May 1966 Electronics World
Wax nostalgic about and learn from the history of early electronics. See articles
Electronics World, published May 1959
- December 1971. All copyrights hereby acknowledged.
It would be more than a decade
after the publishing of this article before the first direct-to-home satellite television
broadcasts would be a reality, so it shows how long plans were being made for such
systems. Rural landscapes are still peppered with the large vestigial C-band (~4 GHz)
satellite dishes, many with faded eyeballs and other clever (and ugly) artwork on
them. Before coaxial cable was strung beyond suburbs, country dwellers who either
could not pull in over-the-air broadcasts from downtown locations or just wanted
more viewing options paid dearly for satellite service. Equipment and installation
costs on early systems could run into the $30k realm. Today's satellite TV systems
use much smaller antennas operating in the Ku band (~12 GHz), with equipment
and installation being free with a 2-year commitment. C-band DBS (direct broadcast
satellite) systems are still available, BTW. This article is chock full of good
Home TV Via Satellite
Our author began his professional career at the Harvard Underwater
Sound laboratory in 1943. From the end of the war until 1949 he was Assistant Professor
of Engineering Research at the Pennsylvania State Univ. and a Group leader of Electronic
Research at the Ionospheric Research Lab. From 1949 to 1959, he was with RCA engaged
in the development of advanced military systems using color television. Since 1959,
Mr. Underhill has been a Senior Staff Member of Arthur D. little, Inc., an industrial
research company with headquarters in Cambridge, Massachusetts. There he has participated
in and directed studies aimed at assessing the technical-economic constraints on
CATV, pay television, and ETV, and has provided consulting services in management,
marketing, engineering, company policy, and corporate assessment to many companies.
By Bradford B. Underhill / Arthur D. Little, Inc.
Can we use space satellites to broadcast TV directly to individual homes? What
are the problems and can they be solved in the immediate future? Here are authoritative
answers to these and other important questions.
Editor's Note: We recently attended a financial seminar presented by the
National Community Television Association, Inc. in New York City, During the all-day
session over a dozen papers were presented dealing mainly with the financial aspects
of CATV. Included were papers by Milton J. Shapp, president of Jerrold Corp., on
the history, size, and relationship of CATV to the broadcasting industry; by Willard
E. Walbridge, representing the National Association of Broadcasters, on the broadcaster's
viewpoint; by Frederick W. Ford, former chairman of the FCC and now president of
NCTA, on industry relationship with FCC and Congress; by E. William Henry, chairman
of FCC, on the FCC view of CATV; and others. Among the most interesting and important
was the talk on satellite broadcasting by Bradford B. Underhill. A portion of Mr.
Underhill's presentation dealt with the possible use of satellites in connection
with a CATV system. Of more general interest, however, were his remarks on the feasibility
of using satellites to broadcast TV programs directly to our homes. This is a subject
of widespread and increasing interest today, especially in view of the successful
use of Telstar, Syncom, Relay, and Early Bird communications satellites. Because
of its importance to our readers, we have excerpted below a large portion of Mr.
In discussing the impact of space satellites on the future of world-wide television
programming, we will consider the technical-economic feasibility of a satellite
system that could, conceivably, transmit 8 to 12 channels of TV to individual homes.
(Ideally, we would, of course, like to postulate over 100 channels which would allow
a wide choice for the individual television viewer; but a discussion of 8 to 12
channels will, I believe, adequately demonstrate the overall situation.)
Clearly, if a television set coupled with a relatively simple, unobtrusive, and
cheap antenna system at each home could, at anyone time, pick up over 100 different
TV programs ranging from comedy through education, news, serious plays, and do-it-yourself
material, we would have reached a very advanced state of communications. Even the
ability to select 8 to 12 different and interesting programs would provide a new
dimension to television, particularly if the selection were on a national or even
Let's consider the practicality of the idea of using space satellites to broadcast
TV directly to individual homes. Any communications system requires a source or
transmitter and a receiver. In a satellite system, whether located on the ground
or in the air, the satellite performs two functions; that is, it must receive signals
from the primary transmitter or transmitters and then amplify and repeat selected
signals to selected audiences.
