I'm building a 4 segment + whip coaxial collinear antenna for
2.441 ghz. The coax I'm using is RG58 with a VF of .66 but
I was told by some of you that the VF will vary on this antenna.

What I need to know is which way and how much should I
adjust the element lengths to compensate?

Also, I followed the instructions on this webpage to the letter:

http://wireless.gumph.org/content/4/...s-antenna.html

But instead of getting sector lengths of 40.5mm, I got 39mm
somehow. Have I created a waste of time because of that?

One last question - with HF antennas, if they are close to other
objects they sometimes need to be shortened slightly to work.
I plan on mounting these antennas on the back of my ThinkPad's
display so maybe they will work being 1.5mm short per sector?

God I love this stuff!
--
73's de Ken KG0WX - Kadiddlehopper #11808,
Flying Pigs #-1055, Grid EM17io,
Elecraft K2 #4913, XG2, 4SQRP Tenna Dipper,
Heath GD-1B, MP-1(X)antenna & HLA-150 amp.

Measure the velocity factor (at or near the operating frequency)! If
you are using solid polyethylene dielectric cable, then the v.f. really
should be close to 0.66. For a flat pattern, you want the
center-to-center spacing of the elements to be 1/2 wave, considering
the net velocity factor. I say net because at the junctions between
elements, it's possible that the effective v.f. is slightly higher than
in the coax itself. You'd cut the elements long by enough to let you
solder them together and end up with that 1/2 wave center to center.
(It's easier and equivalent to measure from top of one element to the
top of the next element, etc.)

Since the _pattern_ depends mainly on the phasing of the feed system,
and the phasing is established mainly by the propagation velocity in
the line, and not by the surrounding environment, putting it next to
something won't affect the pattern, except that if there are conductors
in that display you want to mount it to, they may very well screw up
the pattern by allowing current where you don't want it. Especially if
there is some piece that's resonant near the operating frequency it
would be like putting an unwanted parasitic element in a Yagi: not a
good idea for optimum performance.

OK, so the phasing is designed to get you the right pattern. But the
phasing is independent of the feedpoint impedance. You need to match
to that, and you also need to decouple the antenna from nearby
conductors. Any conductors in the display fall into that category, but
the feedline from the xcvr also does. A choke balun, and other
feedline choking, is very highly recommended.

I used to have a web page with quite a bit of text about the why and
the how of coaxial collinears; it wasn't a construction article, but it
armed you with enough info that you could go out and make one that
would work well on pretty much any frequency you might want. It
included things like why the center-center half wave spacing, and why
it's slightly better to use foam dielectric cable. If I got enough
requests, I'd make a revised version of that available, maybe even with
some pictures this time. Every once in a while I get a request to
re-print it in a club newsletter, so I know there are some of those
floating around out there.

Cheers,
Tom

I saw a construction article in the ARRL Antenna Handbook , 13th Edition,
1980 on page 247-249. I also did some google searchs and there was a thesis
paper on this very same article in the 1990's where two people modeled the
antenna described in the Antenna Handbook and said there were errors in the
article. I could not get the complete paper that would have gone into the
details.

I have built one of these antennas for the 850 mHz range using rigid coax
per the article in the ARRL Handbook (with 7 half wave elements, and two 1/4
wave elements per the article) and saw poorer performance compared to a
small 1/4 whip antenna connnected to the back of the radio. Not sure what
went wrong.

Comments??

Tnx

Jim
"K7ITM" wrote in message
ups.com...
Measure the velocity factor (at or near the operating frequency)! If
you are using solid polyethylene dielectric cable, then the v.f. really
should be close to 0.66. For a flat pattern, you want the
center-to-center spacing of the elements to be 1/2 wave, considering
the net velocity factor. I say net because at the junctions between
elements, it's possible that the effective v.f. is slightly higher than
in the coax itself. You'd cut the elements long by enough to let you
solder them together and end up with that 1/2 wave center to center.
(It's easier and equivalent to measure from top of one element to the
top of the next element, etc.)

