Site: Spar Aerospace
21025 Trans Canada Highway
Ste. Anne de Bellevue, Quebec H9X 3R2, Canada

Date visited: May 6, 1997

WTEC: W.T. Brandon (report author), C. Bostian, K. Bhasin, A. Mac Rae



Spar is Canada's largest space company with 2,500 employees and $594 million in sales. Spar is organized into four major business sectors (% in sales): Aviation and Defense (17%), Communications (37%), Informatics, and Space (43%). Because of the size of the space and communications activities, satellite communications is estimated to comprise over half the total company activity.

While organized to be Canada's prime satellite contractor, Spar has more typically served as a payload integrator. The company buys buses from all of the major manufacturers, produces the payload, and performs assembly, integration, and test (AIT). Spar is transitioning from MSAT and RADARSAT towards Ka-band satellite communications and other applications. Spar is currently integrating INDOSTAR for OSC.

Product areas within the sphere of satellite communications include the following:


Definition is proceeding of a Ka-band multimedia satellite system, originally envisioned for Canada only, and now planned as a North American regional satellite. Spar is leading the systems engineering work for this program, including analysis of packet switching in a mobile multimedia environment.



Spar has been an innovation leader for over thirty years and builds many kinds of spacecraft antennas from L-band to Ka-band. Satellite antennas have been delivered to Hughes, Lockheed Martin, Matra and Aerospatiale. Spar leads in high power feed technology3/4in building antennas with low passive intermodulation (PIM) and high multipactor thresholds. Designing for low PIM is difficult and based on experience. Spar has produced PIM-free sun shields and blankets. Spar is working with Hughes on space deployable "Astromesh" antennas in the 7 to 10 m size range (7 m to 21 m at L-band).

Spar is working on phased arrays. Low cost phased arrays or other tracking antennas are not required for Canadian GEO services, even in the Arctic. Spar has done some work on mechanically steered phased arrays to support wideband mobile systems.

Spar was the prime contractor for Radarsat, which has a phased array, electronically steered in one plane by an 8-bit phase shifter.

Efficiency in the electronics is the key to phased array antennas (i.e., problem of power management or DC to rf conversion efficiency). Beam forming is another important problem and multiple spot beams make the phased array design very difficult. Multibeam antennas introduce reliability problems when each beam is produced by different equipment.

Spar worked on phased arrays with optical beam forming techniques after exhausting other approaches. With continuing interest, Spar has collaborated with a McGill University department specializing in optical techniques. Optical antenna technology is certainly useful in terms of mass and power.

Spar is not working with any one prime contractor on Ka-band antennas. But all the primes are aware of Spar and its capabilities. Some U.S. firms may have an advantage over Spar because of a larger military investment in phased arrays.

Each ODYSSEY spacecraft will have 91 beams. Spar has built a prototype and hopes to build the production antennas.

Rf Products

Spar has produced LNAs, SMTs, microwave hybrids and power amplifiers. Spar buys FETs and HEMPTs from Japan and MMICs from the United States. There are only two foundries for radiation hardened semiconductors in North America.

Spar achievements in power amplifiers (PA) include 11 W for ANIK-E and 80 W peak, 40 W average for M-sat. Currently Spar is not in the PA development business. Japan is their main source for PAs in recent programs.

Improved rf power transistors with higher efficiency and lower cost, and more power at increasing frequencies, are needed. At Ku-band (and above) TWTs still dominate.

Onboard Processing (OBP)

Spar hosts observed that ground sophistication, such as TDMA, migrates to space; hence networking and processing are important to future space communications. Spar has broad knowledge of OBP but concentrates on digital products for OBP that go into the air interface. In other words, it concentrates on demultiplexers, demodulators, and uplink access scheduling (demand assignment). A breadboard of a fast packet switch has been constructed, but is not going into the ASIC development that would be required for flight. The OBP system output would interface to a standard ATM switch.

Spar engineers are interested in baseband processing by digital techniques. This is most important for fast packet switching. They are not interested in X.25 (low-speed messaging) or in circuit switching. They described the latter as an Iridium technique. (Spar used fast packet switching as a generic term that includes ATM.)

Spar is working on a terminal that will work with a bent-pipe transponder but which will also work with later spacecraft that have onboard processing.

Spar has developed a protocol for medium access control (MAC) and is working on a satellite test bed.


We discussed MSATs and AMSCs lack of rapid growth. In part this is thought to be due to cellular penetration happening quicker than anticipated. Also, customers know that Iridium, Globalstar, et al., are coming, and will offer handheld terminals.

