Admin Note: This is the third and final section of the article submitted by Keypounder. EZNEC, the preferred antenna modelling method, is referenced. I highly suggest familiarizing yourself with EZNEC for building and modelling antennas for both efficiency and directivity. Included are several practical examples of antennas in the field, which I might add, are similar to those in use by myself pictured in previous posts in this blog. This is not by accident, it’s what correct looks like.
Antennas and Antenna height
We touched on antennas and antenna height earlier, but at this point let’s peel another layer off the onion and get into this a bit. An 80 meter dipole cut for the center of the band, 3750 khz, and elevated about 25 feet high over average ground (more on ground shortly,) has a radiation pattern that looks like this:
This slice is right across the broadside of the antenna, perpendicular to the wire. You will note that this dipole emits mostly up which is good for NVIS, and also that it has about 5 db of gain, (almost an S unit on your receiver,) over an isotropic antenna. Although you can’t see this, all of this energy is horizontally polarized, but if you took a slice along the wire, you would find that the energy radiated off the end of a dipole is vertically polarized, about which more later. (Yes, a horizontal antenna can emit vertically polarized radiation.) Also note that the half-power gain is 50 degrees from the vertical; this antenna will easily cover a 600 km radius or more, depending on the height of the F layer.
Here is a full sized 80 meter vertical, cut for the same frequency, compared with the dipole we just looked at; the scales are identical.
Notice that there is a null in the vertical antenna’s pattern directly overhead. The vertical antenna is vertically polarized in all directions, so the ground wave will also be propagated outward in all directions. I hope that you can see why the vertical antenna is not well suited to support NVIS, and we’ll move on.
Now, here is a comparison between a 25′ high dipole, and a 5′ high dipole, again, over average ground. Note that the signal is about 10db weaker, so if you were S9 at 25′ high you would only have been ~S7 with the antenna 5′ high.
The reason that the signal is weaker is that more of the RF is being absorbed by the ground, and not bouncing off it and up to the ionosphere. Weaker is not bad as long as you can be heard by the station you want to work, and are NOT heard by the stations you do not want to hear you. To that point, look at how much stronger the higher dipole is at 45 degrees from the vertical. Also, you want to maximize absorption of vertical polarization to reduce the likelihood that anyone can pick up your ground wave. A lot depends on what sort of ground you have; conductive ground helps propagate ground wave and reflect signals for NVIS, while poorly conducting ground absorbs more RF.
Ground absorption and RF surface
At a fixed location, it is probably worth testing to see what the characteristics of the ground at your location really are, (Rudy Severns has published a very handy tool for doing this, see link at http://www.antennasbyn6lf.com/2015/11/determination-of-soil-characteristics-using-a-low-dipole.html or look at the new Antenna Book) but for portable operations in the context of a grid-down event, this will likely be impractical; your transmission location will be driven by other criteria than whether or not the earth is appropriate for radio communications! Another point to remember is that the RF reflecting surface for the RF coming off your antenna may be many feet below the physical surface of the ground, and lower frequencies penetrate more deeply. Also, the ground is not homogenous; ground characteristics can change depending on both location and depth, sometimes in just a few feet.
Salty ground, such as that in a saltwater swamp near the ocean, or near a salt lake or river, is very conductive, and will bounce your signal nicely. Wet organic clays are almost as good. Dry rocks and sand are very poor conductors, and will attenuate signals more than other soils. Most soils are somewhere in between these two extremes, but there can be considerable variation in soil types especially in mountainous areas, where layers of rock have decomposed to form widely different layers of soil.
Bottom line: don’t be surprised if your NVIS installation at your location works much better than you had expected, or worse than you had expected, and be prepared to raise or lower the antenna to compensate for differing ground conditions. Here is one last graph, showing a horizontal dipole antenna 10′ up over excellent ground.
This graph is using the same scale as the other antennas. Although this antenna is only 10′ up, it almost matches the performance of the 25′ high antenna over average ground. Vary the height above ground to compensate for differing ground conditions; raise the antenna if your signals are weak, and lower it if you are getting good solid signal reports and are able to hear well. You can also install reflecting wires on the top of poor ground and this will improve RF reflection. It will also change the impedance of your antenna, sometimes significantly, and it will add to the time required to erect or strike your antenna. May be worth doing for a fixed station location, ( I use them) but reflectors may not worth the effort for portable ops. Try it and decide for yourself.
Simple NVIS antennas
Simple and easy to deploy antennas include:
inverted vee dipole;
Horizontal low height dipole;
Compact dipole using shortened loaded elements;
Dipole on the ground (DOG)
Here is a picture of my 160/80 fan inverted vee dipole.
