Many times when I go portable or mobile I take a variety of antennas with me. Fortunately I drive a lightweight van which gives me plenty of space for antennas and masts.
For some time I have successfully used antenna design software by K6STI Brian Beezley. I use two packages Antenna optimizer AO which is useful for all wire antennas and Yagis. And Yagi optimizer YO which as the name suggests is primarily for Yagi type antennas. From the output of these programs I have constructed many antennas, and using measurements with a reference dipole, verified the results. (How I've made antenna measurements).
I have also used information from Gunter Hoch DL6WU for double optimised long Yagis (VHF Communications 3,4/1977). Also I found that information from the American National Bureau of Standards (NBS) which has been around since the 1950's. Based on that, in 1976 P.Viezbickie the author of 'Yagi Antenna Design'. An article with practical designs by W1JR appeared in the Ham Radio magazine August 1977, well worth looking at if you can get a copy.
The 2 metre antennas I am using at present for mobile and portable are 2 X 2-element Yagis and 2 X 5-element Yagis,
both antennas stacked.
For the mobile I use the 2X2-element Yagis with about 4 ft spacing between the antennas. Fed with a coaxial 'phasing harness' using 75 Ohm coax between the antennas.
And for the portable set-up I use the 2X5-element Yagis and a home-brewpower splitter.
Construction of the antenna is fairly simple and I have used readily available materials from the local TV aerial shop. For the booms I am using 15 mm square section aluminium tube. The driven element is made using a FM broadcast type plastic dipole centre and 13 mm aluminium tube. The reflector is 6 mm aluminium tube.
The reflector is attached to the boom with stainless steel clips, which I got from:
Sandpiper Communications They also supply aluminium tube/rod.
Dee Comm They supply many antenna bits, also useful catalogue.
I have attached elements in the past by drilling through the boom with a hole big enough for the 6 mm element to pass through and fixing the element at its centre point with a stainless steel self-tapper screw. The plastic dipole centre is bolted to the boom.
As I have rear mounted the 2-element antennas. The coax cable is fed from the driven element inside the boom (like a balun) and then down the mast to the centre point. A 'T' piece connector could be used here. But as my antenna system is permanently assembled, I decided to use a small plastic box, making soldered connections with the two antenna feeds and the coax feed to the rig. The vertical spacing between the antennas is 4 ft (1.2m), so the 75 Ohm coax is 5/4 wavelengths (1.716 m) to each antenna feed point. The 50 Ohm feed can be any length to suit.
This method seems to work OK at the 150 Watt level I use mobile or portable, and SWR measurements are about 1.3:1 which is quite adequate. It seems reasonably water proof, but I tend not to use it too much if the weather is really bad.
Details of stacking various short Yagi antennas were found in VHF Communications 2/1986. Here it is shown that maximum gain for a 2-element is achieved at 0.65 wavelength, about 4 ft.
2-element beam
| Otimized frequency | 144.30 MHz | |
| Element length in mm | Spacing in mm | Element diameter in mm |
|
Ref = 1016 |
6.3 |
|
|
De = 940 |
325 |
13 |
| Boom 15 mm square section tube | ||
Click here for polar plot details and analysis
I had thought of using the ZL Special or the HB9CV antennas, but both these antennas have a more complex construction
than the straight forward 2-element Yagis I have used. A comparison of a single ZL Special antenna can be seen
here.
5-element beam
| Otimized frequency | 144.30 MHz | |
| Element length in mm | Spacing in mm | Element diameter in mm |
|
Ref = 1005 |
6.3 |
|
|
De = 956 |
450 |
13 |
|
D1 = 918 |
865 |
6.3 |
|
D2 = 918 |
1280 |
6.3 |
|
D3 = 922 |
1695 |
6.3 |
| Boom 15 mm square section tube | ||
The 5-element Yagi is based on the NBS data. The materials used are similar to those used in the 2-element. However the driven element in this design uses a gamma fed dipole.
Look here for polar plot details.
Gamma match
The gamma match is made using a 25 cm length of 6 mm aluminium tube and a length of the inner conductor and insulation from a piece of RG58 coax or similar cable. This conductor is fed into the 6 mm tube. (You may have to experiment with the length of this cable as it forms the capacitor, which tunes out the reactance of the coupling tube to the driven element). The coupling capacitor/gamma tube is connected to the driven element via a shorting strip at approximately 22.5 cm from the centre of the driven element. (This position may have to be altered). The spacing between the driven element and the gamma tube is 2.8 cm centre to centre. The driven element is bolted directly to the boom at the element centre. Driven element is 13 mm or 1/2 inch diameter aluminium tube. The coax feed is connected to the end of cable sticking out of the 6 mm gamma tube; this connection will need to be weather proof. The end of the conductor inside the gamma tube needs some extra insulation, and a bit of insulating tape works for me.
