If you are space restricted and cannot use a half-wave dipole, you may want to consider a shortened linear loaded dipole called Cobra. This antenna offers about 50% reduction in overall length over a half-wave dipole. Unfortunately, I do not know the name nor the call sign of its originator.
Some authors describe the Cobra as an antenna that can be used over a wide frequency range. I plotted the impedance of a Cobra designed for 20 m band.
The antenna has a clear resonance at 14.2 MHz, where it has quite low resistance (around 10 ohm). When you decrease or increase the frequency, Cobra feed point impedance changes over a wide range. I wonder who started to promote this antenna as suitable for multi-band operation?
Let’s analyze the Cobra as a single band shortened dipole rather than a wideband antenna. In my modelling, I used ideal conductors of diameter 2 mm as wires and I placed the model in free space. The antenna dimensions were designated as in the drawing below.
I set dimension C to 0.2 m and simulated the Cobra for several values of b starting from 0.025 m up to 0.2 m. For every value of b, I found the optimal a resulting in resonance in the center frequency Fc (14.175 MHz). Please note that SWR has been calculated for the source impedance 12.5 ohms, because that was close to the Cobra “native” resistance in resonance. You can see the simulation results below.
You can clearly see in the table that:
smaller b requires greater a to keep the resonance at the same frequency,
smaller b means greater resonance resistance (although still much lower than 50 ohms),
once the C is set, changing b and a (to keep the Fc constant) almost does not affect antenna gain and antenna bandwidth; for C=0.2 m, you get G=1.8 dBi and BW=350 kHz (for SWR=<5),
for 20 m band and C=0.2 m, the optimum value for b is 0.075 m; the antenna resonance resistance is then very close to 12.5 ohm and overall length A equals 5.016 m what is about ¼ wavelength; I assumed 12.5 ohm as a preferred value because this is 50 ohms divided by 4, and you can build a suitable balun relatively easy.
Obtaining SWR=5 at the ends of 20 m band is not a very good result. That would be acceptable for a small multiband antenna (like the Cobweb) but is not great for a single band antenna. The above simulations showed that if the bandwidth is to depend on any dimension, it must be the C. So I designed three versions of the Cobra with different C dimension:
The greater the C, the greater the antenna bandwidth BW (calculated here for SWR=<5). Moreover, greater C requires greater b and smaller a to keep the antenna at the same resonance frequency Fc (14.175 MHz).
To sum up. The Cobra is a rather narrowband antenna, offering you about 50% reduction in antenna length w.r.t. half-wave dipole. Its gain is about 1.8 dBi what is only 0.35 dB below the dipole.
Notes on accuracy of the models. I checked the AGT result for every model. Although it was well below 0.2 dB (what indicates the model is accurate), NEC-2 engine is known for not simulating very well the antennas with very closely spaced wires. On the other hand, the models seem to follow clear trends (changes in Rres, stability of SWR at the band ends, stability of gain) so I believe we can trust the results shown above.
Click here to download Cobra ver. 1-3 models.
I tried the Cobra concept on the quad loop. See its drawing below.
As you can see, the Cobra Quad Loop has reasonable bandwidth (at least for 20 m band) and has nice input impedance. Its gain is slightly larger than a dipole but, of course, a little lower than a full-size Quad Loop. Size matters. The Cobra Quad Loop linear dimensions are about 60% of its full size brother. So, the size reduction is significant.
Click here to download Cobra Quad models.
The next logical step was checking out whether one can design a 2-element version of the Cobra Quad Loop. It turned out that it was possible but, of course, with a little worse performance than the full size version.
As you can see in the drawing below, the driven element and the reflector are the squares of 3.08 x 3.08 m. But they differ in the depths dimension. The reflector is deeper. So, the wire length used for this element is greater. The antenna requires a 2 microhenry coil connected in series with a feed point.
The forward gain changes from 7.2 dBi @ 14 MHz to 4.9 dBi @ 14.35 MHz, what translates to 5dBd and 2.7 dBd respectively. So the gain decreases when the frequency rises. The front-to-back ratio achieves its maximum (15 dB) in the center of the band. SWR changes from slightly above 2:1 to slightly above 3:1. All-in-all, the performance of this antenna is quite acceptable if you take into account size reduction.
Click here to download this antenna models.