When a 9 dBd Collinear Loses to a 0 dBd Whip: Installed Gain and Coverage
When a 9 dBd Collinear Loses to a 0 dBd Whip: Installed Gain and Coverage
A catalogue gain number describes a best direction under stated conditions. Coverage depends on the full elevation and azimuth patterns, mounting, feed system, terrain, clutter and the angles to the stations that matter.
A 9 dBd collinear has a nominal 9 dB advantage over a 0 dBd antenna only when both numbers use the same gain definition and their maximum-gain directions are being compared at the antenna terminals. It does not promise 9 dB at every elevation angle, after different feedline losses, or across an obstructed service area.
First Normalize the Gain Reference
ITU terminology defines dBi relative to a lossless isotropic radiator and dBd relative to a lossless half-wave dipole in its direction of maximum radiation. For the same gain value and direction:
GdBi = GdBd + 2.15 dB
9 dBd = 11.15 dBi
0 dBd = 2.15 dBi
The difference remains 9 dB after conversion. A claim of 0 dBd does not mean “no directivity” or “no gain”; it means the quoted gain equals the half-wave-dipole reference. Always verify that a catalogue number is gain rather than directivity, which polarization and frequency it represents, whether it is peak or typical, and whether mismatch, feeder or mounting losses are included.
ITU-T K.91 explicitly notes that catalogue antenna gain is commonly a maximum value and ordinarily excludes impedance- and polarization-mismatch losses. It also treats feeder and connector attenuation separately when evaluating EIRP. Mixing these reference planes is a common source of fictional system gain.
What Collinear Gain Actually Buys
A vertical collinear phases multiple radiating sections so that their fields add strongly over a narrower set of elevation angles and cancel partly elsewhere. With an approximately omnidirectional azimuth pattern, the extra directivity is obtained mainly by compressing and reshaping the elevation pattern. The result is not a uniformly “stronger doughnut”: it has a main lobe, finite beamwidth, side lobes and nulls.
Current Amphenol Procom data provide a useful scale example, not a universal model. Its 422x.09-445-Txx UHF collinear is specified at 8.7 dBd with an 8° ±1° E-plane half-power beamwidth. Its G-CXL 225-450C 0 dBd antenna is specified with an 80° E-plane half-power beamwidth. Exact patterns, tilt and gain vary with product and frequency, but the directional trade is real.
| Catalogue item | What it can establish | What it cannot establish alone |
|---|---|---|
| Maximum gain | Peak directional gain at the stated reference plane and conditions. | Gain toward a particular user, repeater or terrain sector. |
| Elevation and azimuth patterns | Angular main lobe, side lobes, nulls, tilt and azimuth ripple in the test configuration. | The installed pattern beside a mast, roof, rail, vehicle, cable and enclosure. |
| VSWR or return loss | Impedance match at the measurement plane. | Radiation efficiency, gain, feedline loss, noise pickup or coverage. |
| Input-power rating | A manufacturer limit under specified electrical and environmental conditions. | Gain, linear coverage radius or thermal behavior in a different installation. |
Direction decides which antenna is “better”
If a distant station sits inside the collinear’s installed main lobe, the collinear can deliver much of its advantage. If a station is above or below that lobe, inside a side lobe, or near an installed null, the broad-pattern whip can have more gain in that direction. Electrical or mechanical downtilt, mast orientation and site slope can therefore matter as much as peak gain.
This is especially important for a high site serving nearby users below the antenna, a valley, an elevated repeater, steep urban streets or mixed terrain. A narrow horizontal beam may favor distant near-horizon paths while under-serving some close-in or high-angle paths. A well-designed high-gain installation can use tilt or null fill; a random tall collinear cannot be assumed to do so.
Delete the Universal Clearance Rules
The source proposed fixed 0.25 λ, 1 λ and 2 λ thresholds for choosing antenna gain. Those are not general performance boundaries. Nearby conductors alter current distribution, impedance, polarization and the three-dimensional pattern according to their size, shape, orientation, conductivity, connection and position. A rail parallel to the radiator is not equivalent to a roof below it, and a thin guy wire is not equivalent to a mast through the active aperture.
Amphenol’s own half-wave marine/base antenna guidance warns that nearby parallel metal can change both SWR and radiation pattern. That is useful installation evidence, but it still does not create one safe clearance number for every antenna.
- A quarter-wave monopole needs an intentional return structure such as a suitable vehicle body or radials. Its installed pattern depends on that structure.
