| |
KO6NMZ > TECH 07.06.90 03:09z 320 Lines 14454 Bytes #-13119 (0) @ WW
BID : TG8QHQKO6NMZ
Read: GUEST
Subj: Optimizing the G5RV
Path: SR8BBS<SR1BSZ<EA2RCF<LU9DCE<KC2NJV<N2NOV<KQ6UP<KO6NMZ
Sent: 260607/0253z @:KO6NMZ.#SCA.CA.USA.NOAM [Grand Terrace] DP6.11 $:TG8QHQKO6
From: KO6NMZ @ KO6NMZ.#SCA.CA.USA.NOAM (Maria)
To: TECH @ WW
```
===============================================================================
THE FULL-SIZE G5RV MULTIBAND ANTENNA -- A PRACTICAL GUIDE
Prepared for the Bulletin
de KQ6UP / June 2026
===============================================================================
HISTORY
-------
Louis Varney (G5RV) conceived this antenna in 1946 while looking for a
compact multiband solution for his 100-foot garden in Stony Stratford,
England. He published the design in the RSGB Bulletin in July 1958.
Varney was a former Captain in the Royal Corps of Signals, specializing
in HF interception and direction finding during WWII -- so he knew his
antenna theory cold.
One important historical note that gets lost in the folklore: Varney
ALWAYS specified using an antenna tuner. The idea that the G5RV works
on all bands without a tuner is a myth that came from commercial
marketing decades later. His design goal was a clean match on 20m that
could be pressed into service on other bands with an ATU. Judged on
those terms, it is a remarkably good design for a 1946 hand calculation.
WHAT IT IS
----------
The full-size G5RV is a center-fed doublet -- essentially a flat-top or
inverted-V wire antenna -- fed with a specific length of open-wire
balanced feedline that acts as an impedance transformer. It is NOT a
resonant dipole on most bands. It is a deliberate mismatch that the
feedline partially corrects, with an ATU finishing the job.
Classic dimensions (Varney original):
Doublet wire: 102 ft total (51 ft per side)
Feedline: 34 ft of open-wire line
Feedline Z0: ~525 ohm (original open wire)
Main feedline: 75 ohm coax to ATU (Varney's era)
or balanced line all the way to a balanced ATU
(preferred -- see below)
The antenna covers 80m through 10m with an ATU, and with the right
tuner it will also cover the WARC bands (17m, 12m, 30m).
WHY 102 FEET?
-------------
This is one of the most misunderstood design choices in amateur radio.
A half-wave dipole on 80m needs about 131 feet of wire. A half-wave
on 40m needs about 66 feet. Neither of those lengths works cleanly on
its harmonics when fed with typical balanced feedline.
The 100-105 foot length was actually known before Varney -- it appears
in handbooks as far back as 1937 -- as a length that moderates the wild
impedance excursions you get when a half-wave dipole is used on its
harmonics. On a true half-wave 80m dipole connected to 400-ohm ladder
line, the feedline SWR on 40m can reach 10:1 or worse, presenting
thousands of ohms at the tuner. The 102-foot length avoids the worst
of those excursions across the band range 3.5 to 28 MHz.
What it does NOT do is make the antenna resonant on 80m. At 3.75 MHz,
the 102-foot doublet is only about 78% of a half-wavelength. The
feedpoint impedance on 80m is approximately 19 - j347 ohms -- very low
resistance and very high capacitive reactance. A capable ATU can match
this, but no feedline choice makes it easy. If 80m is your primary
band, you need a longer wire (see recommendations below).
THE FEEDLINE IS NOT OPTIONAL
-----------------------------
Many commercial versions of the G5RV are sold with a short stub of
ladder line or twinlead that terminates directly into coax. This is
where performance falls apart.
The open-wire feedline section is an impedance transformer. It rotates
the antenna feedpoint impedance around the Smith chart as a function of
its electrical length. Different bands see different electrical lengths
for the same physical feedline, so each band arrives at the tuner with
a different impedance. The job of the ATU is to handle whatever shows
up.
If you insert coax after a short stub, you have now fixed the
transformation at whatever the stub length does at each frequency, AND
you have introduced a lossy element into a high-SWR system. On bands
where the impedance is highly reactive -- 15m and 10m in particular --
the SWR on the coax can exceed 20:1. Even a short run of RG-8X loses
several dB at 21 MHz under those conditions. Open wire line at the
same SWR loses a small fraction of a dB. That is not a small
difference -- it can mean the difference between being heard and not.
