I did it a different way too.
Perfect (IMHO) 4 square array SYSTEM, consist of:
1. properly tuned at the design frequency hybrid power DIVIDER (it is
more common to be a 50 ohms design, lets speak of this one) with all
the needed switching and reversing -180 deg. transformer inside;
2. exact 1/4L (at the same design freq.) 75 ohms coax LINES, (and
they are serving as impedance transformers indeed and don`t forget the
current forcing properties of this type of line) with VF
close to 0.8, and a LONG string of the ferrite beads at the antenna ends;
3. all of the ELEMENTS (ELEMENT is a vertical part with it`s radial
system either some wires on the ground or in the air) which is tuned
to be Zant=50+J0 at the design frequency of the SYSTEM.
In such a case Zant=50+J0 transforms to Z=100+J0 at the ends of a
coax lines i.e. switching box inputs.
Side elements is connected in parallel, to the hybrid power DIVIDER
port#1, loading it perfectly with total Z=50.
Front (via 180 degrees reversing transformer) and Rear elements
connected to the hybrid power DIVIDER port#2 loading it with almost
perfect Z=50 too.
The third port should be loaded, again with R=50 ohms resistor
and if everything in the SYSTEM is as sad above , then the
dump power is less then 0.25% at the design freq. and maintains quite
low and acceptable over 3-4% of the design freq. And all the
element is feed with equal current and a 0/90/180 phase shift.
This means by the way, that the perfect loading of hybrid power
divider, i.e. equal power split and zero dumped power,
requires NOT a perfect but quite a mediocre radials system
for a 1/4L vertical!
And it rises the question of what is best -
zero dump power in a 50 ohms designed hybrid or zero loss
(simplifying a lot of course! ) in a radials system
on the ground.
IF your a lucky and have a perfect ground
screen and your are at Zant=36+J0 point with
the full size vertical, then your need 75
ohms design hybrid divider and 50 ohms
lines!!
When freq. changes, the elements impedance is changing too, Z at the coax
lines ends is changing, and the impedance presented at the
hybrid divider ports #1 and #2 is very different from the design 50
ohms!
The loading resistor gets about 15-18% of the input power IF
the design freq. was 3650 and the band is 3,5-3,8.
And this is too much.
Your cant make single radiators lenght CW/SSB system.
The F/B, input SWR however still keeps resonably good by the way.
But never, never depend on a SYSTEM SWR! Easy to be misguided.
Speaking of 3.5-3.8 bandwidth which is widely used here:
- element`s diameter does not change much from CW/SSB point of view;
- you can use a single 1/4 coax lines for 3.5-3.8 cutting them for 3.65, your
need around 15.5 meters of cable lenght at minimum, and this requires foam
dielectric cables with VF close to 0,8;
- 3/4 coax lines makes the system bandwidth more narrow (property
of the 3/4 lines) and your NEED DPDT relays to switch in/out extra
length of cable, making them longer for CW coverage;
- in any case your definitely need a CW loading stub for perfect
coverage from 3,5 to 3.8, I personally prefer them over capasitor`s
for a number of reasons;
your first "tune" the SYSTEM for say 3.725 and then switch a
stub ON to make it "resonant" on say 3.535;
"tune" and "resonant" is a wrong terms here, to be
exact, your are seeking for a condition where the dump
power is minimum at the desired freq. by establishing
radiators needed impedance changing its lenght.
if your are with the ground mounted vertical for 80 (relatively
large), "tuning" is a bit tricky of course - you have to change
verticals length to get Z =50+j0 and to be at this point your
have to find first the number of radials needed.
Sounds complicated but for the first one of 4 elements only :)
And actually your are pruning same things but not for the "old rule:
-100khz resonance" but for more correct criteria for SYSTEM
tuning. IMHO.
With the raised radials your need a good extention ladder :)
I first prune vertical with the single radial together and then
add few more radials, cheking again for Z to be as close to
50 as possible.
Properly tuned SYSTEM, while measuring on the ground level at the
distant point, must be no less then 22-24 db F/B and perfect input SWR.
And on air observation of F/B is a different story of course depending
on the arrival angles.
And in general 4 square array is a forgiving system.
>> I first calculated the radiator length for 100kHz below my target
>> frequency.
>> Then I resonated the first radiator by fine tuning it against a temporary
>> ground rod.
>> The other three radiators I made mechanically the same length.
>> I then installed the elevated radials (have only one each) and adjusted
>> their length one at the time to resonate with the radiator on exactly
>> 6090kHz.
> Very important when using elevated radials: they will affect the resonant
> frequency of the element. When I installed my 4 squares I first installed
> two wires in the form of a 1/2 wave dipole (fed a+t the center) at their
> final elevation above ground. Then trimmed them to length at my target
> frequency of the array. Then I cut the rest of the radials to the same
> length and installed them.
--
Andrey,
RA6LBS/AB2ZB mailto:ra6lbs@volgodonsk.ru
www.RA6LBS.ru
icq 342547543
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|