©1999-2021 F. Dörenberg, unless stated otherwise. All rights reserved worldwide. No part of this publication may be used without permission from the author.

Latest update: 25 Sept 2021 (added ref 1G-1K).

Previous updates: 2 July 2020.


"Loading" is a way to lower the (primary) resonant frequency of an antenna radiator. This technique is typically used in antennas that have radiating elements that are too short for the desired resonance frequency. There are several standard ways to load a radiator (ref. 1A-1F, 2, 3A-3F):

  • Inductive loading: placing a loading coil (inductance) somewhere between the feedpoint of the antenna and the tip of the radiator. This compensates for the capacitive feedpoint reactance of the short radiator.
  • End-hat loading with a "capacitive hat", typically installed at the tip of the radiating element(s). This counteracts the tapering off of the current distribution between the feedpoint and the tip of the radiator. It also raises the radiation resistance of the antenna, i.e., makes the antenna a more effective radiator.
  • Linear loading, by folding a long radiator wire in a zig-zag form onto itself. The result is a radiating element that is three or four times shorter than the overall wire length. The folded wires are parallel and closely spaced. The interaction between the parallel wires is complex, and introduces sub-band resonances (ref. 3C).
  • Helical loading, by winding the radiator into the form of a linear spiral.

Obviously these loading methods can be combined. E.g., a loading coil and end-hat loading. Or: a loading coil combined with linear loading.

The "Cobra" is a dipole antenna with linear-loading: the wires are folded such that it is configured as three or four parallel wires. Its name refers to the way the antenna wire "snakes" around in an zigzag manner. No, the name has nothing to do with the Cobra brand of CB radio products, nor with Coaxial Beam-Rotating Antennas, nor with science- and nature-defying Cobra EH-antennas from a certain Italian manufacturer.

cobra dipole

The folded wires can be kept separate and parallel with spreaders, such as used in cage dipoles. It is more convenient to use multi-conductor wire (not to be confused with single-conductor multi-strand wire). In this case, the wires are very close together: they are only separated by the wire insulation. An example of this is ribbon cable, and flat antenna rotor control cable (which is what I used). 

Just like any dipole, it can be operated as a multi-band antenna - if it is fed with a twin-lead cable or ladder-line and an antenna tuner/coupler. It is also possible to get multi-band operation with a 4:1 balun at the feedpoint (I use an Elecraft BL1 with a "pig-nose" (2-hole) ferrite core, plus a W2DU-type 1:1 current  choke balun) and coax to the transceiver. But  you probably still need an antenna tuner at the rig to keep the transmitter happy.

The standard full-size  Cobra for 160-10 mtrs has a span of 140 ft (≈43 m). The "shorty", "junior", or "half-size" Cobra for 80-10 mtrs spans only half that: 70 ft (≈21 m). The span is about 1/4 λ in the lowest band. I.e., half the span of a full-size dipole.

See ref. 4 for nec-files (antenna modeling/simulation).

MY 40-10 COBRA

At my QTH, I can fit neither a full-size, nor a half-size Cobra. So I decided to try a 1/4-size Cobra for the 40-10 mtr bands. My first version (2008) spanned 2x18 ft (2x5.5 m). This span is about 0.5 λ in the 20 mtr band.

I used 3-conductor rotor control flat-cable (ref. 5). The conductors are AWG #22 (0.64 mm Ø) multi-strand. Adjacent conductors are spaced by 0.1" (2.5 mm). One of the conductors is tinned, which is not good for RF applications. But anyway...

cobra dipole

Flat 3-conductor antenna rotor control cable

Center-insulator and end-insulators are the same as for any old dipole antenna. I usually make them out of chunks of polypropylene kitchen cutting board.

In my experience, linear loading of the Cobra makes the antenna appear about 10-20% longer than its actual span. Obviously nowhere nearly as long as the total wire length. Conversely, it means that a span can be used that is 10-20% smaller than a regular wire dipole for the same primary resonance frequency.

 I found the antenna easier to "tune" (load & match) without a 4:1 balun than with such a balun at the feedpoint of the dipole.

Below is a plot of my miniVNA antenna analyzer:

cobra dipole

Sweep from 1.5 to 29 MHz with my miniVNA antenna analyzer

(configuration: antenna + 300 Ω twin-lead + 4:1 balun + 1:1 choke balun)

In 2011, I rebuilt it with the span increased to 2x21 ft (2x6.4 m), in order to move a resonance dip down to 7 MHz:

cobra dipole

SWR plot for my Cobra extended to 2x21 ft (2x6.4 m)

In 2011 I acquired an EDX-2 external automatic antenna coupler for my Alinco DX-70TH transceiver. The coupler is located very close to the feedpoint of my dipoles. I no longer use a balun. I did some more experiments with the dipole length. In the mean time, I had reduced the span back to 2x20 ft (2x 6m) - don't ask me why (I just don't remember how or why that happened). Then I extended the end of the Cobra wires by 5 ft (1.5 m). Again, don't ask me why, hihi. The SWR plot below was measured at the end of the short 300 Ω twin-lead to the automatic tuner/coupler:

cobra dipole

SWR plot of 2x20 ft (2x6 m) Cobra, extended with 5 ft (1.5 m) wires

Compare that to the SWR plot of a regular wire dipole with the same span, installed at exactly the same place:

cobra dipole

SWR plot of a regular wire dipole with the same span, installed at the same place


External links last checked: October 2015 unless noted otherwise

red-blue line