Conceptually, a space satellite can distribute signals on a grand scale. Consider
a high-altitude space satellite that hovers over essentially the same spot on earth
24 hours per day, 365 days per year. It could receive signals from anywhere on almost
one-half of the world and repeat this to the same portion of the world. One such
satellite could cover all of North America, South America, and much of Europe and
Africa. Two more could cover 95% of the populated world. Compare this with the 35-
to 40-mile television reception available today. The comparison is staggering.
Such a satellite, operating at an altitude of about 22,300 nautical miles, is
called "synchronous" since it keeps pace with the earth's rotation. Another possibility
imagines nonsynchronous satellites operating at a much lower altitude say, 300
miles - which move with respect to the transmitter and receiving stations. In order
to keep each satellite in view of its neighbors and the ground, at least ten such
satellites would be required. Each would have to be tracked by appropriate ground
stations and each would have to transmit to the ground as well as to its neighbors.
In addition, each receiving station would have to track the satellites as they move
across the zenith and acquire the next one as it becomes visible. We do not believe
that such a system will prove feasible for home use, but concede that COMSAT or
some such organization might be able to develop a base for non-synchronous satellite
global communications with heavy government support.
Assuming you will accept the loosely supported contentions as far as non-synchronous
satellites are concerned, let us examine the case for synchronous satellites. Remember
that this is a satellite that remains over essentially the same spot on earth and
can receive signals from anywhere in its field of view and retransmit to that same
wide area; roughly one-half of the world can be seen from such a satellite. A synchronous
orbit is operationally desirable in order to avoid problems of pointing antennas
toward the satellite and making contact with (acquiring) the satellite as it rises
over the horizon.
The satellite must contain a source of electrical power. At the present time,
photovoltaic cells which draw power from sunlight are the commonest source of satellite
power. Conventional batteries are nearly useless for long-time activity. In the
future, nuclear devices capable of generating a substantial amount of electric power
for long periods of time may be generally available for communications satellites.
Until nuclear power supplies are available, transmitted power from a satellite will
be severely limited, for a solar-battery power supply weighs about one pound per
watt of continuous available power.
The satellite platform will contain the receiving antenna and amplifier, modulator,
a rebroadcasting transmitter, and a transmitting antenna. As these antennas must
have high gain and correspondingly narrow beams, the attitude
control of the platform must be sufficiently precise to aim the antennas within
appropriate tolerances. This problem is not insurmountable, but every sharpening
of the attitude tolerances increases the complexity of the attitude sensor and control
system and adds to the amount of energy which must be available for attitude control.
The receiving ground complex consists of receiving antennas and receivers, that
is, an amplifier and demodulator to convert the signal back to the preferred usable
form. With the satellite in a practically stationary orbit, a complex control system
is not needed to aim the ground antenna.
where: dB = 10 log10 P1/P2 or 2x power = 3 dB; 10X power = 10 dB;
100x power = 20 dB.
Down-Link Power Computations:
At first sight, the transmission properties of the link from the ground to the
satellite look very much like those of the link from the satellite to the ground.
The principal asymmetry is that in the up-link the transmitter is on the ground
and the receiver is in orbit, whereas in the down-link the transmitter is in orbit
and the receiver is on the ground. The state of our technology is such that we can
install more powerful elements on the ground than we can in orbit. Practically speaking,
this usually makes up-link performance easier to achieve than down-link performance,
for a ground transmitter can be vastly more powerful than an orbiting transmitter,
whereas a ground receiver can only be a little more sensitive than an orbiting receiver.
Consequently, we shall examine the power budget for the down-link only.
Frequency Assumptions, We will confine our discussion primarily to the existing
television broadcast frequency bands at 54 to 88, 174 to 216, and 470 to 890 MHz.
For specific numerical examples, we chose frequencies of 60, 200, and 600 MHz as
representative of these three bands. Still higher frequencies, i.e., 2000-7000 MHz
might be used but these would require expensive conversion equipment at each home.
Radiated Power Assumptions:
Where there is dependence on solar power, it is likely that the maximum continuous
radiated power will not exceed a few watts, perhaps 10 watts at most. When nuclear
or other power supplies are available, the primary power limitation is removed and
the significant limit will arise from constraints on dissipation and life-time in
the power output stages of the transmitter. We believe 100 watts radiated power
will represent a likely maximum for the next decade. Under extreme conditions, power
of 1000 watts may be attained. However, even with ground-based equipment, current
technology does not attempt to achieve long life-time and high reliability in electronic
circuit elements operating at power levels of even one kilowatt; reliability relies
on replacement. The ordinary techniques of enhancing reliability by overdesign and
derating are ineffective where high power throughout is required. These suggested
limits are consistent with the guidelines cited by G.M. Northrup in a paper prepared
for the Rand Corporation in 1963. Accordingly, I shall assume a total radiated power
from the satellite of 100 watts except where otherwise stated.
