Both antennas are steerable in altitude and azimuth and have detected the sun as well as the more energetic radio sources in Sagittarius, Cygnus, Cassiopeia and Orion. The resolving power of McFarland's telescope is about 11 degrees of arc; it detects the sun as being a disk some 20 times wider than it appears to the eye. The resolving power of the Kansas telescope is about 17 degrees. In contrast, the 250-foot reflector of the radio telescope at Jodrell Bank in England resolves the sun as an object about twice the diameter of the optical disk. Toy spyglasses can disclose much more detail. But resolving power is only one measure of a telescope's performance. Another is the instrument's ability to detect distant objects. The clouds of interstellar dust that block many regions of the universe from view are transparent to some bands of the radio spectrum. The amateurs who built the Winston-Salem and Kansas telescopes set out to have a firsthand "look" at whatever lies beyond the dust, even if the view turned out to be fuzzy.

"I began to work on my telescope," writes McFarland, "during my third year in college, partly as a project for a thesis, and I hoped to finish it before graduation. But a number of bugs developed, and it was not ready for a trial run until the summer following graduation. The telescope has four major components: the antenna and its mount, a high-gain, low-noise receiver, an automatic pen recorder and a noise generator that is used to test the system and as a standard for comparing the strength of radio sources in space. The design and procurement phase of the project took 18 months of spare time and the construction about a year.

"Much of the initial planning went into the antenna. The antenna of a radio telescope corresponds to the objective mirror or lens of an optical telescope, and the performance of the completed instrument depends on it just as critically. In selecting a design for the antenna several configurations of the diffraction type were considered, including a broadside array of half-wave dipoles and an array of helices. These were dismissed in favor of a paraboloid because the complexity of interconnecting a broadside array increases in proportion to the number of dipoles, and the length of each dipole must be changed for each frequency on which the telescope operates. Moreover, I wanted an antenna that would pick up the broadest possible band of frequencies and discriminate strongly against all signals except those that come from a desired direction. A paraboloid best meets these requirements.

"Winston-Salem is a center of intense, man-made electrical disturbance, chiefly from sources such as automobile ignition systems, power lines and harmonic radiation from radio and television stations. By scanning the radio spectrum from 50 to 3,000 megacycles with a short-wave receiver, I spotted a relatively quiet region of the spectrum in the vicinity of 400 megacycles (a wavelength of 75 centimeters, or 29.5 inches). At this frequency a signal equal to a millionth of a billionth of a watt (10-16 watt) would override the noise if the antenna were designed for maximum power gain; that is, if it strongly favored signals arriving parallel to the axis of the parabola. The power gain of a paraboloidal antenna (with respect to the response of a nondirectional antenna) varies directly with the radius of the parabola and inversely with the wavelength of the signal, as shown by the accompanying equation [see Figure 4]. When the focal length of the parabola is equal to half the radius, the maximum power gain in decibels is equal to 10 times the logarithm (to the base 10) of the square of this ratio: 3.14 times the radius divided by the wavelength. At wavelength of 27.5 inches (a frequency of approximately 436 megacycles) and a gain of 20 decibels, this formula yields a radius of 7.5 feet. With this dimension known, the distance from the focus to the vertex of a paraboloid can be calculated. In the case of my antenna it amounts to 3.75 feet. The resolving power of telescope objectives, whether optical or radio, increases in proportion to the diameter of the lens or reflector, and decreases with wavelength as indicated by the second formula [below]. A 15-foot paraboloid operating at 436 megacycles has a resolving power of 10 degrees 54 minutes, which is about 20 times greater than the apparent angle subtended by the sun.

"With the size of the antenna determined, its physical structure was considered next. Aluminum was selected as the most attractive material, from the point of view of both weight and cost. A disadvantage in using aluminum is that all parts of the antenna must be welded. Otherwise voltage may develop across the high-resistance joints between adjacent parts and be detected as noise. The welding can be done most satisfactorily by an electric arc that operates in an atmosphere of helium gas. This is inconvenient but not expensive.

"The paraboloid was formed of sheet aluminum welded to a paraboloidal skeleton of aluminum tubing-a series of concentric rings supported by radial ribs bent so the sheet took the desired shape to within 1/8-inch [see Figure 1]. To build the skeleton I formed nine circles of 3/4-inch tubing, with radii ranging from seven feet nine inches to six inches, on a machine similar to those used for bending model railroad tracks. The rings were nested against 24 radial ribs formed to the approximate final shape on the same bending machine. Many hours were then spent in hand-forming each rib to within 1/64 inch of a master parabolic template cut from sheet aluminum. A jig clamped the parts during the welding operation.

