Feature – Build Your Own satellite Antenna
John Hemming G0UYT describes an experimental satellite antenna for 2m/70cm.
John Hemming G0UYT describes an experimental satellite antenna for 2m/70cm.
Last summer I was recovering from surgery on the brevis tendon in my ankle so I was housebound for around eight weeks. As luck would have it the World Cup was on so I was enjoying my recovery with a beer in hand, football and the wonderful weather.
But of course, as a radio amateur, I wanted to operate outside in the sun. Being unable to travel I decided to work the FM satellites with my dual-band handhelds. Positioning myself in the centre of the garden I was able to hear a few passes of AO-91/2 from my garden chair while my leg was in plaster. I managed, by swinging around my handheld, to make a couple of contacts via AO-91 so the bug had bitten and plans were formed to construct a better antenna.
Time for Research
I researched some designs and decided on a dual-band Yagi using five elements on 70cm coupled with a Moxon two-element for 2m. Of course, this antenna was very good and I was able to make contacts by tracking the satellites across the sky at different elevations. I was able to work through all of the popular FM satellites, including AO-85, AO-91, AO-92 and SO-50. It was a struggle being in a plaster cast and often I would wait for the XYL to return from work and she would hold the antenna under my instructions as to where to point it. This was not ideal so I started working on plans for an omnidirectional antenna to work the birds (satellites) from the comfort of my garden chair while resting my leg!
I had time to research many satellite omnidirectional antennas. These included the ‘Eggbeater’ versions 1 and 2, Lindenblad dipole array, different vertical antennas, phased crossed dipoles, lazy-H, AWX, cloverleaf and double quads.
I managed to make a number of these antennas and some were better than others while attached to a vertical support on a tripod in the garden.
Although I was able to make some contacts with these antennas, I was still not happy because there was lots of fading and loss of signal if the satellite had a high elevation angle. I was sure I could improve on them or design my own antenna.
The design brief I set myself was:
- A single feed dual-band antenna for 2m and 70cm
- No need to use a diplexer
- Gain over a dipole
- Mixed polarisation
The antennas that most impressed me in my research were the Eggbeater and a version of the Lindenblad parasitic antenna for 70cm. I therefore looked more closely at how they worked and tried to incorporate their success into my design.
Because I had previously made a dual-band single-feed Yagi using the open-sleeve method of energising a 70cm radiator from the 2m feed, I decided this would be the heart of my design. I also took note from the parasitic Lindenblad, which uses a single vertical dipole with a number of horizontal dipoles at different angles to achieve close to circular polarisation.
I wondered whether, if I combined some of these techniques but swapped parasitic dipoles with crossed quads similar to the Eggbeater (doing this for both bands but on a single support), would this possibly work? I set about designing and thinking about how I could mount a centre-fed open-sleeve dipole for 2m/70cm with parasitic crossed quads for extra gain and multi-polarisation radiation.
I mention multi-polarisation – both vertical and horizontal radiation because the signal from the satellite changes as it crosses the sky. Fading is often caused by the satellite changing its polarisation along with the addition of multipath distortion where the received signal is bouncing off other objects such as rooftops, trees and the ground. These multipath effects mean that signals reach your antenna at different times, leading to multiple out-of-phase signals rather than one clean signal.
Many successful satellite antennas use circular polarisation but it doesn’t guarantee a perfect fade-free pass and usually has some loss of signal when compared to an antenna with the correct polarisation (bear in mind that as the signal polarisation changes, it will have both horizontal and vertical components so to get the best results, it’s necessary to be able to copy both simultaneously).
Design and Construction
Let me, then, explain the design and construction. The basic configuration is shown in Fig. 1 – a vertical dipole, with crossed full-wave loops for both 2m and 70cm. The various photos should also give an idea of how it is put together. My shopping list for this antenna comprised:
- 22mm PVC pipe (white) cut approx. 1m long
- Reel of PVC coated steel garden wire or hard drawn copper
- Mini RG8
- Trimmer 5-50pF (see what you have in your scrap box)
- 6 ferrite beads
- Electrical tape
- Tie wraps
- Plastic mounting pole
I used 22mm white plastic/PVC tubing found at most DIY stores for the main support of every element. (Black tubing has a higher carbon content and could affect the antenna.)
I used PVC-covered steel garden wire approximately 3mm in diameter but any hard-drawn steel or copper wire could be used.
The first job was to mount a 2m dipole on one of the sides, fixed in place with electrical tape and tie wraps. The next stage was to mount the 70cm radiator at 180° to the 2m dipole so on the opposite side of the tubing.
I used the formula 143/frequency to give me the length of a half wave and simply divided this by two for each leg of the 2m dipole. Thus 143/145 = 92.8cm, divide by 2 = 46.4cm for each leg. However, I would advise starting slightly longer (such as 49.6cm for each leg) and then trimming for best match/SWR once the 70cm radiator is fixed in place. The calculated length of the 70cm radiator is 143/435 = 32.8cm so either cut at 33cm or 32cm − it will make little difference.