The phasing harness is made with an odd number of 1/4 wave-lengths of 75 Ohm coax to each antenna from the centre feed point (' T '), allowing for the velocity factor (vf) of the coax not just the free air wave length. (vf=0.66 for solid polyethylene RG57, UR77,RG11 etc.).
For a frequency of 144.3 MHz the wave length would be 300/144.3 (metres)
2.08 metres 1/4 of that 52 cm.
52*vf (which is 0.66) gives a length of 34.32 cm of coax cable.
The number of odd 1/4 wave-lengths is dependent on the distance between the antennas, but must be equal from the centre to each antenna, this then means that the phase relationship between each antennas feed point is maintained.
A bit of maths may help here.
Transformation between two impeadances Z1 and Z2
All you need is a quarter-wave line with characteristic impedance Z0 = square root (Z1 Z2)
Z0 =The impedance of the quarter-wave line transformer
Z1 =The impedance of the feeder
Z2 =The impedance of the antenna combination
So feeding two 50 Ohm antennas in parallel would give 25 Ohm. If the TX/RX is fed with 50 Ohm coax:
Z0 = square root (50*25)
this would mean that the quarter wave transformer would have to be 35.35 Ohm.
When you use coax again it's like having the two coaxial feeds in parallel which would mean that the coaxial transformer should be 70.71 Ohm. Obviously this value coax is not readily available, but, the more readily available 75 Ohm coax, is a fair compromise and only gives a slight increase in the SWR.
Shown above is the normal arrangement for a two way power splitter. The end that has the two 'N' sockets is where the two antennas are connected with equal lengths of 50 Ohm coax, the other port goes to the TX/RX, again in 50 Ohm coax. Low loss coax in all the cables is important for the best results.
The better method for matching more than one antenna is to use a power splitter/combiner and making your own section
of l/4 line.
I make mine with 1 inch (25.4 mm) square section tube; it has 2 mm wall thickness, and a round centre conductor, using the formula:
Z0 =138log10{1.08*D/d}
Z0 = impedance l/4 line
Where D = inside dimension of outer conductor
and d = outside diameter of inner conductor
Knowing the value for the 1/4 wave-line for two antennas is 35.35 Ohm and the dimension D, it is easy to calculate
the dimension of the inner conductor, then choosing a size of tube closest to the preferred value. I have found
that 12.7 mm or 1/2 inch tube works OK for the inner conductor.
The outer can be copper, brass or aluminium. The inner has to be soldered so copper or brass is used here. It is useful to choose 1 inch square section for the outer because this allows the use of square flange mount N type sockets. The inner has to be a 1/4 wave long and this is in free air so for 2 metres this will be 52 cm. The outer has to be longer to allow for the flange mount sockets and to put waterproof caps at each end of the tube.
Extending this information for power splitters for four antennas. The same principals are involved, now we are looking to match four, 50 Ohm load impeadances combined to give 12.5 Ohms. Using the formula for l/4 line transformer, it is calculated to be 25 Ohms. Using the same outer dimensions as the 2 way splitter would require an inner tube of 14.94 mm or 15 mm.
All four antenna ports are connected at one end of the power splitter.
You must remember that each antenna must have exactly the same length of low-loss coax between its feed point and
the power splitter. Otherwise the phase relationship of the array will not be maintained and cancellation of signal
will result.
It is also possible to make a half wave line fed at the centre, thus forming two l/4 transformers. With this arrangement two loads are connected at one end and the other two loads at the other as shown in the diagram below. This may be more convenient if you are building a large array and need to balance out the positions of the feeders.
For this splitter each pair of antennas combine to give 25 Ohms, and use a 50 Ohm quarter-wave transformer to give 100 Ohms; when the two 100 Ohm points are connected in parallel at the centre, the result is 50 Ohms.
If you make a splitter like the one shown here you will have to drill a large enough hole on one of the sides at the centre so that you can solder the centre connector to the conductor. Then a ' blind grommet ' can be used to weather proof. I usually give the whole thing a couple of coats of yacht varnish, but don't paint the threads of the ' N types '.
Here is a table of values for two way and four way power splitters using a single l/4 line.
Baluns are often a point of discussion and confusion. Some people say that they have used them and found no difference. This may be the case, and I have used antennas with and without baluns. Most antennas are balanced or symmetrical with respect to ground, so connecting unsymmetrical coaxial cable requires some form of transformer or balun between the feed point and cable.
