[TowerTalk] Re: Radials over salt water
n4kg@juno.com
n4kg@juno.com
Fri, 12 Jan 2001 00:27:25 -0600
The questions posed below have been around for a LONG time.
In fact, an extensive study was conducted and published in
the June 1937 edition of the Proceedings of the IRE
(Institute of Radio Engineers, predecessor of the IEEE) by
Brown, Lewis, and Epstein under the title
"Ground Systems As A Factor in Antenna Efficiency".
Field intensity at 1 mile was plotted for various combinatins
of radiator height, number of radials, and length of radials.
Several charts were presented, holding one parameter
constant while varying the others. It is well worth reading
this paper and reviewing the graphs. The results are
especially interesting to amateurs since the study was
conducted at 3 MHz.
More references (72 of them!) can be found in the APRIL 1984
issue of CQ Magazine at the end of an article by Arch Doty,
K8CFU (who I knew as a pre-teen when I was first licensed
as WN8FAW in 1955). The article is entitled "Improving
Vertical Antenna Efficiency".
"Optimum Ground Systems for Vertical Antennas" by
John O. Stanley, K4ERO / HC1 (associated with HCJB)
has a table of "Optimum Ground System Configurations"
showing optimum lengths for varying numbers of radials
(16, 24, 36, 60, 90, 120).
A theoretical analysis of these issues was presented in
QST for JUNE 1985 in an article by Brian Edward, N2MF,
entitled "Radial Systems for Ground Mounted Vertical Antennas".
I did NOT compare these theoretical results with other data
or more modern analyses..."that is left as an exercise for the
student" :-).
de Tom N4KG
On Thu, 11 Jan 2001 "hasan schiers" <schiers@netins.net> writes:
> Now that is a handy observation and helpful formula. It is based on
> Tom's point that the perimeter separation need be no more than .025
> to .05 wavelength for 1/4 wave radials.
>
> Now the question for more compromised situations:
>
> What is the maximum allowed perimeter separation of 1/8 wave
> radials. In
> other words, how many radials 1/8th wave long would be required to
> achieve
> "close to" the same performance of 31 radials 1/4 wave long?
> Or....it is
> just not possible to get "there" from "here". If 31 to 60 radials
> 1/4 wave
> long approach 100% efficiency, how many radials 1/8 wave long will
> approach
> 100% efficiency. (Or more likely, just how many 1/8 wave radials
> will it
> take to reach the point of diminishing return for the number of
> radials,
> period, because no number of 1/8 wave or less radials will ever
> significantly approach 100% efficiency)
>
> The more generic question is something like a table (and I'm sure
> it's out
> there somewhere, but I don't have the source material myself).
>
> Radial Length #Radials for Point of Diminishing Return % or dB
> down from
> 100%
>
> 1/2w ?
> ???
> 1/4w 31
> ??? (not seen value)
> 1/8w ? (obviously lots more)
> ??? (terminal value << 100%)
> 1/16w ? (and even lots more)
> ??? (ditto except <<< 100%)
>
> By terminal value, I mean you really can't approach 100% at all, and
> are
> merely looking for the flat part of the efficiency curve and noting
> how many
> dB down you are from the ideal 100%. << = much less, <<< much,much
> less
>
> To summarize the questions I have from the table:
>
> 1. How many 1/2 wave radials are needed to get to the "flat" part of
> the
> efficiency curve and how far below 100% efficiency is that.
>
> 2. How many 1/4 wave radials are needed to get to the "flat" part of
> the
> efficiency curve and how far below 100% efficiency is that. (We know
> it is
> probably 31 and certainly somewhere between 31 and 62, from Tom's
> excellent
> post)
>
> 3. How many 1/8 wave radials are needed to .....you get my drift.
>
> 4. How many 1/16 wave....the drift continues.
>
> I'm sure it could be questioned as to what really is the "flat" part
> of the
> curve, but I would like Tom's observations using whatever criteria
> he
> chooses. A response to this series of questions would really help
> anyone
> installing a 1/4 wave vertical and ground mounting it.
>
> "I only have room for "1/X" wavelength radials. How many do I need
> before
> I'm wasting copper and effort, and how far down the efficiency curve
> am I
> going to be?"
