> Yes, I don't know why the db-201 has such ridiculously long
> radials, but that is how they were designed. Go figure. I
> would like to know why though if someone knows...
> 73
<Apologies in advance, this is going to wander a little off-topic and ended
up getting long...skip to the bottom if you want the simple answer.>
The venerable quarter-wave vertical antenna seems like such a simple
antenna, but the physics behind it aren't all that simple. It is heavily
influenced by the ground (be it created by radials or the actual earth
beneath the antenna), and, due to that often being a wide-ranging variable,
a lot of myths and ambiguity as far as the "right way" to create the ground
plane or ground system have amassed over the years, especially within the
ham radio realm.
In its simplest form, a ground-mounted quarter-wave can be thought of as
being a dipole with the lower-half of the antenna formed by a ground system,
in whole or in part comprised of the earth. You've probably seen crude
drawings in antenna handbooks showing the vertical radiator, the virtual
"mirror image" half of the dipole buried below-grade, and imaginary
capacitors between the vertical radiator and the ground system to close the
circuit, allowing displacement currents to flow back to the feedpoint.
That's a decent approximation of how it works when the ground system is
actually "the earth", and to get a good ground system when the mirror-half
of the dipole is actually "the earth" usually involves installing buried
radials to improve the conductivity of the earth to lower losses, thereby
improving efficiency.
(Sidebar. What are these "displacement currents" of which you speak? Every
circuit, including antennas, require at least two terminals in order for
current to flow. You can't force current into a single-terminal device - EE
101. And antennas are no exception. Contrary to popular belief, you can't
end-feed an antenna connecting only the center conductor of the coax to the
end of the antenna without a ground system or some other way for current to
flow back to the shield at the feedpoint. Quite often, there is a a part of
the circuit that nobody realizes exists when they claim that they're able to
end-feed an antenna, such as a half-wave "without a ground", and have it
work to some degree. The circuit is closed by stray capacitance between the
coax shield and the antenna, even lacking any direct connection to ground or
a ground system. The currents set up by the E-field of the antenna have to
make their way back to the feedpoint. When these currents flow through a
coupled ground return path, rather than being a hard-wired connection such
as we would have with an antenna like a folded dipole, these are called
displacement currents.)
In the idealized world, the feedpoint Z of an infinitely-thin quarterwave
over a flat perfectly-conducting ground plane of infinite area is half that
of a center-fed dipole in free space, i.e. 36.5 ohms instead of 73 ohms.
Matching 36 ohms to a 50 ohm feedline is fairly trivial, but lacking any
matching, you're still left with a (roughly) 1.4:1 VSWR, the same as you
would have when connecting either a 72 ohm dipole or a 36 ohm quarterwave to
50 ohm line. One simple technique to help improve the match, and often
eliminate the need for any external matching network, is to slope the
radials downward in order to produce a feedpoint Z somewhere between 36 ohms
and 73 ohms, i.e. something closer to 50 ohms.
When you take away "the earth" and synthesize "ground" by using radials,
things change a lot from the theoretical case above. There has been a LOT
of emperical research and computer modeling done on the topic, quite a bit
even in recent years, regarding the performance (efficiency, Z, pattern,
etc.) of antennas with elevated ground radials, most of that research being
specific to MW and HF antennas. To grossly over-simplify, in many cases you
can get as good, if not better, performance out of a quarterwave with a
small number of elevated radials as you can a traditional ground-mounted
quarterwave with a lot of buried radials, especially when ground
conductivity is less than ideal. At HF, the ground still plays a big role
in how these elevated-radial antennas perform. But at VHF and above, with
the antenna mounted very high (in terms of wavelength), the earth has much
less of an effect.
Classic ground-mounted quarterwave antennas (e.g. AM broadcast) use a large
number of shallowly-buried radials (typically 120 or more) to create an
earth-based ground plane. In this case, the radials don't necessarily need
to be resonant. Once you get up in the neighborhood of 120 radials or more,
they act less like individual wires and more like one continuous conductive
disc. In contrast, in the case of a highly-elevated antenna with a small
number (1, 2, 3, 4) of radials, the lengths of the radials are critical;
they make the antenna behave more like conventional dipole in regard to
their effect on feedpoint impedance. You can think of the radials as being
more like resonant counterpoises rather than forming a planar ground if that
helps. If the lengths of the radials are lengthened/shorted, there is going
to be a significant change in feedpoint Z, unlike the case of a
ground-mounted antenna with a lot of buried radials where varying the
lengths of the radials slightly doesn't have much of an effect on Z.
