[CQ-Contest] Contest QTH, hilltop or seaside?

[Steve London]

Since I was mentioned in a previous posting, I'll say a few words about 
my QTH and HFTA.

I live on a hilltop with 400 feet of dropoff within 1/4 mile from NW, 
through N, E, SE and S. However, towards the NE, E and SE, after hitting 
a valley about 1/2 mile away, the terrain rises steeply to another 
ridgeline, 1 mile away, that is at about the same elevation as my QTH. 
There is no soil, to speak of. It's all rock of varying sizes from 
near-soil to boulders. I have been at this QTH for over 10 years, and 
have experimented with a number of different antennas at different 
heights. My summary boils down to this: It is easy to be moderately 
strong from here with low heights and simple antennas. It is much harder 
to be extremely strong.

On 10 and 15 meters, I have 6 element monobanders at 30-35' (on separate 
towers). The 15 meter monobander outperforms another 15 meter antenna at 
75' (on another tower) by about 1 S-unit in most directions. I have also 
compared against a stack of higher KT-36XA's, and, again the 10 and 15 
monobanders outperform the stack.

On 20 meters, I have tried antennas at 30', 45', 60' against a reference 
KT-36XA at 75'. Nothing magic here - the 60' monobander and 75' XA are 
about the same. Lower heights on 20 meters were inferior. Perhaps I need 
an even higher monobander on 20 ?

On 40 meters, I have an M2 40M4LLDD (4 elements, linear loaded) at 105'. 
A killer antenna...maybe. I also have a wire, 2 element, inverted vee 
beam towards the NW, with the apex at 50'. It is only about 3 dB down 
from the much higher 4 elements towards the NW. A good example of how 
easy it is to be moderately strong from a hilltop.

On 80 meters, I have a 2 element wire inverted vee beam, switchable 
NE/SW. The apex is at 100'. Too low for DX by flatland standards for a 
horizontal 80 meter antenna. At this QTH, it is a winner.

While the house was being built, I spent hours playing with HFTA. I'm 
sure that HFTA accurately models relatively simple terrain, such as a 
uniform slope, or a few moderate changes in the slope profile. However, 
I seriously question the validity of HFTA in complex terrain, such as my 
QTH. The experiment that really caught my attention was modeling a 20 
meter beam towards Europe. I had two terrain profiles - one created by 
myself, another created by K0KR. The terrain profiles went out many 
miles, to specifically understand the effect of the steep dropoff, 
followed my a steep rise, and the ridgeline 1 mile away. The HFTA 
terrain profile plots looked identical, lining up perfectly. However, 
the HFTA generated results were drastically different. There was as much 
as 10 dB difference at some important takeoff angles. When I looked more 
carefully at the data points making up each terrain profile, I 
discovered they were not quite identical. However, they were both within 
the range of variability that you could obtain using a topographic map 
or the DEM database. Most of the data points were within 1 foot of each 
other, with no single data point difference of more than 3 feet. That 
would imply that if I was to move a boulder that was a few hundred feet 
in front of my antenna, or even have the driveway regraded, it would 
effect my signal to Europe ! Sorry, but at 20 meters, the real world 
doesn't work like that. It was at that point that I lost faith in HFTA 
to reliably model antennas at my QTH. Now I test antennas empirically, 
always having an unchanged reference antenna. If you are playing with 
HFTA, I strongly urge you to do a sensitivity analysis by changing your 
datapoints, randomly, by 1 to 3 feet. Do this on each band of interest. 
If the results change significantly, then HFTA is not a valid model at 
your QTH.

73,
Steve, N2IC