From Long wire to Yagi

From Long wire to Yagi

Dish antennas (V)

Say some words about "satellite" antennas, these dishes, lenses and other horn that are mainly used in microwave bands (>1 GHz). Dedicated to satellite broadband/multimedia applications from L- to Ka-band, amateur communications by satellites or to amateur EME traffic, dish antennas have multiple applications.

Sometimes very impressive when they exceed a few meters in diameter, not every amateur has the opportunity to use or even to see such antennas in the field. So many amateurs ignore how they are designed and what could be their performances. So let's review their specifications in a few lines.

If we take the example of dishes that show the highest gain compared to other designs, they are characterized by several parameters, most of them being shared with HF antennas while others are simply ignored as they produce very little effect or are non relevant :

- Aperture : it represents the size of the area that captures and concentrates the incident radio wave. It surface is pr², its profile is y = ax² with a=1/4f (or 1/4kD with k=f/D).

The larger the dish, the higher the gain and the weakest the energy captured, hence the use of monsters up to 70m in diameter at NASA (DSN) to record signals from spacecrafts.

Note that a simple dipole antenna shows also an aperture. It is of course very small and shows an elliptical shape with an area close to 0.13l, and a little larger for beams.

- Gain : it is the amount of energy radiated by the main lobe compared to the energy radiated by an isotropic antenna. The larger the aperture, the higher gain. Gain can be calculated using the next relation :

with l the working frequency (m), h the antenna efficiency (%), and A the dish area (m²).

For horns and lenses working at 10 GHz, a 10 dB gain is typical. For a dish used at the same frequency the gain can exceed 30 dB.

Note that the gain is expressed in dBic when the circular polarization is concerned (for example with Yagis and helical antennas)

Easier to use, we can also estimate the gain according to the working frequency (expressed in GHz) and the dish diameter (expressed in meters) :

G_dB = 18 + 20 Log (F_GHz x d_m)

For example, tuned on 10 GHz, a dish antenna 2 m in diameter yields a transmission gain of 44 dB, a power ratio >2.10⁴ times.

- Beamwidth : this is the width of the main lobe at - 3 dB. At microwaves frequencies (~10 GHz) to avoid interferences caused by other satellites, is it recommended to use a beamwidth as narrow as 2° in Europe. Knowing the working frequency l, and the dish diameter D, the beamwidth j_o expressed in degrees is given by the next relation :

j_o = 70l/D

For example, at 1 GHz, a "multi-band" dish of 1 meter of diameter shows a beamwidth of 21° but of only 2° at 10 GHz.

We can also calculate the beamwidth of a dish from its gain. If your dish displays a 20 dB gain over an isotropic, this represents a power ratio of 100. Its beamwidth is thus 100 times smaller than the total surface of a sphere, or expressed in degrees :

Beamwidth = Ö (41253 / 100) ~ 20°

This simple relation could also be applied to Yagis but assuming unrealistic designs where all the energy is uniformly distributed in the beam, with no sidelobes. Example for a 6 dBi Yagi (a power ratio of 4 times) that should offer a beamwidth of about 60° :

Beamwidth = Ö (41253 / 4) ~ 100°

In fact the actual beamwidth is 40% smaller because the beam decreases gradually from the peak, rather than abruptly, and we specify a measure at 3 dB. There is a significant energy in the main beam outside the 3-dB points.

To read : The design and building of a large dish antenna rotor, by KC7PPM


A left a Ka-band (30/20 GHz) suitecase used by Canadian scientists of the Satellite Systems Research Group. At center, an amateur gave his 4.2m (14 ft) TVRO dish when Robert Suding, W0LMD, read his advertisment. He converted it in a tri-band antenna. It has about 21 dBi gain on 435 MHz, 32 dBi gain on 1269 MHz, and 37.5 dBi gain on 2401 MHz. Robert rarely run over 5 W on 435 MHz or 1 W on 1269 MHz. The AO-40 middle beacon comes in with a signal strength of S9+20 dB most of the time ! At right G4CCH's 5.4m EME dish fully home-made.

