Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: Horizontal Full Wave Loop Weird SWR Curve  (Read 3184 times)
AK4SK
Member

Posts: 150




Ignore
« on: December 21, 2012, 07:23:29 AM »

I put up a full wave loop last Saturday and while it seems to work well the SWR curve is peculiar in one place. The fundamental frequency seems to be about 3.240 MHz. On my analyzer coming from the left (lower frequency) the curve sharply dips down to resonance in text book fashion. As the curve goes to the right (toward higher frequencies) it increases from resonance as expected and is roughly symmetrical to left side until about 3.500 MHz. At that point the curve levels out and even dips slightly (back towards lower SWR) until about 3.900 MHz where it resumes increasing up to high non-resonance values until the next resonant harmonic. As frequency is increased even higher there are the to be expected dips at all of the harmonics of the fundamental frequency but they all look normal. If I could post a picture that would help tremendously but that doesn't seem to be a feature of the forum.

Anyhow, I don't think that this is a problem as the antenna works well and tunes easily on the frequencies I use. I'm just curious as to what is going on with it. One thing to note is that the loop itself is irregular. I didn't have the free space to construct a square loops cut for 75 meters so I just followed my property line and house. About half of it, the part in my back yard, is roughly half a rectangle with the feed point higher than the rest of the antenna. The other half follows the roof line on the front of my house at 8 feet above the ground. The front of the house is a series of right angle turns or steps as seen from above. The antenna basically encompasses my back yard and entire house. Also, directly above the feed point and running parallel above part of the loop is a 66 foot terminated folded dipole. My original thought is that maybe this other antenna is acting as a parasitic element to the loop but I doubt that my antenna analyzer radiates enough of a field for the loop to "feel" the presence of the other antenna. 

Any thoughts?

73,
Chris AK4SK
Logged
WB6BYU
Member

Posts: 13005




Ignore
« Reply #1 on: December 21, 2012, 08:39:40 AM »

How are you feeding the loop?  I've run into similar quirks with a dip meter that were due to
the feedline rather than the antenna connected to it.

How close does the loop come to the other antenna?  Or to other wires or metal bits that
might be resonant?  The coupling is a function of the spacing in wavelengths, not the output
power of the signal generator.  An 80m yagi would have elements spaced 30' or more apart,
and they couple sufficiently to make a beam.  One experiment to see if the other antenna
is coupled to your loop would be to short the feedline to the other antenna and see if
that changes the SWR at the unexpected dip.


But remember that any length of feedline, if it isn't perfectly matched to the antenna,
will cause a variation in the impedance seen at the rig end.  It isn't uncommon for
an antenna + feedline to show a resonance at some point where the antenna itself
is not resonant.  Take, for example, the G5RV:  it has resonances on 80m and 40m
at the end of the twinlead section even though the basic 102' radiator itself is not
resonant on those frequencies.
Logged
AK4SK
Member

Posts: 150




Ignore
« Reply #2 on: December 21, 2012, 09:25:58 AM »

How are you feeding the loop?  I've run into similar quirks with a dip meter that were due to
the feedline rather than the antenna connected to it.

How close does the loop come to the other antenna?  Or to other wires or metal bits that
might be resonant?  The coupling is a function of the spacing in wavelengths, not the output
power of the signal generator.  An 80m yagi would have elements spaced 30' or more apart,
and they couple sufficiently to make a beam.  One experiment to see if the other antenna
is coupled to your loop would be to short the feedline to the other antenna and see if
that changes the SWR at the unexpected dip.


But remember that any length of feedline, if it isn't perfectly matched to the antenna,
will cause a variation in the impedance seen at the rig end.  It isn't uncommon for
an antenna + feedline to show a resonance at some point where the antenna itself
is not resonant.  Take, for example, the G5RV:  it has resonances on 80m and 40m
at the end of the twinlead section even though the basic 102' radiator itself is not
resonant on those frequencies.

Good question, I forgot to mention how it is being fed. I am feeding it with 50 ohm coax with a 4:1 current balun at the feed point.

The feed points of the two antennas are right on top of each other on the same mast. The loop at the feed point is about 1 foot away from the lower wire of the T2FD. As the antennas run away from the mast the separation grows to 3 or 4 feet or so. The T2FD extends about 33 feet on both sides of the mast where the loop extends 40 feet before changing directions (one end goes down a fence line, the other end jumps a short distance to the roof). The mast itself is fiberglass. There is a length of solid copper that runs up the mast parallel to the feed lines for both antennas (perpendicular to the antennas).There is a metal pulley, an eye bolt, shackle, and a hose clamp (to hold the solid copper wire to the top of the mast) all at the top of the mast about 2 feet away from feed point of the loop. There are eye bolts on both ends of the 40 foot runs before the loop changes directions and the loop comes within 1 foot or so of these eye bolts. The part of the loop that runs along the roof is right above aluminum gutters and near who knows what kind of metal (nails, house wiring, etc.) inside the roof.

