On Oct 2, 12:23 pm, Jim Kelley wrote:
michel wrote:
I don't know what construction technique you have in mind, but I prefer to
build them from 1/2" copper pipe. If that's what you intend to do, then
just cut the vertical radiator 5/8ths instead of 1/2 lambda. Build the
rest the same (1/4 wave stub). Attach your coax to the stub section using
hose clamps, and tune by sliding the clamps up or down until you achieve
best SWR. Works great.

ac6xg

Should it be that simple?

It can be.

I understood from other posters that it will not
work with the 1/4 wave stub..

For some posters, a view of the forest is obscured by trees. A j-pole
is a quarter wave stub with a 1/2 wave (or longer) radiator attached
to one side of the open end. The feed is connected nearer to the
shorted end of the stub. Build it and they (the QSOs) will come. Get
it up as high above the roof as you can.

ac6xg

I suppose that since there will be an antenna current on the stub, and
the 5/8 section would show a reactive feedpoint if fed against a
ground plane, things are a bit more complicated than just a resistive
matching section (the 1/4 wave stub). I would expect that (1) the
antenna won't behave quite like a 5/8 wave fed against ground, nor
like a 5/4 wave center fed doublet, and (2) the stub will have to be
adjusted in length as well as in transformation ratio to get a
"perfect" match. In addition, if the feedline is not decoupled from
the antenna, the antenna current on the feedline will change both the
pattern and the feedpoint impedance (match) from what it would be if
the stub+radiator were in freespace.

I'd do some NEC simulating to get an idea of a starting point AND an
idea if the pattern was really an improvement over the normal half-
wave over a quarter-wave stub, before trying to build one; and I'd put
some effort into decoupling the antenna from other nearby metal
(including the feedline)--or at least include other elements in the
simulation. Even with simulating, I'd expect to have to do some fine
tuning (of stub spacing or stub tap point, and possibly of stub
length) if I really cared about a good match.

Cheers,
Tom

K7ITM wrote:
On Oct 2, 12:23 pm, Jim Kelley wrote:

michel wrote:

I don't know what construction technique you have in mind, but I prefer to
build them from 1/2" copper pipe. If that's what you intend to do, then
just cut the vertical radiator 5/8ths instead of 1/2 lambda. Build the
rest the same (1/4 wave stub). Attach your coax to the stub section using
hose clamps, and tune by sliding the clamps up or down until you achieve
best SWR. Works great.

ac6xg

Should it be that simple?

It can be.


I understood from other posters that it will not
work with the 1/4 wave stub..

For some posters, a view of the forest is obscured by trees. A j-pole
is a quarter wave stub with a 1/2 wave (or longer) radiator attached
to one side of the open end. The feed is connected nearer to the
shorted end of the stub. Build it and they (the QSOs) will come. Get
it up as high above the roof as you can.

ac6xg


I suppose that since there will be an antenna current on the stub, and
the 5/8 section would show a reactive feedpoint if fed against a
ground plane, things are a bit more complicated than just a resistive
matching section (the 1/4 wave stub). I would expect that (1) the
antenna won't behave quite like a 5/8 wave fed against ground, nor
like a 5/4 wave center fed doublet, and (2) the stub will have to be
adjusted in length as well as in transformation ratio to get a
"perfect" match. In addition, if the feedline is not decoupled from
the antenna, the antenna current on the feedline will change both the
pattern and the feedpoint impedance (match) from what it would be if
the stub+radiator were in freespace.

I'd do some NEC simulating to get an idea of a starting point AND an
idea if the pattern was really an improvement over the normal half-
wave over a quarter-wave stub, before trying to build one; and I'd put
some effort into decoupling the antenna from other nearby metal
(including the feedline)--or at least include other elements in the
simulation. Even with simulating, I'd expect to have to do some fine
tuning (of stub spacing or stub tap point, and possibly of stub
length) if I really cared about a good match.

Cheers,
Tom

Hi Tom -

I've built several of them and achieved a good (1.5:1) match in the
middle of the band.

