The mystery of negative towers

November 1, 2005

Ever since the beginning days of directional antenna usage design, operational engineers have been confronted from time to time by towers that do not behave normally. This situation is probably far less frequent today than it was 25 to 30 years ago. Certainly it is better understood and more easily handled. Back in the ‘60s and ‘70s when a group of radio engineers got together it was not unusual to hear at least one engineer complaining about a negative tower.

Today there are probably fewer negative towers in operating directional antenna systems thanks to the all-powerful computer, but they still exist. In the early days of directional antenna design the designer was usually pretty tired after four or five weeks of work on a Burroughs comptometer. When he found a combination of towers and phases and currents that would produce the desired pattern, he probably didn't look any further.

Figure 1. Block diagram of a negative tower system. Note the direction of power flow in tower 2. Click here to enlarge this image .

Today with a computer program that will iterate until the cows come home, development usually continues until the overall best solution is found. As a result, fewer wild towers are found in modern DA designs and many older systems have been reworked so negative tower problems have been eliminated.

Nevertheless, there are still many original DAs in use that have problem towers whose operating base resistance may swing from positive to negative depending on weather conditions and changes in the mutual impedance of the system. Once the secret of these negative towers is understood they are just as easy to handle as regular towers.

What's normal?

Engineers are generally accustomed to towers that have a normal base operating impedance; for instance 35 +j90 ohms. If we feed current into this tower we radiate an electromagnetic signal. But suppose the sign were -35+j90 ohms; what then?

Such a tower would radiate some RF power and try to get rid of the remainder in some other manner. If, as is more usual, the resistance is on the order of only a few ohms, the operating resistance would probably alternate between positive and negative values, and the DA parameters would go in and out of limits as the weather changed, or something else changes the mutual impedance of the array. Fortunately negative towers, which are caused by combinations of mutual impedance, current and phase, only occur in directional antenna systems. The base operating impedances are determined by the operating parameters required to produce the desired pattern in the directional antenna.

What exactly is a negative tower, and what does it do that is so different from a normal tower? The best definition is a tower with a negative base operating resistance. This means that instead of radiating power it will take power from the field produced by the other towers in the array. This power could be dissipated in a suitable resistance but the Commission will not allow that in most cases. The power that is absorbed from the other antennas must go somewhere, so why not feed it back, in correct phase and magnitude, into the power divider, i.e. the phasor? This is certainly more efficient than dissipating the power in a resistor.

For example

Let's assume that we have a two tower DA system with a licensed power of 5kW. Let's also assume that we need a dent in one side of the pattern with a reduction of 500W in that direction. We could actually drop a wire from one of the guys (in effect another tower), connect a suitable reactance and resistance in series between the tower base and ground and reduce radiation in that direction by 500W. The value of this resistance can be calculated so that 500W is removed from the radiation pattern in the required direction. That, my friends, is exactly what a negative tower does although in this case it is operating as a parasitic — that is, an undriven — element.

That would be an attractive solution for a simple DA pattern requirement. Unfortunately, the FCC does not allow this solution to be used, although one or two such installations have been permitted in the past. One that I know of personally was 50kW WWWE Cleveland (originally WKYC and also called WTAM). The station had a Franklin antenna with a drop wire suspended from a guy in the required direction. As originally designed the drop wire element was included in the three tower DA's system and was designed to be driven by the phasor.

However, it proved difficult to reduce enough power in the desired direction. So the drop wire base was disconnected from the phasor system and a series-dropping resistor used to reduce radiation toward the Canadian co-channel station. This station had been purchased from WKYC, and later it moved the transmitter to a more suitable site with a simple vertical radiator.

Modern computing power makes it easier to find the best solution to designing an array and not just the first one.

Figure 1 is a diagram of a two-tower array in which the number two tower has a negative base operating resistance. Power flow for tower one is normal from phasor to antenna one. However, the power flow for tower two is in the reverse direction. The whole system between tower two base and the phasor is designed to accept the power abstracted from the field of tower one and combine it with the transmitter power in the phasor in the correct phase. It's important to realize that the amount of power radiated by tower one must be sufficient to produce tower one's effect on the pattern. At the same time it must supply sufficient power for tower two to return the correct amount of power to the system. This means that tower one's power will be greater than the nominal power of the station.

Design of the phasor is inevitably bound up with the design of the directional antenna system. During the latter process the consulting engineer adds up the phase shifts introduced by the components between the antenna and the power divider. These values are adjusted to produce the desired phase at each antenna. If during this process someone discovered that a tower with a negative base operating impedance will be required, then he will make another computer run to find a more convenient antenna design.

If it becomes necessary to use a negative tower a change has to be made in the equation for the phasing system. The power that the negative tower will absorb from the array has to be fed in parallel with the correct phase and magnitude back into the power divider system of the phasor.

To do this the transmission line system must present a negative impedance to the tower. In effect the ATU is located in the phasor, line matching must be created and any system tweaking must be performed at the phasor. In essence, the negative tower becomes a generator that must be properly connected to the main generator (the transmitter) with correct phase and impedance so that power taken out of the array is returned in the phasor power divider.

Another factor to consider when dealing with a negative tower is the impact of the regenerative effect of the negative tower feeding power back into the system. Because of this regenerative effect the system Q can become high, sometimes with considerable effect on the bandwidth, which may vary unstably. A point of interest — as the feedback voltage is increased, as the system comes into tune, the ICP current will increase because a negative resistance in parallel with a positive resistance results in a larger resistance value than the positive alone.

E-mail Battison at

Burroughs photo courtesy of David Freeman at

Receive regular news and technology updates. Sign up for our free newsletter here.