Amsat-NA Logo

Amsat-UK's Oscar News, 1993 Apr No.100 p6-109
The Amsat Journal (USA) Vol 16 No. 2, March/April 1993.
Amsat-DL Journal (D), Jg. 20, No. 2, Jun/Aug 1993.
Amsat-VK Newsletter No. 97, 1993 April

A 60cm S-Band Dish Antenna

by

James Miller G3RUH


Introduction

My recent article [1] proposing the abandonment of the 145 MHz band for weak signal amateur satellite downlinks has sparked off much interest. I explained why S-band (2.4 GHz) is unrivalled on every count as its replacement, and described in outline a tiny 60 cm diameter dish that is entirely adequate for Oscar-13. Naturally I have had many requests for more details.

Let me stress once again, you need absolutely no knowledge about S-band to get going, anymore than you need knowledge about 2 metres to get onto mode-B. Just accept my word that S-band is easier by far. I started this project with absolutely no understanding of S-band, designed and built it all in an afternoon's pleasant metal bashing.

Take a look at your 2 metre antenna. It's just a load of mechanical bits and pieces juxtaposed to behave electrically in a special way. So too is an S-band dish antenna.

Don't Panic!

There are three parts to this dish. The parabolic reflector, the boom and the feed. This article is about each of these, and how to fix the reflector and feed to the boom in the right places. Easy. See fig. 1.

 S-band Dish

There are as many ways to do this as there are constructors. I'll tell you my way. You don't have slavishly to replicate every nuance of the design. No dimension is critical except for a small bit of the feed system. No exotic tools or test equipment are needed. You have all you need already; hands, hacksaw, file, ruler. Throw it together, and it will work. I know; I did it. At the end of your toils you will have a 60 cm diameter S-band dish antenna with a gain of about 20 dBi RHCP, a 3dB beamwidth of 18° and, with the suggested converter, more than adequate performance to experiment on AO-13 mode-S.

The Parabolic Reflector

I went into a shop to look for some furniture; the lobby looked perfectly normal. The stairs were closed so I took the lift. I emerged into a wasteland. The shop was in the throes of closing down for good. In the rubbish I spotted a couple of once trendy, spun aluminium parabolic lampshades in a vile buff colour, and bought them for as little as I could get away with. "Might come in handy one day".

The dish measured 585 mm diameter, and 110 mm deep, which translates into an f/d ratio of 585/110/16 = 0.33, and thus a focal length of 0.33 x 585 = 194 mm. An f/d of 0.33 is rather too concave for a simple feed to illuminate ideally, but at $4/dish who cares? It works, and final antenna is much more compact than the usual shallow dish of f/d=0.6

The dish has a 40 mm hole in the middle with three 4 mm convenient holes around it on a 1 inch radius circle, presumably a standard for lighting attachments. So absolutely no modifications were needed to the parabolic reflector.

The Boom

This is made from a length of 12.5 mm square aluminium tube obtainable from any non-ferrous metal supplier or engineering shop. It has to be long enough to stick out the rear of the dish to provide an extension to clamp to the rotator boom (150 mm), long enough to put the feed at the focus (+194mm), and long enough to fix the feed and converter assembly (+186mm). So it is about 530 mm long overall. The clamp is a cheap UHF TV antenna type.

Parabola to Boom Fix

I hacked out a disc of 3 mm thick aluminium, 65 mm diameter and drilled three 4.3 mm fixing holes using the dish as its own template. Then I filed a 13 mm sq square hole in the middle for the boom to pass through. The disc is attached to the boom by four little 15 x 15 x 10 mm wide pre- drilled angle brackets. These items are usually used to mount small boxes to panels etc. You can bend and drill your own, but I used RS Components (UK) part 544-213.

Thus there are four more 3.3 mm holes around the centre of the support disc, and two 3.3 mm holes in the boom. Obviously two angle brackets face forwards, and two face rearwards, otherwise the through-boom screws will foul each other.

The Feed Assembly

This comprises several components. The helix feed, match and N-type panel connector, feed reflector, and S-band converter.

