When a Repeater Antenna Is Too Close to the Roof

A repeater with coverage problems is often blamed on the transmitter, the receiver, the duplexer, the coax, or the antenna gain. Sometimes that is correct. But when the antenna is mounted low above a roof, especially on 70 cm, the real problem may be the radiation pattern itself.

A photo of such an installation showed a Diamond X50 mounted roughly 1.5 m above a roof. At first glance that may look acceptable. On 70 cm, however, 1.5 m is already more than two wavelengths, and the complete antenna is also physically long compared with the wavelength. The roof is not just “nearby structure.” It becomes part of the electromagnetic environment.

The point of this article is not that one antenna type is good and another is bad. A 1/4-wave ground plane, a J-pole, a 5/8-wave vertical, and a collinear antenna can all work well when installed correctly. The real difference is how much each antenna depends on a clean environment to preserve its intended elevation pattern.

First Correction: an X50 Is Not a Simple 5/8-Wave Whip

The Diamond X50 is not just a single 5/8-wave vertical on 70 cm. It is a dual-band collinear antenna. On 2 m it behaves differently than it does on 70 cm. On 70 cm it uses multiple phased radiating sections to compress more energy toward lower elevation angles.

That matters because a collinear antenna does not create gain from nowhere. It redistributes energy. More field strength near the horizon usually means less field strength at higher and lower elevation angles. That can be useful for flat-terrain distance coverage, but it can become troublesome for close-in coverage, valley coverage, or users located below the repeater site.

Wavelength Makes the Roof Look Different on 2 m and 70 cm

At 430 MHz, the wavelength is approximately:

λ = 300 / 430 ≈ 0.697 m

So a 1.5 m mounting height above the roof is approximately:

1.5 / 0.697 ≈ 2.15 wavelengths

That sounds high, but the antenna itself is also long. A 1.7 m antenna is about 2.44 wavelengths at 430 MHz. So on 70 cm, the roof is interacting with a multi-wavelength vertical array, not with a small radiator.

At 145 MHz, the wavelength is approximately 2.07 m. The same 1.5 m roof clearance is only about 0.73 wavelength. The physical installation is the same, but the RF situation is not. This is why a dual-band rooftop antenna can behave acceptably on one band and poorly on the other.

Ground Reference and Roof Reflection Are Not the Same Thing

Two effects are often mixed together in amateur-radio discussions: the antenna’s RF reference and the roof’s reflection or scattering effect.

The RF reference is the return path required by the antenna feed system. A 1/4-wave monopole needs radials or a conductive plane. A J-pole uses a matching stub and an end-fed half-wave radiator. A 5/8-wave vertical needs a suitable reference system to maintain the intended pattern. A collinear antenna includes its own feed and radial structure, but it can still interact strongly with the mast and feedline.

The roof reflection happens because the radiated field encounters nearby conductive or dielectric structure: metal roofing, gutters, rebar, solar panels, foil insulation, ventilation units, wet roofing material, or a roof edge. These objects can reflect, absorb, or scatter energy. That changes the elevation and azimuth pattern even when the SWR looks fine.

Rooftop Rule: All These Antennas Need Feedline Control

When a vertical antenna is mounted low above a roof, feedline-current control becomes important for every antenna type discussed here: the 1/4-wave ground plane, the J-pole, the 5/8-wave vertical, and the multi-section collinear.

The reason is simple. The roof, mast, brackets, gutters, solar-panel frames, rebar, and coax shield are all close enough to interact with the antenna field. If the outside of the coax shield is allowed to carry RF current, the antenna is no longer only the visible radiator. The real antenna becomes the radiator plus the coax, mast, and part of the roof environment.

Pattern Complexity Determines Rooftop Sensitivity

The more an antenna depends on vertical pattern shaping for its gain, the more clearance it needs from the roof. This is the main technical point.

A 1/4-wave ground plane and a properly choked J-pole have relatively broad vertical patterns. That broad pattern makes them more forgiving. The roof may still reflect and distort the field, but the antenna is not relying on a narrow, carefully phased lobe to create its useful coverage.

A 5/8-wave vertical already starts to shape the elevation pattern more strongly. It pushes more energy toward lower elevation angles than a 1/4-wave radiator. That can be useful when the antenna has enough free space around it, but it also means the antenna becomes more sensitive to the roof, mast, feedline, and ground-reference geometry.

A collinear goes one step further. It uses multiple radiating sections and phasing between them to compress the vertical beam and increase gain near the horizon. That narrower beam is useful on a clear mast, but close to a roof it is easier for reflections and scattering to create unwanted lobes and nulls.

Comparison of Common Rooftop Vertical Antennas

Quarter-Wave Ground Plane: Broad and Predictable

A 1/4-wave vertical with a proper radial system is often boring, and that is precisely why it can be good. The pattern is broad in elevation, the feed system is understandable, and the antenna is less dependent on “magic” gain claims.

