more than 30 cellular carriers. It reported the system performed well in all trials, improving traffic call minutes by 100% in some cases.

According to vendors, carriers have been delaying implementation because they are waiting for real-world application data. With field trials under way and successful deployments to their credit, vendors are convinced this technology is finally ready to take off.

"Although we are seeing a small percentage of superconductor system use right now, we should be seeing large-volume commitments by carriers during this next year," said Jim Simmons, Conductus vice president of sales and marketing.

Blake Isaacs, Superconductor Technologies director of marketing, agreed.

"There were only a couple hundred high-temperature superconductor units sold throughout the entire wireless market last year," he said. "All indications are that there will be a market for several hundred units in 1999."

THE TECHNOLOGY In the late 1980s, vendors were looking for a way to filter out distortion, reduce noise on the uplink signal, but still amplify receiver sensitivity. So they began developing new technologies that make it possible for superconductors to carry electrical currents at far less resistance than conventional conductors do. Superconductor filters must operate in temperatures lower than -300*F. They are installed with low-noise amplifiers at the input of base-station receivers.

According to vendors, these filters can fill in the coverage holes that digital deployment creates. Initially, carriers placed towers strategically to overlap coverage for 3W units. As a result, older sites do not have the selective sensitivity to extend as far as today's 0.6W phones require.

For example, TDMA systems have a shorter signal range than analog systems. In most rural and suburban areas, there are dead zones of 7dB to 12dB holes where uninterrupted coverage was once available for high-powered phones.

Vendors claim that superconductor filters also can improve CDMA range and coverage by lowering the noise floor and noise figure. Although testing indicates superconductors reduce dropped calls and extend coverage for digital networks, vendors still are conducting field trials to support that claim with real application data.

"There is a pretty intense technical argument waging about whether the capacity of a CDMA system can be improved by lowering the noise floor," said Hein Moritz, Illinois Superconductor director of systems engineering. "Some of the theory says 'yes,' but we have no data to support or refute that now. We think we can, and we are looking in that direction."

IMPLEMENTATIONS As vendors try to prove their case for superconductor filters, many carriers have been relying on other methods to solve network-coverage problems. Carriers use tower- or rack-mounted amplifiers to help capture some of the signals. They also install new towers to fill in gaps.

Tower-mounted amplifiers cost $5,000 to $6,000 each, and improve some coverage in low-noise environments. Rack-mounted amplifiers, although less costly, boost signal transmissions but also amplify the noise field and create distortion.

Not only are additional tower sites hard to come by, they can cost more than $350,000. Therefore, building towers to fill dropout areas is rarely a practical solution, financially. This is particularly true for small- and mid-size networks that serve low-density populations.

Vendors claim that HTS filters are a more cost-effective solution. Superconductor front-end products begin in the range of $10,000, which makes them an attractive option if you want to maximize coverage.

Superconductor filter systems are ideal for suburban areas where the network's uplink signal and handoff do not perform well, according to Isaacs.

"The suburban area is the most compelling application," he said. "The denser signal environment will create interference issues, but the filter is selective, so amplification works especially well."

Because call volume is higher in suburban areas, you will recoup the upgrade investment quickly. But this solution is not ideal for all carriers. Generally, dense urban cores have a lot of in-band interference, making filters less effective. However, even in those urban areas where extended coverage is not needed, HTS can improve in-building coverage and extend battery life for low-powered phones. In extremely rural areas, superconductor filters can close coverage gaps, but with lower call volumes, you won't recoup the investment as quickly.

Cellular One of East Central Illinois deployed superconductor filters in its network, which serves 280,000 POPs dispersed over 11 counties.

"We tried tower-mounted amplifiers and saw very little improvement in outlying areas," said Kathleen Robbins, network manager. "With the superconductor filters installed, we saw a 5-fold increase in coverage, range and clarity."

The filters extended range three to four miles, according to Robbins, which was beneficial in a region where it installed towers to provide analog coverage for 3W phones.

"There is one area where four towers are positioned so that their coverage was originally intended to overlap in the center, which is about 10 miles from each tower," she said. "This is the area with the weakest signal strength. Superconductor filters dramatically extended coverage, providing a clear signal."

Cellular One installed the filters about six months ago and already has noticed extended call duration.

"At this rate, we project a payback in eight months," she said.

U.S. Cellular purchased 25 filter systems recently. Richard W. Goehring, U.S. Cellular senior vice president of engineering and network operations, said superconductor filters enhance network performance, coverage and call quality, and play a critical role in base-station deployment strategies for its small- and mid-size communities in 25 states.