The centre feedpoint I originally made from electrical choc-bloc connectors but then used a plastic dipole centre piece to make this weatherproof and to house a trimmer. I found I needed the trimmer to tune out any reactance once everything was assembled (more on this later).
I then used the same formula to measure the quad loop lengths, which is 143/Frequency x 2 (for a full wavelength loop), giving the length of wire I needed to trim. I actually rounded this up to make life easier because these measurements were not that critical and I wanted to use the antenna wideband from 144 to146MHz and 432 to 436MHz. I rounded up the figures from the same formulas as above, to give me a quad loop length for the 70cm band of 66cm and for the 2m band a length of 196cm.
You will need to cut two lengths of wire for each band because you will be making crossed loops. Once cut, I simply divided by four for each side of the quad and bent into a square shape with roughly equal sides.
Assembling the Loops
The difficult bit is assembling the loops onto the PVC tubing. Use the centre of the dipole as a reference/datum point. First measure for the holes where the 70cm loops will fit. (I used the tip of my soldering iron to make the holes but you can use a drill if needed.) So, for example, 8.25cm above and below this point but 90° from the dipole elements. The second loop needs to be crossed so turn the tubing 90° and make the second set of holes for the second loop. You will need to make these holes just above the measurement of the first loop by about 5cm or less.
I pushed the partially-formed loop through the wire, turning the tube until it was fitted. I used a small choc-bloc connector to complete the loop. You can, of course, use different methods to complete the loop such as soldering or crimping.
Then it’s the turn of the 2m loops, again using the centre of the dipole for a reference point and making holes above and below this point to insert the loops but, importantly, these loops will be turned 45° degrees from the 70cm loops, as can be seen in the plan view in Fig. 1.
Hopefully now you have completed the dipole centre and fixed the loops in place. To feed the antenna you will need to use a small fly lead (3m) of high-quality coaxial cable. I recommend Mini RG8. I do advise using the best quality and lowest loss coax as possible both for the fly lead and for the feeder to the shack because with satellite work it can be the difference in making a contact or not even hearing them due to the weak signals and feeder losses. The fly-lead is taken to the feedpoint of the 2m dipole.
I mentioned earlier that I used a small trimmer to tune out reactance because I found that, once assembled, I had introduced some reactance. I tuned the trimmer for the best match. Others who have built the antenna didn’t need to and just trimmed the dipole lengths towards the end of the build instead. As long as you have a good match of 70cm and 2m, you should be in business whatever your method.
After trimming, I managed to achieve a 1:1 match on 70 cm and 1.2 on 2m, both of which are perfectly acceptable.
You will also need to use around five or six ferrite beads to slip over the coax feed and fix them in place just below the point where the lower 2m dipole element ends. This is to stop your coax fly-lead becoming part of the antenna and radiating.
Finally, fix the antenna to a suitable plastic mounting pole before using any steel or aluminium mast. The antenna is now complete.
I checked out the antenna with it mounted around 7ft above ground. The ultimate test for me would be to see if I could hear SO-50 on a high elevation pass without moving the antenna or aiming it at the satellite. I mention SO-50 because this satellite has a 70cm downlink and is quite difficult to hear without a beam.
So, at 75° I waited for the bird to come over. Fully quieting signals from the satellite were heard! Compared to a single vertical dipole it certainly showed huge promise. (A vertical dipole has a doughnut-shaped radiation pattern and although good on lower elevation passes fails miserably with a high elevation pass because of the ‘hole’ in the doughnut.)
Next was to test the antenna on transmit. I managed to work through a number of satellites with the antenna still just 7ft from the ground, with good signals across Europe. I’m guessing that the ground acted as a reflector or a large dish to help with the signals.
I then wanted to test the antenna higher up above my roofline, so I could try to work some linear satellites such as AO-73 and FO-29. My son helped with the installation and the experimental antenna was fixed in place just above the guttering on an 8ft mast.
I used around 15m of RG213 to connect to the fly-lead and fed back into my shack – as I said earlier, good quality low-loss feeder is essential.
To my surprise I tuned around and the antenna also worked quite well across the SSB and FM sections of the bands. Compared with my collinear, this antenna certainly sounded different and pulled in stations both horizontally and vertically polarised.
The next test was to try to work FO-29 semi duplex from my Yaesu FT-100D. Using around 20W I managed to work YO5TP, HA6NM, F4DXV, PE1NIL, EA2AZW, UX0FF, OH5LK, UT9NA and R7MU. I also used the antenna with the FM satellites with similar results.
In conclusion, I feel that this antenna, although not perfect and having room for development and tweaks, has a great place in my antenna systems. I am now developing Stage 2 with an added groundplane system, loops 90° out phase from each other and more parasitic elements.
Although the design may seem a little unorthodox and breaks some rules of antenna building, it was a fun solution that happened to work compared with other simple satellite antennas.
Since building the antenna I was requested to submit my papers to the American AMSAT organisation and was bowled over that it was included in talks at the Space Symposium and has been built by some of our American cousins with good results.
Why not have a go? It’s an afternoon’s work. I’m sure you will be impressed.
This article was featured in the April 2019 issue of Practical Wireless