Topband: Hindering factors in the science of back yard 160m vertical installations

Guy Olinger K2AV olinger at
Sat Nov 5 08:24:19 PDT 2011

The Short Version.  I'm not sure there are any list moderators that would
tolerate the long version.  It's a book.

If you already have a dense, uniform 160m radial system, you know what you
have.  If you can and are planning to have dense and uniform, stick with
it.  The rest of this is for you who do not have, will not have or cannot
have.  Reasons for not doing dense radials vary widely from too much work,
to "political" difficulties with the household planning department, to
confining circumstances that simply make it impossible, a whole host of
compelling reasons.

Hams have suffered in the confined circumstances by attempting various
minimalist extrapolations of commercial radial methodology. The rub is that
commercial grade research was never done on non-dense, non-uniform
solutions.  And why would it?  If one builds an AM broadcast station, one
buys property large enough to do the efficient antenna.  The continuing,
year in year out, now until forever, extra power bill expense of running
extra power to an inefficient antenna to maintain an FCC mandated field
strength, economically precludes cutting corners on the antenna system.
It's a total no-brainer to spend the money for the efficient antenna
solution, including the necessary property.

What would we be doing if the FCC mandated that all hams on 160 meters had
to maintain a specific field strength while transmitting, say the field
strength of 100 watts on a commercial BC grade vertical and radial system.
 A poor antenna would require buying and running an amp. Some really poor
installations I have seen would require a linear with 30 amp 240v service
to it to match 100 watts on a commercial BC grade antenna  :>)

Hamdom has made considerable assumptions about extrapolation of commercial
BC antennas into the world of constrained circumstances.  We have been
hindered on all sides coming into proper research on how these
extrapolations actually work.  Hindrances follow, not in any particular

1) Until recently, we have had no readily available way to accurately
measure sky wave. Commercial AM BC is all about ground wave.  Hams are all
about sky wave. The Reverse Beacon Network (RBN) is about a year old.  RBN
is the first practical, universally available measurement device for
skywave that will allow overall accurate comparison of signals from various
installations.  One can compare the signals of two stations graphically
across a contest weekend and compare the transmitted signals of two closely
located stations, to within a dB or two, as heard at locations all over the
country and the world.  A/B tests on an S meter at either end, either RX or
TX on the antenna in question, have proven notoriously unreliable for a
stack of reasons, and have led hams on occasion to discard the better
working solution.  The RBN network has shown 10 dB differences between two
local systems here that popular opinion expects to be in the opposite
relationship.  Folks seem to be slow in embracing these comparisons, with
not much acceptance of conclusions flowing from the comparisons.

2) The raw loss factor of dirt varies hugely.  Some hams can "get away
with" lossy methods because of their superior dirt.  Other hams using the
same installation over poor dirt have their signals sucked dry of strength
like a prune.  This leads to arguments about whether an antenna "works",
that can't be resolved.  The FCC's standard USA map plot for ground
conductivity assumptions for use in AM BC station design and license
applications varies from a sucky sucky 2 millisiemens all the way to an
almost miraculous 30 millisiemens.  If you are one of the lucky ones with
30 millisiemens dirt, you probably can get away with anything, including a
configuration that would be a dummy load on 2 millisiemens dirt.  The real
question at root is how well your DIRT works.  If you're not one of the
lucky 30 millisiemens folks and you can't do dense and uniform, your first
consideration about an antenna should be how to minimize induced ground
loss.  That seems more often to be the last thing commonly thought about.

3) Literature leads us away from the ground loss problem. It has only one
solution: dense and uniform.  But they're NOT telling a lie.  It's true,
and it's the BEST solution.  You SHOULD do dense and uniform if you can
(use your own personal definition of "can").  Those are the killer signals
on the band. But there is just about nothing in the literature like, "How
to do 160 meter vertical antennas when you can't do dense and uniform
radials."  This creates a huge credibility gap for those trying to push
loss mitigation which in turn foments an easily discernible resistance to
progress on the issue.

4) Dirt at MF is a science in it's own right.  And its characteristics that
pertain to radio energy absorption can vary wildly (I use this term with
accuracy and conviction) from area to area, from neighborhood to
neighborhood, and even in the same back yard (measurements, not
speculation). If construction site leveling for house building has not
turned a property into a totally miscellaneous conduction/dielectric layer
cake, mother nature over the millennia may have, as at my house. My back
yard is clay over sand over clay, over ancient forest fire. Out toward the
service road by the creek, it's layers of acidic rotting leaves and pine
straw kept damp under-surface most of the time except late summer. This
numbing variability makes progress by comparison of results difficult.
 One's opinion of the whole thing may be skewed one way or another simply
by the dirt available to those whom one has talked to for advice.

5) Instrumentation for measurement of conductivity at medium frequencies is
not nicely packaged in inexpensive MFJ style boxes. This is in abject
comparison to antenna analyzers, and modeling programs, which have had
wonderful progress in the last decade.

On the whole, we're too busy resisting the idea of starting with loss
mitigation, that if substantially accepted, might generate enough demand
for inexpensive conductivity testers.  We have to roll our own
measurements, with all the data conversion work and opportunity for error
that implies.  How much attention would you pay to SWR if you had to make
your own SWR box to measure it?  In the 50's I never paid any attention.
 My tiny budget for radio barely afforded TX and RX of any kind.  I still
worked all over the world, until I moved my multiband vertical from copper
row house roofs to backyard sod and a ground pipe, and couldn't work
anybody, and couldn't understand why.  I'm certainly not knocking
analyzers, but the demand for such instruments precedes their availability.

