K9AY
Foreward
I have a small contest station,
deliberately limited to CW, simple single-element antennas, and 100W. It is located on 1 1/3 acres of heavily
wooded land. For most of the past ten
years I have been on a constant journey to see how much performance I could
squeeze out of an amateur radio contest station given those constraints. That means I will try nearly anything if I
think I might get a few extra contest QSOs or another multiplier out of
it. After a lot of operator contesting
experience to try to get good at the process and to determine my needs, I have now
assembled what I think is a pretty well optimized station and I have achieved
what I think are reasonable results:
placing at least within the top 10% in my contest class (SOAB CW LP). Sometimes I’ve even got top score in NA.
The most recent addition to this station was added because of its perceived benefits consistent with the above goals and it is a K9AY loop. The purpose of this narrative report is to summarize my experience with this fascinating antenna. Some cautions to the reader, however, are appropriate. Namely:
1. There are so many variables between your place and my place that it is very dangerous to assume that what has happened here is going to happen there. I think that geographic location, soil conditions and the affects of other antennas on my little farm, to name some of the factors, have a big impact on how my antenna works and my results are therefore not necessarily transportable;
2. While I did a lot of research and I received solid engineering advice when I put my K9AY loop together, this report is well removed from being a rigorous engineering analysis of what is right and what is wrong when building a K9AY loop. It is what it is – empirical. I make no claims about this being the optimum design.
3. Results of listening for ten years on a lot of different antenna types have convinced me beyond doubt that the results on any particular antenna can be highly variable from day-to-day. Time of day, atmospheric conditions, propagation conditions, weather conditions, band in use, the season, polarization of incoming signal, and numerous other variables mean that what I experience on any given day on any given antenna may not be repeatable on the next day. Think, for example, about foliage. There is now no question in my mind that the leaves on the maple trees that conceal my antenna wires during the spring and summer have an attenuating affect on my signals. Therefore, I have learned through experience that antennas need to be evaluated for about one year before one can draw ultimate conclusions about their usefulness and their performance. Similarly, in my short period of use of the K9AY, I have found that its behaviour can vary from remarkably good to marginally bad from one day to the next depending on the state of some of these variables.
4. Final caution: If you are an antenna expert or a K9AY specialist, maybe you should go find something else to read.
5. Important: My goal when building the K9AY loop was primarily
to be able to minimize the sorts of atmospheric and man-made noise that most of
us in urban environments experience on the bands used for contesting. A secondary consideration was the ability to
attenuate the very strong signals from north-eastern
Station
Description
I believe that for a reader to assess
whether a K9AY loop is something they might want to attempt based on my
experience, it is important that my circumstances be described.
HF Antennas:
1. full-size 160 meter delta loop with the apex at 95 feet and the base about 250 feet long, 8 feet off the ground on roughly an east-west heading, fed with ladder line and used on all bands;
2. a full-size 80M dipole with one end at about 85 feet and the other at 55 feet at right angles to the delta loop, similarly fed with ladder line.
3. A SteppIR BigIR vertical with a substantial radial field (I like to say nearly a mile of wire. Sounds better to me than saying 5,000 feet),
4. A short (300 foot ladder line) DXE reversible Beverage; and
5. Sufficient relatively clear space left for the K9AY Loop (well removed from transmitting antennas). The loop is hung from a rope strung horizontally between two trees at a 25-foot height.
Geography:
This area was heavily glaciated and the
topsoil is typically two feet or less thick.
It is a suburban sub-division about 15 miles from a reasonable size city
(Halifax/Dartmouth) and in an area with a reasonable amount of light industry. (Regrettably) I’m 20 miles from the
Transceiver
K3 #095 with the KRX3 second receiver, configured for two auxiliary antenna inputs and diversity receive.
Other
N8LP Pan Adapter
Antenna
Details
This is a classic K9AY loop design based on Gary Breed’s original design (easily found on the Internet) of two 85-foot loops at right angles with relays to switch direction, but there are some notable differences which I record here simply (I’m hoping) to add to the body of knowledge about the K9AY.
1.
From my research I concluded that the impedance matching transformer
used in the K9AY is one of two critical components (the other, perhaps, being
the terminating resistor). The goal is
to achieve maximum coupling between the primary and the secondary without
introducing unnecessary capacitive reactance and using ferrite cores that do
not saturate. My transformer (design
stolen from Fred, VE1FA) consists of two Mix 43 ferrite binocular cores glued
end-to-end with a three-turn primary and one-turn secondary using jacketed #24
wire. There are certainly other
designs. They will probably work just as
well.
