Keypounder, a name a few folks may remember, is writing updates to his articles on NVIS that NC Scout published almost 5 years ago at the Brushbeater site.  He has continued his research and study of NVIS, and wants to update  and expand upon his earlier articles on the subject.  This article is being posted as the third of what looks to be now  at least 4 articles on NVIS.

As NC Scout stated 5 years ago-
“…. I will re-iterate that these skills, along with Land Navigation, are among the most perishable and most difficult to learn- under duress, near impossible. So for those of you who feel you’ll do it when ‘the time comes’, you’ll be sadly mistaken.  Please folks, try this at home.”

Part One of this series on NVIS operation focused primarily on the basics of NVIS; what it is, why it is, how it works, and listed some of the major factors involved in successful NVIS operation, briefly touching on these factors. Link here:

Part Two of this series on NVIS operation looked at HF listening and transmitting techniques, some specific to NVIS. Link here:

Part Three discussed how to decide which HF radio to purchase. Several common civilian amateur radios will be reviewed in some detail, and general characteristics desirable in an NVIS station specifically was discussed. Link here:

Part Four will review some basic NVIS antenna characteristics in detail, and discuss different types of operation and touch on the implications of these differences on antenna selection.  This will be followed by Part 5, which will discuss more advanced NVIS antennas and their uses  in non-permissive environments.


Antennas. The single most important feature of any station; apart from the skill of the operator, the antenna largely governs the effectiveness of the station. Your station may have a great brain (rig and computer) but when it comes to putting the RF where it is wanted, antenna systems do the work. To continue the (imperfect) biological analogy, they are the circulatory, musculature and skeletal systems of your station, controlling the strength and direction of the RF your station sends and receives.

When it comes to NVIS, a great many antennas may be made to serve, but some are better suited than others. Before we get into NVIS antenna selection, let’s review some of the basics of NVIS:

  • Intentionally low horizontal antennas to maximize upwards gain.
  • No requirement for high power.
  • No satellites or repeaters required.
  • Use of skywave propagation on lower HF to provide terrain-independent continuous radio coverage out to perhaps 300 miles away, inside the traditional upper HF skip zone and outside LOS.

The key to effective NVIS operation is Signal to Noise ratio; while different modes have different characteristics the key to maximizing the signal and minimizing the noise for any given mode is proper selection and use of the NVIS antenna. We’ve touched on basic NVIS antennas in Part 1, but let’s get a bit deeper into this topic, probably the most important factor in effective, efficient NVIS operation, especially when operating in non-permissive environments.  We’re going to look at some basic antennas, then examine some more advanced NVIS antenna options.

Before we do, consider your expected or intended operational environment. As a tool for emergency communication in a more or less permissive environment, NVIS can be used to support communication from either fixed locations or temporary locations. We talked in Part 3, rig selection, about HF rigs suited for NVIS operation- both fixed, and temporary or portable operation. When we talk about antennas for NVIS, we are going to distinguish between antennas for operating from temporary locations, and antennas for operating from semi-permanent or fixed locations. The operational imperatives are different, and drive different antenna selections.

A temporary operating position must be portable; all of the equipment must be quickly and easily set up and quickly taken down and carefully stowed for the next use by people who must be assumed to be tired, stressed and in suboptimal circumstance and condition. Speed and ease of setup and TAKEDOWN are requirements, when considering antenna selection for portable NVIS operating. Given the size of an NVIS antenna, this is a non-trivial concern. My goal when setting up portable NVIS wire antennas in trees is to get each hoisting point for that antenna in the tree in 15 minutes from a standing start.

This means, string or cord through a good crotch or over a reasonably stout branch, hoisting rope looped into the tree, antenna laid out and connected to the hoisting rope, transmission line connected to the antenna feedpoint and laid out ready to run to the radio. If there are two hoisting points, this takes 30 minutes; three points, 45 minutes, and for 4 hoisting points I allow an hour, minimum. For NVIS, high lifting points are neither required nor desired, but it all takes time and the more hoisting points required the more time required. If you can transport surplus military masts or an extendable fiberglass mast, you can speed this up. Having a trained helper speeds things up too, and if you have an open area in which to put your antenna, you don’t have to worry about tree branches being where your antenna wants to be. Threading a 160/80/40 fan dipole through 250’ of scrub forest is work;  when considering an operating point it is often better to select a spot inside mature forest, or in between trees across a meadow since there will be less brush and undergrowth to compete with.