6) Nobody wants to think they've been "had."  But hamdom, in regard to
extrapolation of commercial MF vertical/radial techniques, and
understanding ground losses and loss mitigation for confined circumstances,
has been "had."  The particulars are largely considered so mushy that the
subject has largely has been reduced to a war of "opinions."  The first
look is not to science, but to the reputation of the author, if even that.
 It has become "schools of thought" when it should have been science and
design.  This introduces an unfortunate irritation factor to discussions,
where people want to think that all opinions should be equal, and don't
want their own opinion stomped on.  Not that science doesn't get into
heated arguments about areas where proof does not yet extend, or the
"proof" itself is a controversy.

7) The math involved is really nasty.  And there is controversy about some
of it.  Small wonder.  Could anything that nasty REALLY have been proved
out all the way?  Some part of my gut screams "Fishy! Fishy!" every time I
get into this.  I do get it, there's a part of ME that remains convinced
there is more than a hint of voodoo in this.  Why should I NOT expect that
from the hinter lands.

8) Even the high-priced, legal agreement, professional grade, USB license
key, modeling programs cannot deal with the vagaries of dirt. The growing
consensus is that models generally UNDER-estimate ground and miscellaneous
dielectric losses.  The reason is simple: noone is going to dig up and
measure the conductivity of every cubic foot of their dirt to a depth of
ten feet across their entire back yard.  And even if you did, the high-end
Norton-Sommerfield ground estimation method was written for monolithic dirt
and to "calibrate" ground estimation to real field strength measurements at
the ground from commercial grade installations. One confounding issue at
160m is that the water table can be visible to the RF, with the ground
increasingly described by knowledgeable authors as actually
semi-transparent at MF.  There is simply no existing way in the current
programming to deal with a VERTICAL two medium stack underfoot.  What will
pay for that expense and trouble?  And would you ever practically have
MEASURED data to drive it for use at a ham level?

9) The physical aspects of a full spread of tests, to combine both existing
dense and uniform, and non-dense and uniform methods, is very difficult and
requires a significant crew some number of days just to do a suite of tests
on only one site with it's unique dirt structure/content data points.
 There is a reason why this kind of work usually gets done with research
money to PAY people to travel and show up.  Then there is the need to go
and do exactly the same tests on different sites to get a range of data to
display the rainbow of results, and see the variation on different sites
with different dirt.

10) The size of a halfwave on 160 complicates matters, eliminates many
sites, and creates difficult construction issues for tests.  To deal with
this, some have made tests on 40m, and extrapolated results to 160m.  We
now know that does not work well because at 40 there is more of  a hard
"skin effect" to the dirt which is easier and more accurately served with a
monolithic ground estimate.  So we have to do our experiments full size,
down at 1.8 MHz, so we get understandings and solutions that work at 1.8
MHz. To run a base comparison on 160 one has to start with a dense quarter
wave radial system, 120 if it's on the ground, and repeat measurements as
one does harmonic reductions of the radial count.  And do that BOTH for
elevated and on-ground. This gives the trusted base measurements to relate
to non-dense, non-uniform counterpoise solutions. This is an awful lot of
work just installing, supporting and removing.  Leading into number 11:

11) The variability of solutions, and the need to include them ALL on a
scale of effectiveness leads to a very large matrix of tests to be run on a
single site.  This in turn needs a clever mechanical progression of tests
to minimize the work moving from one test to the next.  Many of the
INDIVIDUAL tests in such a suite require construction of antenna system
components that would be a fall project for many a ham, and would
thereafter be what was used for a long time because that was all the spare
time (and money?) available for allocation.  We're talking about installing
forty or fifty antenna systems, and then carefully measuring and recording
them, in a few days?

12) At least so far, I have not been able to figure a procedure that would
allow a crew to get through all the needed activity on a weekend of two
twelve hour days.  It appears to need two or maybe even three two day
weekends PLUS corresponding weekends of crews. And that for only one set of
data at one site.

13) The measuring protocol has to be dead on, and in the work laid out in
12) we have to NOT have a killer flaw that requires starting over with the
flaw fixed. That can throw a lot of work down the toilet and be very

14) The weather needs to cooperate, and no rain and mild wind are required.
Some of the tests have some matrix columns for signal sources suspended
from balloons, to positively measure at angles NOT along the ground, since
hams are skywave users. Multiple weekend rentals of helicopter plus pilot
is not being considered.

15) Assuming the above can be worked out, the need to spread out the work
over weekends might require dominating the test site between tests. This
means that a meadow in a public park may be difficult to use, or require
taking everything down after each "session" .

16) If one considers a paid crew of six college students for five twelve
hour days at $100 per day, that's $3000 per data set NOT including food and
transportation. To pull this off at a carefully selected range of ground
type sites, probably scattered around the US, one is talking easily about
25-50 thousand dollars, with the difficulties of rounding up crews LOCAL to
the testing sites.  And from various past experience that expense could be
ridiculously underestimated due to something I'm overlooking.

That's the stuff required for delivering PROOF that people can bet their
hard earned spare cash and time on.  I get that.  I really do.

Bell Labs and all the radio pioneer giants have already been there, but not
for the non-dense non-uniform radial-disadvantaged crowd, and not much for
skywave.  They all had employers who had big money skins in the game that
needed the knowledge.  Adjusting for current value of past expense, we're
probably talking about millions of dollars spent on the original research.
 And even at those levels, that was money considered well spent for needed
answers to questions with a lot of money hanging on them.

This is why we don't have the science long in hand for hams, with MFJ and
Radio Works happily manufacturing decently efficient small site 160
antennas based on those principles plus matching devices that one just
dumps out of the carton, strings up, tunes up, and then goes inside to make
good contest scores on 160.

Just a rant on my part, folks.  Maybe some of this stuff will catch on and
we'll get a little tail wind.

73, Guy.

More information about the Topband mailing list