Other carriers have reported a 44.8% increase in the average monthly minutes of use after deploying superconductor filters. Adding one superconductor receiver front-end system to problem base stations immediately closed 28-mile coverage gaps and improved overall call quality.

One of Illinois Superconductor's commercial installations is at a 5-site cluster near Chicago's O'Hare International Airport. High call volumes and high levels of adjacent band interference caused problems for travelers and airport personnel. Filters cleaned up the spectrum, controlled adjacent band interference, improved in-building coverage and allowed the use of channels that interference previously blocked.

In another application, a co-located antenna on top of a 20-story building with more than 50 additional transmitters served 300 square miles. Performance degraded drastically over the years as subscribers switched to low-power portable units. Superconductor deployment dropped the in-band noise floor to an average of -105dBm.

Superconductor filters also expand B-spectrum channels, which conventional filters do poorly, if at all. In addition, some systems are designed to provide inherent bypass. According to Illinois Superconductor's Moritz, systems are designed to become a conventional filter if a power failure depletes the cooling function. The end user never loses service.

KNOW THE SOLUTION Although there are only a few superconductor vendors in the market today, they offer an array of solutions. For example, filters are composed of thallium barium calcium copper oxide or yttrium barium copper oxide, which are painted in multiple layers on a basic substrate and fired in a kiln to form a 3-D thick film. Or they are created in 2-D wafer-like microchips, called thin film.

Advocates say that thick film is a practical solution that offers performance and size benefits. Illinois Superconductor contends that thick film provides good selectivity and productivity, low material costs, flexible product designs and low intermodulation levels.

According to vendors that use thick film, thin film is susceptible to irregularities in the substrate, which degrades circuit performance and requires engineers to build the circuitry in a special protected, clean-room environment. In addition, Moritz said thick film handles a lot more power than thin filters.

Superconductor Technologies, on the other hand, keeps filters small by using wafer-thin film.

"The big lesson of 1998 was that carriers want a practical solution," Isaacs said. "They want a smaller unit, easy to install and low-maintenance."

According to Isaacs, Superconductor Technologies can compact everything required for a complete sector cell site -- six complete filters, amplifiers and RF pads -- into a 17"x19"x7" unit that weighs less than 50 pounds.

"It is not enough that the product demonstrate benefits," he said. "It must be practical, as well."

Conductus' Simmons agreed that thin film is more practical, compact, highly selective and provides the flexibility to adjust gain and configurations as needed from cell site to cell site. He said thick-film systems require a 4-foot-high system. Vendors can configure thin-film wafers into a 10-inch unit, and install it, along with the cooler, in one 19-inch rack.

But not all vendors offer the same cooling system either.

"The cooler, with its mechanical moving parts, is where reliability is the most important factor," said Charlie Shalvoy, Conductus president. "Stirling cycle coolers can wear out at a much faster rate than Gifford-McMahon coolers operating at higher powers. Our strategy is to select the right cooler depending on the customer's application."

In government wireless products where portability is important and reliability requirements are less stringent, Conductus uses Stirling coolers, which typically have a mean-time failure rate of seven to 27 months. For cellular and PCS applications, where long lifetime is critical, it uses Gifford-McMahon coolers with a 15-year failure rate.

According to Isaacs, however, an advantage of thumbnail-size circuits is that they don't need a lot of cooling power, typically less than 100W, as opposed to the 550W that other superconductor coolers require. Superconductor Technologies uses a small Stirling cycle cooler to maintain HTS applications' low temperatures.

THE FUTURE Robert Hammond, Superconductor Technologies CTE, said the fundamental performance of superconductor technology -- power handling and Q selectivity -- is improving at a dramatic rate.

"HTS is a breakthrough in both sensitivity and selectivity, the two fundamentals of receiving wireless signals," agreed Peter Thomas, Superconductor Technologies CEO.

As these improvements continue, Isaacs said superconductors will become standard equipment for most carriers. HTS could even displace the conventional front-end completely. Instead of a performance enhancer, the filter could become the whole RF system, everything between the antenna and radios.

"Other stages of the receiving equipment, and perhaps even transmit equipment, could eventually upgrade to HTS technology," Moritz said.

With European and Japan-based carriers investigating these solutions and U.S. deployments increasing, vendors expect 1999 will be a key year for superconductor technology.

"There is increasing evidence among some of the larger carriers that HTS will be used broadly throughout wireless networks," Thomas said. "We have every reason to expect this trend to continue. 1999 will become the year of volume deployment of superconducting systems in the United States."