The correct system is:
Doublet wire
|
Open-wire balanced feedline (all the way to the shack)
|
Balanced ATU (link-coupled or T-match with balun)
|
Rig (50 ohm)
No coax anywhere in the RF path.
FEEDLINE CONSTRUCTION
----------------------
Homebrew open-wire line is easy to build and outperforms commercial
ladder line in several ways:
Wire: #14 solid bare copper (NOT stranded)
Solid holds its shape, has lower resistance, better
surface area for skin-effect current, and makes
cleaner solder joints. Stranded wire creeps under
tension and changes spacing over time.
Spacing: 6 inches
This gives a characteristic impedance (Z0) of
approximately 600 ohms with #14 AWG wire.
Z0 = 276 * log10(2*S/d) where S=spacing, d=diameter
Wider spacing = higher Z0 = lower loss at high SWR.
Spacers: Bic pen bodies work excellently.
Polystyrene/polypropylene -- both low-loss dielectrics
at HF frequencies. Remove the ink cartridge.
Drill or notch near each end, thread wire through.
Space the spacers every 12-18 inches to prevent
sag and maintain consistent Z0.
Twist: A gentle half-turn per foot of lay helps cancel
external field pickup and keeps the line balanced.
Not tight like twisted pair -- just a gentle wind.
Entry: Use a proper feedthrough insulator or standoff at
the shack wall. Keep the open wire away from metal,
wet wood, and gutters. A 6-inch air gap to any
conductor is a reasonable minimum.
Why homebrew beats commercial ladder line:
Commercial 450-ohm ladder line has narrow PVC
spacers. In rain, the PVC surface picks up moisture
and the effective dielectric loss increases. Your
6-inch air-gap line with polystyrene spacers is
essentially unaffected by weather. The wider spacing
also gives you higher Z0 (600 vs 450 ohm), which
means lower current for the same power -- and lower
I-squared-R loss in the wire.
ANTENNA TUNER SELECTION
------------------------
This is where most G5RV installations succeed or fail.
The G5RV presents a wide range of impedances across the HF bands.
NEC modeling of the full-size antenna with 600-ohm open-wire feedline
at varying lengths gives tuner-input impedances ranging from roughly
14 ohms resistive on 80m to over 5000 ohms on 15m in some feedline
configurations. No simple L-network or basic T-match tuner handles
that range reliably on all bands.
RECOMMENDED:
Johnson Viking Kilowatt Matchbox (Part No. 250-30)
This is a link-coupled tank circuit with a tapped inductor and
band switch. The design is fundamentally different from a T-match:
-- The input coax side and output balanced side are galvanically
isolated via the coupling link. No balun needed. True balanced
output with no common-mode current path.
-- The band switch selects different tap points on the main tank
coil, so you are always starting from a coil that is roughly
resonant at the operating frequency. The tuning capacitors then
trim the match. This gives coverage across the WARC bands where
T-match tuners run out of range.
-- Rated for kilowatt-level continuous power. The large air-variable
capacitors and tank coil handle the high voltages that appear on
a mismatched balanced line without arcing.
The downside: it is large. Roughly the size of a microwave oven.
This is not a shack-friendly appliance. But it works on bands that
no other common tuner can handle with this antenna system.
ALTERNATIVE:
MFJ-974HB or similar link-coupled balanced tuner
Gets close to the Matchbox performance in a smaller package.
May struggle at the extreme impedance points (80m low-R, 15m
very high-R) that the Matchbox handles with ease.
NOT RECOMMENDED for this system:
-- Any T-match tuner with a balun on the output
(the balun saturates at high-impedance/high-current conditions)
-- Any auto-tuner (range too narrow, unbalanced output)
-- MFJ-949 or similar with a 4:1 balun on the output
(works on some bands, fails on the extremes)
-- The smaller 250W Johnson Matchbox (250-20)
(fewer coil taps, narrower impedance range -- cannot cover
the same bands as the kilowatt version)
RECOMMENDED DIMENSIONS BY PRIORITY
------------------------------------
NEC computer modeling (5,412 runs across the parameter space) gives
the following guidance. Scores are matchability 0-100 where higher
means easier for the ATU. Weights: 80m x2.5, 40m x2.0, 20m x1.5,
15m x1.0, 10m x0.5.
Priority: 40m primary, other bands secondary
Wire: 99 ft total
Feedline: 43 ft of 600-ohm homebrew open wire
Scores: 80m=5 40m=51 20m=13 15m=9 10m=37
Notes: Strong 40m match (tuner Z ~91 ohms, nearly resistive).
Decent 10m bonus. 80m and 20m need a capable ATU.
