I've used a couple programs (vertload.exe and I think LODI) to design
a shortened, single band dipole antenna for 30m. Specifically I'm
interested in 10.149MHz.

The result is:

--1m wire--[~34t 1.5" form 2" long, ~24uH]----2m wire----[center]--...

The center would be a 1:1 current balun, twisting the wires together
and then taking 10-12 turns on FT50-61 (or if that won't fit, fewer
turns on FT50-43).

The predicted resistive input impedance is 22 ohms. Vertload is just
simulating half the antenna (with 0 ohm ground loss) so its match info
is not useful. The other program called for a ".71uH matching coil"
for 50 feed, which I assume is to cancel a calculated -j45 reactive
component??

I would like to pre-calculate an approximate match to Z=50ohms which
I can put directly at the feed point and shrink-wrap with the balun.
Is there a way to do that, or is it too touchy to do without measuring
the actual antenna?

--
Ben Jackson AD7GD

http://www.ben.com/

Ben Jackson wrote:
I would like to pre-calculate an approximate match to Z=50ohms which
I can put directly at the feed point and shrink-wrap with the balun.
Is there a way to do that, or is it too touchy to do without measuring
the actual antenna?

If you were within 20%, you would be lucky. Best to fine
tune it before shrink-wrapping it.
--
73, Cecil http://www.qsl.net/w5dxp

On Tue, 18 Jul 2006 17:37:29 -0500, Ben Jackson wrote:

I've used a couple programs (vertload.exe and I think LODI) to design
a shortened, single band dipole antenna for 30m. Specifically I'm
interested in 10.149MHz.

The result is:

--1m wire--[~34t 1.5" form 2" long, ~24uH]----2m wire----[center]--...

The center would be a 1:1 current balun, twisting the wires together
and then taking 10-12 turns on FT50-61 (or if that won't fit, fewer
turns on FT50-43).

The predicted resistive input impedance is 22 ohms. Vertload is just
simulating half the antenna (with 0 ohm ground loss) so its match info
is not useful. The other program called for a ".71uH matching coil"
for 50 feed, which I assume is to cancel a calculated -j45 reactive
component??

This is not very clear, you say 22 ohms then seem to say it is half
the antenna, is the input resistance to the dipole ~22 ohms or ~44
ohms?

Whichever, it is a relatively simple matter to calculate the
components of an L match, where you detune the dipole to get a small
capacitive reactance and shunt the feedpoint with a coil to match to
50 ohms.

If the R component is around 44 ohms, you need to detune the dipole to
about -18 ohms of reactance and shunt the feedpoint with about 2uH.

If the R component is around 22 ohms, you already know the answer, you
need to detune the dipole to about -25 ohms of reactance and shunt the
feedpoint with about 0.7uH.


I would like to pre-calculate an approximate match to Z=50ohms which
I can put directly at the feed point and shrink-wrap with the balun.
Is there a way to do that, or is it too touchy to do without measuring
the actual antenna?

An adventurous approach. You seem uncertain about the calculated
design, doesn't that suggest trying it before committing it
permanently?

Owen
--

On 2006-07-18, Owen Duffy wrote:
This is not very clear, you say 22 ohms then seem to say it is half
the antenna, is the input resistance to the dipole ~22 ohms or ~44
ohms?

When I use vertload.exe to model one HALF of the dipole, I get a
radiation resistance of about 12 ohms plus wire loss of about 1.5 for
an input resistance of 13.5 ohms. I am assuming that if I put two
of these shortened quarter-wave verticals back to back I get a dipole
with performance (but not radiation pattern) similar to the quarter
wave vertical over a near-perfect ground.

[and I'm going to stick with those numbers and ignore the other
program for this post]

Whichever, it is a relatively simple matter to calculate the
components of an L match, where you detune the dipole to get a small
capacitive reactance and shunt the feedpoint with a coil to match to
50 ohms.

Ok, so the math would go like this:

At 10.149MHz, my input resistance would be 13.5*2 (two verticals back-
to-back) or 27 ohms. It's got no reactive component at all because
it's tuned perfectly.

So if I added -25j capacitive reactance, for example with about 620p
in series with feed point, then transformed the series impedance
Z=27-25j into parallel admittance Y=0.02+0.018j, then I see my
equivalent parallel resistive component is 50 ohms (ok!) and I am
left with a -54j parallel capacitive reactive component, which I can
cancel with a 54j parallel inductive component, which is my ~85uH
inductor across the feedpoint.

Now, to save money on capacitors, I could alternatively detune the
dipole by shortening it until it was Z=27-25j at my 10.149MHz center
frequency, and then shunt as before to get the same effect. Is
that right?

I guess I also don't understand what happened to my resonance point
when the changes were made to a tuned antenna to produce the desired
feedpoint impedance.

An adventurous approach. You seem uncertain about the calculated
design, doesn't that suggest trying it before committing it
permanently?