On the other hand, at a fixed location, a semi-permanent or permanent station, speed of setup is not the primary concern; antennas which have more sharply defined vertical gain patterns, especially those which minimize vertically polarized RF are more of a priority. We will get a lot further into this shortly.

You will need to balance the speed of deployment with reduction in Probability Of Interception and especially DF. If you are part of a small patrol expecting to use NVIS for reporting in from an extended patrol in a non-permissive environment, the less time used and the lower the profile of your radio communication efforts, the better. On the other hand, if you are erecting an NVIS antenna array for a fixed location, time may not be a primary concern. Maximizing your vertical gain, reducing your transmit power, and reducing your vertically polarized RF to the greatest extent possible may be more important. We are going to take a more in-depth look at the basic NVIS antennas, starting with the easiest to erect and moving to the more complex and slower types.

To that point, the easiest antenna to set up and recover is the inverted vee:

  • Single hoisting point;
  • 2 low height and low stress  attachments, one at each end.

You can use a variety of dipoles as inverted vees.  The  Cross dipole,  with one central hoisting point and multiple dipoles cut for different bands at different azimuth angles reduces the interaction between different dipoles, improving bandwidth somewhat.  You also can deploy a Fan dipole (multiple dipoles more or less along the same axis) as an inverted Vee, which has two low height attachments.  If your operation requires instantaneous availability of transmission on multiple NVIS bands, then either of these are good basic choices.  You can receive 80 on a 40 meter antenna, and vice versa, but if your SOI requires rapid shifts in band for transmissions, then the multiband resonant antenna is the preferred option.

Both the cross and the fan can be something of a challenge for man-portable operation and take more time to deploy,  You will have to balance the added time to deploy and recover the antenna against the operational requirement for instantaneous multiband use.  For portable HF operation, especially man-portable operation, including but not limited to NVIS, the linked dipole is a good choice if you are only going to operate on one band per session or if you can take a few minutes to switch bands between contacts.  With a linked dipole, bandswitching requires that you lower the antenna and disconnect the parts of the antenna you are not going to use, but it is a single wire and weighs less than a fan dipole. Mine is made from 18 gage polystealth, and I cut it to allow operation on 80, 40, 30, 20, and 17 meters. I can add extensions to allow 160 operation, as well.

Let’s take a moment and go back to the height of the F2 layer. Remember this graph?


This is a printout of the height of the F2 layer as measured at Wallops Island on May 9th 2021.

The average height of the F layer during daylight drops from over 300 km down to between 200 and 250 km, then rises again as evening comes on and the illumination from the Sun decreases. What this graph does not show is the attenuation from the D layer, which rises from ~0 at night to a max at local noon.

Keep the above graph in mind while we look at a printout of what the broadside radiation pattern of an 80 meter inverted vee with the feedpoint at 25’ AGL over average ground looks like:[pdf-embedder url=””]


  • This pattern is broadside to the axis of the antenna;
  • The gain is about 2.5 dBi;
  • The 3 db points are ~50 degrees down from the peak or 40 degrees up from level.
  • The -10dB point on this graph is more than 70 degrees down from the peak!

This is a pretty broad pattern so your daytime NVIS signal will not be limited by your antenna pattern. What will limit your normal daytime communication is the attenuation of your signal by the D layer.  This is important, bearing both on both your ability to communicate and your vulnerability to DF, so let’s touch on this a bit more.

Normally, the D layer is about 40 km thick, but that varies depending on the sunspot cycle. The density of the D layer also varies depending on the intensity of solar radiation, both visible and ionizing (soft and hard X-rays, among other things.) Normally, D layer attenuation varies more or less predictably, with attenuation low at night, when only cosmic rays penetrate to and ionize the D layer. Because the D layer is low in the ionosphere, with a relatively high atmospheric pressure compared to the rest of the ionosphere, ions don’t last long, unlike the F layer. Once visible light reaches the D layer, however, it immediately starts to ionize, reaching a peak at local solar noon and dropping off again; by an hour after sunset the effect of the D layer is gone in the immediate area. Westward, where it is still exposed to sunlight, it still has effect.