The transformer in use was measured with an MFJ-269 and found to be flat at 55 ohms from 1.8 MHz up to about 8 MHz. ( The load side had a 450 ohm resistor installed) The trifilar design measured about 50 ohms and was flat nearly to 10 MHz. The capacitive reactance in each case was very low. It is apparently true from my experience that the miss-match between 50 Ohms and the 75 Ohm at the antenna feed point is inconsequential.
2.
Because I had read somewhere that if the shape of the loop has a
pronounced vertical component to it, it is apt to be more effective at reducing
local man-made noise. Therefore, instead
of the traditional diamond-shaped loops, I used a pair of fibreglass tent poles
joined end-to-end at each of the corners on the sides to force the shape into a
hexagon. If one ties the lateral support
ropes to the fibreglass rods with the antenna wire loosely secured at each end
of those rods, the rods bow out a little but the wire stays straight and the
whole arrangement looks a little like an archery bow. (Curiously enough, when I carried the
finished assembly to my chosen site, the geometry and location of my selected
support trees would have forced me into that shape regardless of my
preference!). As interesting as it might
be and even if I could, I am quite unlikely to disassemble this thing and
reshape it into a diamond and then try to see if I can detect any
difference. Just in case anyone asks.
Incidentally, because I live in an area of frequent wind and freezing rain build-up on wire antennas, I build all antennas with what in the aircraft industry would be called fusible links. That is to say, the sides of the K9AY loop wires are held in place on the fibreglass poles by light-duty tie-wraps and if the antenna gets stressed into trouble, forces on the loop wires will cause these tie-wraps to fail first, thus protecting the integrity of the antenna assembly. The failure point is located so as to make repair very easy.
3. I did not use bias-Ts or other similar techniques to feed the relay control voltages to the antenna switch box. Rather, I had some left over five-wire SteppIR control cable and used three of the conductors for relay switching (12 VDC garden-variety relays with diodes across the coils) and retained the two spare wires for other uses such as, say, a remote pre-amp if I decided later that I wanted to use one (so far, not apparently justified). There are chokes on this control cable (as well as on the feed line which is 75 ohm RG-6 CATV cable terminated with F-connectors).
Terminating
Resistor
If one pokes around the literature and the vendors’ sites for a while, one will be left with two absolutely inviolable conditions that MUST be met by the critical terminating resistor. These are:
1. Being able to accurately vary the terminating resistor from the shack in real time by means of some device such as, say, a Vactrol, is absolutely critical. This antenna will do nothing for you without that capability; or
2.
It doesn’t make a damned bit of difference. You only ever need a fixed resistor. Typically the best value will be somewhere
around 560 ohms.
Take your choice.
I elected to conduct an experiment to
determine the truth for MY installation.
My friend Jeff, VE1ZAC, has engineering skills that exceed mine by
orders of magnitude. Experience has taught me that whenever I have an odd
requirement, the correct behaviour on my part is to deliberately make such a
nuisance of myself asking questions that he usually not only concocts a
solution but winds up building it for me just to get me out of his hair. In this case, eBay and the remote control
model aircraft hobby and Jeff provided a lash-up that allowed me to physically
turn a 1K pot out at the antenna from the comfort of my shack. This quickly and easily
allowed me to determine the answer to the terminator resistor question by
repeatedly demonstrating that for MY installation, the terminating resistor
value was far from being critical. Indeed, without forcing it that way, I settled
on 560 ohms. That’s where the pot always
seemed to wind up. The value is just
like the book said.
Here is a photo of the temporary variable terminating resistor scheme.
In it one can see the RC control unit
itself, and the wired-up assembly which was connected to the antenna. It consists of the motor-driven pot (top
centre of photo) and, because this system could not execute a full 360 degrees
of rotation, a couple of series resistors that could be switched in and out of
the circuit such that I could cover a range of roughly 200 Ohms to about 900
Ohms. Squeeze the trigger on the RC
control to toggle the series resistors in and out of the circuit; adjust the RC
rotary knob to very the resistance of the pot.
The un-terminated white wire is the remote antenna.