Well, I know relatively nothing about building antennas, and I have
no relevant test equipment (except an SWR meter, but no transmitter
as yet) and no antenna tuner. Getting the antenna built was a side
project while I wait for transmitter parts. Perhaps building a quick
and dirty L-match would be a better use of my time!

--
Ben Jackson AD7GD

http://www.ben.com/

On Tue, 18 Jul 2006 22:23:53 -0500, Ben Jackson wrote:

On 2006-07-18, Owen Duffy wrote:
This is not very clear, you say 22 ohms then seem to say it is half
the antenna, is the input resistance to the dipole ~22 ohms or ~44
ohms?

When I use vertload.exe to model one HALF of the dipole, I get a
radiation resistance of about 12 ohms plus wire loss of about 1.5 for
an input resistance of 13.5 ohms. I am assuming that if I put two
of these shortened quarter-wave verticals back to back I get a dipole
with performance (but not radiation pattern) similar to the quarter
wave vertical over a near-perfect ground.

[and I'm going to stick with those numbers and ignore the other
program for this post]

Whichever, it is a relatively simple matter to calculate the
components of an L match, where you detune the dipole to get a small
capacitive reactance and shunt the feedpoint with a coil to match to
50 ohms.

Ok, so the math would go like this:

At 10.149MHz, my input resistance would be 13.5*2 (two verticals back-
to-back) or 27 ohms. It's got no reactive component at all because
it's tuned perfectly.

So if I added -25j capacitive reactance, for example with about 620p
in series with feed point, then transformed the series impedance
Z=27-25j into parallel admittance Y=0.02+0.018j, then I see my
equivalent parallel resistive component is 50 ohms (ok!) and I am
left with a -54j parallel capacitive reactive component, which I can
cancel with a 54j parallel inductive component, which is my ~85uH
inductor across the feedpoint.


You probably meant ~0.85uH.

Now, to save money on capacitors, I could alternatively detune the
dipole by shortening it until it was Z=27-25j at my 10.149MHz center
frequency, and then shunt as before to get the same effect. Is
that right?

Yes, though the R will change a little... that is one reason why my
response was liberally filled with "about".


I guess I also don't understand what happened to my resonance point
when the changes were made to a tuned antenna to produce the desired
feedpoint impedance.

Resonance of the "radiator" is not a prerequisite of performance. The
changes that you are likely to make to introduce -25 ohms or so of
reactance will not impact the losses in the radiator much or the
pattern.


An adventurous approach. You seem uncertain about the calculated
design, doesn't that suggest trying it before committing it
permanently?

Well, I know relatively nothing about building antennas, and I have
no relevant test equipment (except an SWR meter, but no transmitter
as yet) and no antenna tuner. Getting the antenna built was a side
project while I wait for transmitter parts. Perhaps building a quick
and dirty L-match would be a better use of my time!

With your guestimates of the starting point, why don't you wind the
0.85uH inductor so that you can adjust it a little, build the dipole
and loading coils and tune the dipole for least SWR (length or loading
coil adjustment). If the least SWR is not low enough, tweak the 0.8uH
inductor higher or lower and repeat the process, and follow the clues.

Owen
--

On Tue, 18 Jul 2006 22:23:53 -0500, Ben Jackson wrote:

Perhaps building a quick
and dirty L-match would be a better use of my time!

Hi Ben,

You are on the right track. Put up the tallest stick you can, plant a
dozen radials around it of approximately the same length, and match it
with a gamma match. You don't really stand to gain much by making it
any more "efficient."

73's
Richard Clark, KB7QHC

Ben Jackson wrote in
:

I've used a couple programs (vertload.exe and I think LODI) to design
a shortened, single band dipole antenna for 30m. Specifically I'm
interested in 10.149MHz.

Ah, the good old 10.149 packet net? I used to run a node on there
(VE1EI) a fey years ago.

The result is:

--1m wire--[~34t 1.5" form 2" long, ~24uH]----2m wire----[center]--...

The center would be a 1:1 current balun, twisting the wires together
and then taking 10-12 turns on FT50-61 (or if that won't fit, fewer
turns on FT50-43).

The predicted resistive input impedance is 22 ohms. Vertload is just
simulating half the antenna (with 0 ohm ground loss) so its match info
is not useful. The other program called for a ".71uH matching coil"
for 50 feed, which I assume is to cancel a calculated -j45 reactive
component??

I would like to pre-calculate an approximate match to Z=50ohms which
I can put directly at the feed point and shrink-wrap with the balun.
Is there a way to do that, or is it too touchy to do without measuring
the actual antenna?


How high is this antenna and above what kind of ground?

For 15m high, I get values of 21.21 microhenries and 9.85 picofarads
across the feedpoint (you could make that out of a short coax stub). SWR
would be 1.82 to 1 on 50 ohms, though. That could be improved upon by
altering the length of the outer wires a bit and tuning the capacitor to
match.

--
Dave Oldridge+
ICQ 1800667