73, ac6xg

K7ITM wrote:

On Oct 2, 12:23 pm, Jim Kelley wrote:

michel wrote:

I don't know what construction technique you have in mind, but I prefer to
build them from 1/2" copper pipe. If that's what you intend to do, then
just cut the vertical radiator 5/8ths instead of 1/2 lambda. Build the
rest the same (1/4 wave stub). Attach your coax to the stub section using
hose clamps, and tune by sliding the clamps up or down until you achieve
best SWR. Works great.

ac6xg

Should it be that simple?

It can be.


I understood from other posters that it will not
work with the 1/4 wave stub..

For some posters, a view of the forest is obscured by trees. A j-pole
is a quarter wave stub with a 1/2 wave (or longer) radiator attached
to one side of the open end. The feed is connected nearer to the
shorted end of the stub. Build it and they (the QSOs) will come. Get
it up as high above the roof as you can.

ac6xg


I suppose that since there will be an antenna current on the stub, and
the 5/8 section would show a reactive feedpoint if fed against a
ground plane, things are a bit more complicated than just a resistive
matching section (the 1/4 wave stub). I would expect that (1) the
antenna won't behave quite like a 5/8 wave fed against ground, nor
like a 5/4 wave center fed doublet, and (2) the stub will have to be
adjusted in length as well as in transformation ratio to get a
"perfect" match. In addition, if the feedline is not decoupled from
the antenna, the antenna current on the feedline will change both the
pattern and the feedpoint impedance (match) from what it would be if
the stub+radiator were in freespace.

I'd do some NEC simulating to get an idea of a starting point AND an
idea if the pattern was really an improvement over the normal half-
wave over a quarter-wave stub, before trying to build one; and I'd put
some effort into decoupling the antenna from other nearby metal
(including the feedline)--or at least include other elements in the
simulation. Even with simulating, I'd expect to have to do some fine
tuning (of stub spacing or stub tap point, and possibly of stub
length) if I really cared about a good match.

Cheers,
Tom

I found them to work quite well, though I fine tuned them on the mast
- not in the garage next to the transmitter.

73, jk

On Oct 2, 12:23 pm, Jim Kelley wrote:
michel wrote:
I don't know what construction technique you have in mind, but I prefer to
build them from 1/2" copper pipe. If that's what you intend to do, then
just cut the vertical radiator 5/8ths instead of 1/2 lambda. Build the
rest the same (1/4 wave stub). Attach your coax to the stub section using
hose clamps, and tune by sliding the clamps up or down until you achieve
best SWR. Works great.

ac6xg

Should it be that simple?

It can be.

I understood from other posters that it will not
work with the 1/4 wave stub..

For some posters, a view of the forest is obscured by trees. A j-pole
is a quarter wave stub with a 1/2 wave (or longer) radiator attached
to one side of the open end. The feed is connected nearer to the
shorted end of the stub. Build it and they (the QSOs) will come. Get
it up as high above the roof as you can.

ac6xg

I suppose that since there will be an antenna current on the stub, and
the 5/8 section would show a reactive feedpoint if fed against a
ground plane, things are a bit more complicated than just a resistive
matching section (the 1/4 wave stub). I would expect that (1) the
antenna won't behave quite like a 5/8 wave fed against ground, nor
like a 5/4 wave center fed doublet, and (2) the stub will have to be
adjusted in length as well as in transformation ratio to get a
"perfect" match. In addition, if the feedline is not decoupled from
the antenna, the antenna current on the feedline will change both the
pattern and the feedpoint impedance (match) from what it would be if
the stub+radiator were in freespace.

I'd do some NEC simulating to get an idea of a starting point AND an
idea if the pattern was really an improvement over the normal half-
wave over a quarter-wave stub, before trying to build one; and I'd put
some effort into decoupling the antenna from other nearby metal
(including the feedline)--or at least include other elements in the
simulation. Even with simulating, I'd expect to have to do some fine
tuning (of stub spacing or stub tap point, and possibly of stub
length) _if_ I really cared about a good match.

(Even a half-wave over a nominally quarter-wave stub can benefit from
tuning the stub length to get rid of reactance, if you really care
about a good match [though that in itself is somewhat over-rated].
With the stub attached, it's not a simple half-wave radiator. There
will be antenna current on the stub.)