Finding that Rterm was not critical was double good news to me in that it simplified the design and likely improves antenna reliability. That is so because from my research, the typical amateur radio operator solution to the variable termination involves the use of something like a Vactrol, the resistance of which can be changed by remote controlled from the shack by the application of a variable DC voltage. Everything I heard about Vactrols in the ham radio environment indicated that they were quick to fail in any RF environment, the killer energy coming either from a transmitting antenna or from the sky. I’m not personally convinced that there are adequate designs for circuits intended to protect the Vactrol that don’t, in and of themselves, create the opportunity for spurious signals to be couple into the antenna circuit at a critical point: i.e., where the loop wire goes to ground. Therefore I did not want to have to be driven in that direction. I am told that there are commercial applications that do work properly and reliably but nowhere in my reading about the K9AY did I come across any literature that boasted a reliable variable termination system. That may explain why the manufactures of the commercial version of Gary Breed’s design (apparently) vary the terminating resistor by switching in different resistors from a bank of fixed values at the antenna rather than go for the full control.
However . . . . . .being a pragmatic realist
who wanted to cover his bets, I built my relay box to accommodate a SPDT switch
with the ability to switch in one resistor other than the 560 ohm resistor that
my testing had indicated was optimum. I
chose the alternative value of 800 ohms entirely out of thin air. The toggle switch for these resistors is
outside, of course, and one of these days I’ll walk out there, make the switch
and try the other terminator. Wouldn’t
it be nice if 800 ohms make my K9AY work a little better on 160M and 80M where,
null-wise, it’s not at its best!
I found repeated references to large
voltages arising at the terminating point of a K9AY loop; i.e., across the
resistor. Mine at ½ watt are judged to
be inadequate. My only counter-argument
is that my loop is well away from my transmitting antennas and I only run 100W,
and the toggle switch out there has a centre-off position that I can use in
electrical storms to take the terminating resistor out of the loop. I’ll replace the ½ watt non-inductive
resistors with something heavier if/when experience shows that it is necessary.
Grounding
I have never been able to find a definitive
statement as to the criticality of the ground for this antenna so, as I usually
do, I assumed more is better. Given the
inability to drive any kind of decent ground rod here to an acceptable depth
before hitting bedrock, my various grounds have (rightly or wrongly) always
consisted of several short thick-wall copper water pipe sections, all strapped
together. This technique has seemed to work
adequately for me in other instances. The
only hint I have seen about the requirements for ground for the K9AY was a
statement to the effect that an adequate ground for the can be achieved through
the capacitive coupling between a flat plate of metal and the earth if the
plate is laid on the ground. Since I just happen to have a 2-foot square
sheet of aircraft aluminum, one of these days I am also going to connect that
to my grounding system and/or see if I can reach some conclusions about ground
needs by using ground rods and the aluminum plate singly and in various
combinations. In the mean time, I’m
throwing everything that I have at it based on thinking similar to that which
concludes that if one application of a mind-altering substance is good, then
two would probably be better.
Anecdotal
Experience with the K9AY at VE1RGB
It would be good to reread what I said at
the start of this report: namely, what happens here at my station may be
entirely unique to my circumstances. It would be entirely inappropriate for me to draw hard
conclusions nor state any generalities about the K9AY based on what I have
experienced at VE1RGB. For one thing,
the assessment period has been too short.
What follows, then, are some anecdotal experiences I have had with the
loop over not much more than a week’s experience. [That I should feel comfortable
writing about my K9AY after but one week of use, however, probably telegraphs what
my ultimate conclusion will be].
Use
in Contesting
The first “contest” use of the K9AY was in
the California QSO Party of October 2009.
In that contest (from here), most QSOs are typically made on the first
afternoon of the contest and on 20 meters. When I started the contest on that
band I found that average signals from California from my simple
omni-directional antennas were not much more than about S3, if that, and that
local noise (man-made and atmospheric) was a generous S3. Ordinarily that would essentially been the
end of the contest for me. However, with
the K9AY I was able to improve signal-to-noise ratio to the point where
continuing on was quite easy and comfortable.
That saved the contest for me right there.
Later that evening in CQP after 40M opened,
atmospheric noise crashes from electrical storms made operating only barely possible
on this band, and it was not much fun.
The K9AY fixed that by reducing the noise level to something I could
handle and brought my SNR up to a usable ratio.
Ultimately, propagation shifted to 80M and
there it so was noisy that operating on 80M was simply out of the question
using my usual transmitting antennas.