Since the _pattern_ depends mainly on the phasing of the feed system,
and the phasing is established mainly by the propagation velocity in
the line, and not by the surrounding environment, putting it next to
something won't affect the pattern, except that if there are conductors
in that display you want to mount it to, they may very well screw up
the pattern by allowing current where you don't want it. Especially if
there is some piece that's resonant near the operating frequency it
would be like putting an unwanted parasitic element in a Yagi: not a
good idea for optimum performance.

OK, so the phasing is designed to get you the right pattern. But the
phasing is independent of the feedpoint impedance. You need to match
to that, and you also need to decouple the antenna from nearby
conductors. Any conductors in the display fall into that category, but
the feedline from the xcvr also does. A choke balun, and other
feedline choking, is very highly recommended.

I used to have a web page with quite a bit of text about the why and
the how of coaxial collinears; it wasn't a construction article, but it
armed you with enough info that you could go out and make one that
would work well on pretty much any frequency you might want. It
included things like why the center-center half wave spacing, and why
it's slightly better to use foam dielectric cable. If I got enough
requests, I'd make a revised version of that available, maybe even with
some pictures this time. Every once in a while I get a request to
re-print it in a club newsletter, so I know there are some of those
floating around out there.

Cheers,
Tom

Also, consider that this is a GAIN antenna (a misnomer, as the "gain"
is created, by takeing power from somewhere else!). This means that the
radiation patern is on the horizon, tho, say 15 degrees above, or below
THAT horizon, the radiation from that "gain antenna" may be a LOSS! and
it gets worse, at higher angles! Kinda like aiming a yagi at a (say , 20
degree angle above (or below) the horizon, wouldn't expect to have a
full signal, if the SOURCE were on the horizon, would you)? A 1/4 wave
whip is much more forgiveing at higher angles of radiation0, while your
"GAIN" antenna has a pattern like a doughnut (power concentrated on
the horizon, limited radiation at angles above, and below that horizon).
This can be a problem in mountainous territory,where a 1/4 wave whip
will out-perform a gain antenna for coverage. As info, Jim NN7K


Jim wrote:
I saw a construction article in the ARRL Antenna Handbook , 13th Edition,
1980 on page 247-249. I also did some google searchs and there was a thesis
paper on this very same article in the 1990's where two people modeled the
antenna described in the Antenna Handbook and said there were errors in the
article. I could not get the complete paper that would have gone into the
details.

I have built one of these antennas for the 850 mHz range using rigid coax
per the article in the ARRL Handbook (with 7 half wave elements, and two 1/4
wave elements per the article) and saw poorer performance compared to a
small 1/4 whip antenna connnected to the back of the radio. Not sure what
went wrong.

Comments??

Tnx

Jim
"K7ITM" wrote in message
ups.com...

Measure the velocity factor (at or near the operating frequency)! If
you are using solid polyethylene dielectric cable, then the v.f. really
should be close to 0.66. For a flat pattern, you want the
center-to-center spacing of the elements to be 1/2 wave, considering
the net velocity factor. I say net because at the junctions between
elements, it's possible that the effective v.f. is slightly higher than
in the coax itself. You'd cut the elements long by enough to let you
solder them together and end up with that 1/2 wave center to center.
(It's easier and equivalent to measure from top of one element to the
top of the next element, etc.)

Since the _pattern_ depends mainly on the phasing of the feed system,
and the phasing is established mainly by the propagation velocity in
the line, and not by the surrounding environment, putting it next to
something won't affect the pattern, except that if there are conductors
in that display you want to mount it to, they may very well screw up
the pattern by allowing current where you don't want it. Especially if
there is some piece that's resonant near the operating frequency it
would be like putting an unwanted parasitic element in a Yagi: not a
good idea for optimum performance.

OK, so the phasing is designed to get you the right pattern. But the
phasing is independent of the feedpoint impedance. You need to match
to that, and you also need to decouple the antenna from nearby
conductors. Any conductors in the display fall into that category, but
the feedline from the xcvr also does. A choke balun, and other
feedline choking, is very highly recommended.

