Antenna2

During an emergency amateur radio operators have specific needs for an HF antenna. They need to communicate for a very close range out to say about 500 miles. Normal dipoles don't do this. They usually have a skip zone from zero miles out to about 600 miles. However, if a dipole is configured correctly it will do the job well. Practical NVIS frequencies are considered to be from about 2 MHz to about 10 MHz. In order to get to the lower frequencies, we will use either a 40 meter dipole (66 feet long) for frequencies from about 6 MHz to 10 MHz, or an 80 meter dipole for frequencies between about 3 MHz to 10 MHz. A resonant dipole for 2 MHz would be even better, but would be about 234 feet long which is a bit too long for many installations. Conventional amateur radio wisdom says to have a dipole up about one-half wave length above the ground. This gives excellent performance if you want to do DX, but close in performance is poor. This is because such a dipole can have a low angle of radiation. Low take off angles will cause your signal to refract off of the ionosphere farther away. When your signal comes back to earth it will at a greater distance. The following diagram illustrates a low takeoff angle, which is excellent for DX.
		Diagram 1: Low takeoff angle for dipole 1/2 wavelength above ground
The next diagram illustrates how a low takeoff signal behaves. Note the skip zone. The skip zone is where MARS needs to communicate.
		Diagram 2: Skip zone for low takeoff angles
So what kind of dipole do we usually have? It depends on how high it is off of the ground. For a good DX antenna if you go 1/2 wavelength off the ground then for 40 meters your antenna will be 66 feet high, and for 80 meters it will be 132 feet high. So your antenna may not really be ideal for DX anyway. The lower your antenna is from 1/2 wavelength high, the higher the angle of radiation becomes. If you get the antenna low enough it will direct some radiation straight up. This upward aimed radiation can come straight down if it is reflected back from the ionosphere. If this happens your low end of the range of communications becomes less which is highly desirable. Here is a diagram showing the higher angle of radiation from a lower dipole.
		Diagram 3: High takeoff angle for dipole much less than 1/2 wavelength above ground
This next diagram shows how high takeoff angles of a NVIS dipole fill in the traditional skip zone so local communications is made possible.
		Diagram 4: No skip zone for high takeoff angles
Cebik states that there is a range of heights above ground for a dipole over which its NVIS qualities are good. As it turns out for a 40 or 80 meter dipole, if it is between 30 or 40 feet off of the ground you will have good NVIS performance. This is a very significant piece of information. If your 40 or 80 meter dipole is already in that range of height then you already have a decent NVIS dipole antenna. All that we need to do is assure that the transmission line does not have a lot of loss. A lossy transmission line will make to difficult to hear other stations and other station to hear you. One important thing I must mention. The wire for the dipole has little loss. Even if there is a mismatch in the system with very high SWR all the signal will be transmitted as long as the transmission line is of a low loss type. We are talking of using a non resonant dipole here. If the dipole is resonant on 80 meters we will discuss how it acts as 40 meters.
Feedlines The SWR of a dipole off of its resonant frequency is high. EZNEC just reports it as >100. The ARRL Handbook and Antenna book have the equations to determine the SWR at the feed point of a dipole by using the complex impedance reported by EZNEC. On a typical 80 meter dipole cut for low SWR at say 3.5 mHz, it has a 50 Ohm SWR at 7.0 mHz (2λ) of about 212:1, which is fairly typical of a dipole used off of resonance. This is one fact that is often not realized by hams. If it tunes up at the rig then the antenna and transmission line must be fine. The above books also have the formulas to determine the SWR at the transmitter end of the transmission line and the resulting loss in the line. I made a spreadsheet to calculate this numbers. First see the SWR graph versus frequency below.

Diagram 5: SWR of an 80 meter dipole from 2 to 10 MHz

The green dot is at 7 MHz and the dip in the SWR curve is at 3.5 MHz. At 7 MHz the Impedance at the center of the dipole where the feed line connects is 8733+j4020 Ohms, which translates to a 50 Ohm Standing Wave Ratio (SWR) of 212:1.

