On Fri, 22 Jul 2011 23:13:25 +0000 (UTC), Owen Duffy
wrote:

You are treading down a well worn path.

Yeah, I know the feeling when I'm stuck in a rut.

There is a flaw in thinking that the shield of the coax below the bottom
of your nominal dipole does not carry significant current on the outer
surface. Whilst you are trying to account for the current distribution
on the nominal dipole, you are ignoring the common mode current on the
feedline.

Agreed. The feedline should be decoupled, which one can do by
wrapping a few turns into a coil and tied together with tie wraps.
Another method is a 1/4 wave decoupling sleeve further down the coax
cable.

The antenna will 'work' and you may be delightedm but if you expect that
it will have very low VSWR based on formula dimensions, I think you will
be dissapointed.

Well, putting a 75 ohm antenna onto a 50 ohm system results in a
minimum VSWR of 1.5:1 assuming the coax is cut for some multiple of
1/2 wave electrical length. Also, most HT's are designed to tolerate
the fairly miserable VSWR presented by rubber ducky antennas and body
proximity.

In this case, the problem to be solved was rather mundane. A friend
was in the process of packing for a weekend camping trip with the
kids. His 13 year old daughter apparently had lost the rubber ducky
antenna from her Yaesu FT-60r. Rather than borrow an antenna, or
build an adapter kludge, he decides it's time for her to build an
antenna. They found some SMA cables, and proceeded to make a simple
coaxial sleeve dipole. However, there was some debate over the cut
length of the driven element because of the presence of coax
dielectric. I get a hasty email with the question and discover that I
really don't know the answer. So, I posted the question here. We'll
find out how it worked when they return.

Whether the home made antenna is optimized for best performance
doesn't seem to be important in this case. Any reasonable antenna is
better than a rubber ducky, so she should be fine (assuming she
doesn't lose another antenna).

You might obtain better feedline decoupling by some form of common mode
choke.

Yes, that will help (as previously noted).

The failure of the sleeve to effectively decouple the dipole is the
reason why several antennas don't work as thought. King gives some hints
that the relative diameters of the conductors bears on operation.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

Jeff Liebermann wrote in
:

....
Agreed. The feedline should be decoupled, which one can do by
wrapping a few turns into a coil and tied together with tie wraps.
Another method is a 1/4 wave decoupling sleeve further down the coax
cable.

Those all may reduce common mode current. If you have tried them and
measured the results, you know by how much.

....
kids. His 13 year old daughter apparently had lost the rubber ducky
antenna from her Yaesu FT-60r. Rather than borrow an antenna, or
build an adapter kludge, he decides it's time for her to build an
antenna. They found some SMA cables, and proceeded to make a simple
coaxial sleeve dipole. However, there was some debate over the cut
length of the driven element because of the presence of coax
dielectric.

And the answer is not a simple as put by some. Particularly, if you had
in mind that the sleeve should be an electrical quarter wave to current
flowing on the outside so forming half of a centre fed dipole, and an
electrical quarter wave on the inside to form a quarter wave s/c stub,
the presence of high permittivity dielectric on the inside prevents
those both happening together... so you compromise, like discarding the
centre fed concept, tune the s/c stub for resonance (max decoupling, and
the sleeve will be less than a physical quarter wave), now tune the top
part of the dipole (it will be a bit longer than a physical quarter
wave) for lowest VSWR on the feedline.

I get a hasty email with the question and discover that I
really don't know the answer. So, I posted the question here. We'll
find out how it worked when they return.

To some extent, everything 'works'. Just stripping a quarter wave of
shield off the end of a length of coax probably 'works' about as well.


Whether the home made antenna is optimized for best performance
doesn't seem to be important in this case.

Indeed, it seems that for most modern hams, performance isn't an issue
or of interest.


Owen

Will it work? Yes. Will it be a very good antenna? No, but you
didn't expect it to be, did you? It's possible for it to be better
than the typical rubberducky though. That's the whole point, isn't
it?
- 'Doc

On Sat, 23 Jul 2011 07:15:07 +0000 (UTC), Owen Duffy
wrote:

Jeff Liebermann wrote in
:

...
Agreed. The feedline should be decoupled, which one can do by
wrapping a few turns into a coil and tied together with tie wraps.
Another method is a 1/4 wave decoupling sleeve further down the coax
cable.

Those all may reduce common mode current. If you have tried them and
measured the results, you know by how much.

I haven't. I also know that using the available RG-316 or RG-174 coax
cable is not going to yield optimum results. For a commercial
product, suitable for general consumption, such compromises are not a
great idea. However, for the purpose intended (short notice
replacement of a lost HT antenna suitable for a camping trip), in my
never humble opinion, is totally adequate. As long as it throw RF in
the correct direction, and doesn't cause the transmitter to shut down
from excessive VSWR, it should suffice.

