OS5Z “Dreamer” Antennas

Experimenting with antennas is one of the best parts of amateur radio. Weird shapes, unusual feeds, and “what if I try this?” prototypes are how you learn what matters (height, ground, losses, pattern) and what doesn’t (marketing words, pretty SWR graphs, “it feels louder today”).

Where things go off the rails is when an experiment becomes an “invention” with claims like 1:1 SWR on any band, quasi-omni coverage, rhombic-class gain (9 dBd… 14 dBd “certain”), and “beats other antennas.”

Those are extraordinary claims... and they need controlled measurements and clear definitions. Otherwise, it’s very easy to “prove” almost anything with SWR, reverse-beacon maps, and one good day of propagation (or even a bad one).

This article is a deep dive into what the OS5Z's concept appears to be electrically, why the “nice SWR” is not magical, and why the performance claims (as stated) don’t make physical sense.

Source note: This analysis leans on a publicly available write-up with diagrams and a Q&A interview: La Gigazette n°179 (UBA Waterloo). And a publication in UBA's CQ-QSO March-April 2026

What an “OS5Z antenna” is, according to the published description

In the published description, an OS5Z antenna is presented as a compact wire antenna system with a double lozenge (double diamond / double rhombus-ish) top structure, multiple coils integrated into the wire layout, and... crucially... an additional wire connected to the coax shield that drops down and then runs near ground with band switching along its length.

Specific claims attributed to the write-up include an input impedance around 50 Ω ±1–2% from roughly 1.6 to 28.5 MHz (depending on the selected return-wire length), and a statement that from about 9 MHz upward the braid-connected wire should be kept around 10 cm above ground to obtain a 1:1 match.

The same Q&A describes a quarter-wave approach for the braid-connected wire, with switch placement based on the familiar 300/f/4 rule-of-thumb. The radiation is described as “quasi omni” or a four-leaf clover... but the author also admits there was no modeled/measured radiation pattern plotting via software, relying instead on QSOs and Reverse Beacon observations.

First reality check

A clean SWR curve is not a trophy. It mainly tells you how forward and reflected waves compare on the feedline. It does not tell you how efficient the antenna system is, how much power is heating ground/coils/insulation, whether the feedline itself is radiating, or what the pattern really looks like.

Match and efficiency are different knobs:

• How much power the system accepts (matching / VSWR)
• How much accepted power is radiated (radiation efficiency)

So when someone’s main evidence is “the SWR is fantastic,” skepticism is justified.

Second reality check

OS5Z’s “secret sauce” is not fractal... it’s a tuned return path.

Operationally, the published description reads like this:

• The coax center conductor feeds the top structure
• The coax shield feeds a separate wire that is explicitly made about ¼-wave long (via switching) and then runs close to ground

That is an unbalanced antenna system. The braid-connected wire is not optional... it is the RF return path (a counterpoise by another name). When the return path is deliberately tuned per band, it’s no surprise the feedpoint can be coerced toward something that looks like 50 Ω.

How you can get a perfect 1:1 SWR and still waste lots of power

The classic trap is that the feedpoint resistance is:

R_total = R_radiation + R_loss

If your design increases R_loss, the feedpoint can land nicely near 50 Ω even if a meaningful fraction of your power is being burned as heat (ground loss, coil loss, insulation loss, near-field loss).

Now tie that back to the OS5Z's description: coils are incorporated into the structure, and the return wire is used close to ground (and is even described as height-critical above ~9 MHz). That’s exactly the kind of setup where “amazing SWR” can appear while the system quietly depends on loss and/or unintended radiators.

Why the “fractal rhombic” framing doesn’t hold up

At HF, “fractal” often means “packing wire into space.” That can reshape impedance and create multiple resonances. It does not break the normal trade-offs between size, bandwidth, efficiency, and gain.

A true HF rhombic gets its performance from aperture... physical size in wavelength terms. A compact wire that merely looks diamond-shaped is not automatically a rhombic in the classic sense. Shape words do not create aperture.

The gain claims

The published Q&A contains two statements that are hard to reconcile:

• The antenna is described as “quasi omni” or a “four-leaf clover,” and the author states there is no modeled/measured radiation pattern plotting... just inference from QSOs and Reverse Beacon reports.
• In the same discussion, gain numbers are asserted in the 9–14 dBd range, explicitly compared to “true dBd like a rhombic.”

There’s also a specific 21 MHz example that often gets repeated: a top structure quoted around 42.856 m is said to “resonate” when the braid-connected wire is about 3.6 m, followed by large dBd claims. Without modeled or measured patterns and efficiency measurements, those numbers are not a defensible conclusion.

