Will electrically joined dipoles of different lengths, at right angles, behave as a multiband antenna?

I haven't seen this done, (which, after more than a century of amateur and professional radio development, is probably a bad sign), but I had an idea the other day for simplifying the setup and feed to cover more than one band without the compexity of traps or multiple feeds and RF switching.

If I were to build a pair of dipoles at right angles, on the same feed line, one sized for, say, 40m, the other for, in this case, 20m, would they act as one very poorly tuned antenna, or would they act like a wave trapped dipole (except with two different radiation patterns)? I can think of arguments for either case, and I'm pretty fuzzy on antenna theory.

• There's actually antennas that have dipoles in the same plane instead of right angles and are multi- or even wideband antennas. One antenna, one "relatively constant" impedance over a large swath of frequencies. Is that what you're looking for? Jul 2, 2019 at 12:59
• Yep, potentially along with having different radiation directions (my lot will only support a very long antenna in one direction). Jul 2, 2019 at 13:07
• Like this? Jul 2, 2019 at 13:41
• @PeteNU9W No, not really, at least as far as I can see -- those are high-gain, high-frequency antennas. Think of an X with two opposite legs clipped, parallel to the ground (or a 40m and a 20m horizontal dipole crossing at the feed point). Jul 2, 2019 at 13:46
• @ZeissIkon — you’re focused on the trees and not seeing the forest. Ignore the directors and reflectors. That antenna is two dipoles for two different frequencies, at right angles. Jul 2, 2019 at 13:50

The results depend on the two bands you choose. Frequency ratios of 2:1 are a good choice because the longer dipole, which is a full wavelength at the higher frequency band, will show high impedance on that band, while the shorter dipole, which is only a quarter wavelength at the lower frequency band, will show a high (capacitive) impedance on that band.

Let's use your example of crossed dipoles that are a half wavelength on 40m and 20m, respectively. In EZNEC, this is accomplished by connecting identical very short transmission lines from the center of each dipole to a virtual segment that hosts the single driving current source.

As shown below, the current at the center of the 20m dipole is about 8% of the current at the center of the 40m dipole when operated on 40m:

The situation is reversed when operating on 20m, where the current at the center of the 40m dipole is about 1.4% of the current at the center of the 20m dipole:

One might suspect that interactions between the the currents in the two antennas could adversely affect the SWR on each band, but these plots show that this is not the case:

These SWR curves are virtually identical to the curves obtained when each dipole is swept after the other dipole is completely removed from the model.

One reason this is not frequently done could be that more supports are needed and that the patterns of the two dipoles are orthogonal to each other, which might not serve the desired communicating directions.

Rotating one dipole to be parallel to the other requires the 20m dipole to be lengthened slightly to achieve resonance, but dramatically reduces the 20m SWR bandwidth. This design begins to look like a so-called open-sleeve dipole, in which the longer element is fed directly and excites the shorter dipole through close coupling, restoring much of the useful SWR bandwidth while preserving the directivity on both bands:

Beginning on page 21 of Choosing Your First HF Antenna, the ARRL's Joel Hallas, W1ZR, describes Unfolded and Folded Skeleton Sleeve Dipoles for many band pairs. I have used these antennas at home, for DXpeditions and Field Days and can attest to their simple construction and reliable performance. A matching unit may be required to obtain the SWR required by your transmitter across an entire amateur band. A folded skeleton-sleeve dipole for the 40m and 20m bands would look like this:

• Is that built out of window line, or some kind of metallic framework? Jul 2, 2019 at 18:55
• The antenna is built from window line. Jul 2, 2019 at 19:09
• Sweet! Radiation pattern like a dipole, omni with weak end nulls? Jul 2, 2019 at 19:14
• That's right. To take strain off the window line and the connections, I suspend the antenna from low-stretch Dacron cord woven through the windows. Works like a champ. Jul 2, 2019 at 19:27
• Also, I find that too many beads would be needed to be effective on 80m, which would be heavy and relatively expensive. Since I only run a few hundred watts, I make a conventional choke by winding three turns of coax (usually RG8X type) through two type-43 ferrite cores, resulting in over 1k$\Omega$ impedance on 80m. Jul 2, 2019 at 19:32

What you describe is not far off from a common fan dipole. This is a multiband antenna consisting of several dipoles in parallel, each cut to a different length. Only instead of orienting each dipole at right angles, they are simply spread apart with spacers.

The idea is that the resonant dipole for the band will have a low impedance, while the non-resonant dipoles will have a higher impedance. By the basics of parallel impedances, most of the current ends up on the resonant dipole, and the addition of the higher-impedance non-resonant dipoles don't have a significant effect on the overall feedpoint impedance.

The spacers reduce the interaction between the dipoles which makes the arrangement much easier to tune. Rotating the dipoles to be 90 degrees apart would reduce interaction even more, at the expense of doubling the number of supports required.

The feedpoint current will take the path of least resistance (or least impedance) so it will always flow mainly into the center-fed resonant 1/2WL dipole rather than into any other higher impedance path.

Have a look at Logper antennas!

They are not just "joined dipoles for different wavelengths", but calculated to work in combination (otherwise they'd just be a mismatched piece of metal; still an antenna, but not a good one).

Also of interest to you might very well be Vivaldi antennas, but these will be hard to realize in HF:

What you'll sometimes find in both cases is that you duplicate the antenna, rotate it by 90° along the main direction, so that you get a second antenna in the orthogonal polarization plane; that way, you can generate (and receive!) arbitrary linear, circular and elliptic polarizations by phase shifting the input to either polarization.

• I doubt I could build either one of those for HF. The log periodic would be 60-some feet across the longest elements for 40m, right? My whole lot would be covered, and the RF exposure would be an issue if I ever get above 100 W. Jul 2, 2019 at 13:28
• See the section "shortwave broadcast antennas" in the logper article linked above. But yes, that thing will be size-wise in the order of magnitude of the lowest frequency's resonant dipole. Jul 2, 2019 at 13:31
• Whoa! That makes it look as if one could build an 80-10 antenna that could be managed by a reasonable rotor and stay near resonance over that entire band! Jul 2, 2019 at 13:42
• honestly, it still looks pretty involved, yet possible. Jul 2, 2019 at 13:52
• @ZeissIkon Why build when you could buy…? :-P usantennaproducts.com/antennas/models-lp-1005-lp-1001-lp-1002 Jul 2, 2019 at 17:48

I have done it for years, sometimes 2 bands (40-80m) at a 90° angle sometimes even 3 bands (40-80-160m) at 60° angles all 3 in inverted V dipole settings. It takes a bit of tweaking to get the VSWR OK but it definitely works pretty well.