"Aluminum screening would doubtless have been a better choice from the point of view of wind resistance for covering the skeleton. But the only available material of this sort was ordinary house screening, which is much too light to hold its shape or to weld to the skeleton. The completed structure weighs approximately 260 pounds.

"Incoming signals are focused on a dipole antenna that is supported on the axis of the paraboloid by a short length of coaxial transmission line made of two aluminum pipes [see Figure 2]. The electrical impedance of coaxial lines is equal to the logarithm (base 10) of the quotient of 138 times the inside diameter of the outer pipe (in inches) divided by the outside diameter. of the inner pipe. The inside diameter of my outer pipe is .66 inch and the outside diameter of the inner one is .27 inch. The impedance of my coaxial line is therefore 53 ohms. For maximum transmission efficiency the impedance of the coaxial line and that of its associated dipole antenna must match. The characteristic impedance of a dipole antenna 51 in free space is 72 ohms, but this value is lowered by the presence of a nearby conductor, such as a metal plate. I found that I could match the impedance of my dipole to that of the coaxial line by placing an aluminum disk (16 inches in diameter) a quarter of a wavelength in front of the dipole. Later I found that the disk also helped to shield the dipole from off-axis signals and therefore improved the directivity of the antenna.

"The coaxial line is approximately five feet long and is fastened to the paraboloid through an aluminum plate welded to the skeleton. The outer end is terminated in threaded fittings that take a pair of threaded aluminum rods, each a quarter of a wavelength long, which function as the dipole. The outer pipe of the coaxial line extends approximately two feet beyond the dipole and serves as a mounting for the aluminum disk. The inner end of the coaxial line is equipped with a threaded coaxial coupling for attaching the antenna to the receiver through flexible coaxial cable.

"Sky noise from the cable is fed at 436 megacycles to a parametric amplifier [see "Junction-Diode Amplifiers," by Arthur Uhlir, Jr.; SCIENTIFIC AMERICAN, June, 1959] and gains some 20 decibels in power. The signal is then passed to a conventional converter that divides the frequency to 30 megacycles. After additional amplification the signal is converted to pulsating direct current to run a pen recorder. All the apparatus is powered from closely regulated power supplies, as shown in the accompanying diagram [Figure 3].

"During the initial tests and sightings the paraboloid was mounted on a meridian transit, but as soon as time permits it will be installed on a surplus 36-inch searchlight mount equipped with a motor drive for remote control.

"The parametric amplifier turned out to have substantially more gain than expected, so I was able to use a long section of coaxial cable between it and the converter without introducing excessive noise. The gain was so great, in fact, that I could even install a six-decibel attenuator in the cable to keep the ignition noise of passing cars from driving the converter into overload.

"There has not been time to use the telescope extensively since it has been completed. But test runs prove that its response is satisfactory, considering the comparatively low resolution of the system, as shown by the accompanying graph [Figure 7 ]. One sad incident is worthy of mention because it disclosed that parabolic antennas of sheet aluminum must be painted flat black. While I was observing the sun during the first trial run the 16-inch aluminum disk at the focus of the paraboloid suddenly melted!"

The antenna of the Kansas telescope, according to Houston, Simpson and Mullinix, was constructed primarily for tracking satellites as part of the Moonwatch program during the International Geophysical Year. It consists of a 35-foot spar of pipe that supports a single reflector and a dipole antenna at the back and a series of dipole directors in front. The spar is supported by two braces of pipe and carries an altitude circle [see illustration at right]. The directors and reflector are merely straight lengths of aluminum wire 38 inch thick. The wires stand up well under the Kansas winds, according to Houston, but birds can bend them. The whole business is mounted on an 18-foot telephone pole so that the reflector just clears the ground when the antenna points to the zenith.

"For a given power gain," Houston writes, "Yagi antennas can be built that are lighter and more compact than any other type and that have less wind resistance. These advantages are bought at the cost of a narrow band width. Yagi antennas can be designed for optimum operation at only one frequency-a real disadvantage when they are used for measuring star noise.

"When in operation, the antenna is fixed at a selected altitude on the meridian for 24 hours and picks up noise in varying amounts as the sky drifts past. The sky noise is amplified and converted to pulsating direct current for operating a pen recorder. Normally the antenna is pointed just above the horizon during the first run of a series and then is raised a few degrees higher for each subsequent run until the entire sky has been scanned from the horizon to the zenith. If the recorded traces turn out to be good, corrections are made for instrumental errors and the results are read off and plotted.