Why “quasi omni” + “14 dBd” doesn’t add up: high gain implies high directivity... concentrating energy into fewer directions. An antenna cannot be “nearly omni” and also have beam-class gain in free space unless something else is being counted (measurement error, installation-specific ground reflections, or unaccounted radiators).

A smaller internal inconsistency

The same narrative that emphasizes a height-critical near-ground return wire (around 10 cm above ground for higher HF) also claims little to no influence from dry versus wet soil. Near-ground coupling and ground loss are well known to depend on soil conductivity and moisture, so that “no influence” statement fails a basic physics smell test for a design that leans on ground coupling.

The hidden elephant

Because the system is unbalanced and intentionally uses the coax shield as part of the antenna system, you are always in danger of:

• the outside of the coax becoming part of the radiator
• the pattern changing with feedline routing
• on-air results changing when you add a choke or move the cable

This is not unique to OS5Z's... it’s a general property of many unbalanced designs. If an antenna “works amazingly” but only in one installation where the coax runs a certain way, that’s a classic clue that the system is radiating through more than the part you think.

The Sixpoles / Octopus family

In portable stories you’ll also see “sixpoles,” “octopus,” and clip-on reconfiguration claims... sometimes even framed as “multiple dipoles” or “adding up” element lengths into a longer radiator.

A useful mental model for many multi-arm hub antennas is the crossed-dipole / turnstile family: you often trade peak gain for more even azimuth coverage (fewer deep nulls), which can feel “better” in portable activations without creating free dB out of nowhere.

From an RF viewpoint, these structures often behave like some mix of:

• Fan / cage / parallel-wire effects (effective “fatter” conductors, broader usable bandwidth, impedance changes)
• Hub-fed multi-arm radiators (pattern changes with height and symmetry, often smoothing deep nulls)
• Strong mutual coupling (arms are not independent; the system behaves as a coupled structure)

What these can legitimately do

• Be more forgiving in portable setups where height and orientation are limited
• Broaden bandwidth and make matching “look easy” thanks to effective diameter and coupling
• Change pattern shape enough to be better for a specific scenario (NVIS vs DX, local terrain, nearby structures)

What they cannot do without becoming a true phased array

• They do not magically produce large net gain over a properly installed reference antenna without paying the price in directivity and controlled phasing
• They do not become “3/2 wave” or “5/4 wave” just because someone adds arm lengths as if they were in series. Arms are in parallel at a common feed region and are mutually coupled.

So are these “dreamer antennas”?

If “dreamer” means “made-up and useless”... not necessarily. Many compromise antennas still make plenty of QSOs, and experimentation can produce something genuinely useful for a specific constraint set.

If “dreamer” means “the claims are far ahead of the evidence”... then yes, the OS5Z's narrative (as published) fits that label: it asserts 1:1 SWR across bands while using a band-switched quarter-wave return path, asserts quasi-omni behavior while also asserting up to ~14 dBd gain without modeled/measured patterns, and leans heavily on SWR and reverse-beacon-style inference despite the known traps.

How to test these claims honestly

If you want a fair evaluation (without getting trapped by SWR or one lucky activation), here’s a simple plan that cuts through most illusions:

• Do a choke toggle test... measure SWR and on-air results with and without a real feedpoint choke. If the antenna “changes personality,” the feedline was part of the radiator.
• Do a WSPR A/B test... alternate antennas every two minutes on the same band with the same power and duty cycle, then compare SNR distributions across many receivers.
• Measure impedance at the feedpoint... not “SWR at the shack.” Look at R and X where the antenna is actually fed.
• Compare against a boring reference built to the same constraints (same height, same mast, same environment). If the “weird antenna” wins repeatedly, it’s a decent design for that use case.

Takeaway

• Experimenting is good... universal performance claims need controlled measurements.
• OS5Z's, as described, is best understood as an unbalanced antenna system with a band-switched quarter-wave return path and a compact loaded top structure... not a “fractal rhombic that breaks HF physics.”
• “Nice SWR” is expected when the return wire is deliberately tuned... and it does not prove gain or efficiency.
• The published claims of quasi omni plus up to ~14 dBd (rhombic-class) are not supported by the type of evidence shown and are internally inconsistent as stated.
• Multi-arm “Sixpoles/Octopus” antennas can plausibly feel better than a single dipole in portable work because they reduce azimuth nulls... but that’s pattern smoothing, not free gain.

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 via our RF.Guru contact page for antenna and RF questions.