Antenna Interaction on One Tower: Spacing and Detuning Guide
Put a second antenna near the first and the first one changes. This antenna interaction is one of the least appreciated facts in amateur radio, and it explains many mysteries: the 40 metre dipole whose SWR moved after a new beam went up, the Yagi that lost its front-to-back ratio when a wire was hung from the same mast, the vertical that works well alone and badly beside the tower. Every conductor that is long enough to be near-resonant at the operating frequency picks up energy and re-radiates it, and the re-radiated field combines with that of the antenna you are using. This guide explains the mechanism, shows which combinations are most likely to cause trouble, gives spacing rules of thumb and describes how to locate and cure interaction in a real station.
How Antennas Affect Each Other
An antenna in the field of another one carries a current, induced by the wave. If the conductor is close to resonance, the current is large, and the conductor radiates its own field. In a Yagi antenna this effect is used on purpose. A driven element is the only one connected to the feedline, and the other elements, called parasitic elements, are not, as the Wikipedia article on antenna arrays describes: the director and the reflector shape the pattern through the currents that the driven element induces in them. The same physics applies when the second conductor is an unwanted wire, a guy line or a tower. It works as an accidental parasitic element, with a current and a phase that you did not design.
Two effects follow from this. The first is a change in the pattern: a lobe grows in one direction and a null appears in another, and the front-to-back ratio of a beam may be reduced. The second is a change in the feedpoint impedance, because the induced current in the neighbour flows back into the driven element in the form of a mutual impedance. The result is a shift of the resonant frequency and a change in the SWR. In many cases the effect is small, for example a fraction of a decibel, but when the spacing is small and the neighbour is near resonance it can be large.
What Makes the Coupling Strong
- Resonance of the neighbour. A conductor that is about a half-wave long, or an odd multiple of that, at the operating frequency carries a large induced current. A wire that is much shorter or longer is a weak coupler.
- Distance in wavelengths. The coupling grows quickly as the distance falls below about a quarter of a wavelength. At 14 MHz a quarter wavelength is about 5 m, and at 144 MHz it is about 0.5 m.
- Parallel orientation. Conductors that are parallel to each other couple much more than conductors at right angles. A vertical wire next to a horizontal dipole interacts little, while two parallel wires interact a lot.
- Length of the parallel run. A long parallel section is worse than a short one, so avoid running a guy line or a feedline parallel to a resonant element for a long way.
- Several resonances. Antennas for different bands are related by harmonics. A dipole cut for 7 MHz is also a resonant three-half-waves radiator at about three times that frequency, that is, near 21 MHz, which can disturb a 15 metre beam.
Common Problem Pairs
| Combination | What happens | Typical cure |
|---|---|---|
| Wire dipole under or near a Yagi | The wire works as an extra element in the beam’s field and distorts the pattern on the bands where it is resonant | Move the wire away, make it perpendicular to the boom, or break it up with insulators and traps |
| 40 m dipole next to a 15 m beam | Third harmonic resonance at about 21 MHz re-radiates and affects the beam | Change the length or add a choke or a trap that opens the dipole at the higher band |
| Vertical next to a tower | The tower is a large parasitic element and shapes the vertical’s pattern | Place the vertical at a distance of at least a quarter wavelength, or model the combination |
| Two beams on the same mast, close together | Their patterns change and the gain falls, and receivers hear each other during transmit | Follow the stacking spacings and use filters for multi-station operation |
| Guy wires near resonant length | Guys become resonant elements on one or more bands | Break them into non-resonant lengths with insulators |
| Feedline running along the tower | Common-mode current on the shield makes it an element of the antenna | Run it away from the tower for a distance and use chokes |
Spacing Rules of Thumb
There is no single safe distance, because the effect depends on the geometry and the resonance, but some rules help in planning. Treat them as starting points and check the result with a measurement or a model.
- Keep resonant wires at least a quarter wavelength from a beam or vertical on the band you want to protect, and longer if you can. On 20 m, that is about 5 m.
- Cross, do not run parallel. If a wire has to pass near another antenna, make it cross at a right angle instead of running alongside.
- Stack with known spacings. For identical Yagis on a common mast, the vertical spacing affects the gain and the pattern. Our article on stacked Yagis explains why spacing is chosen for the takeoff angle and how much gain you gain from it.
- Do not place a tower inside a vertical array. A tower with antennas close to its top is itself a resonant structure on one or more bands. If you must use one, model it.
- Break up long conductors. Guy ropes and cables of non-resonant length are the cheapest cure for unexpected resonances. Use a mix of insulators and lengths so that no segment is close to a half wavelength on a band you operate.
The Effect on the Station Receiver
Interaction is not only a matter of patterns. In a multi-transmitter station, for example a Field Day or a contest setup with two or more radios, one transmitter’s signal can enter the antenna of the other and overload its front end. The coupling between the two antennas is a combination of near-field coupling and the pattern, and at the same distance the effect is much stronger when the two antennas work on harmonically related or adjacent bands. The solution is mostly in the receiver: band-pass filters and stub filters that reduce the unwanted signal before the receiver. Our guide to band-pass filters for multi-station Field Day describes them, and the notes on coax antenna switches show how isolation of the switch matters when two radios share antennas.
