Copyright © 2002-2011 by Harold Melton, KV5R. All Rights Reserved. Feel Free to Link to This Page.
How to Make a Good NVIS Antenna
Shop for ready-made Dipole Antennas here.
The best NVIS antenna is one which is simple and effective. One favorite is the dual-band dipole. This antenna uses two dipoles, one for 75 meters (about 122 feet), and one for 40 meters (about 65 feet), both connected directly to 50-ohm coax and supported at 5 points by trees at 10-12 feet. The two dipoles should be well separated at the ends, or they will interact. They may be strung up in an “X” or a “+” shape. The bandwidth of the 75-meter dipole will be quite narrow (<100kc), so it will benefit from using two sets of stagger-tuned wires. Some researchers recommend that the ends of the wires should be a few feet higher than the middle. This will increase gain and raise the feedpoint impedance a bit. If the feedpoint impedance is too low to match, the antenna should be a folded dipole, which will raise the feedpoint impedance by a factor of four. Stringing the antenna over a highly conductive surface, such as salt water or a wet, acidic marsh, will substantially improve the antenna’s performance, compared to stringing it over dry rock or sand.
Since the support points are typically 10-12 feet high, the wires must be both light and pulled tight to remove annoying sag. Appropriate wire ranges from #17 aluminum electric fence wire, ($14 for ¼-mile at farm-supply stores), to #14 insulated stranded copper THHN, ($15 for 500 feet rolls at electrical suppliers, and available in green). The #17 aluminum isn’t very strong, but is almost invisible. The wire may be supported with green nylon string, available at garden centers. The center feedpoint and coax may be built around a simple insulator, waterproofed, and nailed to a tree trunk at 10 feet. Insulators and coax may be sprayed dark green or brown as needed. Antennas below 8 feet should use insulated wire to avoid RF burns. Insulation does not affect the performance of antenna wire, except (1) reduced wind and rain static, (2) lowers the velocity factor a tiny bit, and (3) prevents corrosion.
It is better to use a broadband current balun at the feedpoint when using coax. A simple choke balun made of coiled coax may be used if needed to remove common-mode currents from the line. Try to design the installation so the feedline extends away from the antenna at a 90-degree angle, for at least one-quarter wave. Also, the line should be detuned — that is, it’s length should fall between resonance points. If these are done, feedline RF pickup and re-radiation will be minimized, and a balun should not be needed. Detuning the feedline is also the cure for “RF in the shack” problems. Suitable lengths will depend on how the antenna is fed and whether one side is grounded or not. See the Antenna Book for determining appropriate lengths.
For resonant dipoles, avoid using twin-lead or ladder-line — the feedpoint of these low dipoles will be well under 50 ohms and attaching 300-600 ohm parallel feedline will present a severe mismatch at the feedpoint. However, if the antenna is to be used nonresonant, with a tuner, ladder-line should be used because coax is very lossy when operated at high VSWR.
There is a long-standing myth that dipoles must be resonant to be efficient. Non-resonant dipoles of similar size are just as efficient as resonant dipoles, assuming that (1) impedance mismatches are matched, (2) the matching devices are designed so that losses are insignificant, and (3) feedline losses are minimized (use ladder-line when the SWR is high). It is also important to remember that baluns and matching transformers are quite lossy when operated with a mismatch on either or both ends. The ARRL Antenna Book shows how to make baluns for any ratio of impedance transformation. The myth come from the fact of severe losses in mismatched coaxial line. In the author’s experience, a 160-meter dipole fed with ladder-line will outperform a 75-meter resonant dipole fed with coax, both at the same height, and both operated on 75 meters. This is because the larger antenna, even though not resonant on 75, has an “aperture” twice as large as the smaller one and thus captures, and radiates, more signal. However, it does have 4 partial nulls, while the half-wave dipole has only two.
To connect aluminum or steel wire to copper, make a couple of short #14 solid copper pigtails, twist them tightly into the aluminum or steel elements at the feedpoint, then solder an SO-239, or direct coax feed, to the copper tails. Waterproof the dissimilar metals connections with waterproof grease and Coax-Seal, or silicone caulk. If any moisture gets into the connection, the metals will corrode one another and make a nasty rectification point. Mechanical connectors (split-bolts or set-screw lugs) may be used but they also should be waterproofed.