I used to have a web page with quite a bit of text about the why and
the how of coaxial collinears; it wasn't a construction article, but it
armed you with enough info that you could go out and make one that
would work well on pretty much any frequency you might want. It
included things like why the center-center half wave spacing, and why
it's slightly better to use foam dielectric cable. If I got enough
requests, I'd make a revised version of that available, maybe even with
some pictures this time. Every once in a while I get a request to
re-print it in a club newsletter, so I know there are some of those
floating around out there.

Cheers,
Tom

"Jim" wrote in message
news:4HAXf.10092$dU3.6515@trnddc01...
I saw a construction article in the ARRL Antenna Handbook , 13th Edition,
1980 on page 247-249. I also did some google searchs and there was a
thesis paper on this very same article in the 1990's where two people
modeled the antenna described in the Antenna Handbook and said there were
errors in the article. I could not get the complete paper that would have
gone into the details.

I have built one of these antennas for the 850 mHz range using rigid coax
per the article in the ARRL Handbook (with 7 half wave elements, and two
1/4 wave elements per the article) and saw poorer performance compared to
a small 1/4 whip antenna connnected to the back of the radio. Not sure
what went wrong.

Comments??

Tnx

Jim
"K7ITM" wrote in message

Hi Jim

A properly done colinear array of 8 half wave elements will sure have a
highly directive pattern in the Elevation plane. With some luck and low
loss components you might get 8 or 9 dBi. But, the pattern will have a
pattern max thats broadside to the axis of the colinear array at only one
frequency. That pattern max will squint up and down (with respect to the
horizon) as the frequency varies from that center frequency.
I'd bet that you will be able to realize appreciable dirrectivity from an
array of 7 lengths of RG-6 with 2 quarterwave added elements. But,
developing the array to provide low VSWR and best squint angle at any chosen
frequency will demant some "field testing".
You probably already knew that. I just wasnt sure. Besides I have wasted
alot of time developing a very similar colinear array for commercial use.

Jerry KD6JDJ

Jim,

Thanks for info. You are correct as I do live in "hilly" area. I raised
the antenna up to 30 feet and there was a big difference in signal strength.
I do not have equiment to measure the difference accurately, but on an Icom
R-7000 there was 1 s unit (or so) difference but there was much lower noise
and better readability.

Thanks again

Jim
"Jim - NN7K" wrote in message
et...
Also, consider that this is a GAIN antenna (a misnomer, as the "gain"
is created, by takeing power from somewhere else!). This means that the
radiation patern is on the horizon, tho, say 15 degrees above, or below
THAT horizon, the radiation from that "gain antenna" may be a LOSS! and
it gets worse, at higher angles! Kinda like aiming a yagi at a (say , 20
degree angle above (or below) the horizon, wouldn't expect to have a
full signal, if the SOURCE were on the horizon, would you)? A 1/4 wave
whip is much more forgiveing at higher angles of radiation0, while your
"GAIN" antenna has a pattern like a doughnut (power concentrated on
the horizon, limited radiation at angles above, and below that horizon).
This can be a problem in mountainous territory,where a 1/4 wave whip
will out-perform a gain antenna for coverage. As info, Jim NN7K


Jim wrote:
I saw a construction article in the ARRL Antenna Handbook , 13th Edition,
1980 on page 247-249. I also did some google searchs and there was a
thesis paper on this very same article in the 1990's where two people
modeled the antenna described in the Antenna Handbook and said there were
errors in the article. I could not get the complete paper that would
have gone into the details.

I have built one of these antennas for the 850 mHz range using rigid coax
per the article in the ARRL Handbook (with 7 half wave elements, and two
1/4 wave elements per the article) and saw poorer performance compared to
a small 1/4 whip antenna connnected to the back of the radio. Not sure
what went wrong.

Comments??

Tnx

Jim
"K7ITM" wrote in message
ups.com...

