Project Telstar: Communications Experiment 

29 January 2021


Equipment and operational procedure of The Bell System’s Telstar satellite are described with reference to its communications function. The satellite’s electrical circuits are divided into three parts for purposes of explanation: communications repeater, command and telemetry system, and power plant. Three subsystems comprising the ground station are described: communications, antenna direction, and command and telemetry. Orbital data and the results of linear transmission tests are given.



SMPTE journal cover

Adapted from an illustrated lecture presented on 24 October 1962 at the Society of Motion Picture and Television Engineers convention in Chicago by Hugh P Kelly of Bell Telephone Laboratories and reprinted in the Journal of the SMPTE issue 2, volume 72, in February 1963

THE CHIEF OBJECTIVE of the Bell System’s Telstar Communications Experiment is to test an actual broadband communications satellite. The Telstar satellite was originally intended for experiments in telephony, data transmission and single-channel television. While not primarily designed for two-way telephony, 60 simultaneous conversations can be handled. In addition, the satellite measures the condition of some of its own electronics equipment under the stress of launch and in the space environment, as well as measuring radiation levels in space, a function completely separate from the communications experiments.

The experiment also explores the best technique for accurately tracking a moving satellite. Finally, ground-station equipment is tested under operational conditions.

Equipment used in the Telstar experiment is shown in the functional block diagram, Fig. 1. The sequence of operation followed when the satellite comes into view at the ground station in Andover, Me. [Maine, USA], is listed below. Pertinent data on transmitted powers from the satellite and ground station are shown in Table I.

(1) The command tracker searches for the 136-mc signal and locates the satellite;

(2) The command transmitter turns on the telemetry to evaluate the state of health of the satellite;

(3) The transmitter starts the 4080-mc beacon by turning on the TWT;

(4) Both the precision tracker and the vernier autotrack, which were initially pointed from the ephemeris, locate the satellite and lock-on the 4080-mc beacon;

(5) The precision-tracker data is used to update the ephemeris and, after lock-on, the large antenna is controlled by the vernier autotrack sending corrections to the computer;

(6) The FM communications signal is sent from the ground transmitter to the satellite via the horn-reflector antenna;

(7) The satellite receives the signal, amplifies it, and shifts the frequency to 4170 mc;

(8) The FM signal is then radiated from an omnidirectional antenna to all visible points.

For the first 24 hours after launch, approximately 9 orbits, the positions of the satellite are as shown in Fig. 2. Because “usable” time means the satellite is at least 7.5° above the horizon, the usable time, as shown in Fig. 3, is shorter than visibility time of Fig. 2.


Transmitter Frequency (mc) approx Maximum power (w) Antenna gain (approx)
– Communications 4170 2.25 Isotropic
– Beacon 4080 0.02 Isotropic
– Telemetry and Beacon 136 0.2 Isotropic
Ground Station
-Communications 6390 2000 60 db
-Command 120 200 16 db


Polarisation of Transmitted Waves
4-kmc satellite Left circular
136 satellite Linear
120-mc command Right circular
6-kmc ground Right circular


Table I. Project Telstar Transmitted Signals.


System diagram

Fig. 1. System diagram.


Orbital data

Fig. 2. Orbital data


Wavy graph

Fig. 3. Minimum usable time per day.



System diagram

Fig. 4. Communications repeater.

The electrical circuits in the satellite will be divided into three parts, for purposes of description: communications repeater, command and telemetry system, and power plant. The broadband communications repeater is shown in Fig. 4. FM signals centered at 6390 mc at a —50 to —70 dbm power level are sent to the repeater from the ground station. This signal is converted to 90 mc, amplified and applied to a modulator to convert it to a center frequency of 4170 mc. The drive to the traveling-wave tube (TWT) is constant over the input signal range given above since a diode monitor controls the gain of the IF amplifier. Changes in range and lack of isotropy in the antenna pattern cause a variation of input signal.

Also amplified by the TWT is a weak signal, at 4080 mc, from a string of frequency doublers. Part of the amplified output is fed back to act as the local oscillator for the up converter and part of the output goes to another modulator to produce the 6300-mc signal needed at the down converter. The remainder of this 4080-mc signal, about 13 dbm, goes to the transmitting antenna and acts as the microwave beacon used to track the satellite. The radiated power at 4170 mc is about 2¼ w. It is necessary to operate the tube about one db below saturation, since two signals are amplified by the TWT, to reduce intermodulation of the two signals.


Aspect graphs

Fig. 6. 6-kmc antenna patterns: left, equatorial aspect; right, polar aspect.



Fig. 5

Refer to Fig. 6. Both the microwave antenna for receiving, at 6 kmc, (Fig. 8), and the one for transmitting, at 4 kmc, are very similar. Each consists of a large number of radiating ports uniformly spaced around the equator of the satellite to give an isotropic radiation pattern. Both antennas are circularly polarized; the 4-kmc antenna is to the left and the 6-kmc antenna is to the right. Figures 6 and 7 show the respective antenna patterns.