On a repeater site, this broad vertical pattern can be useful. It may not produce the strongest distant flat-terrain signal, but it often gives better coverage to nearby users, lower users, mobile stations in streets, and stations in uneven terrain.

The condition is that the radials must actually form a usable RF reference. A 1/4-wave whip mounted on a small bracket with a long coax shield pretending to be the missing radial system is not the same antenna.

When a 1/4-wave ground plane is mounted low over a roof, a choke is still recommended. The radials provide the main RF reference, but the coax shield can still become part of the antenna system if it is not isolated.

J-Pole: Good Rooftop Antenna, but Always Choke It

A J-pole is essentially an end-fed half-wave radiator fed through a quarter-wave matching section. It does not need the same radial system as a 1/4-wave monopole, which makes it attractive for compact rooftop installations.

When it is built correctly, mounted clear of nearby conductors, and properly choked below the feedpoint, a J-pole can perform very well. Its relatively broad vertical pattern can be useful for local repeater coverage, especially where users are close to the site or located below the antenna height.

However, a J-pole should be choked in every installation. The matching section, mast, and coax shield can easily become part of the radiating system if the feedline is not isolated. That can produce an acceptable SWR while still distorting the radiation pattern.

So the correct criticism is not that a J-pole is a poor rooftop antenna. It is not. The correct statement is that a J-pole must be installed as a controlled antenna system. With a good choke and sensible mounting clearance, it is often one of the more forgiving rooftop choices.

Half-Wave Vertical or Sleeve Dipole: Often the Predictable Middle Ground

A simple half-wave vertical, sleeve dipole, or commercial dipole-style antenna can be a very good choice for repeater service. The elevation pattern is still relatively broad, and the feed system can be made much more predictable than a poorly isolated end-fed installation.

That does not mean it is immune to roof effects. A nearby mast, railing, metal roof edge, solar-panel frame, or long active coax shield can still disturb the pattern. But compared with a higher-gain collinear, a simple half-wave-style radiator is usually less dependent on a narrow vertical lobe.

For many repeater installations, a controlled half-wave or folded-dipole-style antenna can give more reliable real coverage than a higher-gain antenna mounted in a poor rooftop environment.

5/8-Wave Vertical: Useful, but Less Forgiving Close to a Roof

A 5/8-wave vertical can be a very useful antenna. Its advantage is that it shapes the vertical radiation pattern and pushes more energy toward lower elevation angles than a simple 1/4-wave radiator.

That pattern shaping is also why it is less forgiving when mounted low above a roof. A 1/4-wave ground plane or properly choked J-pole has a broader vertical pattern. A 5/8-wave vertical is already trying to form a more specific low-angle pattern, so nearby roof reflections, mast current, coax current, and an uneven RF reference can disturb the result more easily.

A 5/8-wave antenna should preferably be choked in any fixed installation. When it is mounted low above a roof, the choke should be treated as part of the installation, not as an optional accessory. Without feedline isolation, the coax shield and mast may become part of the counterpoise and alter the intended pattern.

So the weakness is not the 5/8-wave radiator itself. The weakness is that it needs a cleaner environment than a 1/4-wave ground plane or a properly choked J-pole if we want the expected pattern to survive.

Collinear Antennas: Highest Rooftop Clearance Requirement

A multi-section collinear antenna creates gain by stacking radiating sections and phasing them so that more energy is concentrated near the desired elevation angle. On 70 cm, this can produce useful gain toward the horizon.

But that gain is obtained by compressing the vertical pattern. The narrower and more carefully shaped the pattern becomes, the more sensitive it becomes to roof reflections, mast interaction, feedline current, mechanical tilt, and nearby metal.

This is why a collinear generally needs more rooftop clearance than a 1/4-wave ground plane, a properly choked J-pole, or a simple half-wave radiator. It is not a worse antenna. It is simply more dependent on a clean installation environment.

When a collinear is mounted low above a roof, feedline-current isolation becomes very important. The antenna may still show a good SWR, but roof interaction and coax-shield current can move the useful lobe, create nulls, or make coverage uneven by direction.

Repeater Coverage Depends on Elevation Angle

If the repeater antenna is above the user, the signal must leave the antenna at a downward angle to reach nearby users. The required angle is approximately:

α = tan^-1(height difference / horizontal distance)

For example, if the antenna is 30 m above the users:

Those angles look small, but for a high-gain collinear they can matter. If the antenna has a narrow vertical beam and the site is high above the wanted users, the strongest part of the pattern may not be where the users are.

The Roof Creates a Second Radiating Image

A conductive roof can behave like a reflecting plane. In a simplified model, the direct field from the antenna combines with a reflected field from the roof. Depending on phase, those fields can reinforce or cancel.

A simplified expression is:

E_total(β) = E_antenna(β) [1 + Γ e^{-j2kh sinβ}]

where β is the elevation angle, h is the height above the reflecting surface, k is 2π/λ, and Γ is the reflection coefficient.

