First and foremost thank you everybody who took the time to read my post and
provide thoughtful suggestions and additional references. I also realize, based
on some of the comments, that I have not fully described the array.
The array is all wire elements suspended from the top of the tower at about
90'. The wire elements are suspended from two X-arm fiberglass poles, each 23'
long. That is, the rope that suspends the wire elements ends about 11' away
from the tower.
The elements are fed with DX Engineering phasing lines tuned for 3550. The
phasing lines are supplied with ferrite beads to deal with common mode flows
(are these beads enough is a different question). The phasing lines are
connected to the DX Engineering 4SQ box, which presumably is an improved
version of the old Comtek box.
While there were no elements up in the air, we erected and tuned the first
element, SW. Tuned to 1.01 at 3413, resulting in element length of 68.5' and
radial length of 70'. Then we lowered the element and built each and every
element the same way and of course tuning it using the RigExpert analyzer. The
RigExpert itself was calibrated for the 3' of coax we use for measuring.
Note that each and every element on its own has a really nice deep null at
about 3413. Visually it looks exactly what you'd expect from a resonant antenna.
Only after we built and tuned each element we connected the phasing lines and
the DX Engineering 4SQ box. At this point the SWR no longer looks like one
taken from a resonant antenna. The SWR is way too flat and way too wide to be
true (the omni direction of course looks way to wacky). I would have expected
to see nice deep null SWR curves, only shifted up in frequency.
So my working assumption is that the DX Engineering phasing lines do not have
sufficient beads to adequately suppress common mode flows. Later today I will
try to confirm this hypothesis by measuring the SWR curves at the end of the
phasing lines connected to the elements. The thinking here is that if I see
flat and broad SWR curves the phasing line acts as a radiator and essentially
extends the vertical element.
I am also thinking of adding more radials. This however is not that trivial.
ON4UN writes in his book (4th edition, page 11-84) that with 1-4 raised radials
it is very important to make sure the radials are not parallel to each other
(Chapter 11, section 8.2). So my plan is to add 2 1/8 radials (35')
perpendicular to the current radial (this is shown in Figure 11-121-F in the
book, page 11-86).
That said, switching directions yields audible and visually significant (2-3 S
units) results.
Rudy N2WQ
On Thursday, August 31, 2017, 10:10:22 PM EDT, Bob Shohet, KQ2M
<kq2m@kq2m.com> wrote:
My understanding is the mutual coupling between the four 1/4 wavelength
verticals tends to raise the resonant frequency by ~ 5%, (that is what I have
read in numerous places), meaning that you want to start with by cutting the
1/4 verticals for approximately 5% lower in freq. (longer) than where you
actually want it to resonate. My actual experiences with building multiple
4-squares on 40 and 80 is that the resonant freq. tends to be raised by ~ 3 %,
not 5%. Furthermore, if you use pvc coated wire like I do, that has the effect
or lowering the resonant freq. by an additional ~ 2%.
So when I build or rebuild a 4-square, I cut the 1/4 wave elements for about 5%
below the “target” resonant freq. and then the array typically has minimum
reflected power (resonance) within ~ 1% of where I actually want it to
resonate. I use 3 – 4 elevated radials per element. I do not know whether or
not these resonant freq. shifts are the same or different with elements near
the ground and 30 – 60 ground mounted radials, or whether or not there is a
difference in resonant freq. shift with using a 4-square on open ground without
trees. (I am in very dense forest).
I suspect that terrain, forest vs. open field, and ground-mounted vs. elevated
radials, affect the amount of coupling between elements and the freq. “shift”
that one must take into account when cutting their elements. It also goes
without saying that most elevated 4-squares do NOT use a “spider-web” of twine
(box + “x”) to keep the adjacent and opposite element spacing exact –
differences in element spacing will likely affect the resonant freq. shift as
well as any differences in the lengths of the 75 ohm feedlines between the
elements and phasing box.
I suspect that often when a 4-square is “off” from where it is expected to
resonant, it is because either the element lengths, spacing and feedline
lengths are not exact – or there is something NON-equivalent in the coax
(Velocity factors can and often differ in the same cable at different points
within the same roll of coax! – which is why each feedline must be measured
first for physical length and then tuned for electrical length).
73
Bob KQ2M
From: john@kk9a.com
Sent: Thursday, August 31, 2017 9:36 PM
To: towertalk@contesting.com
Subject: Re: [TowerTalk] Tuning raised radial verticals
What is the reason for making the verticals 4% long? I am far from being a
4sq guru however it would seem that you should have a choke, especially with
an elevated feedpoint. It is easy to make one using a short piece of RG-302
and a few ferrites that will leave your phased feedline long enough to still
reach the control box.
John KK9A
To: Bryon Paul Veal NØAH <bryonveal@msn.com>, Towertalk Reflector
<towertalk@contesting.com>
Subject: Re: [TowerTalk] Tuning raised radial verticals
From: Rudy Bakalov via TowerTalk <towertalk@contesting.com>
Reply-to: Rudy Bakalov <r_bakalov@yahoo.com>
Date: Thu, 31 Aug 2017 21:23:25 +0000 (UTC)
List-post: <mailto:towertalk@contesting.com>
I am measuring the impedance at the very base of the vertical with a
calibrated
RigExpert. According to DX Engineering the vertical should be connected
without
any baluns, matching networks...nothing. Furthermore, they are saying that
the
vertical should be resonant 4% of the target array frequency. That is, in my
case, if I want to tune the array for 3.550 each individual vertical should
be
tuned for 3.413.
Rudy N2WQFrom: Bryon Paul Veal NØAH <bryonveal@msn.com>To: Rudy Bakalov
<r_bakalov@yahoo.com>; Towertalk Reflector <towertalk@contesting.com>Sent:
Thursday, August 31, 2017, 4:34:40 PM EDTSubject: Re: [TowerTalk] Tuning
raised
radial verticals
your seeing changes in the FEEDPOINT impedance using a tuned raised radial,
thus your SWR will move if yoy mess with them. Suggest you consider using
UnUns at the feedpoints to match the coax impedence to that of the
verticals. Also, a Comtek hybred coupler.....
Sent with AquaMail for Android
http://www.aqua-mail.com
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|