Fancier (more expensive) NVIS installations include full-wave loops with automatic antenna tuners at the feedpoint. These antennas, if installed at 15-20 feet or more, will provide both excellent NVIS performance on the low bands and DX on the higher bands, where the height of the loop is over ½-wave. However, the pattern of the antenna will have several peaks and nulls on frequencies where it is several waves long.
Two things about loops are worth mentioning: (1) Loops are resonant on every harmonic, not just odd harmonics like dipoles, and (2) the lower the frequency (greater the length) of the loop, the more harmonic points it will have. For example, an 75-meter loop will resonate at about 3.8, 7.6, 11.4, 15.2, etc. But a 160-meter loop will resonate at about 1.8, 3.6, 5.4, 7.2, 9.0, 10.8, 12.6, 14.4, etc. — and the peak SWR arising from imbalanced reactances will be lower between all these points. Therefore, a big loop should be strung up, even if it cannot be used on its fundamental frequency because of low feedpoint impedance.
Carrying this idea further, an operator with acreage might run a really big loop (like 1100 - 2200 feet) atop a perimeter fence and it would have so many resonant points as to be useful as a broadband antenna — although the fundamental and all harmonics below about 3-4 MHz might be unusable for transmitting due to extremely low feedpoint impedance, unless feedpoint matching is used.
Another good antenna is the 3-wire folded dipole. This design may be used on all HF bands, with a tuner. The rules are pretty simple: Make a 2- or 3-wire folded dipole as long as possible (preferably 260 feet). Feed it directly with ladder-line, and match it to the radio with a balanced line tuner. Use a 1:1 current balun at the tuner’s input. The reasons: (1) Feedpoint resistance of low antennas will be very low, typically 15 ohms or so, and the 3-wire folded dipole will raise it by a factor of 9. (2) ladder-line does not suffer any significant loss when operated at high line SWR (unlike coax). (3) Balanced tuners with the balun on the input (the matched side) are considerably more efficient than unbalanced tuners with the balun on the output, because baluns are only efficient when both ends are matched.
Some emergency groups are successfully experimenting with mobile antennas mounted horizontally. For example, pairs of 75 and 40 meter Hamsticks make excellent shortened, portable NVIS dipoles. The mobile antennas are mounted back-to-back and fed in the center just like a dipole. These are oriented horizontally and placed a couple feet above the roof of a vehicle on a short mast.
Other operators carry (1) an autocoupler, (2) a 125′ roll of wire, and (3) traffic cones or fiberglass stakes in the trunk, for rapid roadside NVIS deployment. NVIS antennas have been used as low as 18 inches high. Surprisingly, S9 signals have been received from an antenna mounted 10-½ inches high.
I’m having trouble visualizing the hamstick configuration that you describe. Do you have a pic of this setup, for reference?
Like it says, 2 hamsticks back-to-back fed like a dipole. Here’s a bracket to do that – https://mfjenterprises.com/products/mfj-347
Note that any HF antenna shortened to mobile size is terribly inefficient, particularly on the low bands, but making a dipole with 2 hamsticks is a little better than 1 hamstick worked against the vehicle body; and being horizontal, works much better for NVIS, where you want max signal upwards, whereas a vertical mobile antenna has a null upwards.
Note that nowadays most mobile NVIS are using loops – see https://duckduckgo.com/?q=mobile+nvis+loop+antenna
I’m a brand new Ham, Technician, General next month. Researching antenna’s and found this article very helpful.
Newbie question. Why couldn’t the dipole be run over a tile roof? THHN stranded can be bought w/ orange jacket so it would be camouflaged? Is there a grounding issue? Lots aren’t that big in Arizona
Shouldn’t be a problem long as it ain’t metal.
Yes, be careful about what’s UNDER the tile (or whatever) roof: AC, land-line, and cable-TV lines often run from one end to the other of a building.
In newer buildings, the insulation usually has some sort of metallic reflective barrier: here in Arizona this is practically mandatory to get decent heat-rejection values. Of course, for NVIS this may turn out to be very desirable assuming the antenna is strung outside.
[BTW, all these comments are being written on a horrible keyboard, so sorry for all the typos, etc]
Outstanding article. I am a new ham with a General and intend to mostly work
regional on voice and some DX work on CW in the HF band. The article packed
in an excellent mix of antenna theory and practical considerations that I had not
seen anywhere else — and I have been looking for good information of a month or so.
Thank you for such a useful and enduring article.