This is not a full model of a real rooftop. A real roof is finite, irregular, and often full of metal edges, gutters, masts, rebar, foil insulation, and solar-panel frames. But the simplified model explains why changing the antenna height by even a fraction of a wavelength can move nulls and peaks.

On 70 cm, a quarter wavelength is only about 17 cm. A height change of 20 to 40 cm can visibly change the pattern interaction with the roof. That is why “move it a little and test again” is not a silly experiment at UHF. It is RF reality.

Roof Materials Matter

A metal roof several wavelengths across can act like a strong reflecting plane. If the antenna is centered above it, the azimuth pattern may remain fairly symmetrical, but the elevation pattern can still be strongly modified.

A roof with metal edges, gutters, rebar, solar panels, or ventilation structures is more complicated. It can create azimuth-dependent nulls where one direction is much worse than another.

A tiled or flat roof may appear non-conductive, but wet roofing material, foil insulation, reinforced concrete, or metal support structures can still be significant at VHF and UHF. At 70 cm, small physical details are electrically large enough to matter.

Is 1.5 m Above the Roof Enough?

There is no universal yes or no. On 70 cm, 1.5 m is more than two wavelengths. That sounds like good clearance, but the antenna may also be a multi-section vertical array. The roof is still close enough to influence the resulting pattern.

The better question is not “is the antenna 1.5 m above the roof?” but:

• Is the complete radiator above the roof clutter?
• Is it clear of gutters, parapets, metal edges, and solar panels?
• Is the mast part of the RF system?
• Is the coax shield decoupled?
• Does the antenna have a suitable elevation pattern for the coverage area?
• Is the coverage problem equal in all directions, or only in certain azimuths?

For a broad-pattern 1/4-wave ground plane or properly choked J-pole, 1.5 m may be workable if the roof environment is not too hostile. For a 5/8-wave vertical, more clearance is preferred because the pattern is already more shaped. For a multi-section collinear, 1.5 m may still be too low if the roof is reflective, cluttered, or close to the lower part of the antenna structure.

Better Antenna Choices for Repeater Service

For a repeater, predictability is often more valuable than catalog gain. Better choices may include:

• a commercial half-wave sleeve dipole;
• a folded dipole;
• a stacked dipole array with a published pattern;
• a 1/4-wave ground plane with a proper radial system;
• a 5/8-wave antenna mounted with enough clearance and proper feedline isolation;
• a collinear antenna mounted higher and clear of the roofline;
• a professional array with electrical downtilt if the wanted users are below the site.

A J-pole can work well, but it should always be choked. A 5/8-wave can work well, but it is less forgiving close to a roof and should preferably be choked in any fixed installation. A collinear can work very well, but it needs the cleanest mounting environment because its gain is created by a shaped vertical pattern.

Before Replacing the Antenna, Test the System

A photo is not enough to diagnose a repeater. Poor coverage can be caused by antenna pattern, but also by receiver desensitization, duplexer loss, bad coax, water ingress, connector damage, site noise, wrong squelch settings, or insufficient transmitter power at the antenna port.

A sensible test sequence is:

• Determine whether the problem is uplink, downlink, or both.
• Measure return loss at the antenna system, not only in the equipment room.
• Measure coax loss and inspect connectors for water ingress.
• Check duplexer insertion loss and isolation.
• Measure receiver sensitivity and site noise.
• Map signal strength by azimuth and distance.
• Temporarily substitute a broad-pattern test antenna.
• Change the antenna height by 1/4 to 1/2 wavelength and repeat the field test.

If the weak zones move when the antenna height changes, local reflection is probably involved. If the problem exists only in one azimuth, roof-edge or building scattering is likely. If the downlink is strong but the repeater hears badly, the antenna may not be the first suspect.

The Practical Conclusion

The issue is not that a J-pole is good and a collinear is bad. The issue is pattern complexity and installation sensitivity.

A 1/4-wave ground plane and a properly choked J-pole have relatively broad vertical patterns, so they are usually more forgiving when mounted near a roof. A 5/8-wave vertical is already more pattern-shaped and therefore needs more room. A multi-section collinear needs the cleanest installation because its useful gain comes from a narrower, phased vertical pattern.

All of these antennas should have feedline-current control when mounted low above a roof. A J-pole should always be choked. A 5/8-wave vertical should preferably be choked, and when it is low to a roof, it should be considered mandatory. A collinear also needs good isolation, because coax or mast current can disturb the very pattern that creates its gain.

So for a repeater with rooftop coverage problems, a lower-gain but broader-pattern antenna may outperform a higher-gain collinear. Not because the collinear is inefficient, but because the roof can disturb the shaped pattern that makes the collinear useful in the first place.

Interested in more technical content? Subscribe to our updates for deep-dive RF articles and lab notes.

Questions or experiences to share? Feel free to contact RF.Guru for practical antenna and repeater-system support.