Towertalk
[Top] [All Lists]

[TowerTalk] Delta-loop - long.

To: <towertalk@contesting.com>
Subject: [TowerTalk] Delta-loop - long.
From: jirka@jimaz.cz (Jiri Sanda)
Date: Sun, 4 Jan 1998 00:40:31 +-100
I respond to the discussion on vertically polarized deltaloops on ANT-modeling 
reflector. I respond here since I feel it might be interesting to wider forum 
and as I have seen the "talking" audience is THE SAME

We (me + OK1RF) have done a lot of experimenting with 1, 2, 3 el. deltaloops, 
both parasitic and phased-design during last 20 years. You can hear us with 
those antennas on 80 since about 1980. Also under the call of OK5R in 80s. 
Since this winter also on 160m.

Here are some remarks:

1.We have not found practical and simple way how to feed satisfactorily single 
element delta-loop. The in resonance impedance is always well over 100 Ohms. It 
can be of course transformed down by some cable 1/4 W.L. but......

2.Delta loops are very very sensitive to height of the long bottom wire above 
the ground. For DX operation the bottom wire should not be higher than 4m. As a 
rule of thumb on 80m you can expect that the resonant frequency will move down 
about 100 kHz with each meter of height (going down).

3.the system is not sensitive to exact shape of the loop.

4.the radiation pattern is very sensitive to the position of feeding point. We 
have done in late 80s some field strength measurement with calibrated receiver 
and remote TX (about 5 - 10 km). The lowest angle of radiation was achieved 
when the feeding point is exactly 1/4 wavelength from the APEX. Also OK1DIM 
(while working in 1980s on "INTERCOSMOS" program) has done some experimental 
work on models somewhere over 500 MHz - equipped with professional means and he 
has came to the same results. The optimal real-life height of APEX of such an 
antenna is 0.2-0.25 W.L. It should not be higher. If higher than much much more 
(well over 0.5 W.L.) and fed horizontally i.e. from the TOP or middle of the 
bottom.

5.two elements can be easily adjusted to exactly 50 or 75 Ohm + 0j. Just change 
a bit length of D.E. or REFL or their separation.

6.the performance is influenced (quite a lot) by the tower it hangs from and 
the influence might be both positive and negative, see the AO model and play 
with the position and size of the tower.

7.the loops should be fully vertical if you do not have enough space do not try 
to put them under some angle from the tower - you are going to destroy the 
vertical pattern. Make the bottom section bent - see the model. 

I have pasted here an AO file of our current system on 160m as modeled, you can 
paste them out to AO and play with it. The real life measurement - only 
impedance - was always in "good" or better to say "reasonable" accordance with 
the model, 160m system as described when realized was resonant on 1780 kHz, and 
had to be shorted a little. The projected impedance was 50 Ohm + 0j, the 
measured one after the shorting was 74 + 1j  !!!!!, which made us most happy 
and we just changed 50 Ohm cable to 75 Ohm one and that was it. Of course the 
loops are not above the completely ideal flat land and they are huge and the 
terrain is sloping, on the tower is a monster full size 4Y YAGI for 40m so......

As final remark I would say that those antennas work very well (as would Frank 
W3LPL say) but, it is a lot of work to adjust them for proper operation. Also 
be prepared that if you have 160m ant. and forget reflectometer connected to it 
while transmitting to any other antenna on any band nearby the reflectometer 
will change with a lot of smoke into a piece of carbon.



Delta Loop Beam - 2 el Brezina - more-less as it is in the reality.
over ground
1.835 MHz
14 copper wires, meters
hd = 34.5               ; Apex height of D.E.
hr = 42         ; Apex height of refl.
c = 163.4               ; D.E. circumference
kr = 1.041              ; koeficient relative to D.E.
cr = c * kr             ; reflector circumference
alpha = 45.5            ; half angle at the top - D.E.
alphar = 40             ; half angle at the top - REF.
xd = 25.5               ; driven el. shift from center
xr = -1         ; reflector shift from center


pd = 35         ; horizontal part of lower sect. D.E.
qd = -6         ; how much pd lower compared to angle D.E.
pr = 34
qr = -7

y3 = c/4 * sin(alpha)           ; y of feedpoint
z3 = hd - c/4 * cos(alpha)      ; z of feedpoint
xr3 = xd  
zr3 = z3
y5 = c/4 * (1 - sin(alpha))/(1 + 1/sin(alpha))  ; y of the low corner
y1 = y3 + y5                ; Half-length of D.E. horizontal wire
z1 = hd - y1 * cos(alpha)/sin(alpha)    ; Height of D.E. horizontal wire
xr1 = xd  
zr1 = z1
y2 = cr/4 * (sin(alphar) + (1 - sin(alphar))/(1 + 1/sin(alphar))) 
                                    ; 1/2 of ref horizontal wire
z2 = hr - y2 * cos(alphar)/sin(alphar)  ; Height of D.E. horizontal wire
zr2 = z2
xr2 = xr

y7 = pd/2                       ; y of lower sect. D.E.
zr7 = zr1 - qd                  ; z of lower sect. D.E.
y6 = pr/2                       ; y of lower sect. REF 
zr6 = zr2 - qr                  ; z of lower sect. REF
  
1   xd   0  hd          xr3 -y3 zr3  #12        ; Driven element
1   xr3 -y3 zr3         xr1 -y1 zr1  #12
1   xd   0  hd          xr1  y1 zr1  #12
1   xr1 -y1 zr1         xr1 -y7 zr1  #12
1   xr1 -y7 zr1         xr1  0  zr7  #12
1   xr1  0  zr7         xr1  y7 zr1  #12
1   xr1  y7 zr1         xr1  y1 zr1  #12

1  xr  0  hr            -xr2 -y2 zr2 #12    ; Reflector
1  xr  0  hr            -xr2  y2 zr2 #12
1  -xr2 -y2 zr2         -xr2 -y6 zr2 #12
1  -xr2 -y6 zr2         -xr2  0  zr6 #12
1  -xr2  0  zr6         -xr2  y6 zr2 #12
1  -xr2  y6 zr2         -xr2  y2 zr2 #12 

1  0  0  0              0  0  42     1.5    ; influence of the tower
1 source 
Wire 1, end2


Have a lot of fun !
73 ! 
Jiri  OK1RI            jirka@jimaz.cz


P.S. I hope it is a public answer to K3BU, VE3BMV question.


--
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

<Prev in Thread] Current Thread [Next in Thread>
  • [TowerTalk] Delta-loop - long., Jiri Sanda <=