The current high cost of MSATs is a negative factor. The cost started at $5,000 and may now be about $2,500 (Westinghouse terminal), but people don't want to pay this for a mobile terminal for a "cellular-like capability." The perception of what terminals should cost is based on cellular telephone costs. Regarding a double hop, this has been experienced for many years in Canada and Alaska and it was noted that users learned "not to interrupt each other."

Concerning proposed systems, our hosts indicated that "these global systems have their possibilities, but not at any price."

There is a breakpoint at which satellite communications staves off investment in terrestrial infrastructure. If you go beyond that price or beyond that time, terrestrial infrastructure gets built and then opportunities for satellites become rather limited. The terminal price breakpoint is about $1,000 for Ka-band systems. If these are deployed around the year 2000 at this price then competing terrestrial systems may not be deployed.

There is no one killer application for Ka-band but perhaps a collection of applications. These include high speed Internet access and personal desktop video conferencing (at the right price).

While mainly interested only in ATM and switched services, the telephone companies are finally recognizing that the Internet is here to stay and must be serviced.

The first services for Ka-band will be highly asymmetric. The Europeans are already looking at Ka-band return links for Internet services using a DBS Ku-band downlink. The first will offer 384 kbps back to the hub through a simple repeater on the satellite. It will be implemented using the EUTELSAT HOTBIRD (Ku-band) satellite, which is collocated with an Italian Ka-band satellite.

Digital broadcast video (DVB) and other video-based services are coming. The downlink DVB standard will depend on the forward rate available to the user. The return link will involve trading off the cost of the terminal, transponder characteristics, etc. Spar sees 2 W as the practical upper limit for Ka-band terminal rf power for these applications. The standards will also be determined by the characteristics of IP carried over ATM.

At the design level, Spar is studying optimum implementation of ATM over satellites. It is not going into network management. IP version 6 was mentioned several times, as was the notion that "wireless access" in an ATM environment requires an effort.

The LEO satellite people have not considered digital aspects such as statistical multiplexers, cost of terminals, or OBP; rather the emphasis is on "finance."

The Ka-band filers are changing their architectures in response to changing plans for services. Perhaps the Ka-band FCC filings were assembled by consultants to meet the FCC deadline without the needed systems engineering having been done.

The Ka-band systems are going to have to cope with busy traffic in a multibeam environment. Flexibly distributing power and bandwidth over a wide area is going to require integrating the rf system with onboard processing.

The evolution of DVB was discussed. Phone companies will make short term use of DVB via ATM. There is a question of whether residential or business markets will prevail. If desktop video conferencing becomes popular in business, it will generate demand for Ka-band satellite systems. Success of a "citizens band" video in Europe was noted. This allows transmitting home video to relatives, and a satellite version would expand the residential market. Multi-casting (e.g., to multiple business sites or distributed relatives) reduces asymmetry (contrasted to Microsoft's view of symmetry as "peer to peer").

The major contribution of ACTS may be the propagation experiments. ACTS provides the opportunity to run live experiments while simultaneously measuring the propagation effects. The Crane model is accurate for heavy rain but underpredicts losses in light rain. Clouds, mist, and water film on the antenna all have more effect than anticipated.


The discussion ended with a listing of new technology requirements for envisioned future systems:

Satellite digital front end; digital beam forming; solution to heat dissipation problems in communications payloads; elimination of electrostatic discharge; improved attitude stabilization for controlling beams smaller than 0.25; lighter materials insensitive to thermal distortion; different fading statistics accounted for in communications processor; lens antennas.


par. Annual Report. 1996. 53 pp.

____. Antennas: Design, Manufacturing, Testing, Implementation. 6 pp. [A history of innovation from the 1973 RCA Satcom 1 dual polarization antennas; L through Kuband design, development, test, shaped reflector, phased array and recofigurable antennas].

____. Digital: Design, Manufacturing, Testing, Implementation. 6 pp. [Design, manufacture, test and programming of space qualified digital processors].

____. Media Backgrounder. 80 pp., n.d.; 12 short papers on the company, its organizations and products [Shuttle Remote Manipulator System, Light Duty Utility Arm (hazardous waste derivative of SRMS); Mobile Servicing System, Antenna Contingency System, and Space to Ground Antenna System for International Space Station; high gain satellite antenna for the Earth Observation System; MSAT; Radarsat; and SARSAT].

____. Radarsat: Design, Implementation, Assembly, Testing. 6 pp. [The first Canadian remote sensing satellite features a Synthetic Aperture Radar instrument capable of 9 m x 8 m resolution in the fine resolution mode and a number of unprecedented features; exercise in prime contracting, payload development, integration and test].

____. Thin Film: Design, Processsing, Assembly, Testing. 6 pp. [Wide ranging design and foundry through test of MHMIC and other devices to space quality].

Published: December 1998; WTEC Hyper-Librarian