As previously discussed, D layer signal attenuation is affected by both the distance the signal travels through the D layer and by the wavelength of the signal. The longer the path through the D layer, the greater the loss; this is a direct linear relationship. Twice the path length means twice the attenuation. D layer attenuation for different wavelengths, however, goes as approximately the square of the wavelength.

To go back to our Inverted Vee broadside pattern above, our 80 meter signal is down 3 db at 40 degrees above the horizon. Let’s say that we are operating NVIS right at local noon, here in 2021. Our vertical attenuation from the D layer is about 28 dB, but because of the slope and the increased distance the signal travels through the D layer at a 40 degree signal path the actual D layer attenuation is about 44 dB. If we do the math, we can reach about 320 miles at that angle, on 80 meters. If our signal was S9 at a station just over the mountain a few miles away, it would be down 19 dB or about an S6, still readable if the noise floor was below the signal.

If we want to work a station a bit farther away, say 470 miles or 780 km away, our signal departs at about 30 degrees, and the pattern says we’re down about 5 or 6 dB in radiated signal. D layer path losses are going to be greater at the lower angle- 1/(sin(30))=2; 2 times the 28dB of a vertical path, or 56 dB or another 28 db above the base loss. 28 plus 6 db is 34 dB or about 6 S units. We have gone from an S9 signal to an S3 signal, which may or may not be readable above the noise floor.

This loss gets progressively worse as the angle drops- when you get to 20 degrees, the antenna pattern is down 10 dB and the loss from the D layer would be another 54 dB, taking your S9 signal down to -10dB, even neglecting the effects of the E layer, which start to become an issue at low angles during solar max. If you were getting S9 with 5 watts CW or digital, you’d have to boost your power to the legal limit to have a chance of getting a readable signal over S3 noise.

This discussion about the D layer attenuation so far has just been about 80 meters; as noted previously, attenuation goes as the square of the wavelength, so base losses for 160 (double the wavelength) will be about 4x the power loss or about 6 db more loss, for a total of about 34 dB, about an S unit greater than on 80. The power loss would be about 6 dB less on 40 meters or around 22dB.

What this means is that if you needed to use 20 watts for reliable communications on 80 over a given path, you’d expect to need 80 watts on 160 and 5 watts on 40 meters for the same path, assuming that the foF2 supported operations on these three bands and that noise levels were the same on all three bands. They aren’t; we’ll peel another layer on this in a bit.

For emergency operation in a permissive environment, this is a good thing; if you are operating where you need to be concerned about DF, in a nonpermissive environment, not so much. Of particular concern is the vertically polarized radiation, since this will propagate by ground wave.  Any dipole emits vertically polarized RF off the ends of the dipole, but an inverted Vee radiates more low-angle vertically polarized RF than a flat-top or a Vee, as we will see which is one reason the Inverted Vee is used as a low band DX antenna. Here is the elevation pattern for the same inverted Vee off the ends of the antenna–

[pdf-embedder url=””]

If you zoom in and read the fine print, you will note that this is a shot taken at about 10 degrees off the end of the antenna, or about 80 degrees from broadside. I did this to show you that almost all of the radiation is vertically polarized here, shown by the red trace almost as high as the total field line. The pattern, while slightly less wide than the broadside pattern, is still wide.

As discussed, the primary concern for interception and DF is vertically polarized low angle RF, which propagates by ground wave. Across smooth terrain, low frequency ground wave can propagate for many miles; it is attenuated by vegetation, by abrupt shifts in topography, and by changes in soil type as well as distance. Here is an azimuth or overhead plot of the radiation pattern from the same antenna, showing the radiation pattern at 6 degrees elevation. This will give you a reasonable approximation of the relative magnitude of the groundwave vertical radiation pattern to be expected from this type of antenna.

[pdf-embedder url=””]

A few important things to note:

  • At 6 degrees elevation, the vertical RF off the ends of the inverted Vee is 3 dB stronger than the horizontally polarized RF off the broadside, TWICE the power;
  • The signal is still strong even at these low angles, only -12.4 dBi.
  • If you were running 100 watts output into this antenna you’d still get 6 watts of ground wave ERP off the ends of your antenna, which is a significant amount of power, propagating on level ground for a number of miles, depending on ground type.
  • Radiating in the broadside direction, the RF is both half the power and unlikely to propagate via ground wave. This low angle section shows a very different profile from the high angle profile.