Not much to look at, is it? In fact, it is not at all easy to see, which was the point; I used 14 gage copper weld for the 1.90 mHz dipole, which has turned a nice greenish brown, and 14 gage brown insulated THHN house wire for the two 80 meter legs. I managed to get the choke balun in front of the tree trunk in this picture, which is that little white blob. It is MUCH higher than a good NVIS antenna needs to be, and as a result it is also much noisier! The peak is about 60′ high and the two ends are about 15 feet off the ground. This antenna works well for long haul communication on the low bands, but would not be my first choice for NVIS in a grid-down situation, for precisely that reason. If I were to use this antenna for NVIS, I’d drop it to about 25 or 30 feet at the peak and keep the ends just high enough to keep them out of reach of children. This would be easy to do.
Here is a photo of the feedpoint of an NVIS antenna designed for 160, 80 and 40 meters.
This is a fan dipole- you can see one of the spreaders for one side above and to the left of the gray rectangular balun. Again, it is pretty hard to see, but you can see the coax feeder running up to the feedpoint (rectangular gray box with stainless hardware on the sides.) The feed is about 25′ up arranged in a Vee configuration with the feed point lower than the two ends, and with discontinuous terrain at both ends of the antenna to limit ground wave propagation; there is a 200′ high cliff about 300′ off one end, and a steep 150′ high hill with a 400′ cliff on the other side dropping down to a river. The ground is poor and there is a reflector wire underneath to enhance horizontally polarized NVIS signals. This antenna does exceptionally well for NVIS, and being placed in a valley with terrain discontinuities at both ends, and heavy growth all around it is much less noisy than the inverted Vee shown above.
Here is a photo of a low dipole that can be used for 40 or 80 meter NVIS; it is a cross dipole and can have both 40 and 80 available at the same time.
In the above picture it is deployed as a horizontal dipole, with the 80 meter legs extended, and the 40 meter legs left coiled up. The feedpoint is suspended from a low tree branch at 6′ elevation at the edge of a clearing, with the insulated brown THHN wire legs draped over convenient shrubs and branches. but it could easily be elevated either by hoisting from a tree or by using a mast system to raise it up. GI surplus masts are relatively light, and if you have the ability to transport them can be used to elevate a compact dipole antenna to 25 feet or so for NVIS. Once elevated, both the 80 and 40 meter dipoles could be tied off at 90 degree angles to each other, and used for 80 meter NVIS and 40 meter intermediate range skywave skip communication. I have successfully made CW contacts on 80 meters using 2 ½ watts to stations 100 to 200 miles away using the antenna pictured above and an FT 817.
A few things to note about the antennas pictured above:
All of the antennas above use choke baluns to ensure that the radiation pattern of the dipole is not distorted by RF on the outside of the transmission line being re-radiated. My experience is that choke baluns do make a difference, but you will have to judge whether or not the added weight is worth a couple db of signal. All of the baluns shown are rated for a kilowatt of RF or more.
All of the antennas above use brown insulated stranded 14 gage THHN, at least in part where possible and when stretch is not a concern. It is cheap, readily available, hard to see and easy to use. It is heavier than bare wire, but the insulation allows one to thread these antennas through vegetation without changing the resonance of the antenna, which bare wire in contact with moist vegetation will do.
Where strength is required, I use 30% or 40% copperweld antenna wire, at least 14 gage. For longer antennas, those more subject to wind and icing, or those permanent antennas more subject to tree sway stress, I use 12 gage copperweld. It may be overkill, but so far I have not lost an antenna due to wire breakage. Rope, yes, copperweld, no.
If I were to build an antenna specifically for backpacking and use in an off-grid scenario, I’d make it from 18 gage polystealth wire, I would order smaller toroids suitable for lower power level baluns, and I’d use a smaller enclosure and lighter hardware. When backpacking, “ounces is pounds!”
Here are some pictures of my compact dipole setup I mentioned; the dipole is made up of Hustler masts and resonators with a center bracket with 2ea 3/8×24 threaded fittings and a U bolt to hold the bracket on to a mast. One of the fittings takes a standard VHF coax fitting and the other is a simple coupling nut. The Hustler masts screw in to these two fittings and the two resonators of choice (same band!) screw onto the Hustler mast. Like most compact antennas, this has a narrow bandwidth, but it works very well, is very quick to put up and is handy when there are no tall trees. Here are the components, have fiberglass and aluminum surplus mast sections in the rubberized bag, and the Hustler masts in the tan cloth bag. The tripod fitting I bought from GoVerticalUSA.com. As you can see from the photographs, I have no shortage of trees, but the tripod is extremely useful; I have used it to put up satellite antennas, VHF yagis, cross dipoles, and for a number of other uses.
