DX Is Not Always Low Angle
DX Is Not Always Low Angle
DX describes a distant or difficult contact in its operating context. It does not assign one elevation angle to the path, and an antenna does not launch every useful signal at one magic takeoff angle.
“DX means low angle” is a useful shortcut until it becomes a design rule. Long paths often reward energy near the horizon, but the useful angle belongs to the propagation mode, frequency, path length, ionosphere, terrain and antenna patterns at both ends. My receive antenna that wins on 80 metres does not have to win on 40 metres—and a genuine A/B reversal is not a contradiction.
My practical rule: choose an antenna for the azimuth and elevation distribution that reaches the station at that moment, then judge it by usable SNR. Distance alone does not tell you the angle, and an S-meter alone does not tell you the result.
DX Is an Operating Description, Not a Ray Angle
Amateur radio has no single worldwide distance at which a contact becomes DX. On one band or from one country, a path may be routine; from another station it may be rare. What remains technically useful is the separation: distance and difficulty describe the contact, while elevation angle describes the direction of departure or arrival at a terminal.
For a particular ionospheric mode, a longer single hop generally needs a shallower terminal angle than a shorter hop through a comparable virtual height. Real HF paths are not one fixed mirror. E- and F-region modes, changing layer height, one-hop and multi-hop solutions, ionospheric tilt and long- or short-path geometry can place usable energy at different angles.
ITU-R P.533 therefore calculates elevation angle as part of each predicted HF mode instead of assigning one angle to “DX.” It then combines the path with transmitting and receiving antenna gain in the appropriate directions. That is the right mental model: the propagation mode supplies a distribution of opportunities, and the installed antenna pattern weights them.
An Antenna Has an Elevation Pattern, Not One Takeoff Angle
A vertical plot contains lobes, nulls and gain over a range of elevations. Calling the main-lobe peak “the takeoff angle” can hide useful energy above and below it. It can also hide a deep null at the angle a path needs.
The installed pattern depends on frequency, height in wavelengths, conductor geometry, current distribution, ground parameters, terrain profile, nearby structures and unintended common-mode current. At very low elevation, modest changes in terrain or the land–sea boundary can matter. A free-space pattern, a flat-earth model and the antenna in your garden are not interchangeable evidence.
Low elevation is often valuable for long-haul F-region work, especially when the path and noise environment support it. It is not automatically better. Gain placed at an angle with no supported mode does not make a contact, and lowering a main lobe can trade away energy needed for shorter or higher-angle paths.
Why the Fixed Angle Table Had to Go
A universal table that maps 3–10 degrees to intercontinental DX and 25–60 degrees to regional work looks precise but leaves out the variables doing the work. The same ground distance can be supported by different virtual heights and hop counts. The same band can change mode during the hour. Great-circle distance alone does not uniquely determine terminal elevation.
Use a propagation prediction as a hypothesis, not a certificate. Record the station coordinates, frequency, time, solar and geomagnetic conditions, short- or long-path choice and the modes the model predicts. Then compare those angles with the measured or modelled installed pattern. A prediction using an optimistic 3-degree antenna gain is not evidence that the real site delivers it.
| Question | Evidence that helps | What it does not prove |
|---|---|---|
| Which angles can the path support? | Mode-aware HF prediction for the path, time and frequency | That one predicted mode will dominate every minute |
| Which angles does the antenna favour? | Installed pattern model or calibrated pattern measurement | That the main-lobe peak is the only useful angle |
| Which antenna gives better copy? | Fast or simultaneous calibrated A/B data with signal and noise separated | That the winner has more gain at every angle |
| Did the seaside site cause the result? | Terrain/ground model plus repeatable inland/coastal or restored-baseline comparisons | That proximity to salt water guarantees low-angle gain |
Why a 40-Metre Winner Can Lose on 80
A customer can report that antenna A won on 40 metres while antennas B and C gave better copy on 80 metres at the same seaside site. I take that observation seriously. I do not turn it into a universal ranking.