Cheers,
Tom

michel wrote:
Should it be that simple? I understood from other posters that it will not
work with the 1/4 wave stub..

A 1/2WL + (1/4WL series stub) is a resonant Zepp.
When the stub is shorted at the bottom and tapped
for 50 ohms, it becomes a typical J-Pole.

A 5/8WL + (1/5WL series stub) is a resonant Extended
Zepp. There is only one resonant point and it may
not be at 50 ohms but it may be close enough. You
won't know till you try it. The feedpoint impedance
can be varied by varying the spacing between the
stub elements.

A 5/8WL + (1/4WL series stub) is non-resonant.
Anywhere you tap on the stub will result in some
reactance. There's really no reason to build
this inferior design.
--
73, Cecil http://www.w5dxp.com

"Cecil Moore" schreef in bericht
t...
michel wrote:
Should it be that simple? I understood from other posters that it will
not work with the 1/4 wave stub..

A 1/2WL + (1/4WL series stub) is a resonant Zepp.
When the stub is shorted at the bottom and tapped
for 50 ohms, it becomes a typical J-Pole.

A 5/8WL + (1/5WL series stub) is a resonant Extended
Zepp. There is only one resonant point and it may
not be at 50 ohms but it may be close enough. You
won't know till you try it. The feedpoint impedance
can be varied by varying the spacing between the
stub elements.

A 5/8WL + (1/4WL series stub) is non-resonant.
Anywhere you tap on the stub will result in some
reactance. There's really no reason to build
this inferior design.
--

Cecil,

That is some clear information!! The 5/8 wave version needs a shorted stub
as well?

michel wrote:
That is some clear information!! The 5/8 wave version needs a shorted stub
as well?

What we are looking for is the length of a series
matching stub section that will bring the system
to resonance, i.e. we are looking for the current
maximum point.

The graph at http://www.w5dxp.com/majic.gif
gives the length of the series matching stub
section for dipoles of various lengths from
1/2WL to 1.5WL. The graph can be used for
monopoles by doubling the monopole length.

For instance, doubling the 5/8WL monopole length
gives a 10/8WL dipole length which is 1.25WL.
A 1.25WL dipole needs a series matching stub section
of 0.19WL, i.e. the bottom stub on a 5/8WL J-Pole
needs to be 0.19WL. Make it a little too long and
trim for resonance.

If you double the 1/2WL monopole length of the
J-Pole you get a 1.0WL dipole. From the graph,
a 1.0WL dipole needs a 0.25WL series matching
stub section, i.e. the bottom stub on a 1/2WL
J-Pole needs to be 0.25WL. Of course, that is
the standard J-Pole design.

If you happened to want to design a 0.35WL J-Pole,
the series matching stub would need to be 0.33WL
long. 2(0.35WL) = 0.7WL

Let me know if you understand the graph.
--
73, Cecil http://www.w5dxp.com

michel wrote:
That is some clear information!! The 5/8 wave version needs a shorted stub
as well?

After my first cup of coffee, I canceled my previous
posting. The matching method I described works well
for balanced dipoles but is probably not applicable
to monopoles with no ground plane because the current
in the other leg of the series matching section has
no place to flow in a J-Pole configuration. So a 0.2WL
matching stub is not a good feed design for a 5/8WL
monopole and will generate common-mode problems unless
there is a ground plane into which the current can flow.
Best to stick with the standard 1/2WL J-Pole design. I
apologize for my fuzzy thinking - it made sense until
I woke up. And it would work for a 5/8WL monopole if
it already had ground plane radials.
--
73, Cecil http://www.w5dxp.com

WELL, I WOULD DO A 3/4 WAVE BY A 1/4 WAVE.

THIS WORKS REAL WELL ON 10 METERS WITH A LOT OF GAIN.

CHECK OUT MINE ON MY WEBSITE.

http://www.dxradioworld.com

Gary #203 wrote:
WELL, I WOULD DO A 3/4 WAVE BY A 1/4 WAVE.

A 3/4WL vertical monopole has a take-off-angle of 47 degrees.
5/8WL is the length limit for good monopole DX performance.
--
73, Cecil http://www.w5dxp.com