Priority: 20m primary (closest to Varney's original intent)
Wire: 92 ft total
Feedline: 42 ft of 600-ohm homebrew open wire
Scores: 80m=4 40m=27 20m=70 15m=4 10m=8
Notes: Excellent 20m match (tuner Z ~59 ohms, nearly perfect).
Other bands possible with the Kilowatt Matchbox.
Classic 102 ft Varney, homebrew 600-ohm feedline, optimal tap:
Wire: 102 ft total
Feedline: 42 ft of 600-ohm homebrew open wire
Scores: 80m=5 40m=52 20m=10 15m=7 10m=34
Notes: Best 40m performance from the 102-ft wire.
Surprisingly close to Varney original on composite score.
Classic 102 ft Varney, original dimensions (525-ohm open wire):
Wire: 102 ft total
Feedline: 34 ft of ~525-ohm open wire
Scores: 80m=9 40m=22 20m=41 15m=18 10m=5
Weighted composite: 19.9 / 100
Notes: Varney's original scores well for a 1946 hand calculation.
Primarily a 20m antenna as designed. All other bands
require a capable ATU. Holds its own against computer-
optimized designs when judged on the same metric.
If 80m is a serious operating priority:
The 102-ft doublet is fundamentally limited on 80m.
Antenna feedpoint impedance at 3.75 MHz: ~19 - j347 ohms.
Best achievable match score with ANY feedline: ~39/100.
SOLUTION: Extend the wire to 130-135 ft.
At 131 ft the doublet is a half-wave on 80m and the feedpoint
impedance jumps to a manageable 50-75 ohms resistive.
Everything else (40m through 10m) still works with the ATU.
If space allows, this is the better antenna.
INSTALLATION NOTES
-------------------
Height: Higher is always better. Minimum useful height is about
35 feet for the center (apex if inverted-V). The antenna
models assume 35 ft apex height throughout.
Geometry: Flat-top is marginally better than inverted-V electrically,
but an inverted-V is often more practical. The inverted-V
with 35 ft apex and ends at 10 ft gives a real horizontal
span of about 89 ft for a 102-ft wire -- workable in most
suburban lots.
Ends: Keep the wire ends at least 10 ft off the ground to avoid
ground losses and contact hazards. Higher end height is
better and also improves the take-off angle slightly.
Wire: #14 solid bare copper is the standard choice.
Copper-clad steel (Copperweld) is stronger for long spans
but slightly higher loss. Fine for most installations.
Center insulator:
Use a good quality center insulator rated for the feedline
tension plus wind load. The open-wire feedline hangs from
this point so it sees mechanical stress as well as RF.
Egg insulators at the ends, proper strain insulator at center.
Common mode:
With a properly balanced ATU like the Kilowatt Matchbox,
common-mode current on the feedline is inherently suppressed
by the balanced topology. If you experience RF in the shack,
add a few ferrite beads (Fair-Rite #31 material) around
the power cord and any audio cables near the rig -- the
problem is almost always conducted, not radiated.
WHAT VARNEY GOT RIGHT
----------------------
Computer analysis (NEC2c, 11,480 simulation runs) confirms that
Varney's original design scores 19.9 out of 100 on a composite
matchability metric -- solidly middle of the pack against modern
computer-optimized designs. For a 1946 hand calculation targeting
primarily 20m DX performance, this is a remarkable result.
The ZS6BKW (1980s computer-optimized variant, 93 ft wire / 36 ft
feedline / 450-ohm ladder line) scores 18.7 on the same metric --
actually slightly LOWER than Varney's original. The ZS6BKW was
optimized for a different goal (minimum SWR into 50-ohm coax without
a tuner on specific bands) and wins on that metric. But as a system
for use with a capable balanced ATU, Varney's intuition from 1946
holds up extremely well 80 years later.
The bottom line: the G5RV is exactly what Varney said it was.
A space-efficient 20m antenna that works on other bands with a tuner.
It delivers on that promise just as well today as it did in 1946.
Pair it with a proper balanced ATU, run open-wire feedline all the
way to the shack, and you have a capable multiband HF antenna that
will serve you well for decades.
===============================================================================
73 de KQ6UP
Modeling: NEC2c (Numerical Electromagnetics Code)
Ground model: Sommerfeld/Norton, medium earth (sigma=0.005, er=13)
Feedline model: lossless transmission line, Z0=600 ohm, VF=0.97
All scores: matchability index 0-100 (100=perfect 50-ohm match)
===============================================================================
Read previous mail | Read next mail
| |