"The readings are somewhat fictitious, of course. Just as an optical telescope shows spurious disks around the stars, so do radio telescopes. By running the sun across the antenna we found that the disk of our instrument is about 17 angular degrees in a horizontal direction. We did not measure its vertical width but it should be about the same. 'Radio stars' and other discrete sources of radio signals are below the resolution of the system and are lost in the general background noise.

"Our telescope consists of the antenna, a preamplifier, a frequency converter, a wide-band intermediate amplifier, a detector and a recorder [see below]. The sensitivity of the system is limited by the noise generated in the vacuum tube of the preamplifier. (Disturbances from man-made sources are not serious in our locality.) Of the several tubes that we tested, the Western Electric 416-B generated the least internal noise. We used it in a conventional grounded-grid preamplifier circuit of the type described in The Radio Amateur's Handbook.

"To our dismay, however, we found that the 416-B is quite sensitive to changes in temperature, and we could not regulate our room temperature closely enough to overcome the difficulty. So we devised a simple indicator for measuring variations in the tube's performance. The solder-tipped hand of an alarm clock was rigged so that once an hour it closes a contact that actuates a relay. The relay transfers the input of the preamplifier temporarily from the antenna to the terminals of a 72-ohm resistor. The set then measures and records the noise developed in the resistor. The output of the resistor remains relatively constant during slight changes in room temperature. Hence the record of the resistor's amplified signal is a measure of the performance of the 416-B changes with temperature, and recordings of sky noise can be adjusted to correct for it. The value of the resistor was selected experimentally so that its amplified noise would be lower than recordings of sky noise.

"The output of the frequency converter is fed directly to the first intermediate amplifier stage of an old television receiver. The band width of a correctly aligned television receiver is about five or six megacycles, whereas that of our Yagi antenna is on the order of two to four megacycles. Reducing the band width of a radio telescope is comparable to filtering the reds and blues from the ends of the optical spectrum: the energy of the signal is reduced, and in the case of radio telescopes the maximum amount of sky noise is wanted. The weak link in our telescope, however, is the narrow band width of the antenna, so the television amplifier is adequate.

"The output of the television amplifier is rectified by a 1N64 diode and, after passing through a resistor and into a capacitor, it is fed to the pen recorder. The time required for the capacitor to charge is about a tenth of a second. This delay tends to smooth the recorded graph because the pen does not respond to current pulses of less than a tenth of a second, such as bursts of lightning.

"Our equipment-most of it salvaged from the scrap box-can only be characterized as crude. It operates on 108 megacycles, far below the 1,421-megacycle hydrogen line that is so widely observed by radio astronomers. But, like most amateurs, we have a fondness for making our initial forays with the help of salvaged junk and in areas that are neglected by the hunters of bigger game.

"We set out to make a radio map of the sky-and we made one that agrees broadly with those compiled by the best radio telescopes, considering differences in frequency and antenna resolution. In plotting the map we made no effort to 'compute away' the influences of discrete radio sources in space. Although these powerful cosmic radiators do not show up as spots on our map, they do tend, like an overexposed star on a photographic plate, to distort our contour lines. This is particularly apparent in the lines that dip to the south from Cassiopeia. Even more striking is the influence of the two radio stars in Cygnus near 40 degrees declination and 20 hours right ascension. Another major difference is the hourglass configuration in Orion. One map made by a large radio telescope at 250 megacycles shows two local areas in Orion, one at the Equator around the 'belt' and one 10 degrees north and somewhat toward the east. The pinch in the hourglass of this map falls at about 10 degrees south declination, whereas ours lies directly on the Equator. This difference worried us, and we made a dozen extra runs in the region of Orion, even making runs in declination by hand. But our results were always the same. We now suspect that the appearance of this region of the sky may be different at 108, megacycles from its appearance at 250 megacycles.

"In spite of the fact that we ran traces at each declination for several days, and in some cases for many days, we never succeeded in recording a full 24-hour run at any one time. The sun was usually evident, and if it wasn't, we did not trust that portion of the curve. To map such regions of the sky we simply had to wait until the sun moved on. On other occasions the electronic equipment went psychopathic and awed us by recording giant forces that never reappeared; now and then thunderstorms ran the pen out of ink. In general we got one good trace out of 10 tries. But it was fun."

Bibliography

THE RADIO AMATEURS HANDBOOK. The American Radio Relay League, 1961.

RADIO ASTRONOMY. Rote Greber in Scientific American, Vol. 181, No. 3, pages 34-41; September, 1949.

THE SCIENTIFIC AMERICAN BOOK OF PROJECTS FOR THE AMATEUR SCIENTIST. C. L. Stong. Simon and Schuster, Inc., 1960.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.