Measure the velocity factor (at or near the operating frequency)! If
you are using solid polyethylene dielectric cable, then the v.f. really
should be close to 0.66. For a flat pattern, you want the
center-to-center spacing of the elements to be 1/2 wave, considering
the net velocity factor. I say net because at the junctions between
elements, it's possible that the effective v.f. is slightly higher than
in the coax itself. You'd cut the elements long by enough to let you
solder them together and end up with that 1/2 wave center to center.
(It's easier and equivalent to measure from top of one element to the
top of the next element, etc.)

Since the _pattern_ depends mainly on the phasing of the feed system,
and the phasing is established mainly by the propagation velocity in
the line, and not by the surrounding environment, putting it next to
something won't affect the pattern, except that if there are conductors
in that display you want to mount it to, they may very well screw up
the pattern by allowing current where you don't want it. Especially if
there is some piece that's resonant near the operating frequency it
would be like putting an unwanted parasitic element in a Yagi: not a
good idea for optimum performance.

OK, so the phasing is designed to get you the right pattern. But the
phasing is independent of the feedpoint impedance. You need to match
to that, and you also need to decouple the antenna from nearby
conductors. Any conductors in the display fall into that category, but
the feedline from the xcvr also does. A choke balun, and other
feedline choking, is very highly recommended.

I used to have a web page with quite a bit of text about the why and
the how of coaxial collinears; it wasn't a construction article, but it
armed you with enough info that you could go out and make one that
would work well on pretty much any frequency you might want. It
included things like why the center-center half wave spacing, and why
it's slightly better to use foam dielectric cable. If I got enough
requests, I'd make a revised version of that available, maybe even with
some pictures this time. Every once in a while I get a request to
re-print it in a club newsletter, so I know there are some of those
floating around out there.

Cheers,
Tom

"Jerry Martes" wrote
A properly done colinear array of 8 half wave elements will sure have a
highly directive pattern in the Elevation plane. With some luck and low
loss components you might get 8 or 9 dBi. But, the pattern will have a
pattern max thats broadside to the axis of the colinear array at only one
frequency. That pattern max will squint up and down (with respect to the
horizon) as the frequency varies from that center frequency.
_______________

The change with frequency in the angle of the elevation pattern maximum
field is the result of the varying relative r-f phase vs the operating
frequency that is applied to the radiating elements in this design.

This can be eliminated by using a vertically stacked array of identical
radiators fed via an n-way power divider, whose output connects to feed
cables of equal electical lengths to each element of the array. The
directivity (gain) and sidelobe distribution of this type of array will
change with operating frequency, but the elevation angle at which the peak
field is directed will remain the same, regardless of the input frequency.

This approach is used in broadcast transmit antenna designs giving good
signal coverage over large sections of the UHF TV band, or the entire FM
band.

RF

Do you have a sketch of the harness a guy might use to feed that 8
dipole colinear array? Remember, this colinear antenna is mounted on the
back of a Lap Top Computer for WiFi use.
____________

Sorry, not for that application.

Richard,

The antenna that I am looking to build is not for Wi-Fi, but for 840mHz
range and will be mounted outside on a pole about 25 feet high.

Tnx

Jim
"Richard Fry" wrote in message
...
Do you have a sketch of the harness a guy might use to feed that 8
dipole colinear array? Remember, this colinear antenna is mounted on
the back of a Lap Top Computer for WiFi use.
____________

Sorry, not for that application.

But if you feed a coaxial collinear at the center instead of at an end,
the top and bottom portions will have tilts in opposite directions, and
the maximum of the pattern will remain perpendicular to the axis of the
antenna, at least over a broad range of frequencies. The gain will
change, but not drastically; matching properly to the feedpoint over
the range of frequencies is likely going to be the bigger issue.

And if the OP puts the gain antenna on a panel which can be tilted at
various angles, the performance may well be far from what he really
wants. I'd suggest a small stand of some sort that will keep the
antenna in a fixed position, if the panel can move.

Cheers,
Tom