My TX Antenna Strategy from 160 Metres to 4 Metres

When people ask for my go-to transmit antenna on each band, I do not start with giant arrays or towers everywhere. Even when space is available, I prefer antennas that are mechanically realistic, electrically understandable and strategically complementary. That last word matters most: the station is the antenna system.

No single geometry is best for low-angle DX, higher-angle regional work, short skip and every changing ionospheric path. A layered station lets me choose a different current distribution when the operating job changes. This is transmit thinking; my receive strategy is deliberately separate because wanted-signal SNR, local noise, overload and diversity impose different priorities.

The Band-by-Band DX Layer

The list looks mixed, but the reasoning is consistent. I change antenna family when another geometry becomes mechanically simpler or gives me a more useful installed current distribution. I do not change because a label promises DX.

Why I Keep the Low Bands in Inverted-L Territory

On 160 and 80 metres, a full-size ground-mounted quarter-wave vertical and a low-loss radial field demand serious land and wire. A dedicated inverted-L lets me put part of the radiator vertically and fit the remaining conductor horizontally. On 40 and 30 metres, the same family remains practical while giving me a deliberate mixture of vertical and horizontal current.

That does not make every inverted-L automatically efficient or low angle. Height, the vertical-to-horizontal proportion, bends, ground, nearby conductors and current on the coax exterior all affect the result. The horizontal section may contribute useful higher-elevation field, particularly when it is low in wavelengths, but “inverted-L” alone does not specify the pattern.

I also do not describe an EFHW as a no-counterpoise antenna. The installed end impedance must be measured, the matching network must suit that load, and the return current must close through an intentional conductor or local reference rather than being left accidentally to the feedline and shack. An impedance transformer and a common-mode choke do different jobs. The choke position follows the measured exterior-current path; it is not copied from a universal fraction-of-a-wavelength rule.

Twenty Metres Is Where the Raised Quarter-Wave Becomes Attractive

At 20 metres, a full quarter-wave radiator and elevated radial system are physically manageable. I can place the feedpoint clear of immediate clutter, route the feedline away from the radial plane, and inspect the current on each radial and on the coax exterior. That makes a raised quarter-wave a clean engineering starting point for an omnidirectional DX layer.

“Raised” is not a magic height. The radial plane, soil, mast, feedline route, nearby roofs and terrain all change the realised pattern. I choose the height from a model and the site, then measure. ITU-R BS.705-2 explicitly separates theoretical antenna characteristics from the deviations introduced by ground, topography and surrounding structures; that distinction matters in an amateur garden too.

Why Seventeen and Fifteen Metres Keep Their Half-Waves

For 17 and 15 metres, my preferred compact DX layer is a raised end-fed half-wave with an LC matching network and a controlled return arrangement. These radiators are short enough to elevate cleanly without needing a large ground-mounted radial field. The LC network is selected from the measured complex load at the installed feedpoint—not from a copied turns count or a generic “EFHW” recipe.

The network must be checked for circulating current, RF voltage, component Q, loss, temperature and voltage spacing at the intended power and duty cycle. A good transmitter-side SWR says that the network transformed the load. It does not say that the network is cool, the feedline exterior is quiet or the far-field pattern is useful.

Why I Use Five-Eighth-Wave Verticals from 12 Metres Up

A full-length five-eighth-wave radiator becomes mechanically reasonable on 12, 10, 6 and 4 metres. That is the range where I use the geometry as a simple omnidirectional TX option with a useful low-elevation component. It is not a one-lobe quarter-wave: the longer current distribution can produce more than one elevation lobe, and the balance between them depends on the driving point, return system, height and surroundings.

The current RF.Guru VertiCore five-eighth-wave models implement this personal choice on those four bands with a band-specific full-length radiator, an offset driving point and an elevated four-radial plane. The nominal input is established by that radiator/radial geometry; the 1:1 balun controls unwanted common-mode current rather than supplying the impedance transformation. The finished assembly is measured, but no model name can guarantee one take-off angle at every site.

On 6 and 4 metres, a five-eighth-wave vertical remains my omnidirectional choice. When the job becomes extreme weak-signal work on a known bearing, a directional array is another system category entirely. My vertical is not presented as a substitute for that aperture or directivity.

The Regional Layer Uses Loops

I do not force the DX layer to cover every closer path. For regional work, short skip and occasions when higher elevation angles are useful, I keep a loop layer available: the multiband DeltaRex, or a band-specific SolidDelta where a monoband installation makes sense.

This preserves the original reason those antennas are in my station plan: they give me a different installed pattern to compare with the inverted-L and vertical layer. It does not mean every delta loop is an NVIS antenna or that a low loop always beats a vertical for regional work. Loop perimeter, shape, feedpoint, height in wavelengths, ground, surrounding conductors and feedline current determine the pattern.

DeltaRex is the multiband closed-loop option in this architecture. SolidDelta is the monoband loop option. I retain those roles without turning product-page dimensions, impedance, power, choke position, gain, efficiency or DX/NVIS wording into a guaranteed station result.

Propagation Chooses the Useful Lobe

My labels “DX layer” and “regional layer” describe operating intent. They are not fixed ionospheric boundaries. ITU-R P.533 calculates useful HF elevation angle from path length, propagation mode and effective reflection height. The required angle changes with frequency, time, ionospheric state and path.

That is why the station needs complementary patterns rather than one slogan. A low-elevation lobe is useful only when the path supports it. A higher-elevation lobe is useful only when it places energy into a supported mode. The antenna and the ionosphere must meet at the same angle.

How I Verify the Layered Station

NEC can calculate currents, input impedance, coupling and radiated fields for a declared geometry and ground model. ITU-R BS.705-2 also documents the importance of ground and the practical environment. Neither removes the need to compare the installed station.

• Document every radiator, radial or return conductor, support, mast and feedline route.
• Measure complex impedance at a declared reference plane and account for feedline loss.
• Map exterior-shield current at repeatable positions on every operating band.
• Inspect matching-network temperature and voltage/current margin at realistic duty cycle.
• Model the complete azimuth and elevation pattern, including nearby conductors that matter.
• Compare antennas with the same transmitter power reference using A/B/B/A field or received-signal measurements across several paths.
• Verify antenna-switch isolation, unused-feedline behaviour and transmitter interlocks before multi-radio operation.

A single contact proves that a path existed. It does not isolate antenna gain, efficiency or take-off angle. The strategy earns its place when repeated measurements show that the selected antenna supplies the useful field for its assigned job.

Questions, antenna-factor records or height trials to share? Contact RF.Guru.