In a permissive environment, that vertical RF is is not a big issue although it can contribute to fading and loss of signal for close in stations, but in a non-permissive environment, this could be very important. We’ll get more into this.

We can summarize the Inverted Vee, whether fan, crossed or linked configuration, as follows:


  • Relatively easy to deploy and recover, using only one hoisting point, and less impacted by low branches and undergrowth;
  • Impedance can be adjusted by raising or lowering the ends of the antenna, or by raising or lowering the feedpoint as well.
  • Broad pattern, allowing good continuous coverage out to 400+ miles on 80, perhaps more;


  • Emits significant vertically polarized radiation at low angles off the ends of the antenna;
  • Requires a feedline;
  • Lower gain than a flat top dipole (we’ll see that shortly);

For operation in permissive environments at either fixed or portable stations, the Inverted Vee is a common choice. For operation from permanent or semi-permanent locations in non-permissive environments, it carries some risks.


Now we’ll peel the onion on the Vee antenna.  The Vee antenna requires two lifting points, one on each end; the feedpoint in the middle is typically not supported and is kept just high enough to keep the wire out of reach, say about 10’ high. This makes the feedline required quite a bit shorter, perhaps 15 or 20 feet at most, and it is possible to directly connect the radio to the Vee antenna with a cobra head adaptor if the feedpoint is lowered. The Vee has about the same gain as the inverted vee, or maybe a trifle more. Although there are two lifting points to be set, in most areas with substantial trees, since the required height is low the location of the lifting points is not critical as long as the hoisting rope is long enough and there is little intervening brush.

The broadside pattern of the Vee is very similar to the inverted vee, but the endfire pattern is different, as shown here:

[pdf-embedder url=””]

Note that the magnitude of the vertical radiation off the ends is notably lower than the Inverted vee.  If we look at the azimuth view taken at 6 degrees, you can see that not only are the low angle broadside and endfire emissions about the same, but they are notably lower than the Inverted vee at -18 dB max or about 3 dB lower than the Inverted Vee.  This is still enough RF to be a concern, but less is better.

[pdf-embedder url=””]

The Vee takes more time to put up but does reduce the vertical Rf off the ends of the antenna by a modest amount, about half.  What the model does not show is that the Vee, because it has less gain for low angle RF, is less likely to pick up local noise which is commonly vertically polarized, and can improve your Signal to Noise ratio slightly.

We can summarize the Vee, whether fan, crossed or linked configuration, as follows:


  • Impedance can be adjusted by raising or lowering the feedpoint of the antenna;
  • Broad pattern for broadside RF, allowing good continuous coverage out to 400+ miles on 80, perhaps more;
  • Shorter feedline and allows direct connection of the antenna to the rig if a cobra head is used;
  • Less vertical radiation off te ends than the Inverted Vee.
  • Less receive noise than the Inverted Vee.


  • Takes ~double the time to erect as the Inverted Vee, and is more susceptible to interference from brush and low branches;
  • Again, as with the Inverted Vee, lower gain than a flat top dipole.

For operation in permissive environments at either fixed or portable stations, the Vee is a good choice, offering modest improvements in S/N ratio, and shorter feedlines, reducing potential weight for man-portable ops.  While slower than the Vee to set up and take down, it is better than the inverted Vee for locations in non-permissive environments, but still carries some risks.

Now we get to the more traditional “flat top” dipole, a staple of radio operators for the past hundred years at least.

For antennas cut for  lower HF and upper MF, the resonant flat top dipole requires three lifting points, one at each end and one in the middle because of the length of the antenna, and the resulting sag of the wire and depending on the feedline used. On 40 meters if you are using RG8 type transmission lines, then the center support will be required to make the antenna effectively flat; if you are using 14 gage copperweld, feeding with ladder line, and have sturdy supports, you may get by with supporting the ends only, but note that it does not take much deflection to start to lose gain. This takes more time, as previously discussed, and also requires more coordination to find three hoisting points more or less in line, with no obstructions up to about 25 feet AGL anywhere along about a 300’ path, if you are using a 160/80/40 fan, or about 150’ if you are using an 80 meter antenna.