One of the antennas is used by the command receiver and the VHF beacon. When isolated from the satellite, this antenna is circularly polarized. However, the satellite surface acts as a ground plane, so when the antenna is mounted it is almost linearly polarized, as can be seen in Fig. 5. Figure 8 shows the final radiation patterns. Those at 136 and 123 mc are very similar.

Depending on range and satellite orientation, the 123-mc signal from the command transmitter arrives at the satellite at —80 to —100 dbm. For reliability, the system includes two radio receivers which receive the pulse-coded commands and two decoders which translate the received pulses into usable instructions. These are shown in Fig. 9. The command switching unit, comprised of 15 commands which are described in Table II., controls nine relays which turn the appropriate circuits on and off.


Aspect graphs

Fig. 7. 4-kmc antenna patterns: left, equatorial aspect; right, polar aspect.


Aspect graphs

Fig. 8. VHF antenna patterns: left, equatorial aspect; right, polar aspect.


Command Function
D Turns on telemetry and energises radiation experiment circuits.
A Turns on TWT filament voltage.
B Turns on TWT helix and collector voltages. Energises all transistor circuits associated with communications experiments.
C Turns on TWT by applying TWT anode voltage.
CC Turns off TWT anode voltage.
AA Turns off TWT helix, collector and filament voltages. De-energized transistor circuits used in communications repeater. There is no BB command.
DD Turns off telemetry and de-energizes radiation experiment circuits.
T1 Turns off command receiver and decoder #2 for 15 sec so command receiver and command decoder #1 can be tested.
T2 Similar to T1, used to test command receiver and decoder #2.
E Turns on current in the orientation loop.
EE Turns off current in the orientation loop.
F Connects telemetry encoder #1 to circuit.
FF Connects telemetry encoder #2 to circuit. F and FF also control direction of current in orientation loop.
SS Performs duties of CC, AA and DD plus removing all load from the batteries.
S Connects batteries back in the circuit.


Table II. Telstar Commands.


System diagram

Fig. 9. VHF beacon, telemetry and command diagram.

The PCM-FM-AM telemetry system provides information once a minute on 112 channels. Using an eight-bit word code, seven information bits and one sync bit, the system generates binary words. These words are used to frequency modulate a 3-kc subcarrier which, in turn, amplitude modulates the 136-mc beacon signal. Although the 136-mc signal is modulated only when telemetry is used, it is always transmitted except when the batteries are disconnected by command.

The solar-cell plant (Fig. 10), is the primary source of power for the satellite. 50 groups of cells are connected in parallel. Each group contains 72 cells in series. Through diodes, they charge 19 nickel-cadmium batteries. The batteries are necessary because maximum power to all satellite components is about 35 w and the solar cell can furnish only about 14 w. In the event the low-voltage cutoff operates, the batteries can be charged, but they cannot furnish any load until commanded to do so. With the exception of the traveling-wave tube, the 16-v output of the main regulator powers all units.


System diagram

Fig. 10. Power plant.



Ground Station

System diagram

Fig. 11. Ground transmitter.

Three subsystems make up the ground station: the communications systems (6-kmc transmitter, 4-kmc receiver, horn-reflector antenna), the antenna direction system and the command tracker and telemetry system. To reduce radiation damage the cells are covered with a 30-mil layer of sapphire.

The ground transmitter is shown in Fig. 11. Operating at 6 kmc, it provides an FM signal of 2000 w maximum with a peak-frequency deviation of ±10 mc. The output power can be programmed or varied manually. When the transmitter is used for TV experiments, the center frequency is 6390 mc but it can be shifted ± 5 mc for two-way message experiments. The output of the transmitter is connected, through the microwave diplexer, to the horn-reflector antenna.


System diagram

Fig. 12. Ground receiver.


Figure 12 shows a block diagram of the 4-kmc low-noise communications receiver. The 4170-mc signal arrives at the maser input at a level of —80 to —100 dbm. The signal is amplified by the maser, shifted to a 74-mc intermediate frequency for further amplification, shifted again to 6123 mc, and then fed into the FM feedback receiver and baseband amplifier. The FM feedback receiver reduces the effective noise bandwidth of the receiver to give a 4- to 5-db advantage in breaking margin over a conventional FM receiver. When the antenna is at a 7.5° elevation angle and the 3-db receiver bandwidth is 25 mc, the over-all receiver noise temperature of the system is 42 K.


The horn reflector

Fig. 13. Horn-reflector antenna.


Rectangular horn reflector

Fig. 14. Geometry of rectangular horn reflector antenna.

