[CQ-Contest] Coax Stubs for SO2R

[Jukka Klemola]

Thanks Rudy ..


In case you really split the hairs, there is a connection between the two;
placement and length; but it is weak.

In case you really want to experiment, make the stub a little long.
  -if you want to be precise, I would suggest to tune the stub to harmonic
band lower edge
Second, find a place where it attenuates the harmonic more than elsewhere.
  -you will likely notice the maximum attenuation placement is not critical
  -you will likely notice finding the spot where the attenuation is not too
great is narrow

Disclaimer: this is depending on the output stage and antenna; but
typically it is like that with tube amps and monobanders.

After finding the adequate placement, then prune the stub length to get the
compromise for the attenuation over the harmonic band.

Please note it is possible in a case of a 40m stub that the stub placement
is not at all the same for 20m attenuating or 15m attenuating stub.


** Warning, scientific content:
Then, if you really measure the stub carefully, you will notice the 20m
attenuation maximum is not the half of the attenuation maximum on 10m.
It is because the speed of radio wave changes in the coax dielectric.
You will definitely notice this if you build a 160 or 80m stub and tune it
to attenuate 80 or 40.
Then measure it on 20 or 10 .. it is not centered on 20 and the maximum
attenuation may be just outside the 10 band.


About all generalizations we see written and drawn:
All rule of thumb gut feelings are approximations because the equipment
output stages vary.
Some amplifiers have Pi outputs on some bands and Pi-L on others.
That means the absolute maximum result stub placement requirements differ
for that amplifier when changing bands.


Encouragement:
I have found it is easier to get a stub working than failing.

Go, experiment, discover.


73,
Jukka OH6LI


2016-07-18 22:17 GMT+03:00 Rudy Bakalov <r_bakalov at yahoo.com>:

> Hi Jukka,
>
> Thanks for the detailed response. Your original email was clear...at least
> to me :-)
>
> I assumed from the very beginning that stub tuning and placement are two
> different steps. What I was wondering is if, in the case of random or
> unknown stub placement, the stub can/should be optimized after it has been
> placed. Sounds like the answer is NO
>
> Rudy N2WQ
>
> Sent using a tiny keyboard.  Please excuse brevity, typos, or
> inappropriate autocorrect.
>
>
> > On Jul 18, 2016, at 2:57 PM, Jukka Klemola <jpklemola at gmail.com> wrote:
> >
> > Umm .. I did not exactly finalize the email before sending.
> > The missing content is here:
> >
> >
> > There are two things to recognize when making RF filtering; in this case
> > with stubs:
> > Stub tuning
> > Placement of stub
> >
> >
> > Stub tuning:
> > Stub may be a little off the exact frequency and it still gives you
> > benefits.
> > The attenuation it will give, is more than 20dB at the worse band edge
> even
> > if the stub is not exactly centered on the band.
> > The condition to provide more than 20dB is, if the stub is placed at the
> > correct place on the feedline.
> >
> >
> > Stub placement:
> > The stub placement should be on the feedline at a point where the
> impedance
> > is high on the frequency we want to attenuate.
> > How to determine where that impedance is high:
> > As the amplifiers differ and transceiver outputs differ, it is better to
> > experiment the placing.
> >
> > Decide a place where you want to place the stub.
> > Rudy wants to place the stub after the remote antenna switch.
> > Likely, the stub will not fit exactly at the switch output.
> >
> > Make a coax, maybe 2 or 3 ft long and try. If it helps so much you feel
> you
> > have what you needed, let it be.
> > If you want to test and try different lengths, go ahead. It will take
> time
> > and effort to improve to perfection.
> >
> >
> > If you have a possibility, to avoid the need for optimizing, you might
> want
> > to build a double stub.
> >
> >
> > I hope this email is a better answer to Rudy.
> >
> >
> > 73,
> > Jukka OH6LI
> >
> >
> > 2016-07-18 21:24 GMT+03:00 Jukka Klemola <jpklemola at gmail.com>:
> >
> >> I will try and answer each question below:
> >>
> >> 2016-07-18 20:37 GMT+03:00 Rudy Bakalov via CQ-Contest <
> >> cq-contest at contesting.com>:
> >>
> >>> If we are learning that stubs cannot be placed randomly,
> >>
> >>
> >> They can be placed randomly.
> >> Even if the stub is not exactly on the hertz .. if it is within, say
> 50kHz
> >> of the frequency you want to damp / attenuate, it will provide to your
> >> system.
> >>
> >> Being like less than 50 ft of coax, the cost is very reasonable, even if
> >> the contribution is 10dB.
> >>
> >> The cost (USD or EUR) per dB is very good if not excellent.
> >>
> >>
> >>
> >>> is there a method of tuning a stub that is placed randomly along the
> feed
> >>> line?
> >>
> >>
> >> Yes.
> >>
> >> Build a station with more than one radio.
> >> Build a stub to one of the radios. Then try some lengths of coax between
> >> the amp output, coax switch, some panel or where ever you have been
> >> thinking to place the stub.
> >> Then change the coax length a little.
> >>
> >> I have made male-female coaxes.
> >> One is about 2 or 3ft, other is about 5ft or so and third is maybe 10ft.
> >> I can test any length from about 2 to about 17ft in a matter of minutes.
> >>
> >> It is a very quick and effective method to produce actual measurable
> >> results.
> >>
> >>
> >>
> >>> For example, for an existing installation, there is the existing coax
> run
> >>> from the shack to the remote antenna switch. It is not practical to
> change
> >>> the length of the coax run. What can be done in this situation, which
> I'd
> >>> assume is very common, to squeeze the best performance out of a stub?
> >>
> >> After the remote antenna switch, you have coaxes to monoband antennas, I
> >> presume?
> >>
> >> Change the length from coax switch to antenna i.e. place some cable
> >> between the switch and the stub.
> >> No need to touch the coax from station to switch.
> >>
> >> Basically .. try.and.measure before applying full power.
> >> This approach usually avoids the .err -phase that is typical to us hams
> >> when we go to deep experimenting mode.
> >>
> >>
> >>>
> >>> Rudy N2WQ
> >> 73,
> >> Jukka OH6LI
> > _______________________________________________
> > CQ-Contest mailing list
> > CQ-Contest at contesting.com
> > http://lists.contesting.com/mailman/listinfo/cq-contest
>
>