70 cm collinear antenna: calculation and building from coax

When a quarter-wave whip is no longer enough, but you don't want to put up a bulky array — the collinear steps onto the stage. It's a vertical antenna with gain that you can build literally out of a piece of coaxial cable. The price you pay for that gain is how finicky it is to reproduce and tune: you simply can't manage without an analyzer here. Let's break down how a collinear works, how to calculate section length with the cable's velocity factor, and when it actually pays off.

What a collinear is and why you'd want one

A collinear is several in-phase half-wave sections lined up in a single vertical line and radiating in phase. By adding their fields together in the horizontal plane, the antenna delivers more gain than a single whip: typically 5–9 dBi depending on the number of sections.

Physically, the gain doesn't come "out of thin air" — it comes from the shape of the radiation pattern. The collinear presses the lobe down toward the horizon: the energy a whip wastes radiating into the sky and into the ground gets gathered into a flat "pancake" along the earth. That's why a collinear is especially good for long-distance work over flat terrain — plains, fields, water. In the mountains or strongly broken terrain, however, the narrow vertical lobe may shoot past your correspondent, and there's no advantage to be had.

How a coaxial collinear works

The easiest version to reproduce is a coaxial collinear made from lengths of cable. The antenna is assembled from half-wave sections of coax in which, at every joint, the center conductor and the braid swap places (re-phasing): the center conductor of the lower section is soldered to the braid of the upper one, and vice versa.

This "crossover" is needed so that the currents on the outer surface of the braid flow in the same direction in every section — that is, in phase. Half a wave in the cable provides exactly the required phase shift, and with the swap-over every section radiates in step. A quarter-wave whip (the end radiator) is usually added on top, and the feed point and decoupling go at the bottom.

Calculating section length: velocity factor is critical

The main trap of the coaxial collinear: inside the cable the wave travels slower than in air. How much slower is set by the cable's velocity factor (VF). For cheap RG-58/RG-174 it's around 0.66, for foam PTFE — 0.80–0.85, and higher for quality cables. So a half-wave section of cable is calculated with a VF correction:

Section length (mm) = 150000 / F(MHz) × VF

where F is the working frequency and VF is the velocity factor of your specific cable (take it from the datasheet, not "from memory"). The end quarter-wave whip, by contrast, runs in air and is calculated without VF — like an ordinary 1/4λ.

Example for the center of the 70 cm band and a cable with VF = 0.66:

Number of sections and gain

Each added in-phase section narrows the vertical lobe and adds gain — but the return diminishes as the number of sections grows:

The figures above are a guideline in dBi and for ideal addition. With a coaxial collinear, some of the gain is "eaten" by losses in the section cable itself, so the real improvement is usually more modest than the rated value: the classic 8-section N1HFX antenna is quoted as "9 dB" (relative to a dipole), while its shortened 4-section variant loses another roughly 3 dB. In practice people usually fit 2 to 4 sections; more than that is rarer and only when the height allows and you have the means to tune it. The longer the antenna, the narrower the vertical lobe — and the more precisely it has to be kept strictly vertical, otherwise the "pancake" tilts sideways and misses the correspondent.

Matching, the choke and decoupling

The feed point of a coaxial collinear is at the bottom, and its impedance depends on the number of sections and is usually above 50 ohms. In short 2–4-section versions, with careful geometry, it can be brought close to 50 ohms, while longer antennas use a quarter-wave matching section (stub) — it works as a transformer and pulls the impedance down to 50 ohms. You can fine-tune the match by shifting the feed point a couple of centimeters: higher raises the impedance, lower drops it.

A mandatory element is a blocking choke at the input. Without it, RF current "creeps out" onto the outer braid of the feed cable, the cable itself becomes part of the antenna, the pattern wanders, and the SWR starts to depend on how you hold the feedline. The choke is made like this: a few turns of the feed cable wound into a coil at the antenna base, or ferrite rings/beads threaded onto the cable. This decouples the antenna from the feedline and stabilizes the tuning.

Tuning by SWR

Here a collinear is noticeably harder than a simple whip: there are more points where "something is off", and the interaction between sections isn't obvious. So an analyzer is essential — you can't tune this by ear or with a two-needle SWR meter.

1. Build the antenna strictly to the calculation with VF, leaving some spare on the end whip.
2. Connect the NanoVNA and sweep SWR across the 420–450 MHz band — look for the minimum (resonance). Never feed transmitter power into the analyzer — it instantly burns out the measurement bridge. An antenna is tuned on a NanoVNA only with the radio switched off.
3. If the minimum is below the working frequency, the antenna is "long" — trim the end whip by 2–3 mm at a time. Above it — the sections are too short (most often a VF miss; it's easier to recalculate).
4. Fine-tune the match by shifting the feed point, aiming for SWR 1.2–1.5 over the desired range. Don't expect a perfect 1.0 on a collinear — that's normal.
5. Check that, once the choke is fitted, the SWR doesn't change when you wiggle the feedline. If it changes, the choke is insufficient.