Saturday, November 17, 2012

Comparison Between Different Antenna Systems "in Free Space"

Table 1: Comparison Between Different Antenna Systems "in Free Space"
 ANTENNA FORWARD GAIN FRONT / SIDE REJECTION RADIATION  RESISTANCE 1/4 Vertical 1.45 dBi 0 dBi 21 ohm Vertical Dipole 2.13 dBi 0 dBi 72 ohm Delta Loop (Equilateral) 2.96 dBi 2.4 dBi 115 ohm Quad Loop (Square) 3.30 dBi 3.8 dBi 126 ohm Delta Loop (Compressed) 3.55 dBi 5.2 dBi 31 ohm Rectangular Loop 4.55 dBi 8.3 dBi 32 ohm Half Square 4.83 dBi 8.3 dBi 32 ohm ADR Loop 5.25 dBi 26.1 dBi 63 ohm Bobtail Curtain 6.46 dBi 99.9 dBi 34 ohm SDR Loop 6.57 dBi 19.3 dBi 50 ohm

1/4 VERTICAL

Quarter wavelength vertical antennas are widely used in phased arrays, expecially in those which cover more than two directions. The famous "four square" array is a good example of it. The down side of the ground plane is that it can be a very inefficient antenna. Unless one uses "elevated radials" with sufficient height, the antenna return current will travel through the ground, and back to the feed point. This will cause RI2 losses. Unless special cares are taken, there is a heavy price to pay in terms of losses when the radial system is located "in the ground" or "on the ground".

It is not uncommon to see the efficiency of a 1/4 wavelength antenna to be well below 50%. There is a number of experiments and publications which demonstrate how the efficiency of a ground plane is sensitive to its radial system. One experiment which can be easily conducted is to measure the feed point impedance of the antenna while adding more radials to the ground system. The difference of impedance observed represents the "loss resistance" which in practice translates into loss of power into the ground itself. Knowing the feedpoint impedance of the antenna and the theoritical radiation resistance also allows calculating the efficiency of the antenna.

Here is an example of data obtained after having performed the experiment described above. I started the experiment with 21 radials, the antenna was tuned for 14.200 Mhz. The vertical section of the antenna was only 1 foot above the ground and the 1/4 wavelength long radials were laying on the ground. The feed point impedance was measured using an antenna analyzer model MFJ-259B. Here are the actual results obtained:

Table 2: Antenna Efficiency versus # Radials Used
 # Radials Feed Point Impedance Antenna Efficiency Power Lost into Ground (1 kw output) Loss (ohm) (%) (watts) (dB) 0 94 27% 734 -5.8 1 84 30% 702 -5.3 2 69 36% 638 -4.4 3 61 41% 590 -3.9 4 55 45% 545 -3.4 5 49 51% 490 -2.9 6 46 54% 457 -2.6 7 42 60% 405 -2.3 8 39 64% 359 -1.9 9 39 64% 359 -1.9 10 38 66% 342 -1.8 11 36 69% 306 -1.6 12 36 69% 306 -1.6 13 32 78% 219 -1.1 14 31 81% 194 -0.9 15 31 81% 194 -0.9 16 31 81% 194 -0.9 17 30 83% 167 -0.8 18 30 83% 167 -0.8 19 30 83% 167 -0.8 20 30 83% 167 -0.8 21 30 83% 167 -0.8

These results are quite stunning; going from 3 to 14 radials decreases the feed point impedance by a factor 2, which in turn translates into a 3 dB improvement! In other words, the efficiency of the antenna would be doubled simply by using 14 radials instead of only 3. As mentioned previously, knowing the radiation resistance of the antenna allows calculating the efficiency of the antenna. The simple formula is shown below:

% Efficiency = Radiation Resistance (ohm)

Feed Point Impedance(ohm)

The experiment shown in table 2 has been conducted by many ham operators, often times measuring the field strength as opposed to the feed point impedance. The trend observed is always consistent with the results shown above.

The conductivity of the ground is also a key factor here. Poorer the ground conductivity, more dramatic the losses will be. Over a very good ground, a smaller number of radials will suffice to obtain a good antenna efficiency. Over salt water, only 2 radials will yield satisfactory results!