And the answer is not a simple as put by some.

With all due respect, the answer is NEVER as simple as the common
wisdom and practice. I just went through a similar mess in
rec.bicycles.tech, where I contended that many of the common design
practices found in todays bicycles are not sufficiently optimized.
When success is measured in milliseconds and profitability in tenths
of a penny, then good enough is never really good enough. Antennas
should all be silver painted, made from Litz wire, use Teflon
insulators, suspended air dielectric coax cable, and gold plated
connectors. That should squeeze the last 0.000001dB out of the
design. One can always make an antenna a little bit better.

However, I don't think the 13 year old will notice. Similarly, 90% of
the hams using a sleeve dipole coax antenna will notice. The real
question is "what is good enough". In this case, it's the ability of
a 13 year old to build an antenna on short notice, that works at least
better than the stock rubber ducky, with available parts, and probably
keep her out of the parents way as they load up the camper.

Particularly, if you had
in mind that the sleeve should be an electrical quarter wave to current
flowing on the outside so forming half of a centre fed dipole, and an
electrical quarter wave on the inside to form a quarter wave s/c stub,
the presence of high permittivity dielectric on the inside prevents
those both happening together...

Huh? I thought (err... guessed) that the dielectric only has an
effect if it were between the conductors. The grounded half of the
dipole formed by coax braid has the vinyl outer jacket forming a
dielectric. I presume this is what you consider to be a problem.
Worse, the jacket can be various grades (flooded, plenum, buriable)
with various dopings (UV proofing, gel filled). In general, it seems
to shorten the electrical wavelength. This is usually not a problem
because expanding the diameter of the coax braid, as it's folded back
over the shield, also mechanically shortens the braid. In the 2 or 3
such antennas I've built over the years, I've randomly shortened the
braid. What makes such arbitrary design decisions function is that
the dipole has a rather large operating bandwidth. One can be
incredibly sloppy and still end up with an acceptable antenna.

so you compromise, like discarding the
centre fed concept, tune the s/c stub for resonance (max decoupling, and
the sleeve will be less than a physical quarter wave), now tune the top
part of the dipole (it will be a bit longer than a physical quarter
wave) for lowest VSWR on the feedline.

I must confess that I haven't bothered to try optimizing such an
antenna, while you obviously have made the attempt. What I usually do
with crude antennas is to attach one of several reflection coefficient
bridges in my pile or a directional coupler and some diode detectors,
sweep generator, and scope, and just play with the dimensions until it
works. Calculations, NEC models, and theory are great, but I prefer
to see the results on some test equipment. (For 2.4 and 5.7Ghz
antennas, I have a rather marginal test range and transfer standard,
which are suitable for estimating the gain of the antenna).

I get a hasty email with the question and discover that I
really don't know the answer. So, I posted the question here. We'll
find out how it worked when they return.

To some extent, everything 'works'. Just stripping a quarter wave of
shield off the end of a length of coax probably 'works' about as well.

That's probably what she ended up building. If she can talk into the
local repeaters, it's good enough. In this case, the 13 year old
antenna builder is probably not up to doing much more than measuring,
removing the outer jacket and peeling back the braid. I was going to
suggest adding a decoupling sleeve (using aluminum foil duct tape),
but decided that would be too much for a quick build. Maybe when they
return.

I forgot who rammed this down my throat when first got started in
engineering.... "The secret to engineering is to know when to stop
engineering and start building". It's a moving target, that has many
implications, but my method is that when it meets the "suitability for
purpose" criteria, you're done.

Whether the home made antenna is optimized for best performance
doesn't seem to be important in this case.

Indeed, it seems that for most modern hams, performance isn't an issue
or of interest.

That's also true for 90% of bicyclists. 90% of the bicycling public
do not race their bicycles. Similarly, 90% of the hams probably are
not optimizing their antennas and designs to what can theoretically be
achieved. It doesn't take much to look at an existing design of
literally anything, and think of a better way to do it. I do this all
the time. (Former ladyfriend called it the engineering mentality and
claimed it drove her nuts). Whether the average ham wants or needs
that last dB squeezed out of the design is problematic. If I were
doing moonbounce or microwave DX, where every fraction of a dB is
important, I would certainly be into such optimizations. However, for
yacking on the local repeaters, almost anything will work.

I'll probably steal a few of your ideas and comments in order to build
a "better" coax cable dipole, if she returns with an interest in how
it works, or how to make it better.



--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558