Frequency changes the antenna's electrical size, current distribution, pattern, nulls, feed behaviour and common-mode sensitivity. The propagation paths available on 40 and 80 metres can also differ at the same time. A 40-metre signal described as DX by the operator may arrive at an intermediate angle while an 80-metre target uses another mode and azimuth.
The site adds another layer. A coastline can reshape the elevation pattern, but the result depends on antenna polarization and height, shoreline bearing, beach slope, land and water conductivity, terrain behind the antenna and the path azimuth. “Tested near the sea” is not a calibration.
That is why the honest conclusion from an A/B reversal is simple: the antennas sampled the combination of path, pattern and noise differently. To identify the cause, we need more than one signal and one day.
On the Low Bands, Copy Often Beats Gain
Receive performance is not the same as transmit efficiency. On 160 and 80 metres, atmospheric and man-made noise can be large enough that a physically small or intentionally lossy receive antenna still keeps external noise above the receiver's internal noise. If it rejects more unwanted field than wanted field, it can improve copy despite lower absolute output.
The boundary must be measured. ITU-R P.372 separates external radio noise from receiver-system noise; it does not say every low-output antenna has adequate sensitivity. Check antenna noise at the receiver reference plane, cable loss, preamplifier noise and gain, filtering, overload and intermodulation. If the receive chain becomes internally noise-limited, further attenuation does cost SNR.
This is why I call a receive antenna a spatial filter. Its pattern weights wanted and unwanted fields by direction and elevation. A Beverage, loop, flag, phased array or active element can be useful because of what it rejects—not because its S-meter number is impressive.
One Antenna Cannot Cover Every Path and Null
A single installed antenna has one instantaneous relationship with bearing, elevation, polarization, local noise and common-mode coupling. HF signals do not share one relationship:
- short path and long path arrive from different bearings;
- one-hop and multi-hop modes can use different elevations;
- the supported mode changes with frequency, season, hour and ionosphere;
- local noise sources occupy their own directions and polarizations;
- an antenna null can help one signal and erase another; and
- feedline current can distort the pattern and import noise from the station.
A serious receive station therefore keeps genuinely different views: another azimuth, a different elevation response, another polarization or a quieter location. The number of antennas is not the achievement. The useful difference between them is.
A Field Comparison That Can Explain the Result
- Identify the path. Log callsign, coordinates, bearing, distance, short/long path, frequency, time and signal source.
- Describe both antennas. Record geometry, orientation, height, ground, terrain, feedline, common-mode control and receive electronics.
- Use simultaneous channels or fast switching. Slow comparisons confuse fading with antenna behaviour.
- Calibrate the branches. Match receiver bandwidth, AGC, attenuation, gain, filters and ADC scaling.
- Record wanted signal and noise separately. Compare SNR or decoding outcome, not S-meter peak alone.
- Check more than one station. Use targets at several bearings, distances and predicted mode angles.
- Map common mode. Clamp the complete feedlines and restore the original routing after each intervention.
- Repeat A/B/B/A. A restored baseline exposes changing propagation and accidental station changes.
- Keep the claim local. A repeatable site result is valuable without pretending to rank every antenna everywhere.
Bottom line: low-angle energy is an essential part of long-haul HF work, not the definition of DX. Match the installed elevation and azimuth pattern to the modes the path can support, and keep enough receive views to escape the wrong lobe, null or noise direction.
Primary sources checked
Mini-FAQ
- Does DX always mean low angle? No. DX describes a distant or difficult contact in context; the terminal elevation depends on the supported propagation mode and path geometry.
- Does an antenna have one takeoff angle? No. It has an elevation pattern with gain, lobes and nulls over a range of angles.
- Why can one antenna win on 40 metres but lose on 80? Frequency changes both antenna behaviour and the propagation modes available at that time.
- Is the strongest S-meter reading the best receive result? No. Compare wanted-signal SNR or decoding with identical receiver settings and calibrated branches.
- Does a seaside test prove a saltwater advantage? No. Shoreline bearing, terrain, polarization, height, ground and path direction must be included and compared with a baseline.
- Why keep several receive antennas? Different patterns, bearings, polarizations and noise pickup give the station another view when one antenna has the wrong lobe, null or noise direction.