FULL WAVE LOOPS

Rectangular Loop, Compressed Delta Loop)

Full Wave Loops are very simple antennas which can be extremely efficient if a minimum of precautions are taken in their design. For some reasons, only a few variations of these loops have been thoroughly documented; mainly the Quad and the Delta Loops. Interestingly, from a performance and “ease to feed” standpoints, these models are clearly not the most appealing ones. It is still a mystery to me why the other variations have not been better explored and documented.

One of the appealing characteristics of the Full Wave Loops is that they can actually exhibit a very interesting GAIN and even offer some directivity. The Table above shows the gain and directivity of different loop designs and other common antennas. Another interesting aspect of the Full Wave Loops is that they do not require radials for the return current. In a sense, they are “self contained”. In fact, a vertically polarized loop could be represented as being a pair of phased verticals, each with its own elevated radial.

Just like it is the case for all vertically polarized antennas, the quality of the ground is an important factor and will have a determining effect on many characteristics such as take off angle, gain, near field losses, etc. It is also important to note that the horizontal wire of the loop will tend to couple with the ground itself so one has interest to elevate the loop as high as possible and/or improve the quality of the ground underneath the antenna itself.

SHAPE OF THE LOOP

The shape of the loop itself will greatly influence its gain, radiation resistance and bandwidth. For example, the quad and the equilateral delta loop have both relatively high radiation resistance and similar gain. By compressing the vertical sides of these loops and widening the horizontal component, the gain will increase significantly and the radiation resistance will decrease to a level that can be optimized to perfectly match a 50 ohms coax line. The price to pay for these advantages will be some bandwidth. Following this logic, going from a quad to a rectangle or going from an equilateral delta loop to a compressed delta loop would provide the advantages mentioned above. Once again, the comparison table shown above testifies of these very facts.

EFFECT OF THE GROUND

Just like it is the case for all vertically polarized antennas, the quality of the ground is an important factor and will have a determining effect on many characteristics such as take off angle, gain, near field losses, etc. It is also important to note that the horizontal wire of the loop will tend to couple with the ground itself so one has interest to elevate the loop as high as possible and/or improve the quality of the ground underneath the antenna itself.

PERFORMANCE

Even though it is almost difficult to believe, going from a ¼ wavelength vertical to a simple rectangular loop can lead to 3-4 S-units improvement! The video clip below shows an example of this fact…. (copy email explanation here).

VERTICAL DIPOLE

The Vertical dipole...

DELTA LOOP (EQUILATERAL)

The Square Loop, often refered as "Cubical Quad",...

DELTA LOOP (COMPRESSED)

RECTANGULAR LOOP

The Rectangular Loop is the most performant antenna I have tried so far, not mentioning other interesting advantages like:

- No BALUN required (can be fed directly with coax).

- Radiation Resistance can be adjusted by changing the ratio of the height / width

- It is by design a dualband antenna where a 20m loop for instance will resonate as a 2 wavelength loop on 10m. Using a 75 ohm 1/4 wl. section cut for 10m will give a perfect match for both band.

-
HALF SQUARE

(The example shown in the picture is an "inverted" Half Square)

This is one of the best antenna I have tried. Compared to the 4BTV, signals were either equal or had an advantage of 1-3 S-Units. The inverted version does not seem to work as well as the standard version. However, this conclusion is questionable since tests made were not extensive enough.

It is obvious that in the case of the inverted version, the horizontal wire is closer to the ground which implies more ground losses.

In all case, the there was a significant effect of the feed line on the SWR which indicates a strong issue with RF feedback. A BULUN is clearly a MUST.

On advantage of the Half Square is that it can easily be made multiband by adding relays on the vertical sections. By choosing a compromise dimension of the horizontal wire (like 29 feet to cover 17-20m), good performance can be obtained. The model I tried was:

Horizontal wire: 29 feet  Vertical wires: 19 feet

The litterature recommends to connect the ground of the coax to the vertical section and the positive to the horizontal wire. Apparently, that should help minimizing the RF feedback issue.

I tries it and did not noticed any significant improvement. Reversing the polarity also had the effect to increase a bit the resonant frequence from 14.200 MHZ to 14.600 MHZ.