- Sidelobes : there is no antenna that radiates all its power in its main lobe, hence its qualifier of "main" lobe. Indeed, there are always more or less extended secundaries lobes, at rear and on both sides of the antenna. Usually a very directional antenna like a dish can easily display sidelobe 50 dB down from the main lobe. The more extended sidelobes, the less energy will be radiated in the main lobe. In other words the amplitude of these sidelobes affect drastically the antenna efficiency. In this context we observe that a dish that has a beam as narrow as 1° wide, but sidelobes 40 dB down from the main lobe display an efficiency of only 20%.

- Polarization : to get a maximum gain using three-dimensional antennas like dishes and other horns at microwaves wavelenghts, it is essential to consider both polarization planes (E and H). If a dish, horns and lenses utilize both polarizations, the installation is sometimes huge. Therefore in some cheap installations amateur working in EME or by satellites prefer using helical antennas (helix-shape) that take advantage of both E and H polarization planes.

To read : UHF-Satcom


At left, a Quorum S-band tactical dish of 2.3m, 45 kg, offering 10.5 dB gain on 1691 MHz (for weather geostationary satellites reception). At center close up on a L-band RHCP feed/downconverter sold by Quorum and used with a 2.4m dish for NOAA polar orbiting satellite reception. At right F2TU's 8m f/D=0.51 home-made dish for the least impressive for EME operations.

- Phase : unlike HF antennas which radiation patterns show only the mean power strength (amplitude), at very high frequencies (VHF to microwaves), it is mandatory to know the signal phase, its variation as a function of time. If you remember how varies a sine curve vs. time like voltage and currents vs. time, while signals are synchronized, the motion is regular, in phase. But as soon as one signal is delayed or faster than the other, you get out of phase, and you lost energy. This is a similar phenomenon that also happens between voltage and current when you get a high SWR.

But this dephasing is a problem because like optical rays (think to polarizing filters), two radio waves arriving at the dish in phase opposition at the same time will see their peak cancel. In other words, the combination of their voltage will create a deep fading. Conversely, in they arrive in phase, their combination will add together and you will get a high voltage. In between you will get what engineers call an interference pattern; the resulting waveform will depends on the the phase difference between the two signals. We can thus conclude that the phase difference depends on each point of the space, depending on the relative distances and signal amplitudes at each point.

Therefore a well designed feed for a parabolic antenna must use a single phase center so that the radiation is transmitted from a single point source in order to avoid dephasing and signal loss.

To join : Amateur DSN Group (Yahoo!)


From left to right, typical wave reflections on a dish. The feedhorn placed at center (hot spot) doesn't really see a point source what affects the signal of possible interference and thus losses. At center, a typical radiation pattern with its secundaries or sidelobes that induce up to 60% loss over a system showing already an efficiency not better than about 50%. At right, the spillover loss (red) and illumination loss (green) at the edge of the illumination pattern (dashed black sector). In this latter case the f/D = 0.5 and the edge taper (illumination taper) reaches 6 dB. Only a modification of the feedhorn can reduce this loss.

- Illumination : like a bulb placed at the focus of a parabolic dish should only lighten the parabolic surface but neither more or less to be efficient at 100%, and not spread its light in all directions, a dish antenna should work in the same way. The design being theoretically reciprocal, what is good for receive is good for transmit. So the point-like source located at the focal plane (feedhorn) must radiate or "illuminate" uniformely its energy on all the surface without loss or waste. We speak of "spillover loss" or "illumination loss" when the feed illumination does not match exactly the dish shape (typical with feedhorn), and of "edge taper" when there is a energy or power difference between the center and the edge of the dish. A typical energy loss at the edge of the dish is 10 to 15 dB, the highest the loss the lowest efficiency (see below).

- Inverse square law : speaking of signal amplitude, we know that an emitting source radiates its energy in a sphere centered on the source and extending to infinite. Its intensity or amplitude decreases as it moves away according to the inverse square law (1/r²) like gravity and other fields. In other words, when distance double, the receiver experiments (1/2²) or 4 times less power, a 6 dB loss. It is due to the increase of the illumination area by the beam that follows the famous Newton's Law. Consequently, the power density decreases by the same ratio. This loss becomes rapidly consequent mainly in EME traffic where we experiment a loss that exceed 240 dB !