I guess things get complicated, or at least non-ideal, when it comes to the feed line not being matched to the antenna. Does the twin lead radiate on a G5RV or does it just help the antenna become resonant at certain frequencies?

Thanks for the reply.
73,
Chris
Logged
WB6BYU
Member

Posts: 13005




Ignore
« Reply #3 on: December 21, 2012, 06:38:45 PM »

Quote from: AK4SK
I put up a full wave loop last Saturday and while it seems to work well the SWR curve is peculiar in one place. The fundamental frequency seems to be about 3.240 MHz...

Was it designed as a full wave loop for 3.240?  Or as a full wave loop on 80m?

Are the harmonic resonances of 3.24 MHz or of 80m?

How long is the coax?

If the loop wire runs very close to the T2FD then likely there is a lot of coupling.
That might result in some of the loop power being dissipated in the T2FD termination
resistance.  If you see the SWR changing on the loop on 80m when you apply an
alternating open and short circuit across the end of the T2FD feedline, then you
know you have close coupling between the antennas.

If the harmonic resonances are of the 3.24 MHz fundamental, then the loop is
too long:  shortening the wire should move the resonant frequency up into the
80m band.  (A good way to set it is to get the 20m resonance in the band where
it belongs, and that should put the rest about as close as you can reasonably
get them.)

What is the SWR without a tuner at the minimum point on 80m?  I would expect
about 2 : 1, maybe a bit higher if you are using a 4 : 1 balun.  If the other
resonances are in the 40m and 80m bands, then you may have it adjusted
properly and the 3.24 MHz resonance would be due to the combination of
loop and feedline length.  If you get the resonant frequency pretty close
to 20m then the coax losses should be low on most bands when you use
a tuner in the shack.
Logged
WB6BYU
Member

Posts: 13005




Ignore
« Reply #4 on: December 21, 2012, 07:10:29 PM »

I modeled a loop for 3.24 MHz with an 80m dipole nearby.  I can see a distinct flattened
area in the SWR curve with 20' of separation between the dipole and the loop.

It sounds like your antenna is much closer than that.
Logged
AK4SK
Member

Posts: 150




Ignore
« Reply #5 on: December 24, 2012, 11:06:50 AM »

Quote from: WB6BYU
Was it designed as a full wave loop for 3.240?  Or as a full wave loop on 80m?

Thanks for the reply. My apologies for a being a few days late in replying myself. In one respect it was not designed for any particular frequency. I knew from rough measurements that I would be putting up enough wire for it to be in the 80 meter band or a bit below it. So I just put it up and 3.240 is what I got. That number actually varies a bit from day to day, sometimes as much as a few hundred kilohertz, usually on the lower side of 3.240 but I've seen it as high as 3.300. A good length of it is in the wind so it flexes and move around a bit. I'm actually using it for MARS so it works out that it is resonant a bit lower than the 80 meter band. It also tunes well anywhere in the 80 meter band and on the other frequencies I need it for. For everything higher than 80 meters I use the T2FD.

Quote from: WB6BYU
Are the harmonic resonances of 3.24 MHz or of 80m?

They are of 3.24 MHz.

Quote from: WB6BYU
How long is the coax?

The total run to the shack is somewhere between 40-50'.

Quote from: WB6BYU
If the harmonic resonances are of the 3.24 MHz fundamental, then the loop is
too long:  shortening the wire should move the resonant frequency up into the
80m band. 

Right now I'm happy with the way it performs so I'll probably leave the length alone. I would like it a little higher but because of the way I have it routed around my yard and house it would be difficult for me to shorten it easily. Anyhow, I'm ok with the harmonics not falling in on the higher bands because I have the other antenna. Also, I put it up to replace a 80 meter linear loaded dipole that was electrically lengthened. I don't have the space for a full length dipole so I tried the linear loaded dipole first. It didn't seem to work very well on the lower end of 80 meters or below the band so I decided that I'd try a loop knowing that I'd at least have a full length antenna.

Quote from: WB6BYU
What is the SWR without a tuner at the minimum point on 80m?  I would expect
about 2 : 1, maybe a bit higher if you are using a 4 : 1 balun.

It is actually very close to 1:1, 1.1:1 or 1.2:1 usually.

Quote from: WB6BYU
If the loop wire runs very close to the T2FD then likely there is a lot of coupling.
That might result in some of the loop power being dissipated in the T2FD termination
resistance.  If you see the SWR changing on the loop on 80m when you apply an
alternating open and short circuit across the end of the T2FD feedline, then you
know you have close coupling between the antennas.