- An end-fed half-wave may not require a traditional quarter-wave ground plane, but it is not environment-independent. Its matching network, mounting hardware, coax exterior and equipment can carry return/common-mode current.
- A common-mode choke is not an automatic cure. Select its impedance and placement from measured exterior current and the intended current boundary; on an end-fed system, adding a choke can also change the match and pattern.
- A dual-band antenna is a different electrical array on each band. Gain, lobe structure, match and feedline loss must be compared separately at 2 m and 70 cm.
The defensible rule is simpler: follow the manufacturer’s mounting envelope where one exists, model the actual structure when the decision matters, and verify the installed current, match and pattern or coverage.
Why the Single “Roof Null” Formula Fails
The source assigned a first elevation null from antenna height alone. A real reflected path needs at least transmitter height, receiver height, horizontal range, reflection-point geometry, polarization and the reflection coefficient’s magnitude and phase. Roof dimensions, slope, material, edge diffraction and surrounding clutter can be decisive. Nearby coupling is not even the same mechanism as a far two-ray ground reflection.
For a smooth flat two-ray geometry, the exact direct and reflected path lengths can be written as:
rdirect = √(d² + (hT − hR)²)
rreflected = √(d² + (hT + hR)²)
Δr = rreflected − rdirect
Whether Δr produces reinforcement or cancellation also includes reflection phase and the antenna patterns in the direct and reflected directions. The source table therefore cannot predict a universal null at 14.5°, 9.6° or any other angle. Use it neither as a mounting rule nor as evidence that a specific antenna will lose.
Feeder and Mismatch Loss Belong in the Same Budget
Compare at one reference plane. A directional downlink budget in decibels can be written as:
PR = PT − LT,feed + GT(θ,φ) − Lpath + GR(θ′,φ′) − LR,feed − Lother
All gains must use the same reference, and each pattern value must be taken in the direction of the path. If ordinary antenna gain is used, add mismatch separately. If realised gain already includes mismatch, do not subtract it twice.
For reflection coefficient Γ, the accepted-power mismatch loss is:
Lmismatch = −10 log10(1 − |Γ|²)
|Γ| = (VSWR − 1) / (VSWR + 1)
A VSWR of 1.5:1 gives about 0.18 dB mismatch loss; 2:1 gives about 0.51 dB. Those values do not prove that match is unimportant—transmitter foldback, cable heating and multiple reflections can matter—but ordinary moderate mismatch alone does not explain a missing 9 dB. Nor does a low VSWR prove gain or efficiency.
A named cable can erase the paper advantage
“Thin coax” is not a specification. Belden’s current 8259 RG-58 data list nominal attenuation of 5.4 dB per 100 ft at 100 MHz and 12.4 dB per 100 ft at 400 MHz. A 30 m run is about 98.4 ft, so the published 400 MHz value corresponds to roughly 12.2 dB before connector and mismatch effects. Other RG-58 constructions and temperatures differ, but this product-specific example is already larger than a 9 dB antenna advantage.
Measure or calculate the actual cable type, length, frequency, temperature, connector count and age. A taller mast can improve propagation clearance while simultaneously increasing feedline loss; the net result belongs in one budget.
Installed Pattern Is the Missing Specification
The pattern in a catalogue fixture is not the pattern after installation. Important mechanisms include:
- mast and bracket scattering, especially through or close to the active aperture;
- roof, rail, vehicle body, tower, guys and nearby antennas;
- unwanted current on the feedline exterior and support wiring;
- azimuth tilt from asymmetric mounting;
- pattern and match change across a wide amateur band;
- connector, duplexer, lightning-protection and feeder loss;
- polarization mismatch and local multipath.
ITU-R BT.1195 notes that theoretical and measured VHF/UHF antenna performance can differ for practical reasons and calls for verification of both element and complete-system patterns. That is the right evidence boundary here.
Coverage Is a Map, Not a Peak-Gain Number
ITU-R P.1812-8 is the current path-specific point-to-area method for 30 MHz to 6 GHz. Its inputs include actual terrain profile, representative clutter, terminal heights, frequency, location and radio-climatic information. That list explains why “9 dBd reaches farther” is not a complete site prediction.
For a real 2 m or 70 cm service area, evaluate:
- terrain and clutter clearance in each azimuth sector;
- the elevation angle from the antenna to each target area;
- diffraction, reflection and time/location variability;
- base and mobile antenna patterns, heights and polarization;
- both downlink and uplink budgets;
- receiver noise, interference and required signal-to-noise or signal-to-interference ratio.
