I was looking through some of my old antenna books and came across an
article from K6TS about building a 20 Meter Collinear antenna. Has
anyone ever built one, or have any input on if it would be worthwhile to
pursue. Weather is getting warm here, so I was looking for another
antenna project. 20 meters, vertical and ground mounted. So I was
thinking about 1/2 Wave, 5/8 wave or the collinear.

Ken

On 24 feb, 17:45, Ken Slimmer wrote:
* I was looking through some of my old antenna books and came across an
article from K6TS about building a 20 Meter Collinear antenna. *Has
anyone ever built one, or have any input on if it would be worthwhile to
pursue. *Weather is getting warm here, so I was looking for another
antenna project. *20 meters, vertical and ground mounted. *So I was
thinking about 1/2 Wave, 5/8 wave or the collinear.

Ken

Hello Ken,

I would not use the 5/8 wave antenna, unless you can make a dipole of
1.25lambda. The 5/8 wave vertical only gives the published gain over a
large good conducting ground plane. 3 or 4 quarter wave radials may
provide a reasonable floating ground for feeding the antenna, but it
is not a large ground plane.

Using a half wave has the disadvantage of the more complicated feeding
network. You may expect impedances up to kOhm range (depending in
thickness of the radiating element), so you need some high voltage
evaluation of your structure in case of 100W input power.

The advantage is the low requirement for the (floating) ground at the
feed point. Just 1 or 2 quarter wave radials are sufficient. These
radial wires may also slope down, as they carry low current, hence do
not have large influence on radiation pattern. When you have some
metal structure around you, you can use that as ground, eliminating
the need for radials. When you look to half wave CB antennas, most
ones do not have radials at all.

When you want to use horizontal polarization, a full wave center fed
dipole or 1.25lambda center fed antenna can be nice. Of course you
have to make something to rotate it….

When you want to design a vertical HW antenna from the ground up, I
have a document on my website dedicated to HW end-fed antenna design.
It also addresses high voltage issues. http://www.tetech.nl/divers/HWmonopoleNL1.pdf.
It is in Dutch language, but all comment in illustrations and formulas
is in English, so it can be helpful.

Best regards,

Wim
PA3DJS
www.tetech.nl
without abc, the mail is OK.

Wim;
Thanks for your input, I just noticed that I had my finger on the
shift key when I posted the subject, so it came out as @0 meter.. :-)
I don't ever plan on running high power, 100 watts is about my
limit. I do want to stick to putting a vertical up. I have an old 1/4
wave 40 meter vertical that I was going to scrounge parts from.

Ken


On Tue, 24 Feb 2009 10:19:14 -0800, wimabctel wrote:

On 24 feb, 17:45, Ken Slimmer wrote:
Â* I was looking through some of my old antenna books and came across
Â* an
article from K6TS about building a 20 Meter Collinear antenna. Â*Has
anyone ever built one, or have any input on if it would be worthwhile
to pursue. Â*Weather is getting warm here, so I was looking for another
antenna project. Â*20 meters, vertical and ground mounted. Â*So I was
thinking about 1/2 Wave, 5/8 wave or the collinear.

Ken

Hello Ken,

I would not use the 5/8 wave antenna, unless you can make a dipole of
1.25lambda. The 5/8 wave vertical only gives the published gain over a
large good conducting ground plane. 3 or 4 quarter wave radials may
provide a reasonable floating ground for feeding the antenna, but it is
not a large ground plane.

Using a half wave has the disadvantage of the more complicated feeding
network. You may expect impedances up to kOhm range (depending in
thickness of the radiating element), so you need some high voltage
evaluation of your structure in case of 100W input power.

The advantage is the low requirement for the (floating) ground at the
feed point. Just 1 or 2 quarter wave radials are sufficient. These
radial wires may also slope down, as they carry low current, hence do
not have large influence on radiation pattern. When you have some
metal structure around you, you can use that as ground, eliminating the
need for radials. When you look to half wave CB antennas, most ones do
not have radials at all.

When you want to use horizontal polarization, a full wave center fed
dipole or 1.25lambda center fed antenna can be nice. Of course you have
to make something to rotate it….

When you want to design a vertical HW antenna from the ground up, I have
a document on my website dedicated to HW end-fed antenna design. It also
addresses high voltage issues.
http://www.tetech.nl/divers/HWmonopoleNL1.pdf. It is in Dutch language,
but all comment in illustrations and formulas is in English, so it can
be helpful.