Here are both the Broadside and off Ends patterns:

[pdf-embedder url=””] [pdf-embedder url=””]

You can see that the FT dipole gives about another 2 1/2 dB of vertical gain, a modest difference, but in situations where transmit power is at a premium and every watt counts, the flat-top does give you the most gain of the single dipole family. In practical terms, it means you can reduce your transmit power by half and still have the same readability as if you were using an Inverted Vee.  Here is the 6 degree elevation azimuth view:

[pdf-embedder url=””]

Note that the flat-top and the Vee have about the same amount of vertical radiation off the ends of the dipole, in absolute terms, while the flat-top gives more overhead gain and also more horizontally polarized RF output at low angles, a possible advantage for longer haul night-time operation.  If you have the time and the supports available it is a good option, possibly a good option for a longer term operating post.

We can summarize the Flat Top dipole, whether fan, crossed or linked configuration, as follows:


  • has about 2.5 dB of added gain over the Inverted vee and the Vee, almost doubling the ERP;
  • As with all low horizontally polarized dipoles, it has a broad pattern of broadside RF, allowing good continuous daytime coverage out to 400+ miles on 80, and more at night;
  • Similar to the Vee, it radiates less vertical radiation off the ends than the inverted Vee.


  • Takes ~3 to 4 times as long to erect as the Inverted Vee, and is the most susceptible to interference from brush and low branches;
  • Requires use of a feedline;
  • Will receive more horizontally polarized noise at low angles than the Vee;  during summer operations the flat-top may be noisier at night than the Vee.

For operation in permissive environments at fixed stations, the Flat-Top is a reasonable choice, offering minor improvements in gain at the expense of more effort.  It is noticeably slower than either the Inverted vee or the Vee to set up and take down, but if the antenna is not relocated frequently this is a minor issue.  In non permissive environments it carries some risks but the increase in overhead gain means that transmit power can be cut almost in half to decrease vertically polarized radiation.


I’m going to digress a moment and talk about another option to get multiband operation, the tuned doublet. This antenna is a center fed antenna, but uses ladder or window line at the feedpoint and requires a balanced feed tuner. For 160 and 80 meter NVIS operation, each leg should be identical in length and about 90’ long, 180’ overall, so it is somewhat shorter than a resonant 160 antenna; an 80 and 40 meter version would be about half that, or around 45 feet for each leg, 90 feet overall. Window line is lighter than RG8X and does not require a balun, so it is possible to erect this antenna with only two lifting points and still get a more or less flat antenna; you can also configure it in a Vee or inverted Vee, just as with the resonant dipole. It does require a tuner, and window line, while light, can be bulky, and while it is faster to adjust the tuner than to lower a linked dipole, it is still not suitable for instantaneous cross band operation. The patterns for the balanced tuned doublet are similar to those of the resonant dipole, but a couple of dB down.  This option is listed primarily for those operators wanting both 160 and 80, or 80 and 40 meter NVIS in the lightest and easiest to deploy antenna package possible.


So, we’ve delved fairly deeply into the specific characteristics of the basic NVIS antennas, and gained some insight into the strengths and weaknesses of each of these three antennas all of which have been used for decades by knowledgeable radio operators for military and civilian NVIS operation, as far back as World War Two, and all over the world since then. They work. Most amateurs have no concerns about radiolocation or direction finding as it applies to NVIS, and until recently, neither did the US military.

Given the ongoing improvements in computer technology, the advent of widespread use of drones for both intelligence gathering and delivery of high-explosive ordance, and the integration of sigint into computerized artillery counterbattery systems, this is changing. Even in a down-grid scenario, it is a bad idea to assume that “Bubba” cannot locate your transmissions. Both the probability of and the adverse consequences of having your transmissions being located are drastically increasing. It therefore behooves the intelligent radio operator to concern himself with minimizing his HF NVIS signature. The primary means for achieving this goal is improved NVIS antennas.

In our next installment, we will discuss the following types of improved NVIS antennas, again, organized by ease of setup:

  • Circularly polarized antennas (less fade)
  • Dual dipole arrays
    • In-line antenna and patterns
    • Shirley and Jamaica antennas and patterns
  • Loops and arrays of loops
  • NVIS-specific Listening antennas
    • DOG- Dipole on the Ground
    • LOG Loop on the Ground

Remember, folks, this information is not academic, and if you are serious about learning how to communicate in emergencies and in non-permissive environments, you need to try this out, practice it and learn how it works.
Hope to hear you on the air,