A broader whip can improve some directions while a collinear improves others. It can also receive a different mix of wanted signal, local interference and environmental noise, so “stronger S-meter reading” and “better SNR” are not interchangeable. Avoid reporting uncalibrated S-units as engineering dB.
A Fair A/B Test
- Define the objective. Choose the routes, repeaters, buildings or grid cells that constitute coverage and the minimum usable metric.
- Record both datasheets. Normalize dBi/dBd and capture pattern files, frequency, tilt, connector, input-power and mounting requirements.
- Characterise the feed systems. Measure insertion loss for each cable/accessory chain or establish delivered feedpoint power. Do not credit the antenna for a better cable.
- Measure the match at the antenna connector. Calibrate to the intended reference plane and sweep each band. Treat match as one diagnostic, not as gain.
- Control installation geometry. Use the same mounting position, polarization, azimuth reference and comparable antenna phase-center height where practical; document unavoidable differences.
- Check exterior feed current. Measure before and after any choke change, because the feedline may be part of the installed radiator.
- Use a stable signal. Hold transmitter power and modulation constant, or use a stable remote beacon. Swap antenna order to expose propagation drift.
- Sample the service area. Log calibrated received level plus noise/interference at repeatable positions, heights and headings. Use medians and percentiles rather than one favorable report.
- Test uplink and downlink. Receiver sensitivity, remote antenna pattern, power and interference may make the two directions different.
- Compare by sector and reliability. Keep the coverage distribution and outage locations; one average can hide a deep installed null.
Choose by Geometry, Not Antenna Height Alone
- High, clear site serving distant near-horizon users: a characterized collinear with the correct beam tilt can be an excellent choice.
- High site serving close users below it: verify the downward elevation pattern, side lobes and null fill; peak gain is insufficient.
- Balcony, attic or rail: no gain class is automatically best. Model or A/B test the actual structure, current path and angular coverage.
- Vehicle roof: a frequency-appropriate monopole near the center often produces a more symmetric azimuth pattern, but roof size, body geometry, mounting and dual-band behavior still matter.
- Long UHF feeder: solve line loss first or place suitable equipment closer to the antenna; paper gain cannot recover power already dissipated in the line.
Engineering verdict: 9 dBd is a valid maximum-gain specification when correctly referenced. Whether it beats 0 dBd is a directional, installed-system and site-coverage question. Normalize the units, subtract real losses, inspect the pattern at the required angles and verify the map.
Authoritative sources checked
- ITU Radio Regulations, 2024 edition and ITU-T K.91 (2024)—gain references, ERP/EIRP and feeder-loss boundaries.
- NTIA/NIJ Antenna System Guide—gain versus directivity, dBi/dBd and monopole ground structures.
- ITU-R BT.1195-1—VHF/UHF array patterns and practical pattern verification.
- ITU-R F.1336-5—reference omnidirectional/sector antenna patterns and gain-beamwidth relationships from 400 MHz upward.
- ITU-R P.1812-8—current terrain- and clutter-aware point-to-area prediction from 30 MHz to 6 GHz.
- Amphenol Procom 422x.09-445-Txx, G-CXL 225-450C and MA 450—manufacturer gain, beamwidth, pattern and mounting examples.
- Belden 8259 RG-58 technical data—frequency-specific nominal feeder attenuation.
- Rohde & Schwarz antenna design guide—VSWR, reflection coefficient, accepted power and mismatch loss.
Mini-FAQ
- Is a 9 dBd collinear always 9 dB stronger than a 0 dBd whip? No. The 9 dB difference applies between their quoted maximum-gain directions at the same reference plane and conditions. Directional pattern, installation and feed-system losses decide the path result.
- Is 0 dBd the same as 0 dBi? No. A gain of 0 dBd equals 2.15 dBi because the dBd reference is a lossless half-wave dipole rather than an isotropic radiator.
- Does low SWR prove that the collinear has its advertised gain? No. SWR describes impedance match at a reference plane. It does not measure radiation efficiency, directional gain, installed pattern, feedline attenuation or coverage.
- Is an end-fed half-wave independent of its coax and surroundings? No. It may work without a traditional ground plane, but its matching network, mounting, coax exterior and attached equipment can affect return current, match and radiation pattern.
- Will adding one or two metres of height always improve coverage? No. Height often improves clutter and horizon clearance, but it also changes path angles, reflections and feeder length. Predict and measure the site-specific net result.
- How should a collinear and whip be compared fairly? Normalize dBi and dBd, hold delivered feedpoint power and mounting geometry constant, measure match and line loss, then compare calibrated signal and noise across the intended service area.