I haven't tried this yet but I will when I get a chance just to see. Because of the way things are I can't do much about it so I'll just have to deal with it. I'm mainly just trying to understand what is going on so I can learn, not necessarily fix a problem. Also, there is a good chance that the linear loaded dipole was coupled with the T2FD because they were arranged in the same fashion except the T2FD was the lower antenna. My goal with the loop was to put up an antenna for 80 meters that had a broader bandwidth and higher efficiency than the linear loaded dipole. At this point I don't know about the efficiency part but it has better SWR response on the frequencies I use.

Quote from: WB6BYU
I modeled a loop for 3.24 MHz with an 80m dipole nearby.  I can see a distinct flattened
area in the SWR curve with 20' of separation between the dipole and the loop.

It sounds like your antenna is much closer than that.

Thanks for looking into that, it is neat that the model shows similar behavior. I have modeled both the loop by itself and the loop with the T2FD. I've seen flattened areas in both models and I have seen normal behavior in both models, it just depended on how I ran the calculations. I will say though that I have very little experience with modeling so there is a good chance that I don't know what I'm doing with the programs. I used both EZNEC Demo and 4nec2. I think that the physical representation of the models are good, concerning only the wire heights and lengths that is, but I don't know about the rest (segments, ground, etc.).

Thanks again for all of the help.

Merry Christmas and 73,
Chris AK4SK
Logged
KB4QAA
Member

Posts: 2259




Ignore
« Reply #6 on: December 24, 2012, 01:39:13 PM »

Quote
One thing to note is that the loop itself is irregular. I didn't have the free space to construct a square loops cut for 75 meters so I just followed my property line and house. About half of it, the part in my back yard, is roughly half a rectangle with the feed point higher than the rest of the antenna. The other half follows the roof line on the front of my house at 8 feet above the ground. The front of the house is a series of right angle turns or steps as seen from above. The antenna basically encompasses my back yard and entire house. Also, directly above the feed point and running parallel above part of the loop is a 66 foot terminated folded dipole.
Hold it. Stop. Do not pass GO, do not collect $200.

-Non-standard shape, non-ideal height, non-ideal earth, interfering objects.

-No need for further discussion.  You won't get text book curves, with non-text book installations.  Close the thread please!   Wink
Logged
WB6BYU
Member

Posts: 13005




Ignore
« Reply #7 on: December 24, 2012, 03:02:47 PM »

I've used similar loops, string however possible around the house and yard.  They work fine
as long as you don't fold them back on themselves too much.

Your efficiency should be fine on 3.3 MHz.  With 60' of RG-213 the total loss will be close
to 2dB at the top end of 80m due to the high SWR on the coax (over 10 : 1).  But that
should still be usable.

In fact, I'd suggest that you compare the loop with the T2FD on each band and see - you
may find that the loop does better in some cases, depending on the distances involved.
Logged
AK4SK
Member

Posts: 150




Ignore
« Reply #8 on: December 26, 2012, 11:05:14 AM »

Thanks again for the info. I'm using LMR400 so that should help with the loss. I'll give it a try on the other bands.
Logged
N8CMQ
Member

Posts: 353




Ignore
« Reply #9 on: December 26, 2012, 05:24:19 PM »

Sounds like interaction between your loop and
external structures and other antennas.
I have seen such interaction with many antennas,
but it isn't all bad, it just shows how nearby
structures can cause problems.
Now, I wonder what the pattern of the loop looks like.
Logged
AC2EU
Member

Posts: 328


WWW

Ignore
« Reply #10 on: December 26, 2012, 06:16:04 PM »

Using the LMR vf=.85 ,and a center frequency for the antenna at 3.5 mhz
the coax should be a multiple of 119 feet for minimum influence on the SWR.
So, the short run is probably part of the issue as well beyond geometry and nearby interactions.
The half wave multiple will give a closer representation of what is actually happening with the antenna itself.

There's a lot of moving parts to this one!
Logged

AK4SK
Member

Posts: 150




Ignore
« Reply #11 on: December 27, 2012, 06:55:36 AM »

Using the LMR vf=.85 ,and a center frequency for the antenna at 3.5 mhz
the coax should be a multiple of 119 feet for minimum influence on the SWR.
So, the short run is probably part of the issue as well beyond geometry and nearby interactions.
The half wave multiple will give a closer representation of what is actually happening with the antenna itself.

There's a lot of moving parts to this one!