Before I go any further, I should mention that radials on some VHF/UHF
antennas can serve two functions, sometimes with mixed success: a) forming
the groundplane for the radiator to "work against" (to complete the
circuit), and/or b) to decouple the feedline from the radiator. Quite
often, b) isn't done correctly, which results in current flowing on the
shield of the coax, altering the radiation pattern of the antenna, as well
as having an effect on feedpoint Z. B) is often ignored completely in poor
antenna designs, especially in the amateur realm. Everybody's favorite
(ahem) 2m antennas, the Ringo and Ringo Ranger, are classic examples of
improper decoupling of the antenna from the feedline (the later Ringo Ranger
II attempted to correct this). And the J-pole articles you find all over
the Internet and even in well-known antenna books often aren't properly
decoupled from the feedline, which is why many people have a hard time
getting the match right, and/or concluding that the antenna just doesn't
seem to work as well as it should.
And to dispel another myth - VSWR does NOT cause current to flow on the
shield of your coax, or cause your feedline "to radiate"! Improper
decoupling of the feedline from the antenna does. That's one of many
misconceptions that raise my ire whenever I hear it being repeated...
But I digress...
To get back to the topic of the DB-201 and why the radials are so long, keep
in mind that the classic textbook donut-like radiation patterns you see for
various vertical antennas (and by that I don't only mean quarter-wave
verticals) assume an *infinite* ground plane. For our quarterwave, in order
to put the main lobe on the horizon where we want it, the ground plane needs
to be infinite, textbook-style. Nothing we fabricate in the real world
comes close to being an "infinite" ground plane, although arguably you can
make the ground plane large enough, and efficient enough, to approximate
perfect ground as far as measurable results are concerned. Getting back to
the handbook-style VHF/UHF quarterwave groundplane (you know, the kind made
out of an SO-239 with four sloping radials), the radials are resonant, half
the length of the nominal dipole length, just like the vertical radiator.
At the risk of contradicting my good friend Skipp, I haven't found any
factual basis to back up the notion that the radials should be 5% longer
than the vertical radiator. I don't know if that's a myth that morphed out
of yet another often-misapplied rule-of-thumb that the reflector on a yagi
has to be 5% longer than the driven element, or if the guy that came up with
that rule experimentally found that, for HIS PARTICULAR ANTENNA, making the
radials 5% longer caused a slight change in feedpoint Z that resulted in a
better match to his coax, or where it came from. But I can tell you from
the bottom of my heart that I've never seen a factual treatise that proves
that radials 5% longer than the radiator are necessary for the proper
operation of a quarter-wave groundplane.
Do radials HAVE to be resonant in order for the antenna to work? The simple
answer is "no", but nothing's ever that simple when it comes to antennas, is
it? For ground-mounted antennas or those with radials only small distance
(in terms of wavelengths) above the earth, the ideal length of the radials
varies a whole lot depending on ground conductivity, distance above the
earth (for elevated radials), the desired feedpoint Z, and even the desired
take-off (elevation pattern) angle. But again, for the topic at hand (VHF
antenna mounted high above earth), a lot of the earth effects are
eliminated.
So, to get back to the DB-201. Because the antenna is elevated so far
above earth, the geometry of the ground radials plays a big role in
determining the pattern and feedpoint Z. By making them longer, we're
moving closer (albeit slightly) to having an infinite ground plane, thereby
helping to keep the main lobe on the horizon where we want it; we definately
wouldn't want to make them shorter or we'd be almost certain to have
undesirable up-tilt. But by making them non-resonant (i.e. something longer
than 1/4 wave), the feedpoint Z changes significantly from the theoretical
value. Again, without knowing the details of the DB-201 you're working
with, I'm guessing the radials are longer than 1/4 wave but shorter than 1/2
wave. The net result is the resistive component of the feedpoint Z ends up
being easier to match to 50 ohms. Not having modeled one, the length of the
radials may also be chosen to cancel, in whole or in part, a reactive
component that would otherwise be present at the feedpoint. When altering
the length of the vertical radiator and radials to scale the antenna up/down
in frequency, it's quite possible that the radiator and radials won't both
scale by exactly the same ratio due to the fact that the non-resonant
radials are critical to the feedpoint Z.
So, the simple answer to the simple question originally posed - why are the
radials so long on the DB-201? It's a) to get the feedpoint Z right, with
the added side effect of also b) increasing the area of the ground plane to
improve the elevation pattern slightly.
Now, don't you wish you had just skipped to the end of the message? :-)
--- Jeff WN3A