Converter

As Ed Krome points out in his excellent survey of S-band [2], there is no such thing as an ICOM/KENWOOD/YAESU FT24000 transceiver". He might have added, nor is there ever likely to be. This is not a marketing problem, it's an electrical one. Any antenna necessarily has to be some distance away from the radio. But coax less than 1" hardline attenuates 2.4 GHz signals like crazy. This wipes out the advantage of S-band, which is the almost total absence of ambient noise.

So there has to be some receive electronics at the feed, such as a receive pre-amplifier or a converter, 2.4 GHz to 144 MHz for example. This puts the low noise receiver and its gain right at the masthead, and the downlead can be inexpensive lightweight RG58/u coax.

I used a model UEK-13-P3C from SSB Electronic. Note the suffix -P3C. This denotes 2400-2402 MHz, as distinct from the 2304 MHz version. [This has now been superceded by the model UEK-2000SAT ].

There are also converters available from Down East Microwave, and a number of specialist designers "advertise" regularly in magazines like VHF Communications and Dubus. See also VHF/UHF/Microwave handbooks (ARRL and RSGB) and the UHF Compendium from Germany.

An S-band helix has a radius of approximately 21.25 mm so, if fed at its periphery, the connector would be 21.25 mm from boom axis, or about (21.25- 6.25) = 15 mm from the boom edge. This time Murphy was on my side. It just so happens that the corresponding connector on the converter is 15 mm from its edge too. So if the converter is clamped directly to the boom, the two connectors line up perfectly. Since an N-type connector has the mating strength of two elephants, it can also support the entire feed assembly! Thus the helix, connector and converter form one sub-assembly which is then clamped to the boom with two large cable wraps. This arrangement also allows it to be slid up and down to peak up the receive signal.

Of course if you use some other converter, the reflector assembly can simply be fixed with two angle brackets in a similar way to the dish attachment.

Reflector

I made this from a 125 mm square of 1.6 mm thick aluminium. It looks prettier if the corners are cropped off 10 mm but this is not essential. The reflector could also be 125 mm circular. It's just easier to cut straight lines than round ones. See fig. 2.

There is a 13 mm square hole in the middle once more, and the N-type panel connector is affixed approximately 21.25 mm from the middle as described above. Get the exact dimension needed by carefully measuring the the converter. The N-type panel connector is actually mounted on a 1.6 mm thick spacer so that the back of the connector is flush on the helix side. Drill and fix with three screws only, as the fourth will foul the start of the helix.

 S-band Dish Feed

Helix

This is 3.3 mm diameter copper wire. In fact it's the solid inner stripped out of a length of H100 coax (similar to Belden 9913). You need about 1/2 metre. Wind it all, about four turns, around a 40mm diameter former. I found a socket in my car tool kit that was perfect. It will spring out slightly after winding. Wind it left handed, that is to say in the opposite sense to a conventional screw thread. As you look at it from one end, it will recede anti-clockwise.

Now carefully stretch the helix to a turn spacing of 28 mm ( +/-1 mm). Make sure it is nice and even.

Cut off and discard the first 1/2 turn. Now, with careful use of pliers bend the first 1/4 turn about 10° so that it is almost parallel to the reflector again. This will form part of the matching section.

Next you need a very thin brass strip, about 0.2 mm thick, 6 mm wide, shaped to follow the curve of the 1/4 turn you just bent. It will go flat between that turn and the reflector. Use the 1/4 turn as a template onto a piece of paper.

Since the copper is 3.3 mm wide, and the strip needs to be 6 mm wide, a line 1.4 mm away can be drawn both sides of the copper with a fat pencil. Judicious use of clamps, weights, blu-tack and adhesive tape will provide the six hands needed to do this! Spend plenty of time to get this drawing just right. Then sellotape the paper to the brass, and with great care use a pair of very sharp scissors to cut out the shim you have sketched.

With more fiddling as above, trap the shim between copper and a flat heat- proof surface, and with a 60 watt iron and large tip, solder the shim to the copper all the way along its length. If you make a mess, dismantle, wipe clean and start again.