Best regards,

Wim
PA3DJS
www.tetech.nl
without abc, the mail is OK.

On Tue, 24 Feb 2009 10:19:14 -0800 (PST), wrote:

Hello Ken,

I would not use the 5/8 wave antenna, unless you can make a dipole of
1.25lambda. The 5/8 wave vertical only gives the published gain over a
large good conducting ground plane. 3 or 4 quarter wave radials may
provide a reasonable floating ground for feeding the antenna, but it
is not a large ground plane.

This and other points are deceptive. First, the performance you site
is indeed due to the plane of ground (not to be confused with our
usage of the term ground plane) to the extent of the conductivity of
ground out about 5 to 10 wavelengths away from the antenna. No
practical ground system is going to impact that.

The ground system placed below the antenna WILL impact gain, only
insofar as it shields the ground's loss contribution. Hence the large
number of radials.

Using a half wave has the disadvantage of the more complicated feeding
network. You may expect impedances up to kOhm range (depending in
thickness of the radiating element), so you need some high voltage
evaluation of your structure in case of 100W input power.

The advantage of the half wave is exactly for its high impedance in
relation to the loss of ground. The far ground still dominates low
angle launch characteristics, but if (like the large number of radials
offers) you lose less to ground, you have more in the air in all
directions.

The advantage is the low requirement for the (floating) ground at the
feed point. Just 1 or 2 quarter wave radials are sufficient. These
radial wires may also slope down, as they carry low current, hence do
not have large influence on radiation pattern.

If there is just 1, or if the 2 are not symmetrical, then the DO
contribute to the radiation pattern lobe shape. As to the degree or
notice, that is variable to the user/listener.

When you have some
metal structure around you, you can use that as ground, eliminating
the need for radials. When you look to half wave CB antennas, most
ones do not have radials at all.

They probably rely on the coax shield as a return path, which makes it
notoriously unreliable in its state of tune.

73's
Richard Clark, KB7QHC

On Feb 24, 2:44*pm, Ken Slimmer wrote:
Wim;
* * *Thanks for your input, I just noticed that I had my finger on the
shift key when I posted the subject, so it came out as @0 meter.. *:-)
* * *I don't ever plan on running high power, 100 watts is about my
limit. *I do want to stick to putting a vertical up. *I have an old 1/4
wave 40 meter vertical that I was going to scrounge parts from. *

If you already have a 32 ft radiator, I would go with the half wave.
Feeding one is simple. I prefer the "gamma loop" type of feed.
I've built many of those, and they are simple to get going, and work
well.
The antenna as it is does not require radials to function as a
"complete" antenna. Most will work fine as is, with no radials.
But if one wants to further improve one, they can be further decoupled
from the feedline using decoupling sections.
The way I usually decouple a base fed half wave is to use a 1/4 wave
length of coax dropping down the supporting mast to a union which
I clamp a set of quarter wave radials. The shield of the coax is
bonded
to the radial set.
That will do a pretty good job of decoupling the feed line, if common
mode current ends up a problem.
But like I say, I've never had any problems using one with no radials.
I consider the decoupling as optional.
A 5/8 wave GP would be nice, but that will take a pretty tall radiator
on 20m, and may be unpractical.
Also, I don't use 1/4 wave radials under 5/8 verticals.
This page explains my position on that issue..
http://home.comcast.net/~nm5k/acompari.htm
If you want to avoid the funky high angle lobe you often
see on poorly designed 5/8 wl antennas, use 5/8, or 3/4
wave radials.
I've done very extensive testing of all the usual types of verticals
on 10m, and the 5/8 GP is the best performer of the bunch,
even with it's supposed warts.
Even the 1/2 waves I used with decoupling sections were never
as good as the 5/8 GP when working far off space wave stations.
But like I say, a 5/8 GP on 20m could end up being a real pain
to deal with. You are talking a 41 ft radiator, and that needs a
supporting mast under it. Radials under it too.
BTW, a 1/4 wl GP is not a bad antenna if you want to take the
easy way out. Needs radials though.
I haven't looked at the collinear design, but sounds like more
trouble than it's worth. Would be tall too.
Most higher gain collinear antennas used on 20m would
probably be horizontal wire affairs.