The word balun means (Balanced to Un-balanced) transformer. And of course coaxial cable, which is the most common feeder, is unbalanced electrically. Connecting coax directly to an antenna could mean that some current will flow in the opposite direction on the outer or braid of the feeder. A correctly designed balun can prevent this and any of the problems that can arise when RF flows on the outer of the coax. The balun can also help preserve the best radiation pattern for the antenna.
Obviously some antennas can be fed directly with coaxial cable e.g. antennas that do not have a balanced feed point, such as a gamma matched Yagi.
At VHF baluns are used both as impedance transformers, as in the case of a 4:1 balun to feed a folded dipole which would have an impedance of 200 Ohms, this would give a good match into 50 Ohm coax. The construction of such a balun is straight forward, but must be made such that it is low loss. Look here for typical construction of a coaxial 4:1 balun.
1:1 baluns come in several forms. At VHF it is easy to make a sleeve balun which is a shield surrounding the final l/4 of coax at the antenna feed point.
I use the MFJ-259 Antenna analyser; this is a very useful tool when you are building your own antennas.
I did find however that it was not quite so accurate at 144 MHz. After looking at the design of the analyser I decided to take out the SO 239 connector and replace it with a ' N type ' connector. Making sure that the ground connection to the case was good. I then used some fixed value resistors to re-calibrate the instrument. This has improved the accuracy of the 259 especially in the 2 metre band.
Accuracy is important particularly when making up phasing harness, or any other line measurements when cutting for resonant lengths. I would say it has definitely been worth the effort of carrying out the modifications. Now the analyser gives much more reliable result than it did in it's original state.
When I make a new antenna it is nice to find out how it performs. The usual method I use is to set up the test antenna and a dipole. At some distance away (I usually measure over a distance of 1/2 to 3/4 of a mile) I get a colleague to set up a low power signal. Over a clear line of site path with no nearby obstacles. The low power beacon provides a reliable reference point to carry out the measurements. At the test side I have a receiver connected via an accurate stepped attenuator to either the test antenna or the reference dipole. First I set up the receiver with the dipole connected, and I adjust the S meter, with the aid of the attenuator, to a SET point (a rig with a large analogue meter is useful here), to which I can be sure I can re set to, say S9. Then I replace the dipole with the antenna under test, and assuming all is well and the antenna has some gain over the dipole, readjust the S meter reading with the attenuator till it achieves the SET point. Then it's just a case of reading off the difference on the attenuator to give the gain of the antenna in this situation.
That gives the gain of the antenna over the ref. dipole. Now a similar method can be used to give FB and overall polar pattern if you want to take a few more readings. This relies on calibrating the 'S meter' with the stepped attenuator, and again using the distant signal source to take readings.
So far with this method I have achieved good correlation with computed data and also with published figures for known antennas.
Lastly testing on air over a much greater distance, although this can be difficult especially if there is QSB to contend with. When I first used the 2X2-element array I found that it was giving extremely good results considering the size of the antenna. Tests over a path of 300 miles with GU3EJL, on the Island of Alderney in the channel Isles, showed considerable improvement over a single 2-element beam although results just relying on S meter readings.
I have also tried a single 4-element beam in comparison with the 2 over 2 stacked and again the results seemed to come out on the side of the 2X2-element.
In my opinion the most ideal antenna for normal tropospheric propagation, is an antenna that has plenty of forward gain; but not too tight a beamwidth in the horizontal plane; but with a narrow beamwidth in the vertical plane. So a horizontal fan shaped beam pattern. I believe this is best achieved by stacking beams rather than going for a very long Yagi. But it's 'horses for courses and it is more complicated to put together large antenna arrays.
The largest system I have made for the 70 cm band was 8 X 18-element long Yagis. Sadly I no longer use it but I believe that at least half of the array is still in use at the QTH of G4MPH, there still must be some mileage in them.
For information on 23 cm antennas link to 'double quad'.
A table of comparative 2 metre and 70 cm antennas.
Stacking antennas link here.
6 metre Beam.
Sandpiper Communications
Unit 5/6 Enterprise House
Cwmbach Industrial Estate
Canal Road, ABERDARE
MID. GLAMORGAN CF44 0AE
Tel. 01685 870425
Dee Comm
Amateur Radio Products
Unit 1 Canal View Ind. Est.
Brettell Lane
Brierley Hill
West Midlands DY5 3LQ
Tel. 01384 480565
Brian Beezley, K6STI
3532 Linda Vista
San Marcos,
CA 92069
USA
This Page is ©
1999-2001