You should now have a helix which has a brass shim soldered to its first 1/4 turn essentially flat to the reflector. Cut off the surplus turn(s) from the helix so that you have a total of 2 1/4 turns including the matching section.

Finally solder the helix to the pip of the N-type. The spacing of the matching section should be 1.2 mm at the start by the N-type, and 3.0 mm at the end. Use scrap bits of metal sheet to jig this up, carefully bending here and there to get the whole edifice just so, match spaced properly, helix axis perpendicular to the reflector. This also needs six hands.

Finally apply a very generous blob of solder to the end of the helix and N-type. This will be the only fastening for the helix. It may have a propensity to twist. Sealant will cure this later.

Don't get too paranoid about precision. I guessed the prototype's matching, and it worked great. Then I took it to a friend's 3 GHz network analyser. It showed a lousy match. So I designed the new perfect one as above. On-air performance was unchanged!

Final Assembly

You can now couple the converter to the connector. As I used a male panel connector, it coupled directly to the female on the converter. However male panel connectors are very difficult to find, so you will probably have to use a female panel connector, plus a male-male coupler.

Slip the assembly over the boom, and slide it to the correct position. The middle of the helix should be at the dish geometric focus, 194 mm from the parabola on my own model.

Attach a BNC connector to a suitable length of RG174/u coax, and pass it down through the boom, together with the +12v wire, to emerge at the rear and link up with your system. The coax outer provides a return, so only one power wire is needed.

Testing

You need a 2.4 GHz signal generator, an RF detector and a power meter.

What? You don't have any of these? Wrong! The Sun is your signal generator, the 144 MHz SSB radio is your detector, and an ordinary moving coil AC meter is the power meter.

Indoors, connect everything together and switch on. If you now switch off and on power to the converter, you should hear a substantial change in noise output from the radio in SSB mode. The converter is "converting" something at least, so the wiring is probably sound.

Now set up shop outdoors. Choose a location clear of obstructions. BEAR IN MIND THE POTENTIAL DANGER OF USING MAINS POWERED EQUIPMENT IN DAMP CONDITIONS.

Power up the system once more. Holding the boom by the rear and above your head, point the antenna towards the zenith, that is straight up. This will be the quiet sky. Listen to the noise level and make a mental note of the volume. Headphones are useful here. Now point the antenna at the ground. The noise should increase substantially, approximately doubling in loudness (+3 dB).

Now point the dish at the Sun. If it is shining, this is easy. If it is obscured by cloud, you will have to use your skill and judgment. Either way, the noise level should rise by about 1 dB compared with the quiet sky. This is only very slight, and is why a meter is helpful. Note: if the Sun is less than 10 degrees above the horizon, your antenna will "hear" the gound and this will swamp the Sun noise.

Connect the moving coil AC voltmeter to the loudspeaker terminals. Set the voltmeter onto its most sensitive range, probably 3 volts full scale, and adjust the volume control to about half travel. Note the reading on the meter. Now aim at the Sun, and the voltage reading should increase by about 10%.

This method of measuring noise power is quite accurate provided two conditions are met. First, the audio amplifier must not be limiting through overdriving. You can hear this as a bad crackling noise, or look on a scope. Secondly, the receiver AGC must not be operating. If the S- meter is showing a deflection, the AGC almost certainly is in action, and the AC output voltage will not be proportional to signal strength. You will need an attenuator to reduce the 144 MHz signal. A couple of junk box resistors will do. 1/4 of the signal voltage (-12 dB) is sure to suffice.

Finally wait for an Oscar-13 mode-S session. Just listen around 2400.720 MHz. That will be 144.720 MHz on your 2 metre SSB radio. (Note 144, not 145 MHz). The conversion accuracy won't be very precise, and even with doppler compensation may well be 30 kHz off nominal. If the mode-B transponder is ON too, you will hear many interfering LSB signals corresponding to mode-B's 145.880 - 145.920 MHz.

If the B transponder is OFF, you will hear genuine mode-S stations on USB loud and clear.

There is no such thing as a poor mode-S signal. Anyone who can get a decent mode-B uplink can overload mode-S with ease! The beacon, 70 kHz lower in frequency, is not on at the same time as the transponder. It will be very strong, and decode perfectly at all times.

