I often see discussions about NVIS antennas describing mounting heights that are too low. Dipoles must be mounted greater than 1/4-wavelength above ground before maximum radiation beings to move from straight-up (NVIS) to lower angles. As the antenna is moved lower, ground losses begin to increase, reducing the amount of signal radiated upwards. In the "Photos" section I have posted EZNEC elevation plots comparing dipole antennas mounted at 10ft, 20ft, and 30ft for both 80 and 40 meters.
On 80 meters, the difference between 10ft and 20ft is greater than 4dB, meaning more than half your power is being lost in the ground and not radiated anywhere. This figure rises to more than 6dB when one compares 10ft vs 30ft. In reality, one must go above 60ft before an 80 meter dipole begins to move from NVIS to lower angles of radiation.
On 40 meters, the difference between 10 and 20ft is nearly 3dB, but only slightly more when comparing 10ft to 30ft. Note that 30ft is very close to 1/4-wavelength on 40 meters.
I like to compare NVIS to a 2-element yagi. Picture the dipole as the driven-element, and the ground as the reflector; 2 elements pointing straight up. How far do you mount the reflector from the DE for max. forward gain? Usually on the order of 1/4-wvelength.
The conclusion is, if you can get your antenna up 20-30 feet above ground, do so. Don't lower it thinking you will get better NVIS performance. If, however, you can only get it up 10 feet or so, go for it. When operating portable, or in neighborhoods with antenna restrictions, any height that gets you on-the-air is good.
Good info Mike,
I might add a couple of observations on antenna height. When the antenna is in the inverted V configuration, the height to be concerned with is the apex feedpoint and follow Mike's suggestions. The ends can be anywhere hi. HOWEVER, with the inverted V you have several operational options. First, most of the polarity of the antenna in high angle and vertically polarized. Good NVIS, but but for noise. As you raise the ends up to the height of the feedpoint, the NVIS stays pretty much the same, but the noise drops off!!!!
Second, any horizontal antenna is less prone to have electrical noise regardless of height.
So, while the attempt is made to make a contact, you have to weigh and adjust your situation accordingly.
Like Mike says, any port in a storm... WORKS sorta.
A few weeks ago, Mike wrote:
That works OK for a moderately ideal reflecting surface: one with infinite size and very good surface conductivity.
1/4-wavelength is not best for a reflector wire. (Isn't the optimum separation for a Yagi around 0.15 wavelength?) That's because the reflector wire is most often cut longer than resonant length, making it inductive, with consequent phase shift in its current. The phase shift causes an additional "delay", requiring it to be closer than 1/4-wavelength to the driven element so that the reflected signal lines up in-phase.
For a dipole above ground, the reflecting surface is not just the top millimeter of topsoil. NVIS antenna users find varying performance versus height, depending on the type of soil and rocks below the antenna. A possibility is that the wave penetrates some distance into the earth before reflecting, maybe reaching the water table or other substance that makes a good reflector. If you measure your antenna height above this reflecting material, and the soil above it is dry, your antenna might be several feet closer to the earth's surface. The earth is not a mirror.
EZNEC doesn't show this. To EZNEC, earth is a thin reflecting surface with nothing below it.
I don't know if NEC4 can do non-homogeneous ground layers.
The problem is only thinking that better gain = better NVIS performance. They aren't synonymous. Better performance comes from better signal-to-noise ratio.
Some experienced NVIS users have found that a lower antenna works better for them (in their soil environment) because the lower antenna knocks down the low-angle "noise" (QRM/QRN) significantly, allowing them to hear the other station better. The implication is that the upwards-going lobe narrows as well as shrinking. Sacrifice a little gain to get noticeably better S/N.
Of course it is a 2-way street, and it only works if the other guy can hear you with your antenna's higher loss. Some people who looked into this found that everyone in a net got better performance (more reliable contacts) when all their antennas were mounted lower.
But the overriding rule is: YMMV. You may find otherwise. Do what works, for YOU.
These are actual empirical results. Theory is nice and may tell you one thing, but when the results you experience do something else, then the conclusion is that either the theory or the model was insufficient. Antenna modeling is great for wires, but lousy when it comes to soil.