- Path loss : the path loss between the two stations concern mainly operations "in sight" at distances between 10 and a few hundreds kilometers. This attenuation also occurs in loss-free medium. It depends on the distance, working frequency, power, and the gain of the two concerned antennas according to next relation :

The next table displays some typical attenuation levels (expressed in dB) for various distances and frequencies :

Distance (km)	Frequency (MHz)
Distance (km)	145	435	1250	2350	5700	10250	24000	47000
10	95.7	105.2	114.4	119.9	127.6	132.7	140.0	145.9
15	99.2	108.7	117.9	123.4	131.1	136.2	143.6	149.4
25	103.6	113.2	122.3	127.8	135.5	140.6	148.0	153.8
40	107.7	117.2	126.4	131.9	139.6	144.7	152.1	157.9
60	111.2	120.8	129.9	135.4	143.1	148.2	155.6	161.4
100	115.7	125.2	134.4	139.9	147.6	152.7	160.0	165.9
150	119.2	128.7	137.9	143.4	151.1	156.2	163.6	169.4
250	123.6	133.2	142.3	147.8	155.5	160.6	168.0	173.8
400	127.7	137.2	146.4	151.9	159.6	164.7	172.1	177.9
600	131.2	140.8	149.9	155.4	163.1	168.2	175.6	181.4
1000	135.7	145.2	154.4	159.9	167.6	172.7	180.0	185.9

Below 10 GHz the attenuation is mainly due to bad weather conditions but is not very important. The signal strength can be increased by ducting or affected by rain.

- f/D ratio : like any optical system (lens, parabolic mirror, etc), an antenna dish is also defined by its focal length f and the diameter D of the disk. A typical dish shows a f/D ratio between 0.25 and 0.65, what 'd envy all photographer, Hi ! Only drawback, as f/D becomes smaller, the illumination pattern broader and we need to use appropriate feed horns.

If you don't know the focal length of a dish, from its diameter D and the depth in the center of the dish, you can calculate the focal f as follows :

f = D²/16d

- Efficiency : it reveals the real performance of the antenna. It is the ratio of power received over the power arriving. It mainly depends on the working frequency, aperture, gain, illumination, focal length, and directivity. The efficiency is defined by the the k-coefficient . For horns and lenses as well as small size dishes a typical efficiency is 50% or k = 0.5. At JPL the 34m DSS has an efficiency of 74.5% !

An efficiency ranging between 50-75% is roughly equivalent to an edge taper ranging between 15-10 dB.

At last any dish antenna must be assembled : the feedhorn must be attached to the dish, and this latter mounted on a tripod or a pylon. These are two critical mechanical problems to entrust to experimented people, not to forget that you need some free space in your backyard...

Æ (m)	Frequency (MHz)
Æ (m)	435	1250	2350	5700	10250	24000	47000
0.4	-	11.8	17.3	25.0	30.1	37.4	43.3
0.6	6.1	15.3	20.8	28.5	33.6	41.0	46.8
0.8	8.6	17.8	23.3	31.0	36.1	43.5	49.3
1.2	12.2	21.3	26.8	34.5	39.6	47.0	52.8
1.6	14.7	23.8	29.3	37.0	42.1	49.5	55.3
2.4	18.2	27.3	32.8	40.5	45.6	53.0	58.8
3.2	20.7	29.8	35.3	43.0	48.1	55.5	61.3
4.8	24.2	33.4	38.8	46.5	51.6	59.0	64.9
Gain of various dish antennas for an efficient of 55% (k=0.55). You have all chance to confirm an EME QSO when your antenna gain exceeds 20 dBi. For another efficiency use the next relation : k (dBi x 100 / 55). For example, for k=0.55 on 435 MHz, a 0.6m-dish gives 6.1 dBi, for k=0.75 you get 8.3 dBi.

Voilà, this close our review of main antennas that you can use on all amateur bands. Of course tens if not hundreds of other models derivated from these basic designs are available. See next links for additional information, mainly ARRL books about antennas as well as websites of manufacturers.

For more information

Satellite reception (on this site)

Space communications with Mars (on this site)

Wire antennas (incl. for Top band), M0MTJ

How High should my Dipole Antenna be?, QRZNow

Attenuation & Power Handling Calculator (for cabling), Times Microwave systems

The design and building of a large dish antenna rotor, by KC7PPM (on this site)

UHF-Satcom

Amateur DSN Group (Yahoo!)

The World above 1000 MHz, by G3PHO

Paul Wade, W1GHZ (all you need to know about dish antennas)

A 14' Dish for AO-40, by W0LMD