[Elecraft] Portable antenna

Tue Oct 1 07:44:00 2002

Bill and All,

Thank you for your comments.  See my comments below:

At 09:33 PM 9/30/2002 -0400, Bill Coleman wrote:
>On 9/30/02 10:18 AM, Charles Greene at [email protected] wrote:
>
> >Three points.  1. The vertical performs a little better than both the
> >Hustler 6BTV with 22 radials and the G5RV on both receive and transmit
> >which it would not do if it were that much worse (5 and 8 dB respectively)
> >that those two on receiving noise.
>
>I have no answer to this other than to ask exactly how you determined
>that the antennas "perform" the same?

I used Spectrogram, measured a received PSK31 signal using the cursor to 
which is accurate to 1 dB to measure the amplitude of the received signal, 
then quickly switched antennas and measured it again.  The readings were 
nearly always within 3 dB of each other, and all cases, the new vertical 
was receiving better.  The biggest difference between the antenna is the 
noise floor.  On Spectrogram the noise floor was -68 dB for the new 
antenna.  For the Hustler 6BTV, it was -58 dB, and for the G5RV, it was -52 
dB.  The readings are dependent on the gain settings of the receiver and 
the sound card, and were made at 900 hz.  Spectrogram reads in voltage dB 
which uses the 20 log function.  For transmit, I made contact and asked the 
receiving station to give me a signal report on two different antenna.  A 
few times I used low power and switched the antennas hot.  The signal 
reports I received from other stations were higher when using the new 
antenna in all cases.

> >  2. There are many studies that indicate
> >that using more than two (or four for symmetry) radials does not improve
> >the performance of elevated radial verticals.
>
>This is true when the radials are elevated 1/2 wavelength above the
>ground. Near the ground, things get more complex. Of course, we're
>talking about 1/4 wave radials (or longer), not two foot pieces of wire
>as were suggested in the original posting.

I'm sorry if I led you to believe the two radials were two foot long pieces 
of wire.  I ran tests on radials of #12 wire, 15' long with a radiator of 
18' varying the length of both the radiator and radials some, radials 17'6" 
to 20' with a radiator of 16' varying the length of both some, and finally 
settled on radials which are 17' 1", and a radiator of 16'6". The radiator 
is made from 1 1/4" telescoping tubing, which I could easily vary the 
length.  I ran tests both with an earth ground connected and not 
connected.    (Results seem a little more consistent with the earth ground 
disconnected).  The feed line is perpendicular to the radials for about 1/2 
wavelength, and has a 1:1 bead balun at the antenna to keep the feed line 
from acting as an unintended radial.  Tests I ran were measuring SWR, Z and 
R+/-jX across the 20 meter band and sometimes above and below.  I used a 
Autec RX Vector Analyst Model VA1 (commonly known as an Antenna Analyzer) 
and an Autec WM1 SWR/Power.  I ran tests mostly at the base of the antenna 
meter, but also some at the operating station.  The feed line is 84' of 
LMR-400 and 16' of RG-8X.  I verified results of various lengths of 
radiator/radials using EZNEC ver 3.0.42.

>There were some computer models that suggested that a small number of
>radials elevated 1/32 wave or so above the ground had the same effect a
>few years ago. Although such installations "worked", when attempts were
>made to compare actual field strength measurements of the ground and
>elevated systems, the ground systems showed stronger field strengths.
>
>I can also point to the study that LB Cebik published recently in the NCJ
>trying various models of ground radial systems in various modelling
>software. The most intriging thing about this report is how widely the
>results varied.
>
>Bottom line: I wouldn't trust the computer modelling of radial systems
>within 1/2 wave of the ground as producing 100% accurate results.
>Corroboration with actual measurements is needed.

I agree.  The model does not produce results 100% accurate.  It doesn't 
completely describe the ground system, and doesn't model the interaction 
between the wire radials and the actual ground completely.  However, models 
are getting better, and produce surprisingly accurate results.  I made 
every effort to verify model results, and usually it was the other way 
around.  I took measurements first then ran the model to compare them.

> >  3. When I modelled the
> >antenna using EZNEC, going from two to 16 radials did not improve the
> >performance at all. In fact, the model showed that 4 radials was less
> >efficient than two.  If you like, I will send you my EZNEC file.
>
>It does not surprise me at all that two 2 foot radials is
>indistinguishable from 16. Make your radials at least 1/4 wave long, and
>you'll see substantially different results.

I didn't use 2' radials and didn't model them.  My offer to send you the 
EZNEC files still stands.

>And again, the model is just that -- a model. It isn't reality.

With the last dimensions given above, model shows a resonant frequency of 
14.35 MHZ, an impedance of 38 ohms, and an SWR of 1.25:1.  Measured 
resonant frequency is 14.19 MHz, impedance is 49 ohms and the SWR is 
1.02:1.  Measured Z at 14.0 MHz is 44 ohms and SWR 1.11:1, at 14.4 MHz Z is 
59 ohms and SWR is 1.20:1.  My next action is to determine the efficiency 
of the system then I will run tests with the radials at 6' or 8'. I 
understand I the system has ground losses, just like a 1/4 vertical with 
buried radials.  That's the principal reason for the differences between 
the model results in impedance and the actual results.

>Bill Coleman, AA4LR, PP-ASEL        Mail: [email protected]
>Quote: "Not within a thousand years will man ever fly!"

73, Chas, W1CG
K2 #462