Figure 13 shows the horn-reflector antenna. It has an aperture of 3600 sq ft, a focal length of 60 ft and is covered by a radome 210 ft in diameter. The focusing surface of the antenna is a section of a paraboloid of revolution with the feed at its focal point. If, instead of being fed by the feed and large conical section, this section of the parabola were fed by a small horn feed at the focal point, the antenna gain would not be changed appreciably. However, the antenna noise temperature would be greatly increased because thermal noise from the earth would enter the horn feed. The horn-reflector antenna with a rectangular aperture is shown in Fig. 14. The horn-reflector antenna is driven from ephemeris on tapes which are corrected by information from the vernier autotrack system. Using previous data furnished by the precision tracker, the ephemeris data are calculated. The angular error is less than 0.05° and smooth tracking is achieved. In this normal mode of operation the precision tracker is not part of the antenna direction system. In another mode of operation it is because the horn-reflector antenna can be slaved to the precision tracker antenna. The large antenna can be driven directly from tapes or from analog data from the vernier autotrack only, to provide other operating modes. Only the 4080-mc microwave beacon signal is received by the precision tracker and the vernier autotrack system. The precision tracker uses a Cassegrainian antenna with a monopulse-type feed. The vernier autotrack has couplers in the throat of the horn reflector antenna which extract incoming energy from both the TE11 and TM01 modes. The magnitude and direction of the pointing error can be obtained by comparing the magnitudes and phases of the signals from these modes. The precision tracker and vernier autotrack are phase locked to the microwave beacon, which has a short-term stability of ±5 parts in 108.

Three functions are performed by the command tracker and telemetry system:

(1) The command tracker acquires and tracks the 136-mc beacon signal from the satellite and gives this information to the precision tracker when needed.

(2) A 120-mc 200 w transmitter is connected to the antenna to send commands to the command receiver in the satellite.

(3) The telemetry information, which is carried by the 136-mc beacon, is received by the system.



Linear Transmission Tests

Three baseband transmission characteristics are measured from the test bay in the Control Building in Andover and looped back either in the upper antenna room (IF) or at the boresight tower (RF), as shown Fig. 15. The IF loops have a number of video amplifiers; however there are no audio diplexers and no 2-mc roll-off filter on the transmitting side. The signal is looped at IF from the output on the FM modulator in the ground transmitter to the input of the receiving IF amplifier. The gain-frequency characteristic with the standard FM receiver in the circuit is shown in Curve 1 of Fig. 15. Transmission is very flat up to at least 5 mc since the IF equipment in the transmitter and receiver adds very little shaping. The curve for the video equipment alone would be very similar to 1. Curve 1 would again not be noticeably altered if the signal were sent to the boresight satellite and back (RF loop). As seen by Curve 2, when the FM with feedback receiver is substituted for the standard receiver, the bandwidth is a little greater than 3 mc.


Graph of gain in decibels

Fig. 15. Baseband transmission characteristics.


The transmission characteristic over the boresight tower, with the 2-mc roll-off filter and the audio diplexers inserted, is shown in Curve 3. This transmission characteristic, which has been used for many of the television demonstrations, represents a preliminary design. By redesigning the diplexers and the roll-off filter a substantial improvement is possible. Both the spectrum of the transmitted signal and the amount of high-frequency energy entering the FM feedback receiver can be limited through use of the roll-off filter.

The quality with which black-and-white television pictures are transmitted over the system is determined to a great extent by the baseband transmission curves of Fig. 15. The reduction of the high-frequency content of the “after” picture is obvious on all three patterns.


Graph of loss in decibels

Fig. 16. Baseband frequency response, Pass 44 (RF Loop with standard FM receiver, no diplexer or roll-off filter).


Figure 16 shows the transmission through the Telstar satellite during pass 44. The curve is similar to that of Curve 1, Fig. 15, the only noticeable difference being a peak of about 0.5 db at 1.8 me. This characteristic, peculiar to this particular model of satellite, is believed to be due to an internal feedback path in the satellite discovered beforehand on the ground. This irregularity causes no noticeable degradation of transmission. Except for introducing additional noise, television pictures can be transmitted over this wideband satellite system without degrading the video quality. Color TV signals were transmitted across the Atlantic using this arrangement and TV pre-emphasis and de-emphasis.


Three graphs

Fig. 17. IF-RF transmission characteristics taken during pass 188, July 30, 1962.


Figure 17 shows the gain-frequency characteristic of the IF and RF equipment, including the Telstar satellite, taken during pass 188. The maser tuning is adjusted before each pass so the transmission curve looks like the one in the center. When the satellite first appeared the transmission was like that shown in curve at left; it gradually changed to the center curve and then to the one at right. Since the earth magnetic field either adds or subtracts from the maser field, the maser tuning shifts at the rate of 2.4 mc/gauss. To transpose the shapes of Fig. 12 into quantitative terms, a two-fold increase of the earth magnetic field in the maser gap has to be assumed. This change results from a change in the orientation of masers in the earth’s field as the antenna moves in azimuth. When superimposed on the maser equalizer characteristic, this shift introduces the transmission slopes shown. These do not affect the system detrimentally.



A Dick Branch presentation


Your comment

Enter it below

A member of the Transdiffusion Broadcasting System
Liverpool, Thursday 18 July 2024