Table 1 - Diameter of various high-gain transmitting antennas.
Table 2 - Effective capture area of omnidirectional antenna.
Satellite Transmitting Antenna Directivity:
The effective radiated power from the satellites can be enhanced 80 times (19
dB) with a directional antenna whose beam just illuminates the near-hemisphere visible
from the satellite. An antenna 10 times as large will illuminate, in the temperate
zone, an oval about 2500 x 4000 nautical miles, and will have a gain of 800 (29
dB). An antenna 10 times larger yet will have a gain of 8000 (39 dB) and illuminate
an oval about 800 x 1200 nautical miles. At the frequencies under consideration,
however, these antennas are quite large, as shown in Table 1.
Although very large antennas have been considered, we have as yet no experience
in the erection and operational use of antennas even as large as 60 feet, to say
nothing of antennas 10 times this diameter. Furthermore, a 29-dB antenna has a beamwidth
of about 6°, and a 39-dB antenna has a beam width of less than 2°. The pointing
of such a sharp beam adds a burden to the attitude control system of the platform.
For the moment, let us assume the 19-dB antenna on 1 satellite and defer consideration
of higher gain antennas until later.
Because of the way certain standards of television signal strength have been
developed, comparisons are facilitated by computing signal strength at the earth's
surface directly in terms of received power density per unit area for one channel.
Let us assume that the 100 watts which is radiated through a 19-dB gain antenna
is equally divided among 12 channels, each having a bandwidth of 6 MHz. The slant
range from the satellite to the temperate zones is approximately 45,000 kilometers.
On the ground the power density is: PD = 2.6 x 10-14 watt per square
meter or 26/1000th of a micromicrowatt/meter2.
Receiver Antenna Directivity: The effective capture area of an omnidirectional
antenna is shown in Table 2 for the various frequencies under consideration. A half-wave
dipole antenna has 1.6 times (2 dB) more directivity, provided it is appropriately
aimed. Combinations of reflectors and multiple dipoles can give gains of several
decibels, but more substantial gains require structures whose size is comparable
to those shown in Table 1.
An antenna with more than a few dB gain is likely to be large and expensive.
Such antennas will probably be ruled out in applications where one would be required
for each home television user. For the sake of comparison, further computations
are based on the use of a non-directional receiving antenna. This factor can easily
be modified by assuming some other antenna gain figures.
From: Television Engineering Handbook," Donald G. Fink, ed.; McGraw-Hill, N.Y.
1957, pp. 2-35.
Table 3 - Grades of television service and the FCC regulations
Table 4 - Equivalent-noise assumptions for various sources.
Table 5 - Deficit received power from a satellite system. Received
power density in dB relative to various service grades.
Noise assumptions: We have two bases for estimating noise limitation. On the
one hand, we can use the grades of television service from the FCC regulations,
as quoted in Table 3.
For convenience, the power density in watts per square meter has been computed
and exhibited alongside of the field strength. On the other hand, we can base signal-to-noise
relationships on thermal noise and an assumption about the noise temperatures of
the receiver and the associated environment. A set of such assumptions is shown
in Table 4.
In order for the television receiver to function properly, the signal strength
must exceed the noise by a substantial margin. The exact amount depends on the grade
of service required and on the character of the noise. It will be assumed that a
signal-to-noise ratio of 26 dB is required for satisfactory service and that 5 dB
additional margin is required to allow for system degradation other than that arising
in the down-link from the satellite. If the noise is white thermal noise this will
result in a high-quality black-and-white television picture but one that is not
totally free from visible noise effects. The sensitivity of color television to
noise interference is somewhat greater.
Resulting Margins: Table 5 compares the received signal strength, as computed
above, with the FCC standards and with the performance of a low-noise receiver with
a 100°K effective noise temperature connected through a non-directional antenna.
The resulting figures are expressed in decibels and are all negative, showing the
deficit in dB between the signal as it would be received from the satellite and
the desired grade of service indicated at the head of the column.
