Traveling wave tubes and the process required to manufacture them have changed little over the past half-century. But these archaic vacuum-electron devices, known as TWTs, generate high-powered radio-frequency signals on some of the most advanced telecom satellites orbiting today.
Over the years, TWTs have become smaller, lighter, more powerful and efficient. But as satellite technology races ahead, production of these key components remains largely stuck in the past.
With global demand for faster and more affordable broadband on the rise, large satellite operators in Europe and the U.S. are starting to populate their fleets with high-throughput Ka-band communications satellites that can deliver high-speed bandwidth virtually anywhere on the planet. Over the next few years, the world's two principal TWT suppliers—and Electron Devices—anticipate a corresponding increase in TWT unit orders, in particular for high-frequency Ka-band tubes that generate huge amounts of spectrum power in the 18-40-GHz. range.
With close to 11,000 TWTs in space, Thales has more than 500 million hours in orbit, “and 50 million hours being added each year,” says Jean-Francois Auboin, director of Thales space programs and engineering. When mated with a regulated power supply, TWTs are ideal for sending large amounts of data over long distances, providing more power and efficiency than solid-state radio-frequency transistors, he says.
Both L-3 and Thales have seen an increase in orders for Ka-band wave tubes in the past few years. At Thales, roughly 50% of TWT units delivered in 2011 were in the Ka-band frequency spectrum, Auboin says. Although he anticipates the trend will continue, comparably ramping up TWT production capacity is easier said than done.
Built one at a time by hand, in production facilities redolent of the 1950s, Ka-band TWTs are assembled in a process that is part engineering, part alchemy, a series of painstaking steps that can take up to a year to produce a single, functioning device.
Hunched over microscopes, tweezers in hand, workers at Thales's Velizy factory look less like radio-frequency engineers than jewelers soldering metal or setting precious stones. Rudimentary as the process may seem, the skill of assembling high-frequency-band tubes is costly and difficult to master. Once trained to assemble C- or Ku-band tubes, a worker cannot simply switch to building higher-frequency Ka-band TWTs.
“They are absolutely not interchangeable with some of the other bands,” says Jerry Ozovek, president of L-3 Communications Electron Technologies. “You can do training, obviously, but there is a certain knack to the Ka-band. It is a difficult product to build.”
Once a wave tube is assembled, a lengthy “burn-in” cycle ensues to confirm the stability of its cathode, focusing, power and gain. Thermal tests then monitor the device's performance across temperature cycles that simulate the vacuum of space, a test and verification process that can take months.
Most time-consuming, however, is the build, where individual piece parts and their precise configuration can differ depending on frequency and power needs. In Ka-band, “you're talking about literally what could be an infinite number of combinations and permutations in terms of configuration that are all driven by the platform designers,” Ozovek says. “It's not as though you can buy the parts, put them on the shelf, wait to see who wins and then build their configuration.”
Although a TWT is comprised of roughly 1,000 parts, only the vacuum tube is standard, leaving suppliers unable to stock parts in anticipation of demand. David Bair, technical director for satellite fleet operator Eutelsat, says component standardization at the prime level could remove a major obstacle for suppliers that is causing “choke points” in the timely delivery of Ka-band satellites.
“Ka-band tubes came into play very recently and put a real strain on the system,” Bair says. “It slows things down and some [manufacturers] have bought tubes on speculation.”
Martin Halliwell, chief technical officer at SES S.A., cautions that satellite manufacturers and suppliers need to react in unison for standardization to work. “It's the correct way to go, but we shouldn't underestimate how difficult this is going to be to get the entire industry to sign up for it,” Halliwell says.
David Bernstein, senior vice president of program management at Space Systems/Loral, says harmonizing designs at the prime level could prove daunting, though some TWT components offer the potential for standardization. One example is the TWT base plate that serves as a mechanical interface between tube and spacecraft.
Bernstein says TWT base-plate designs vary in size and configuration and are determined largely by a spacecraft's mode of heat dissipation. Because heat dissipation is integral to the spacecraft architecture, a fundamental change in its thermal design would be costly, given the time needed to develop, test and requalify a satellite.
“If you change your fundamental approach to thermal design, you're going to have to prove to the customers that it still works,” a process Bernstein says can take several years.
Less drastic changes may be possible, however, such as harmonizing bolt positioning for base plates on competing satellite platforms.
“It's really understanding what flexibility each of us has in that interface and what things we can change and move and what things we can't,” Bernstein says. “Certainly, if the tube manufacturers could improve their cost and schedule, we're willing to talk.”
David Thompson, CEO of., says a projected dip in orders for satcoms in the next two years, when large operators complete fleet replenishment cycles, could afford a window to harmonize equipment standards, “which might be more difficult to achieve in a period of high demand.”
Arnaud de Rosnay, executive vice president for telecom satellites atAstrium, says while standardization is a worthy goal, suppliers could do more to boost production.“I think it's more of a yield problem that I hope will improve with time,” he says, referring to the fact that around 50% of Ka-band wave tubes are rejected during test and verification due to manufacturing defects.
“You have to scrap the tube if you can't get it focused,” Ozovek says. “There's very little you can do in terms of rework.”
Stephen O'Neill, vice president for commercial satellite systems, space and intelligence systems at, says he recently visited facilities at L-3 and Thales to assess production capacity.
“In both cases we were quite pleased to see the effort they've made to reduce lead times in the Ka-band area,” O'Neill says, noting that the newness of Ka-band technology affords a slight learning curve for TWT engineers that is leading to higher yield rates.
Ozovek agrees: “We're closer to two out of three now, but it's still a big number that falls out of the process.”
In the meantime, TWT manufacturers have remained leery of high-volume production since the early 2000s, when anticipated large numbers of telecom satellites failed to materialize, leaving a glut of unused production facilities. In the past six years, however, Thales has increased TWT manufacturing capacity 2.5 times to shorten production cycles that can take up to 18 months to deliver a full set of tubes.
“Our goal here is to increase capacity and to be able to start work on dedicated tubes more quickly,” says Michel Cazaux, head of marketing for microwave and imaging subsystems at Thales Electron Devices.
Long-term supply agreements with primes that guarantee a certain level of volume could also ease bottlenecks and allow suppliers to invest in added capacity. “It may be a year before I have the improvements that investment is going to yield, but at least I have agreements that say I'm going to have product to push through that increased capacity,” Ozovek says.
TWTs provide more power and efficiency than solid-state transistor amplifiers. To see an animation of how a TWT works, check out the digital edition of AW&ST on leading tablets and smartphones, or go to AviationWeek.com/twt