On 25 feb, 00:10, Richard Clark wrote:
On Tue, 24 Feb 2009 10:19:14 -0800 (PST), wrote:
Hello Ken,

I would not use the 5/8 wave antenna, unless you can make a dipole of
1.25lambda. The 5/8 wave vertical only gives the published gain over a
large good conducting ground plane. 3 or 4 quarter wave radials may
provide a reasonable floating ground for feeding the antenna, but it
is not a large ground plane.

This and other points are deceptive. First, the performance you site
is indeed due to the plane of ground (not to be confused with our
usage of the term ground plane) to the extent of the conductivity of
ground out about 5 to 10 wavelengths away from the antenna. No
practical ground system is going to impact that.

The ground system placed below the antenna WILL impact gain, only
insofar as it shields the ground's loss contribution. Hence the large
number of radials.

Using a half wave has the disadvantage of the more complicated feeding
network. You may expect impedances up to kOhm range (depending in
thickness of the radiating element), so you need some high voltage
evaluation of your structure in case of 100W input power.

The advantage of the half wave is exactly for its high impedance in
relation to the loss of ground. The far ground still dominates low
angle launch characteristics, but if (like the large number of radials
offers) you lose less to ground, you have more in the air in all
directions.

That advantage of high impedance is also the disadvantage, 1500 Ohm
end-fed impedance, or higher, is not uncommon. With 400W input power,
this leads to 1100Vp voltage (at 1500 Ohms). Without careful
construction, E-field at sharp edges will exceed 3000V/mm easily. This
will not result in full air breakdown (due to strong nun-uniformity of
E-field, but will result in undesired corona discharge.

The advantage is the low requirement for the (floating) ground at the
feed point. Just 1 or 2 quarter wave radials are sufficient. These
radial wires may also slope down, as they carry low current, hence do
not have large influence on radiation pattern.

If there is just 1, or if the 2 are not symmetrical, then the DO
contribute to the radiation pattern lobe shape. As to the degree or
notice, that is variable to the user/listener.

End fed impedance for 3cm thick radiator is about 1500 Ohms, hence
radiator current (middle) is about 5 times higher then feed current to
the quarter wave radial. Therefore current*length product for radiator
is 10 times as high as for the radial. When the radial is vertically
oriented (worst case situation) influence on field from radiator is
+/- 10%. So very worst case you are talking of 1dB. When the radial
runs horizontally, the effect on the vertical component under low
elevation angle is negligible. As the original question relates to
amateur service, mentioning: "hence do not have large influence on
radiation pattern" is justified, in my opinion.


When you have some
metal structure around you, you can use that as ground, eliminating
the need for radials. When you look to half wave CB antennas, most
ones do not have radials at all.

They probably rely on the coax shield as a return path, which makes it
notoriously unreliable in its state of tune.

Mostly CB antennas are mounted on a metal mast, so part of the return
current goes through the mast. You are right, in some cases this may
lead to significant common mode current, but looking to my experience,
this seldom resulted in untunable systems when lambda/dradiator is
high. I did experience problems in antennas for VHF where thickness
of radiator is no longer thin to wavelength. Fortunately, at such
frequencies a simple ground is easy to make.

73's
Richard Clark, KB7QHC

Wim
PA3DJS

On 25 feb, 09:10, wrote:
On Feb 24, 2:44*pm, Ken Slimmer wrote:

Wim;
* * *Thanks for your input, I just noticed that I had my finger on the
shift key when I posted the subject, so it came out as @0 meter.. *:-)
* * *I don't ever plan on running high power, 100 watts is about my
limit. *I do want to stick to putting a vertical up. *I have an old 1/4
wave 40 meter vertical that I was going to scrounge parts from. *