Thanks. I though that the balun took the coax out of the equation, is that not true? I have read about people using 1/4 wave sections of 75 ohm coax in place of a balun. Are you saying that what I'm seeing on the analyzer is not representative of what is really going on (which wouldn't be a surprise as all these things have limitations) and that the SWR curve at the 1/2 wave point (about 7 MHz) is probably what the "true" SWR curve looks like at 3.5 MHz?
73,
Chris
Logged
AC2EU
Member

Posts: 328


WWW

Ignore
« Reply #12 on: December 27, 2012, 07:31:15 AM »

Using the LMR vf=.85 ,and a center frequency for the antenna at 3.5 mhz
the coax should be a multiple of 119 feet for minimum influence on the SWR.
So, the short run is probably part of the issue as well beyond geometry and nearby interactions.
The half wave multiple will give a closer representation of what is actually happening with the antenna itself.

There's a lot of moving parts to this one!

Thanks. I though that the balun took the coax out of the equation, is that not true? I have read about people using 1/4 wave sections of 75 ohm coax in place of a balun. Are you saying that what I'm seeing on the analyzer is not representative of what is really going on (which wouldn't be a surprise as all these things have limitations) and that the SWR curve at the 1/2 wave point (about 7 MHz) is probably what the "true" SWR curve looks like at 3.5 MHz?
73,
Chris

 It's more about the how the coax works. At the half wave multiple the coax the input impedance is equal to the load impedance. It's a special case where the coax impedance does not matter! Of coarse that is only true exactly at the half wave point, but very close on either side of resonance as well. It will be a whole lot closer to the actual impedance presented at the balun.
The 4:1 baluns are for impedance transformation of say, for an antenna that has a known impedance of around 200 ohms to match the coax .
If you can remove the balun without a lot of trouble for testing, I would suggest sweeping the antenna with a half wave multiple coax to determine what the impedance actually is.  If you get readings around 200 ohms, the balun makes sense, otherwise you may need to rethink the match at the antenna.

If you have a specific frequency of interest, it might be a good idea to make the half wave length at that that frequency for more accurate results. I don't know what the MARS frequencies are, so i centered on 3.5, thinking that you may want to get into the low end of 80 as well.

Logged

WB6BYU
Member

Posts: 13005




Ignore
« Reply #13 on: December 27, 2012, 08:45:18 AM »

Quote from: AC2EU
Using the LMR vf=.85 ,and a center frequency for the antenna at 3.5 mhz
the coax should be a multiple of 119 feet for minimum influence on the SWR.


No, this isn't true.  You're confusing the impedance transformation property
along a feedline with SWR.

The SWR will be the same regardless of transmission line length, except
for some reduction due to line losses.  If the load impedance is 100 ohms, for
example, the SWR along a lossless 50 ohm feedline will be 2 : 1 regardless
of length
.  The only thing that using a half wavelength does for you is that
the impedance at the rig end will look like 100 ohms rather than some other
value that also gives a 2 : 1 SWR.  For example, if you used a quarter wave
of feedline, the rig would see 25 ohms, or for 1/8 wavelength it is about
40-j30 ohms, but both will still give an SWR of 2 : 1.

So if you are using an SWR analyzer to measure the antenna feedpoint
impedance, a half wavelength of feedline comes in handy (as long as it is
cut correctly for the frequency at which you are taking the measurement.)
But it does not change the SWR as seen by the transmitter.
Logged
AC2EU
Member

Posts: 328


WWW

Ignore
« Reply #14 on: December 27, 2012, 10:24:16 AM »

Quote from: AC2EU
Using the LMR vf=.85 ,and a center frequency for the antenna at 3.5 mhz
the coax should be a multiple of 119 feet for minimum influence on the SWR.


No, this isn't true.  You're confusing the impedance transformation property
along a feedline with SWR.

The SWR will be the same regardless of transmission line length, except
for some reduction due to line losses.  If the load impedance is 100 ohms, for
example, the SWR along a lossless 50 ohm feedline will be 2 : 1 regardless
of length
.  The only thing that using a half wavelength does for you is that
the impedance at the rig end will look like 100 ohms rather than some other
value that also gives a 2 : 1 SWR.  For example, if you used a quarter wave
of feedline, the rig would see 25 ohms, or for 1/8 wavelength it is about
40-j30 ohms, but both will still give an SWR of 2 : 1.

So if you are using an SWR analyzer to measure the antenna feedpoint
impedance, a half wavelength of feedline comes in handy (as long as it is
cut correctly for the frequency at which you are taking the measurement.)
But it does not change the SWR as seen by the transmitter.

Agreed. However I was thinking in terms of impedance and impedance analysis so that he may get raw data from the antenna with an antenna analyzer. I should have made that clear.
Also, it's important to know that coax does skew the impedance at the shack end according to length is the match is not 1:1.
After a base line is established, he can proceed from there. As you pointed out,there just isn't enough information in SWR alone.
Logged

Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!