Topband: BOGs and dipoles on the ground

[K4SAV]

I may as well give you the results of my measurements with BOGs and 
dipoles on the ground.  It seems to be a lot different from the data I 
am hearing from other people.

This experimentation started when I tried to correlate the current in a 
BOG as predicted by EZNEC to an actual measurement.  The results were 
worlds apart.  It seems the current on a BOG disappears much faster down 
the wire than NEC predicts.  It really dissipates in a hurry as 
frequency as increased.  If NEC can't correctly predict the current in 
the wire, it can't correctly predict the pattern.

So then I tried a dipole on the ground to try and estimate the velocity 
factor of that wire.  The dipole was 119 ft total length (because that 
is what the length of the piece of wire I picked up was).   I left that 
wire on the ground for a month and periodically measured it.  The 
results were highly unpredictable, at least by NEC.  I can only 
attribute this variablilty to ground moisture variation, although I 
never measured it right after a rain.  To give you a feeling of the 
variability, here are the results of a few of the measurements.

1. F = 2.75 MHz, X = 0, R = 186 ohms,  also f = 6.25 MHz, X = 0, R = 522
2. F = 3.37 MHz, X = 0, R = 182, also close to resonance at frequencies 
above 8 MHz with R of about 440 ohms
3. F = 3.00 MHz, X = 0, R = 178 ohms, also F = 4.6 MHz, X = 0, R = 710 
ohms, also F = 7.0 MHz, X= 0, R = 562 ohms'
4. F = 2.70 MHz, X = 0, R = 167 ohms, also F = 4.5 MHz, X = 0, R = 675 ohms

You won't be able to predict those numbers using NEC.  NEC says the 
resonant frequency of this dipole should be about 2.8 MHz with R = 97 
ohms.  You can get some variation in that depend on the ground constants 
you select but you will never be able to come close to the measured data 
no matter what ground constants you choose.  (Actual ground is typical 
Alabama red clay with dipole 1 to 1.5 inches above it.)  You may be able 
to get NEC to give agreement to the feedpoint impedance at the lowest 
resonant frequency by selecting a ridiculous ground constant, but 
everything above that frequency will be way off.  NEC says the impedance 
should rise rapidly above the resonant point and and there are no other 
resonant points until you get to about 9.1 MHz.  So once again, like the 
BOG, NEC can't predict the pattern of this dipole on the ground.  Degree 
of error is unknown.

Incidentally, N6LF's pattern for a BOG computed with NEC4 I was able to 
duplicate.using NEC2.  So I have no confidence that NEC4 can predict the 
pattern of a BOG either.

So what to do?  If I can't compute the pattern of a wire on the ground, 
the only thing left is measurements.  I was also concerned with all the 
rules generated by other people for building BOGs because I know that 
data was generated from NEC analysis.  I duplicated those rules by doing 
the same analysis.  I also generated some new rules for improving BOGs 
generated from NEC but when I implemented that and made actual 
measurements, there was no improvement.  Thankfully I measured the 
results before publishing those rules.

So I set up an experiment using a 366 ft BOG and a 250 ft BOG both 
pointed in the same direction (to EU)  I spent a month taking data on 
signals from all directions, both close stations and DX stations, and 
compiling the differences in performance between the two.

A brief summary of these two antennas was that the forward gain of both 
was the same and the front to back was the same, at least within a 
degree of accuracy that made any practical difference. (It was not 
measurable.)  That was the same on 80 and 160.  The forward pattern of 
the shorter BOG was wider so the response at 90 degrees off forward was 
stronger.  The response for high angle signals was greater for the short 
BOG.  The shorted BOG has a wider front lobe both in the azimuth and 
elevation directions. The response at 120 degrees off forward was 
greater for the long BOG. That means the resulting signal to noise ratio 
should depend on where the noise sources are located.  In general with 
low noise sources, which is usually the case at my location, there was 
no detectable difference between the two.  Both of these antennas played 
well on 80 and 160.  The shorter antenna was better on 40.

If you do a NEC analysis of 250 and 366 ft BOGs, the beamwidth 
difference between the two will be as I measured, although the degree of 
difference may or may not be the same.  I didn't try to make an actual 
dB difference measurement, which is very difficult with over the air 
signals.  I also can't speculate the error in pattern produced by NEC 
when it has big errors in the current in the wire.

I also don't believe the RDF number for a BOG produced by NEC.  I can 
get an RDF for about 11 for a 366 ft BOG.  A BOG is a good antenna but 
it's not that good.  Many years ago I chose 366 ft based on a NEC 
analysis and it has worked well.  I will arrive at a better RDF for a 
366 ft BOG later but based on crude measurements (comparison with my EWE 
array), I'm estimating it's probably between 9.5 and 10.  NEC shows that 
a 366 ft BOG has 2 dB better RDF than a 250 ft BOG.  I don't believe 
that either (based on measurements).

An interesting observation was that the 366 ft BOG self terminated at 
about 4.5 MHz and the 250 ft BOG self terminated at about 5 MHz.  The 
antennas were still usable at higher frequencies but probably not as 
good as they could be if they were shorter (not tested).

I have done some testing with reversible BOGs also, but all the data 
above is for a single direction BOG.

I also compared the performance of these BOGs to my phased EWE array.  
The EWEs have a better S/N but not by a huge amount. Pattern for the 
EWEs is much cleaner with much deeper nulls. Incidentally, I just 
modified my 4 direction EWE array to 8 directions and it will be 
interesting to see how that performs this winter.

Jerry, K4SAV