If you already have a 32 ft radiator, I would go with the half wave.
Feeding one is simple. I prefer the "gamma loop" type of feed.
I've built many of those, and they are simple to get going, and work
well.
The antenna as it is does not require radials to function as a
"complete" antenna. Most will work fine as is, with no radials.
But if one wants to further improve one, they can be further decoupled
from the feedline using decoupling sections.
The way I usually decouple a base fed half wave is to use a 1/4 wave
length of coax dropping down the supporting mast to a union which
I clamp a set of quarter wave radials. The shield of the coax is
bonded
to the radial set.
That will do a pretty good job of decoupling the feed line, if common
mode current ends up a problem.
But like I say, I've never had any problems using one with no radials.
I consider the decoupling as optional.
A 5/8 wave GP would be nice, but that will take a pretty tall radiator
on 20m, and may be unpractical.
Also, I don't use 1/4 wave radials under 5/8 verticals.
This page explains my position on that issue..http://home.comcast.net/~nm5k/acompari.htm
If you want to avoid the funky high angle lobe you often
see on poorly designed 5/8 wl antennas, use 5/8, or 3/4
wave radials.
I've done very extensive testing of all the usual types of verticals
on 10m, and the 5/8 GP is the best performer of the bunch,
even with it's supposed warts.
Even the 1/2 waves I used with decoupling sections were never
as good as the 5/8 GP when working far off space wave stations.
But like I say, a 5/8 GP on 20m could end up being a real pain
to deal with. You are talking a 41 ft radiator, and that needs a
supporting mast under it. Radials under it too.
BTW, a 1/4 wl GP is not a bad antenna if you want to take the
easy way out. Needs radials though.
I haven't looked at the collinear design, but sounds like more
trouble than it's worth. Would be tall too.
Most higher gain collinear antennas used on 20m would
probably be horizontal wire affairs.

Nice simulations, and nice results also!

Best regards,

Wim
PA3DJS
www.tetech.nl
the mail is ok when you remove abc.

On Wed, 25 Feb 2009 04:04:39 -0800 (PST), wrote:

The advantage of the half wave is exactly for its high impedance in
relation to the loss of ground. The far ground still dominates low
angle launch characteristics, but if (like the large number of radials
offers) you lose less to ground, you have more in the air in all
directions.

....
this leads to 1100Vp voltage (at 1500 Ohms). Without careful
construction, E-field at sharp edges will exceed 3000V/mm easily.
....
This
will not result in full air breakdown (due to strong nun-uniformity of
E-field, but will result in undesired corona discharge.

I would have thought they were the same. (Corona discharge is NOT air
breakdown? Is there some distinction to "full?")

However, this is not an exclusive disadvantage of a half-wave radiator
when a quarter-wave radiator can exhibit similar problems through
similar poor building practices. The same high potential "problem"
exists at the quarter-wave's distant end where the half-wave's is at
the near end, feed point. As the original poster posed this as a
single band antenna, a half wave has practical solutions if the
additional gain is deemed sufficient for the effort.

On the other hand, going from quarter-wave to 5/8ths does yield
benefit that exceeds +/- 10% or 1dB.

73's
Richard Clark, KB7QHC

Richard Clark wrote:
On Wed, 25 Feb 2009 04:04:39 -0800 (PST), wrote:

The advantage of the half wave is exactly for its high impedance in
relation to the loss of ground. The far ground still dominates low
angle launch characteristics, but if (like the large number of radials
offers) you lose less to ground, you have more in the air in all
directions.
...
this leads to 1100Vp voltage (at 1500 Ohms). Without careful
construction, E-field at sharp edges will exceed 3000V/mm easily.
...
This
will not result in full air breakdown (due to strong nun-uniformity of
E-field, but will result in undesired corona discharge.

I would have thought they were the same. (Corona discharge is NOT air
breakdown? Is there some distinction to "full?")

One often makes a distinction between a corona discharge which exists as
a steady state sort of thing and the streamers which precede a "spark".
Both are air breakdown phenomena, but qualitatively different, and
both are different from a low pressure discharge like that found in a
fluorescent lamp or neon bulb, or from phenomena like St Elmo's Fire.

Field uniformity is only part of it, of course.

Bazelyan & Raizer, "Spark Discharge" from CRC Press, 1998, is probably
one of the better books on this.

Jim Lux wrote:
breakdown? Is there some distinction to "full?")

One often makes a distinction between a corona discharge which exists as
a steady state sort of thing and the streamers which precede a "spark".
Both are air breakdown phenomena, but qualitatively different, and both
are different from a low pressure discharge like that found in a
fluorescent lamp or neon bulb, or from phenomena like St Elmo's Fire.
. . .

Most interesting. I've always thought that St. Elmo's fire was a corona
discharge, and a quick web search indicates that it's apparently a very
widely held misconception. What's the difference?

Roy Lewallen, W7EL