The K9AY, however, almost entirely eliminated the noise and operating
then became easy.
Ultimately – and yes, I know, this is only
one contest and it is too early to form long-lasting opinions – my score in CQP
for 2009 was just inches away from beating my best ever which was achieved
during the much better times of 2003, and therefore I would keep this antenna
based on this single experience.
30M
and 17M
As much or more than on any other bands,
the nulls on these two bands have been the most noticeably pronounced. My early observation is that nulls on the
bands above and below that are more subtle.
Yet there have been notable exceptions on all bands (see below).
Perhaps the most interesting of all results
is what the K9AY does on 30M. Put
simply, it is now my best receiving antenna by a considerable amount. Not only is noise suppressed on 30M by the
K9AY, signals coming into this station are measurably stronger on the receiving
loop than on any other antenna. Better
than the delta loop. Better than the
vertical. Better than the dipole. Better than everything. Always. Worth
the investment for the combined benefits on these two bands alone at least at
my station.
40M
Depending upon the direction and strength
of the signal (i.e., coming from EU vs coming from NA) I can obtain nulls. The big advantage of the K9AY on this band,
as on the two below it, is its ability to quiet down a band which could
otherwise be unusable. And I may also be able to turn my back on QRM by nulling
some of the big NA contest signals aimed at me when I’m trying to look
elsewhere in internationally contests.
80M and 160M
I have conducted only one test on these
bands in which I have tried to quantify the impact of using the K9AY, but they
proved interesting. Using the N8LP
Panadapter as a piece of test equipment and listening to W1AW, I could watch
the noise on 160M running at about -120 db on my display. That was with
As for 40M, based on what usually seems to
be my worst problem during contesting, the K9AY very significantly reduces the
atmospheric noise of electrical storms. Dramatically. It sometimes
renders the band from unusable to usable.
Broadcast
Band
I have demonstrated the ability to tune out one or the other of two stations occupying the same frequency if one of the signals happens to be coming at me from a different angle (surprisingly difficult in these remote parts of the world – all AM transmitters seem to exist in New England and therefore always on the same heading from me). Having said that, it was not easy and I had to look around for a long time before I could execute such a conquest. I view this capability as a party trick and I have the recorded the frequency upon which I found these particular targets so that I can show off this neat feature in the future. It wasn’t what I was looking to achieve out of the antenna. In fact my K9AY is generally not a particularly good BC Band receiving antenna at all; my other wires are better.
Future
Trials
Diversity
Receive
With the second receiver in the K3 and now
with the K9AY in addition to the Beverage, I am well positioned to evaluate
both receiving antennas side-by-side (A/B) as well as see if advantages can be
achieved by using a mix of receiving and/or transmitting antennas of different
types. You know, vertical vs horizontals vs receiving. From a quick try I do not see that using the
receiving antennas with the second receiver will often be a useful contest configuration. Finding the optimum mix of
antennas and rig settings takes too long, usually. This will be useful for DX only.
Different Transformers?
I have the materials and the design for two impedance matching transformers different from one I used. One uses ferrite EMI beads and 6 turn/2 turn windings. The other is a trifilar-wound design using a toroid. I may investigate these alternatives. I designed the enclosure at the antenna with an internal terminal strip such that swapping out components for trial purposes would be a trivial job.
Ground
Experiment
I am still curious as to whether grounding
is or is not important to my K9AY installation.
It would be easy to methodically disconnect what I currently have out
there at the loops and attempt to assess the difference.
Pre-Amp
I am far from convinced that an external
pre-amp is necessary in this station.
Indeed I have rarely found it necessary to use the existing K3 internal pre-amp. More evaluation is required and I do have
temporary access to a suitable pre-amp for evaluation purposes.
Conclusions
My sole purpose for adding the K9AY to this
station was to improve my contesting capability. There is no question that I have been able to
achieve that, mostly through atmospheric and man-made noise reduction. The antenna appears to have attributes well
in excess of those associated with noise reduction and those come from its
apparent ability to generate useful nulls.
It can even demonstrate superior received signal strengths on some
bands. If not for contesting, I will capitalize on
those other features during DXing. It is impossible to believe that I will not
one day make a valuable QSO by means of having added this antenna to the farm.
This is just the right kind of project for
someone who likes wire antennas but no longer sees any joy in hanging them from
the top of 90-foot towers. And worth the effort.
Contact me if you have practical questions
that you think I might be able to answer.