[Antennaware] EZNEC how to choose number of segments
David Gould
dave at g3ueg.co.uk
Sun Jul 14 12:10:06 EDT 2019
Hi Guy,
Thanks for the detailed information.
I don't know how to say this any different way, that the base of the
vertical element (38mm tube) is 10m above ground. The top 1.5m of the
support mast is fibreglass the remaining 8.5m is aluminium pole. The
coax will just run down the pole to ground level.
In implementation, the centre of the coax goes to the base of the
vertical. The braid of the coax is connected to two tuned (to 14.175)
wire (2mm) radials drooping down at 45 degrees (I got this dimension by
modelling separately an inverted V dipole with the two legs drooping 45
degrees)
I would implement the design with a common mode current choke at the
feed point, but I did not know that it needed to be modelled, and I have
no idea how to do this. Is there a reference I can read how to do this?
If the antenna is 10m in the air with tuned radials I do not see how the
ground type would have much effect. I am using real/MININEC medium -
but could easily change it if there is a better option.
My source placement does break the first three of your rules. My
thought is to move the source into the 2nd segment up in the vertical
element. This would mean that the bottom segment of the vertical then
effectively becomes part of the radial system, then the effective length
of the radials will be increased and hence tune to a lower (and unknown)
frequency. I can only think that my best bet to get the radials back on
frequency would be to shorten the radial lengths by the length of the
bottom segment - does this sound reasonable?
I am not bothered about gain, I am mainly interested in getting the feed
impedance.
73,
Dave G3UEG
On 13/07/2019 14:33, Guy Olinger K2AV wrote:
> Hi Dave,
>
> Question A) From your description, I need to know the height above
> ground.
>
> Question B) Exactly what did you mean by tuned radials? Do you mean
> radials carefully adjusted to 1/4 wave, or smaller lengths and some
> series device to tune them to resonance. If the latter, what is the
> circuit?
>
> Question C) Are you using a common mode current block at the
> feedpoint. If not, you MUST literally model the coax shield, and
> placement lengths at your site, including places where the coax lays
> on the ground. Again specifics in counterpoise make a huge difference.
>
> Essential issue D) Source placement issue. See below.
>
> Whether a ground type, counterpoise configuration or ground
> description matters depends on what you have decided to do for the
> vertical's counterpoise, and the variation in modeling issues and
> results vary enormously depending on just exactly what you are doing
> with counterpoise.
>
> It is all too common for a vertical antenna modeling project to
> completely ignore the counterpoise and variations, and presume that
> all issues proceed from the vertical conductor. Practically, the
> starting answer is solve a vertical's counterpoise efficiently for the
> target situation and only then start monkeying with the vertical.
>
> You had not mentioned anything about the counterpoise, and the
> counterpoise is the number one issue for verticals 95% of the time in
> correspondence I get. That is why I asked about the counterpoise and
> ground. For all I knew it was ground-mounted, and a plethora of
> considerations apply.
>
> Essential issue D) Segment placement rule: Never place a source in a
> given segment if either end of the segment 1) connects to more than
> one wire, 2) connects to a wire at an angle, 3) connects to a wire of
> a different diameter, or 4) connects to a wire with a large difference
> in segment length. Does not always cause a problem, but can, depending
> on whatever. Do NOT depend on geometry checks to warn you off. DO IT
> YOURSELF, EVERY TIME. Discipline.
>
> One good way to deal with that in advance, BEFORE problems pop up, is
> to use a larger count of smaller segments everywhere, and always use
> segment #2 instead of #1 or segment n-1 instead of segment n. If a
> single one segment wire has to contain a source, break the wire into
> three segments and place the source in the center segment.
>
> Breaking the segment placement rule will often give you gain AND/OR
> impedance errors of some degree. IF you break that rule you need to
> test for sensitivity to the rule in the specific model to see if
> changing to the rule makes a difference. So you had to create the
> compliant model to see if the non-compliant model caused a problem. I
> finally figured out doing the non-compliant (on-purpose) was stupid me
> (slow learner), and consciously go to compliant placement to start with.
>
> A test you can do, IF you are using small segments, is to run Z and
> max gain with source placed in segment one, then segment two, then
> segment three. If the gain varies at all, you can't use segment one.
> If the Z diff 1 vs. 2 is different than Z diff 2 vs. 3. You do it this
> way because the movement up the wire will vary the Z normally. The
> three segment test says that the difference as you move should be the
> same or very close for small segments. I find that source in segment
> one often erroneously changes the gain figures. Not so cool if you are
> putting together comparisons of differing antenna solutions.
>
> If you are worried about fractions of a dB, or at least somewhat
> accurate feed impedance estimates, go to small segments and stay there.
>
> Hope this has helped. Remember questions A) B) and C) above.
>
> 73, Guy K2AV
>
> On Fri, Jul 12, 2019 at 9:39 AM David Gould <dave at g3ueg.co.uk
> <mailto:dave at g3ueg.co.uk>> wrote:
>
> Hi Guy,
>
> Thanks for quick reply,
>
> If I just start with the 20m vertical. It is 38mm tubing with the
> base
> 10m off the ground, the source is in the first (bottom) segment.
> Then
> there are two tuned 1/4 wave 2mm wire radials for 20m connected to
> the
> bottom of the vertical and drooping down at 45degrees.
>
> Is the ground type that important when it is so far off the
> ground? For
> reference it is real/MININEC medium - would an alternative be better?
>
> Does that give you what you need?
>
> 73,
>
> Dave G3UEG
>
>
>
> On 12/07/2019 14:16, Guy Olinger K2AV wrote:
> > Hi Dave,
> >
> > What are you using for the vertical's counterpoise? What are you
> using
> > for the the ground type? Where is your source placed?
> >
> > These are essential to answer your question.
> >
> > 73, Guy
> >
> > On Fri, Jul 12, 2019 at 8:02 AM David Gould <dave at g3ueg.co.uk
> <mailto:dave at g3ueg.co.uk>
> > <mailto:dave at g3ueg.co.uk <mailto:dave at g3ueg.co.uk>>> wrote:
> >
> > Most of my modelling has been with wire antennas but now I am
> > modelling
> > some verticals with elements having diameters of between
> 38mm and
> > 25mm
> > for 20m and 40m (and using drooping elevated radials made of
> wire)
> >
> > I usually use around 9 or 11 segments for a 1/4 wave element. I
> > noticed
> > that when I changed the number of segments the results for
> things
> > like
> > feed impedance changed quite dramatically.
> >
> > Are there some guidelines for the number of segments for a
> 1/4 wave
> > straight wire element?
> >
> > How is the choice of segment length affected by the diameter
> of the
> > element tubing? Is there a limit on the ratio of segment
> length to
> > segment diameter?
> >
> > 73,
> >
> > Dave G3UEG
> >
> > _______________________________________________
> > Antennaware mailing list
> > Antennaware at contesting.com <mailto:Antennaware at contesting.com>
> <mailto:Antennaware at contesting.com
> <mailto:Antennaware at contesting.com>>
> > http://lists.contesting.com/mailman/listinfo/antennaware
> >
> _______________________________________________
> Antennaware mailing list
> Antennaware at contesting.com <mailto:Antennaware at contesting.com>
> http://lists.contesting.com/mailman/listinfo/antennaware
>
More information about the Antennaware
mailing list