Thanks all for your comments.
Ive been modeling antennas for years and was looking for a trick, but there
must not be one. The numerical instability of a short segment compared to
the wavelength may pop up too.
Using EZNEC+ V6 with its 2000 segments, I'm at 1890 segments now. If I have
to add sections for the swages, I'm hoping not to use up more than all the
remaining segments.
An alternative would be to examine the swage, and split the largest portion
off to the large tube in the model and then the smallest portion to the
small tube.
For this project Ive ignored the swaging for one of the yagis (an M2) in
the middle of a 3-stack of 6ele 20M OWAs and its had a noticeable impact on
gain, FB and SWR.
All center freq's of those curves have moved down due to dismissing the
swage and creating a fatter element in several places along each element.
Even modeling the resonant freq of each individual elements of existing
middle yagi at the middle height (either in FS or real ground) and applying
those freqs to my M2 model still results in these shifts.
So that leaves only the errors from the my dismissal of the saging as the
last probable error.
I'll let you know what does resolve this.
-Charlie N1RR
On Tue, Sep 29, 2020 at 2:05 PM K9MA <k9ma@sdellington.us> wrote:
> I haven't tried it, but you could do a simple model experiment. Model a
> dipole with various different swaged sections: ignored, average
> diameter, etc. Use lots of segments, and see how much the resonant
> frequency changes. If those changes are really small, like 0.1 percent,
> it's probably not going to make any difference in performance of, e.g.,
> a yagi.
>
> 73,
> Scott K9MA
>
>
> On 9/29/2020 12:10 PM, jimlux wrote:
> > On 9/29/20 10:00 AM, Mark - N5OT wrote:
> >> Back when I would model stuff all the time (K6STI's YO) I just put in
> >> the swaged parts as separate segments that were the correct diameter
> >> and length as the swaged parts. That does not really compensate for
> >> the tiny bit of transition from larger to smaller, but my segments
> >> were chosen arbitrarily to be halfway through that tiny bit of
> >> transition (i.e. the combined lengths added up to the actual phisical
> >> length of the HyGain part I was modeling. I figured no matter how
> >> inaccurate my method was,
> >
> > but does the model actually show much difference? One can get way down
> > in the weeds with this - put a tapered segment in that's 1 cm long,
> > etc. But if the wavelength is 20 meters, a 1 cm transition is 0.0005
> > wavelength. I'd worry more about numerical instability than model
> > accuracy at that point.
> >
> >
> > _______________________________________________
> >
> >
> >
> > _______________________________________________
> > TowerTalk mailing list
> > TowerTalk@contesting.com
> > http://lists.contesting.com/mailman/listinfo/towertalk
>
>
> --
> Scott K9MA
>
> k9ma@sdellington.us
>
> _______________________________________________
>
>
>
> _______________________________________________
> 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
|