Here is the base of the tripod assembled with the top mast section (non-conductive fiberglass) attached and the dipole assembled on the fiberglass mast section; the coax is attached to the bracket.
If needs be, I can attach a length of 550 cord to the top of the mast and support the Hustler resonators to help prevent sagging, but for brief use in calm conditions, I have not found this to be needed.
Now, here is the whole assembly with 5 surplus mast sections through the tripod. It takes a practiced person less than 10 minutes from a standing start with all the components at hand to get this antenna up and on the air.
No description of NVIS antennas would be complete without mentioning the lazy man’s special, the dipole on the ground. If I were stranded in the desert, and needed to build and deploy an NVIS antenna, I’d cut a doublet antenna of insulated wire, lay it out straight on the top of the nearest sand dune and it would be “comms up!” For extremely poor soil (like sand dunes) where there are no trees or masts available it is a possible option.
Most people don’t have a dry sand dune to use as an antenna support, though, and loss increases rapidly in most soils as you approach the ground surface, so it is worthwhile placing the NVIS antenna a foot or two off the ground by any means available if you don’t have trees. Things I have used in the past to get antennas a couple of feet up; fiberglass electric fence stakes, sticks and dry tree branches, shocks of grass gathered into a sheaf and tied into a cone, and brush and shrubs among other things. Basically any non-conductive material that can get the antenna a foot or so off the ground would be preferable to the DOG in terms of signal strength, but DOGs do work if conditions are favorable and more power is available.
Now, if you have time and resources, there are more complicated antennas for improving NVIS signals. I won’t go into them in any detail because they are a LOT more effort, they are frequency specific, and in my opinion not worth the effort under present circumstances. However, the day may come where added gain and maximum reduction in POI is worth the investment of time and resources.
There are two variants of the venerable Lazy H that have been used by the British Empire extensively; one is a folded dipole varient of the Lazy H called the Shirley, and the other is the Jamaica. Basically, both of these antennas are two dipoles at the same height above ground, about 25 to 30 feet up, about ½ wavelength apart, and fed in phase in the center of each dipole. They are reputed to be VERY effective, and modeling shows that they provide at least 3 db of added gain to the vertical signal.
The other is the horizontal loop, which is also reputed to be effective. I cannot comment on any of these more complicated antennas from personal experience (yet!) but they are of interest, and literature and EZNEC indicate that they do provide advantages. If you do try any of these I would be interested in hearing about your results, especially any “A to B” comparisons between these antennas and a reference antenna.
Geography, geology and your signal.
I live on the west slope of a mountain, about halfway up from the valley. The slope is steep, averaging well over 10%, and the top of the mountain creates a barrier to incoming radio waves, especially low angle radiation on the upper HF bands. I have made one contact with Africa in the past two years of operating from this location. However, to the west, I have been able to make many contacts using 100 watts or less into Australia, New Zealand and other far off places, because the long slope down to the west reduces the takeoff angle by about the amount of the slope. While this has nothing directly to do with NVIS, the point is that the configuration of the area around you can have a significant effect on your signal.
In particular, locating your NVIS transmitting antenna on a long slope will change how the antenna radiates RF, and will shift the direction of the peak gain in the direction of the downward slope. A westerly slope will increase the signal strength slightly to the west. In my case, the effects are minimal because the area to the east of me is relatively small, so the fact that I have lost about 50-100 miles of distance in that direction makes little difference to me. If I were in the central US, however, it might become significant. Something to keep in mind.
If my station were located on the top of the mountain, and the peak was a sharp ridge, I would see much less signal propagated upwards, even with excellent soil underneath, and few mountains have nice deep conductive soil beds. For long haul communications, being high up on a mountain top can be good, but for NVIS it is not optimal.
Valleys on the other hand can be very good places to locate an antenna. Antenna theory says that the Fresnel zone, where the RF radiation pattern is formed, lies within about a wavelength of the antenna, so a deep valley about 500′ to 1000′ wide and about 250′ to 500′ deep would be ideal for 160 or 80 meter NVIS. Not only will the valley slopes enhance NVIS by tending to reflect signals upward, but the soil in the floor of the valley is typically more conductive and a better reflector. Such a geologic structure would enhance vertical signals and reduce unwanted vertically polarized sidelobes, especially if the ends of the antenna were pointed at the steepest and highest walls of the valley. Vertical cliffs, up or down, are better still for attenuating vertically polarized ground wave RF. Fresh water rivers and lakes, being sharp discontinuities also act to attenuate ground waves; any sharp change in ground characteristics or angle attenuates ground wave.
