I meant to sent this post to the group but somehow only sent it to
K2AV. So second try.
The system has messed up my data formatting to the point where it is not
readable. The data has been reformatted below. Maybe this will be better.
A little more info on the tests. The source was a small rig operated
from a battery. The feedline length was about 3 feet. There was also a
transformer included to match the rig output to the impedance of the
BOG. SWR was less than 1.3 to 1. I know the BOG is too long but the
object of the test is the comparison between measured data and what
EZNEC predicts. If there is no model where EZNEC can predict currents
that are close, then any patterns generated will be wrong. It also
becomes painfully obvious that published BOG patterns by others are wrong.
There are no buried wires anywhere near the wires in these tests.
When monkeying with EZNEC parameters to get anything that might be give
results close to what I measured, I tried all possible ground
parameters, including those that were absurd and could not possible
exist. None produced close results. I tried all heights, regardless of
actual height. None were close. I tried some wilder stuff too. I
even tried my own ground using distributed resistors, capacitors, and
inductors over a MININEC ground. I tried to SPICE it but that was
doomed due to not modeling the mutual inductance between segments. I
planned on transferring those parameters to EZNEC but that SPICE
simulation was wrong. In summary, nothing came close to the right answer.
I have not tried the plane wave analysis, but I will. Thanks for the
suggestion.
EZNEC models of the BOG included radials for terminations, two at each
end, at right angles to the BOG. Radial length was adjusted such that
they were resonant exactly on the frequency being plotted. Their length
varied between bands and varied for different heights. Incidentally,
non-resonant radials of the right length can improve the front to back,
according to the simulations. This is where the idea of a resistive
plus reactive termination came from. Simulation shows that for a long
BOG the front to back can be improved significantly, but I suspect that
is just wishful thinking since the simulations are in error.
The dipole-on-the-ground to determine the velocity factor was 118.25 ft
long, insulated #16 PVC jacket 0.019 inches thick, and resonated at 2.25
MHz. Just by varying the ground parameters in EZNEC I was unable to
predict this. By setting the wire height to 0.2 inches and average
ground EZNEC came close. When using that height for the BOG, results
were not close. I also did a test for 80 meters. Duplicating that
result required placing the wire height to 0.25 inches. Wire actual
height was 1.5 to 2 inches.
Jerry, K4SAV
DATA:
Current measurements on a 364 ft BOG. Height above ground estimated to
be 1.5 to 2 inches, lying on the top of short grass mowed just before
the test, dry high ground, red Alabama clay.
Load data taken at the end of the BOG:
Freq MHz ___ Source ma __ Load ma ___ EZNEC predicts ma
1.84 ________120__________25_________99
3.52 ________150 _________5 _________67
7.01 ________150 ________<1 ________ 22
10.11 _______130 ________<1 _________9.9
Measured at the 212.25 ft (58.3%) point from the source:
Freq MHz ___ Source ma __58.3%, ma __ EZNEC predicts ma
1.84__________120 ________70__________102
3.52 _________150 ________73 __________71
7.01 _________150 ________35 __________44
10.11________ 130 ________15 __________2.7
I also measured the velocity factor of a wire in the same place where
the BOG was. On 2.25 MHz it was 0.67. At a second place the
measurement was close to the same.
_________________
Topband Reflector Archives - http://www.contesting.com/_topband
|