W3EDP antenna--an analysis


From: Charlie Lofgren (clofgren@BENSON.MCKENNA.EDU)
Date: Sat Mar 07 1998 - 21:41:31 EST

There's been a lot of recent interest on qrp-l in the W3EDP
antenna. This is an antenna with a 84 foot wire on one side of
the output port and a 17 foot wire on the other side. The 17
foot side is sometimes called a "counterpoise." Here's my "take"
on the antenna's operation, along with a suggestion or two:

To begin, the "counterpoise" isn't a real counterpoise, or at
least it doesn't function like the usual counterpoise (except on
20 meters). The normal function of a counterpoise is to
establish a point of minimum rf voltage or "ground" potential. A
quarter-wave length of wire does this when the far end is open.
(That's assuming it isn't detuned by nearby objects or by the
ground itself.) Except on 20 meters, where it is approximately a
quarter-wave long, the "counterpoise" part of the W3EDP antenna
doesn't fit the bill here. The best approach is to forget about
the short wire as a counterpoise.

To understand the W3EDP, instead conceive of the short side as
one side of a feedline that's been separated or pulled apart from
the other side of the feedline.

Now in your mind move the short side so that it's parallel to the
first 17 feet of the long side and anywhere from several inches
to a foot or so away. What you have is a section of feedline.

This gives you a standard *end-fed zepp* (in other words, a true
zepp antenna, as in zepplin flying), with a fundamental frequency
of about 7 MHz. The flattop portion is 67 feet long, and the
feedline is 17 feet long. The impedance of the feedline, which
is not critical, is somewhere in the vicinity of 500-800 ohms,
depending on wire diameter and spacing.

An end-fed zepp will work on its fundamental frequency and on odd
and even harmonic frequencies (that is, where the flattop is an
odd or even multiple of a half wave). With our W3EDP-derived
end-fed zepp, the antenna will work satisfactorily on 40, 20, 15,
and 10 meters.

The principle of operation is this: At the feedline end of the
half wave flattop (or multiple half wave flattop), the impedance
is *very* high. The impedance at the antenna end of the *open*
side of the feedline is also *very* high. (Were it not for
capacitive coupling to space and various objects, the impedance
at the antenna end of the open side of the feedline would be
infinitely high, but in reality the current never quite falls to
zero.) If the length of the flattop is properly adjusted, then
the currents on the two sides of the feedline are *roughly* in
balance, but out of phase, so not much radiation occurs from the

(If you're familiar with a j-pole antenna, it's the same
principle of operation. And with both the end-fed zepp and the
j-pole, there's controversy over whether they work quite as
they're alleged to. But that's another story, getting into fine
points of feedline balance and what happens at the antenna end of
the open side of the feedline. Also, the resonance of the
flattop on harmonic frequencies isn't exactly an odd or even
multiple of the fundamental frequency, owing to differing impacts
of nearby objects and of end effects. But that, too, can be put
aside in understanding why the W3EDP works.)

Now, if you start pulling the feedline apart, you start to get
more radiation from the two sides of it. If you move the short
side out of the vicinity of the long side, you don't have a
feedline effect at all and you have a W3EDP antenna. But--and
this is crucial to the W3EDP design--the currents at the
transmitter/tuner ends of the two wires are still roughly equal
and roughly out of phase. This means that you can connect them
to a link-coupled tuner without serious imbalance. And this
remains true on harmonic frequencies.

This also explains why, if you try to operate a "standard" W3EDP
on 30 meters, you may find significant hand-capacitance effects
as you try to adjust the tuner.

If you follow through the foregoing analysis, you will see that
there is nothing absolute about the 84 foot and 17 foot
dimensions. In general, what you need is a short side of x feet,
and a long side of (x + y) feet, where y is a half wave at the
lowest frequency. For example, you could set the two sides at 22
and 89 feet, for fundamental operation on 40 meters, and harmonic
operation on 20, 15, and 10 meters. This would likely not be a
good length for 15 meters, however, because the impedance at the
tuner would be quite high. (Think of the input impedance of a
half-wave length of feedline that's terminated in a high
impedance.) Indeed, the "standard" W3EDP is likely to have a
high feedpoint impedance on 10 meters, just as in the case of an
end-fed zepp fed with a half wave feedline. For 10 meters, the
22 and 89 foot dimensions would probably be better in terms of
keeping the impedance at the tuner within a satisfactory range.

An idea for including 30 meters in the W3EDP design is to have a
second short side to switch to. With the standard 84 foot length
for the long side, a short side of 38 feet should give reasonable
balance on 30 meters (the rough conceptual equivalent of a 46
foot flattop and a 38 foot feeder).

There is one *important* point to keep in mind. Unless the short
side is run fairly close to and parallel to lower portion of the
long side, there will be significant radiation from the short
side. This means that it is best *not* to have it on or close to
the ground (in the fashion of the usual counterpoise). If it's
close to ground, it will be radiating into a rather lossy
environment. In fact (without having modeled the different
configurations for comparisons, or having done actual comparative
tests), I'd say there is at least a theoretical advantage in
running the short side in true "feedline" fashion, within several
inches or a foot of the lower portion of the long side and
roughly parallel to it. (Or get the wires close to this
configuration.) The result will be more of the total radiation
occurring at a greater height and thus lower near-field ground

I've used a W3EDP with one of my z-matches, with no difficulty on
the indicated bands. I'd imagine the W3EDP design should pose no
problem with Roy Gregson's ZM-1 or ZM-2, either, although the two
output links in my design give a little more flexibility. A
standard link-coupled tuner (which L.B. carefully describes on
his Web site) should be fine, too. A single-ended tuner--L
network, T network, etc.--requires a balun (and that may present
its own problems.)

Charlie, w6jjz

Search QRP-L Archives

[ QRP-L Archive | ]
[ 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 ]


This archive was generated by hypermail 2b29 on Fri Jun 02 2000 - 11:36:49 EDT