Vegetation is a great absorber of RF, more so at higher frequencies than at lower HF. Moist vegetation is more absorptive than dry; attentuation in the summer is more pronounced than in the winter, but in any season, keeping your antenna away from living trees can be very helpful, and if you cannot, it is worthwhile to keep the antenna at right angles to larger limbs. Again, for NVIS, having vegetation around the location of the transmit antenna can also help by absorbing the low angle RF and reducing the likelihood of DF. Unless you are in triple canopy jungle, or perhaps mature sequoia forests, the difference is a db or two at most, but if I had the opportunity, I would take every advantage I had.
Power and NVIS
Most ham operators take every chance to improve their ability to send and receive signals. More power, often through the use of an amplifier, is a common goal, because they want to be heard and make more contacts over a wider area. Experienced NVIS operators, however, try to improve their signal to noise ratio, through intelligent selection of antenna and antenna height, time of day, and geographic location (when possible.) There may be times when a power increase is the answer, but more often, the needed communication can be made with changes in antenna height, or by a change in operating mode. In a grid-down event, with grid power out, and charged batteries a precious commodity, increasing power may not be an option in any case.
Single Side Band or SSB is a significant improvement over AM or FM, in terms of ‘talk power’, especially with modern radios that have integral voice processing. Processors can double or even quadruple the average power output of a SSB radio depending on the voice characteristics of the operator, which means a 3 to 6 db stronger signal. 6 db is an S unit on a calibrated receiver, and sometimes that difference is just enough to make the signal rise up out of the noise. However, that signal is still spread over a wide section of the band, usually about 2500 hertz. (Hz) This means that the average power per hz of bandwidth is very low. Let’s look at the math-
Most modern non-qrp amateur tranceivers have 100 watts of output available. 100 watts divided by 2500 hz is 1 watt for 25 hz, or 0.04 watts per hz. If instead of 100 watts, all you have is 5 watts, a normal QRP radio max output, then your SSB signal is 0.002 watts per hz. 2.5 watts, which is what my FT817 will do on internal battery, drops the average talk power to 0.001 w/hz. Contrast this with a CW signal, which is about 50 hertz wide, or less. 5 watts divided by 50 hertz is 0.1 watts per hertz, 2 ½ times better than the 100 watt SSB transmission, and if your CW signal is narrower, as I often observe with most modern radios, the difference is still greater. Put another way, a CW signal has 100 times the punch of the SSB signal, a 20 db difference in signal strength. That is over 3 S units. CW will get through with minimal power and minimal equipment when nothing else will.
Digital modes vary in bandwidth, but a common one is PSK31, which is 31 hz wide. A 5 watt PSK31 signal is about 0.16 W/Hz, or 4 times more punch than a 100 watt SSB signal. There are better digital modes for NVIS, which again vary in bandwidth, but I have been experimenting a bit with FSQ, designed to be an NVIS mode for HF communications, and it has great potential. The new Fldigi (HIGHLY recommended!) distributions now include FSQ as an option. BTW, while it is worthwhile to learn Morse, if you can’t or haven’t yet, there are other options, including using a computer to read the code for you. Code that is machine generated (computer or keyer) is easily read when there is little interference. Code that is manually created, using a straight key or a bug, is much harder for a computer to read. I can run a bug or a keyer, and I do use computer readers especially during contests, when many stations are running computer generated code at 40 or 50 words per minute (WPM), but most times when I run CW I usually operate with a straight key at 15 to 20 WPM to keep those skills polished.
I hope those who have stuck with me on this long download have not had too much trouble dealing with the pile of onions we’ve peeled! So, to wrap this LONG article up, here are my recommendations for the NVIS operator:
In a grid down situation occurring in a solar minimum, such as we’ll have for the next several years, I would operate NVIS during the day on 160 or 80 meters, using a dipole antenna(s) mounted low to the ground, say 5 to 25 feet up, varying the height to accommodate the ground, weather conditions and space weather. My preference would be to mount this antenna in a vee, with the center lower than the ends, ideally between two trees about 150′ apart, in a valley with heavily forested steep slopes or cliffs and with the ends of the antenna pointing at the slopes. I would run the lowest power I could, and if the other operators could use it, my preferred mode would be CW sent with a straight key. If I did not know CW, and had access to a laptop or tablet with Fldigi, I’d use Olivia or Contestia or FSQ and communicate digitally.
That’s it! Good luck. Questions, comments or concerns can be posted below, or email NC Scout.
-ARRL antenna book (recent editions, #21 to #23(current edition), are all useful. Try to get the computer disk that comes with it so that you can get a copy of EZNEC, which is extremely useful)
Web sites to check out-
Hope I will hear you on the air soon!