Figure of Merit

The overall performance of any receive system is a function of only two things. The antenna gain (more = better), and the noise level (less = better). Thus the ratio of gain to noise is a concise indication of the system sensitivity. This is known as the G/T ratio.

The system you have built has a Gain of about x100 (20 dB), and a receive noise Temperature of about 200 K. Thus the G/T ratio of your dish is roughly 100/200 = 0.5. And this is the minimum requirement for AO-13 mode-S.

So if someone else is boasting say a 7 metre diameter dish (Gain x13600) and an ultra-low noise preamplifier with an overall receive noise temperature of 100K, then their G/T ratio is 13600/100 = 136. This is 24 dB better than our little dish, so AO-13 signals will be louder. But they won't be better, because the transponder noise floor will be louder too. And their antenna beamwidth will be only 1.5°, which is irritatingly small for satellite operation. Most az/el rotators rattle more than that.

So how do you measure G/T? Carefully measure the Sun noise increase as described above. Call the noise power ratio Y = (Vsun/Vzenith)2. Then G/T = (Y-1)/0.64 at 2.4 GHz!

Now the "0.64" factor includes a measure of the noisiness of the Sun, and this fluctuates from time to time by a dB or so [3]. However, I find the test is mostly repeatable, and is a reliable indicator of performance for better or worse if you have tweaked something. So make it a regular part of the station checkout routine.

Weatherproofing

When you are happy about the system performance, obtain some Dow Corning RTV clear sealant for electronics (NOT bath-tub sealant), and seal the pip of the N-type connector. When set, this will prevent any rotation of the helix. Also lightly seal the lip, shim and screw heads on the other side of the reflector. This will keep moisture out of the connector.

The UEK-13-P3C is built into a tinplate box with press-fit lids. It is absolutely not weatherproof or rustproof. The easiest way to keep the rain out is to use a polythene bag. Obtain a clear tough freezer food bag, and pull it over the converter assembly, stretching it around the reflector periphery to make a tight fit. This works well, but is not too robust. There are variations on this theme which will suggest themselves as you contemplate the problem.

A more subtle idea is to keep the antenna indoors. It works through a window, open or closed, though the noise level is higher due to the proximity of the warm house. Or it can be put outside on a tripod when needed. It is so small that it is highly portable, passing through doorways with ease. With its 3 dB beamwidth of 14°, it can be aimed once only for a mode-S session on AO-13.

Conclusion

This is a very easy project. In fact I found it the easiest thing I have ever done in ham radio, especially since I knew nothing about dishes and S-band other than the tit-bits one can read in articles. You can have practical fun with AO-13, as well as be ready for P3D from day one. When you have suceeded, your appetite will be whetted to absorb even more knowledge about 2.4 GHz.

S-band will no longer be a mystery, apparently accessible only to a favoured few. You will acquire a painless introduction to microwave operation - without even realising it's microwaves.

References

1. Miller J.R. G3RUH, Mode S - Tomorrow's Downlink?, Oscar News No. 97, October 1992. Also The Amsat Journal (USA), September/October 1992, and Amsat-DL Journal, December 1992.

2. Krome Ed KA9LNV, Satellite S Band: How to Become QRV on Our Highest Band, The Amsat Journal (USA), November/December 1992.

3. Shuch H.P. N6TX and Wilson P.M. W4HHK, Calibrating the Signal Generator in the Sky, QST November 1992.

Reading

Kraus John D. (W8JK), Antennas, 2nd Edition, McGraw-Hill 1988. ISBN 0-07-035422-7.

Reference Data for Radio Engineers, 6th edition, Section 27-39. Howard Sams 1982, ISBN 0-672-21218-8.

Dixon M.W. (Ed) G3PFR, Microwave Handbook Vol 1, RSGB 1991. ISBN 0-900612-89-4.

The ARRL Antenna Book, 15th Edition, Chapter 18. ISBN 0-87259-206-5.


Feedback on these pages to KB5MU. Feedback on the article should be sent to James Miller

Created: 1995 Jan 15 -- Last modified: 2005 Oct 31