A Worked Example: A 40 m Dipole Beside a 15 m Beam
Suppose you have a half-wave dipole for 7.1 MHz. Using the common approximation of 468 divided by the frequency in MHz, the length in feet is about 66, which is just over 20 metres. That wire is about one half wavelength on 40 m, but at three times the frequency, around 21.3 MHz, the same wire is about three half wavelengths long and can carry a strong current again, with a pattern of several lobes. A 15 metre beam operates at 21 to 21.45 MHz, right in that region. If the dipole hangs 3 or 4 metres from the beam, within a quarter wavelength on 15 m (about 3.5 m), the dipole carries a significant induced current on 15 m and re-radiates it. The effect on the beam may show as a shifted resonance, a lower front-to-back ratio or a distorted pattern, even though the dipole is never used on 15 m.
The remedies follow from the diagnosis. You can shorten or lengthen the dipole by a few per cent so that its third-harmonic resonance moves out of the 15 m band, which has a small effect on 40 m if you also retrim the antenna, or you can feed the dipole with a choke or a trap that makes it non-resonant at the higher frequency. The best solution is geometric: move the dipole so that its wire is far from the beam, or arrange it at a right angle to the boom. The example shows why a wide look at the whole site, and not at one antenna at a time, is needed when you add something new.
Finding the Culprit
If an antenna behaves strangely after a change, test the neighbours one by one. The plan below uses an antenna analyzer or a NanoVNA and a little patience.
- Record a baseline. Sweep the SWR across the band, and save the curve or write down the resonant frequency and the minimum SWR.
- Disconnect the suspect. Open the feedline of the other antenna, or lower a wire, and repeat the sweep. Then connect a dummy load in place of the suspect antenna and repeat once more, because a connected radio can change the situation.
- Compare. If the resonance moves by a noticeable amount when the neighbour is removed, you have found the coupling.
- Try a detuning change. Add an insulator to a guy wire or a wire, change the length by 5 to 10 per cent, or move the conductor by a metre, and sweep again. Record each change.
- Repeat on other bands. An interaction that is cured on 20 m may remain on 15 m, because the same wire resonates on several bands.
A nearby antenna may also show itself through a change in the pattern, which you do not see on an SWR sweep. In that case, a comparison of received signals from a distant station or from a beacon, while switching a neighbour on and off, gives the answer. The guide to antenna radiation patterns, lobes and front-to-back ratio explains how to read these changes.
Modelling Before You Build
Antenna modelling software lets you place a second antenna, a tower and the guys in the same model and see how the pattern changes before you climb. A model is not perfect, but it shows the trend, which is often all you need: whether a wire parallel to a boom is a problem, how far to move a vertical from the tower and which guy segment is resonant. The beginner’s walkthrough in our article on antenna modelling with EZNEC and 4nec2 is a good place to start. Include the tower and the guys in the model as soon as you can, since they are the most common sources of surprise.
Practical Habits That Prevent Trouble
- Plan the site on paper. Draw the antennas, the tower, the guys and the feedlines to scale, and check the distances against the quarter-wavelength rule for each band.
- Add one thing at a time. Measure the old antennas before and after each addition, so that you know which change caused which effect.
- Use chokes on feedlines that run near the elements, so that the shield does not behave as part of the antenna.
- Keep a log of the SWR curves of each antenna. A drift from the old curve is the first sign of a new interaction, a broken connection or water in a cable.
The Bottom Line
Any conductor near resonance close to an antenna becomes a parasitic element that changes its pattern and its impedance. The coupling is strongest for resonant, parallel, nearby conductors, including wire dipoles, guy lines, towers and feedlines, and it can involve harmonically related bands. Keep resonant wires at least a quarter wavelength away, cross instead of running parallel, break up guys and use chokes on feedlines. When a problem appears, record a baseline, remove the suspect antenna, compare the results and change the geometry one step at a time. With that discipline you can put many antennas on one site and keep each of them working as designed.
Frequently Asked Questions
What is antenna interaction?
It is the effect of one antenna or conductor on another nearby. The second conductor carries an induced current and re-radiates, which changes the pattern and the impedance of the first antenna.
How far apart should two antennas be?
There is no universal figure. As a rule of thumb, keep resonant conductors at least a quarter wavelength from the antenna you want to protect, and avoid parallel runs. Check with a measurement or a model.
Can a tower affect a wire antenna or a vertical?
Yes. A tower is a large conductor that may be resonant on one or more bands, and it can change the pattern and the SWR of a vertical or a wire placed close to it.
Why does my 15 metre beam change when I add a 40 metre dipole?
A 40 metre dipole is also resonant near the third harmonic, close to 21 MHz, so it can act as a parasitic element on 15 metres. Change its length, move it or add a trap.
How do I find which antenna is causing the problem?
Measure a baseline SWR curve, disconnect or remove the suspected neighbour and measure again. A shift in resonance points to the coupling.
Can interaction cause receiver overload?
Yes, in multi-transmitter stations a strong signal from one antenna can enter another. Band-pass filters and good isolation between antennas reduce the effect.