[Elecraft] Diversity in CQ 160 contest [updated]

[Guy Olinger K2AV]

I was asked to post this on the Elecraft reflector.  You may have seen it on
TopBand.   I have modified some of the earlier text for clarity and added
commentary.

------------------------------

The just past CQ 160 was my first 160 contest using diversity (K3) and an RX
antenna.

Using an extension of some successful experiences with BOG arrays I designed
for W0UCE, I used a pair of balanced, ungrounded electrical fullwave
Loops-On-Ground phased for the NE (LOG array ? :>)), and my TX antenna (3/8
wave endfed L).

I was (and remain) somewhat stunned by the degree that the noise from BOTH
TX and RX antennas is spread around the "audio horizon" in diversity, while
the CW (from either or both antennas) remains discrete in some "direction"
on the audio horizon even if that one direction varies a lot.  I have a few
persistent weak birdies which are convenient for s/n comparison, which prove
that directional discrimination IS in effect. My favorite one at 1816.67,
daytime, shows a steady and persistent 13 dB improvement in s/n on the LOGs
over the L on Spectrogram.

My RX to EU has gone from my being afraid to call CQ QRO because of simply
not hearing people calling me on the TX endfed L, to being as sensitive or
more sensitive as anyone in the area lacking W8JI/W4ZV/W3LPL style beverage
arrays.

The real (and unexplained) bonus is what using RX + TX antennas under
diversity has done for hearing off the TX antenna.  Case in point, VP8ORK.
 They were not copiable on either separately, STRONGER in the noise on the
TX vs the LOGs.  With diversity VP8ORK was well in the clear, as if some
broad noise cancellation was in effect.  Certainly more than I had expected,
going from a 229 on EITHER ear separately to a very clearly heard 549 on
BOTH in diversity.  Since all the noise cancellation is going on in my
BRAIN, not in electronics, to be able to hear it that way was quite the
surprise.

To be complete, the night was unusually clear of QRN (will we ever have
quiet winters again?) and it was the more "hissy" kind of 160 ambient noise
controlling.

This clarity was also in effect for the CQ 160.  I was running 100 watts (a
long story) SOLP for the contest.  At one point a UT5 came on my run
frequency with his run, up about 80 Hz.  I was able to hear him well for
over two hours, but doubt I bothered him running the K3 barefoot. About half
the time I could clearly hear the stations calling him. The diversity and 80
Hz clearly separated stations from the US calling me and stations from EU
calling him, making his presence on the frequency a non-issue.  This would
have been a maddening distraction listening without diversity, and loss of
my run frequency.

The root of diversity's extraordinary ability to spread the noise in my head
seems to be a qualitative difference between the noise on the two. Just to
the ear, the LOG noise is more hissy and missing a "harshness" in the noise
on the L.  The LOG does reject the buzzy stuff from the south on the L that
I hear during the day, but that buzz seems to be covered by the nighttime
ambient noise on the L, so the buzz would not seem to control the noise at
night on the L. But the difference in the "harshness" remains at night.

In any event, the difference between the two creates a perceived spatial
spread that vastly improves a discreet CQ signal on the L, even one that is
poorly heard on the NE phased LOGs. (Remember that we are talking about a
process in my brain, not electronics and firmware.)

To this point all the talk on RX antennas typically has to do with brute
force improvement of signal to noise in a single signal stream, and I do not
question the results at stations the likes of W8JI, W4ZV, W3LPL et al.  But
these power plants are not feasible at most QTH.  I am wondering if
diversity RX can improve urban reception significantly with a single on
ground RX antenna whose MAIN purpose is to provide "anti-noise" for
brain-based diversity spread of noise in a diversity RX, any directional
improvement in same appreciated, but not the essential reason for laying out
the antenna.

I have the severe feeling (unproven) that Ground Low Velocity Factor (GLVF)
antennas, being down on the ground, are hiding "under the radar" of
immediately local noise, with the very lowest incoming angles attenuated by
ground and pattern. This gives the ears, roughly, a choice between higher
angle non-local noise vs. higher angle non-local noise PLUS local noise. My
brain spreads these two out and leaves the discrete CW as an audio "point
source".  The other phenomenon of hearing through antenna-specific fades was
still in effect, extremely useful, but not the strongest advantage of the
combination.

If the reason for the spread IS the "under the radar" effect, this will be
an advantage to GLVF devices not shared by pennants, flags, K9AY and other
above ground RX antennas whose essential advantage is discrimination by
pattern.  This is NOT a tested and confirmed conjecture.

The audio noise spread is reduced by use of very narrow bandwidths, and
using 350 or 450 roofing filters (aka 8 pole 250 and 400) was best for the
noise spread.  A width of 250 seemed to be about the smallest that kept a
full noise "spread" in my head. At lower bandwidths than that the difference
"mushes out", and the noise spread narrows in from the sides toward the
center.

The effects above are most pronounced in headphones, but still obvious
enough on speakers.  This makes for a very comfortable casual tuning use of
diversity without the phones.

The LOG's are ROUGHLY 70x70 loops on the ground, which can be notched into
the ground like invisible fences for dogs.  Smaller variants of these GLVF
antennas will be tried, including phased Dipoles On Ground, with some
space-saving bends to make a double DOG directional fit in a smaller
footprint.  GLVF antennas DO have some design constraints which will be
explained in publication.

In the 160 contest I also used a single wire 5/16 wave folded counterpoise
(FCP) which is a +33, -33 feet linear footprint replacement for inevitably
lossy lot-constrained faint imitations of properly dense radial systems.

This CQ160 CW certainly was my personal best in a 160 contest, by far.
 Thanks to Elecraft for putting true diversity within my means.  Whether or
not these small lot techniques can put a competitive station on in a contest
is in your own evaluation of the contest scores.

But as of now IMHO, it appears that with the best L someone can get up into
their trees, a 5/16 FCP, some form of GLVF RX antenna that fits the property
as a source of "anti-noise", perhaps with a directionality bonus, and a K3
with a KRX3 for true diversity, one can make the SOLP class on 160 VERY
competitive from a lot that was heretofore considered too far gone for 160,
even if you have neighbors with devices that shut down when you use an amp,
and you have to run barefoot.

The LOG array GLVF antenna and the 5/16 folded counterpoise will be up on a
web page very shortly having gone through a 160 season-long shakedown of
some number of variations of the FCP, and all but one deep-sixed.  It is a
good thing we waited to publish, some attempts at the FCP in hindsight
invoking now obvious losses :>)

73, Guy K2AV