I have computed the losses for this antenna at 7 MHz system using three commonly available types of feed line. This is summarized below. Each feed line is 100 feet long. RG8 is at the left, 450 Ohm window line is at the center and RG8X is to the right.

Diagram 6: Feed line loss of 80 meter dipole at 7 MHz

There are two types of loss in a feed line. On the second line from the top is the "matched" feed line loss for 100 feet of line. This value is normally quite low. It is just less than 1 dB for the RG8X, and a half dB for the RG8. Both these values are quite low. Note the 450 Ohm line has a loss of only .06 dB. "Matched" line loss is the loss in the feed line when the SWR is 1:1. All the power goes out the end of the feed line and is radiated out the antenna the first time it travels up the feed line and none of it returns. However we are talking about a non-resonant dipole where the SWR can be over 200:1 into a 50 Ohm feed line. In a non-resonant situation only part of the power is radiated during the first traverse to the dipole. A large portion of the power is reflected back to the transmitter at very high SWR. Once it gets back to the transmitter it is reflected back into the feed line for another trip to the antenna. Each time this happens some additional power is radiated from the antenna, but some power is lost when the signal goes back up or down the feed line. This happens until all the power is dissipated either in heat loss in the feed line or as radiation from the antenna. With all the traverses back and forth along the feed line, a lot of power can be lost in lossy feed lines. How much power is lost? The fourth line from the bottom lists "unmatched" line loss. This is the additional loss in signal due to high SWR in a lossy feed line.

For the RG8 feed line the mismatched loss is 10.7 dB for a total line loss of 11.2 dB. This is almost 2 S units of loss. For the RG8X it is even worse with 13.9 dB of loss. All of this loss is in the feed line. None of it is from the antenna.

On the right table, for a typical 100 feet of RG8X, the resulting SWR at the transmitter end of the feed line was only 8.8:1. This can be adjusted by a manual tuner. So the average ham says : "Yeah, the SWR was a bit high (8.8 to 1), but my tuner took care of the mismatch and everything is fine". Well there is a reason why the SWR is 8.8:1 at one end of the coax and 212:1 at the other end. It is called loss. Your coax is acting like a dummy load (dummy loads have an SWR of 1:1). This relationship is similar for other 1λ-2λ dipole relationships. If you have always wondered why your coax-fed 40 meter dipole was so poor on 20 meters, this is why. The situation is not much better for the RG8 on the left column (or LMR 400, which would give similar unmatched line loss). The SWR at the radio is 16.6:1 for the RG8. It is a dummy load just like the RG8X.

So what should the concerned ham do? Well if you could put your auto tuner at the center of the dipole, the coax mismatched line loss would be eliminated since the tuner would transform the impedance to 50 Ohms for the feed line. This is done by some hams. Its a bit complicated but the details are available at the Texas Army MARS web site which you should Google. Lew, AAR6UK, has done a lot of work in this area and has a very good position paper on this approach.

However, there is another solution. Look at the center column in Diagram 6. It is for 450 Ohm window line used as a feed line. The mismatched line loss is only .6 dB for a total line loss of only .66 dB. This is a negligible amount of loss. So if you use 450 Ohm window line as a feed line your feed line losses will be very small. Window line does not radiate like some hams fear. Just keep it 4 diameters (4 inches for 450 Ohm line) away from large metal objects. Note also the SWR at the dipole end of the feed line is 23.5:1 and at the other end 20.2:1. Although this SWR (20.2:1) for the 450 Ohm line is greater than the SWR for the RG8X (8.8:1), it is the difference in SWR (percentage change) at the two ends of the feed line that is important. For the 450 ohm line the difference in SWR is small indicating that there is much less loss in the feed line. In other words, the feed line is not acting like a dummy load.

A have photos of such antenna installations here: NVIS Antenna