>
> Thanks to Tom and Bill for stimulating my questions on this matter.
> It is
> not just an academic exercise. I will be putting my full size 1/4w
> 40m
> vertical back up in the spring and will be ground mounting it. Where
> I
> choose to mount it is somewhat limited and knowing the compromises
> involved
> (with an approximate numerical value), will be most helpful.
>
> 73,
>
> hasan schiers, N0AN
> schiers@netins.net
>
> ----- Original Message -----
> From: Tom Rauch <w8ji@contesting.com>
> To: <towertalk@contesting.com>; Bill Coleman <aa4lr@arrl.net>
> Sent: Thursday, January 11, 2001 12:17 PM
> Subject: Re: [TowerTalk] Re: Radials over salt water
>
>
> > Hi Bill,
> >
> > > Hmm. 1/4 wave radials means a circle of 1/2 diameter. That's a
> perimeter
> > > of pi/2 wavelength, which divided by .05 yeilds about 31
> radials.
> > > Similarly, .025 spacing at the ends is about 63 radials.
> > >
> > > If >63 radials is like a solid screen, I wonder why there's so
> much in
> the
> > > amatuer literature about installations with 120 or so radials.
> > >
> > > Bill Coleman, AA4LR,
> >
> > Good observation Bill.
> >
> > You hear that because people are parroting the FCC requirement,
> which is based on "twice as much as near-perfect is good enough".
> > The FCC not only wants a system to work, they want it to have
> > "headroom" if something happens over time and it deteriorates. If
> > you read the RCA study that the FCC based its decision on, it
> > confirms the rule of thumb I offered.
> >
> > As another point, BC stations almost always use ground screens
> > near the base of the tower. Yet it does absolutely nothing for the
> > signal, as long as the radials are sufficient in number. You can't
> > make people not use ground screens, no matter how much proof
> > you offer. It just "looks good".
> >
> > Another wive's tale is four elevated radials are noticeably more
> > efficient than 120 radials on the ground. How can anything be
> > "noticeably more efficient" than something that is already almost
> > 100%??
> >
> > If they know that much about it, why don't they know 60 radials
> > are an immeasurable amount less than 120 when the radials are
> > 1/4 wl long? The truth is 30 radials on the ground are barely
> > noticeable as being worse than perfect, and a lot less headache
> for
> > RFI and lightning damage than any sparse counterpoise system.
> >
> > If you follow the general rule I outlined, you will be way out on
> the
> > flat part of the efficiency curve. You would only use 120 radials
> if
> > they were about 1/2 wl long, and you chose the narrowest end
> > separation.
> >
> > Use more than that, and you are wasting copper. I know a couple
> > people with 360 1/4 wl radials! They could have used 100 1/2 wl
> > radials, and actually have picked up a couple tenths of a dB or so
> > and used less wire (and spent less time installing the ground).
> >
> > I don't tell them that, cause it's a done deal. I listen to them
> telling
> > people that, and hear the admiration in the other stations voice
> > when they reply "no wonder you're so strong".
> >
> > Maybe we'd better keep this a secret Bill.
> >
> >
> > 73, Tom W8JI
> > w8ji@contesting.com
> >
> > --
> > FAQ on WWW: http://www.contesting.com/FAQ/towertalk
> > Submissions: towertalk@contesting.com
> > Administrative requests: towertalk-REQUEST@contesting.com
> > Problems: owner-towertalk@contesting.com
>
>
> --
> FAQ on WWW: http://www.contesting.com/FAQ/towertalk
> Submissions: towertalk@contesting.com
> Administrative requests: towertalk-REQUEST@contesting.com
> Problems: owner-towertalk@contesting.com
>
________________________________________________________________
GET INTERNET ACCESS FROM JUNO!
Juno offers FREE or PREMIUM Internet access for less!
Join Juno today! For your FREE software, visit:
http://dl.www.juno.com/get/tagj.
--
FAQ on WWW: http://www.contesting.com/FAQ/towertalk
Submissions: towertalk@contesting.com
Administrative requests: towertalk-REQUEST@contesting.com
Problems: owner-towertalk@contesting.com