In a message dated 98-07-01 03:17:32 EDT, you write:
<<
1. What is a recommended hight for the base of the vertical
(assuming to have it mounted on a 'pole' and the radials
going horrizontal to other poles)<<
I used six feet on 80 meters.
<< 2. Should the radials resonant for the desired frequency or is
that not necessary?<<
The radials should be resonant. I resonated my radials by tuning them as a
dipole pairs, then connecting them as radials with no major change in height,
length or location.
<<3. any other comments or suggestions?<<
Be sure to isolate the feedline and radials from ground with common mode
chokes. I measured a 1 dB additional shortfall when the radial common point
was earthed to three six foot copper rods!!!
>>
Hi Carsten,
I've done this several times now, most recently a few months ago. I've written
a letter to Communications Quarterly about making radial measurements that
will be included in the next issue, or I can e-mail it to you. It explains
some flaws in current radial measurement concepts.
Even though long, I'll post this additional information because it applies to
other antennas as well. You are probably going to be measuring a four to six
dB change at the system extremes of a few elevated radials to a full blown
system (about eight to ten dB change if you use the very short radials
recently recommended to be resonated by inductors). Most instruments, because
you have to depend both on exact power transmitted as well as received power,
will give you a few dB of accumulated error. You need to use the instruments
in a way that minimizes the errors, instead of allowing them to accumulate!
1.) An A-B test does not require an accurate ABSOLUTE value of field strength
(given in mV/m), but does require an accurate relative change indictator
(given as voltage, such as displayed by fairly accurate RF voltmeters). It is
necessary to modify ONE system and measure at the same points the same way
with a vertically polarized antenna. You can NOT do a two antenna ground wave
field strength comparison between two systems because of path differences and
interaction.
You can A-B two system over the air, but it takes MANY averages over a long
period of time to reduce the variables. It takes a lot of care to do this, but
I've found the results closely track groundwave measurements.
(Someone misunderstood my posting about A-B tests on 160 with a high dipole
and verticals, and suggested rotating the four square. The antenna NULLs from
my vertical and four square is primarily caused by the feedline for the dipole
and the 260 foot tower 500 feet west of the four square and 600 feet from the
omni vertical. Even with 500 feet of spacing and the tower detuned, the sheer
size of the dipole feedline, dipole, and tower STILL affects the patterns a
few dB. To do an accurate test on 160, I'd need over 1000 feet of
separation!!! On 80, I was able to get away with just over 700 feet of
separation for over the air A-B comparisons. We can expect two similar
antennas, even verticals, to NEVER have symmetrical patterns even when a
wavelength from surrounding structures!)
2.) The most accurate results occur when the indicating device is operated at
the same area of it's scale, and accurate attenuators are used to set the
indicator back to the same scale area for every reading. The attenuator
requires an exact load for maximum accuracy, but source impedance does not
matter. I use a CATV type RF FS meter calibrated in dB and mV, along with a
precision pad.
The attenuator should be properly terminated if a fixed ten dB or larger pad
is used between the measuring device and the attenuator. Everything MUST be
the same impedance.
As a bonus the same pad setup can be used to verify the power meter and
measuring instrument!
3.) When you measure POWER, remember the limits of the meter. ALWAYS use a
slug near full scale for power measurements. A Bird 43 has an accuracy of + -
5% of full scale anywhere on the scale. The full scale error is almost 1/2 dB,
and if you use the Bird at half scale the power error is almost 1 dB!
4.) When you measure power, be sure to subtract the reflected power from
forward power to get true forward power. Use a low power slug to measure
reflected power, keeping the meter near full scale for both readings.
Otherwise, you have to add in the error in reflected power, since it directly
affects the true forward power and SWR readings! RF bridges are much more
accurate than a Bird for measuring SWR.
5.) The best way to measure power is to build a very low loss tuner low
operating Q tuner using high quality components (efficiency above 98%) and use
that matching network to insure the load Zo is always 50 ohms. This eliminates
reflected power meter errors and the accumulated errors of reflected power and
forwrad power subtraction.
6.) A regular receiver with a meter that functions the AGC OFF, such as occurs
in an FM mode, makes a good detector when the attenuator is used because the
scale is so expanded.
7.) Be sure you are in the far field, but it isn't necessary to get too far
away. Several wavelengths is far enough, and a better indicator than changes
measured at large distances (where uncontrolled variables can change results).
This is a tough thing to do correctly, but it can be done. If I can help let
me know.
--
FAQ on WWW: http://www.contesting.com/towertalkfaq.html
Submissions: towertalk@contesting.com
Administrative requests: towertalk-REQUEST@contesting.com
Problems: owner-towertalk@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|