It is easy to see that the FCC Grade B Service is closely parallel in signal
strength requirements to that required by a low-noise 1000°K receiver with an
isotropic antenna. Those of you who have ever tried to operate a television receiver
in an area with Grade B Service, so-called fringe-area reception know that with
off-the-shelf home television receivers, the performance is very unsatisfactory
without a directional antenna.
In one way or another, the deficit of 33 to 70 dB must be made good. If a 1000-watt
transmitter can be used aboard the satellite 10 dB can be gained at once.
The transmitting antenna gain may possibly be increased 20 dB at 600 MHz and
10 dB at 200 MHz. At 60 MHz it is hard to see how this could be increased at all.
A receiving antenna six feet in diameter will have a gain of 20 dB at 600 MHz
and 10 dB at 200 MHz. At 60 MHz it is fruitless to attempt antenna directivity because
galactic noise at 60 MHz is greatly in excess of 1000°K noise assumed for the
receiver. The highly directive antenna would focus on sources of high galactic noise
periodically during the earth's rotation and nullify the desired improvement.
The receiver noise figure could be reduced somewhat provided a high-gain, low-side-lobe
receiving antenna were used. This leads to a complex and expensive system, and would
probably be useful only at the 600-MHz frequency; for even at 200 MHz, galactic
noise is a significant source of interference. Let us assume a possible 5-dB improvement
at 600 MHz. Table 6 shows the total improvement achievable by these means. It is
clear that success is possible only at 600 MHz and that the margin is slim. Even
this success is bought at the cost of a 1000-watt transmitter, a 60-ft diameter
antenna on the satellite, a 6-ft diameter receiving antenna with low side lobes,
and a receiving system operating at an effective noise temperature of 300°K
on the ground. We believe that such a low-noise receiver antenna system would add
a minimum of $200 to the homeowner's cost. Even if we are overly conservative, a
home conversion cost of at least $6 billion is indicated to convert current TV households
to satellite receivers. Without further analysis, it appears that this is hopelessly
expensive for the ultimate users of such television signals.
Table 6 - Additional system improvements that are available.
Sources: "A Study of the Sources of Noise in Centimeter-Wave Antennas," D. C.
Hogg, 1961. "Aids for the Gross Design of Satellite Communication Systems," G. M.
Northrop, Rand Corporation.
Synchronous-altitude satellite radiating 100 watts divided equally among 12 channels
through a 19-dB gain (full-earth coverage) antenna. (assumes 2.6 x 10-14watt/m2
available; Table 3 power required, and Table 2 antenna gain) *19-dB transmitting
antenna -26 dB S/N Into 1000°K receiver, 5-dB margin.
Although the technical problem is made easier by going to frequencies above 2000
MHz, this would add the burden of conversion cost to the individual home owner and
would still provide only marginal signal strength unless only a small portion of
the earth were covered by the satellite.
The 9-dB deficit at 600 MHz could be made up by reducing the number of channels
transmitted to only one or by increasing the power to 10,000 watts. The first is
not attractive and the second is beyond our capability at this time.
There is no doubt that television transmission via satellite relay is technically
feasible with a larger permanent ground antenna and receiver installation, but such
installations cost many millions of dollars to build and hundreds of thousands of
dollars per year to staff and maintain.
Such installations would allow economic world-wide transmission and reception
of a single TV channel and several radio channels and would be attractive to the
broadcasters. Retransmission to homes would be achieved by relaying the received
information to the individual homes via cable or local transmitters, Certainly governments
might find such a system valuable for transporting information to other world centers
or to developing nations.
To sum up, we at Arthur D. Little, Inc. believe that:
• Non-synchronous satellites would require trackers at each home. Large-scale
CATV operators could conceivably develop a receiving system to track multiple satellites,
but this would be an expensive undertaking.
• Synchronous satellites broadcasting twelve channels to the home are not
technically feasible below 200 MHz and are only marginally feasible at 600 MHz.
Here expensive receiving equipment is required.
• At frequencies between 2000 and 7000 MHz, trunking is possible but this
would require expensive ground equipment that only the most affluent CATV systems
• A feasible system would require at least a 20,000-watt continuous power
source (more probably 30,000 to 40,000 watts of prime power would be required) and
the development of high-power transmitting equipment that could operate reliably
for many years. Neither development appears possible for many, many years.
Posted May 18, 2020
(updated from original post on 12/4/2014)