Call Search

New to Ham Radio?
My Profile

Friends Remembered
Survey Question

DX Cluster Spots

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

donate to eham
   Home   Help Search  
Pages: [1] 2 Next   Go Down
Author Topic: Feeding a horizontal loop  (Read 5281 times)

Posts: 2

« on: January 19, 2007, 09:21:26 AM »

I'm planning on putting up a horizontal loop. I plan on feeding the loop with coax and tuning it with the internal tuner on my 756Pro2. I'm hoping to get somewhere in the neighborhood of 274' feet of wire up. I have read many conflicting opinions about whether I need a balun or not and if I use a balun, whether to use a 4:1 or a 1:1 balun. Get I get some opinions please?

Posts: 21757

« Reply #1 on: January 19, 2007, 09:36:54 AM »

Impedance will vary with frequency and height above earth; but nominally your loop will likely resonate about 3.67 MHz on 80m and then also on about 7.34 MHz (above 40m); 11 MHz (well above 30m); 14.68 MHz (above 20m); 18.35 (above 17m); 22.02 (well above 15m), etc.

Before suggesting anything else, I'd have to ask what frequency or frequencies do you intend to use?  If you're mostly into 80m CW, the 274' length is pretty close; but if you're more into phone operations, and especially if you intend to use the antenna on other bands above 75m, you're making this loop too long.

I'd make it shorter and adjust final length with an open wire tuning stub if I wanted to use coax feed; or, I'd use open wire or ladder line from the loop to the shack and deal with it there.

Tricky to make a loop work multiple bands if you're using coax to feed it.  


Posts: 2

« Reply #2 on: January 19, 2007, 11:38:03 AM »

I was hoping to use it for multiple bands. I also wanted to use coax because I don't know how to terminate ladder line into my 756Pro2. Thanks.

Posts: 21757

« Reply #3 on: January 19, 2007, 11:57:31 AM »

I'd recommend you go here and read this:

Problem with using a loop as you describe on multiple bands is its Z will be in the 50-70 Ohm range on 80m but much higher on all the other bands.  So, while you might want a "1:1" balun at the feedpoint for 80m work, you'd want at *least* a 4:1 balun for the higher bands, and actually about 9:1 would probably be better.

If your coax feedline is short, even with a substantial (8 or 9:1) mismatch its loss might not be too high; but if it's long (hundreds of feet), it can get quite high and make ladder line start looking better all the time.

Various ways to tackle this problem, but one of the easiest is to use an external antenna tuner set up for balanced line, and then just use ladder line all the way from the loop to the tuner, with a coax jumper from the tuner to the rig.


Posts: 351


« Reply #4 on: January 19, 2007, 01:12:30 PM »

Good questions, and since I've just gone through much of the research on horizontal loops, maybe I can help.  But I warn you, I get downright evangelical about the loop now that I have one!

The previous link to W4RNL's article is a good one.  So are:

There's a lot of data there and it may be baffling if you try to absorb it all, but you should be able to glean more from it each time you refer to it.  The horizontal loop makes a good multi-band antenna for the band for which you make it at least one wavelength long and for all those that are higher in frequency.  But to make it work best, you simply have to feed it with ladder line.  The previous comments were correct.  The impedance will vary all over the place from frequency to frequency so it is not as simple as putting a 4:1 or 6:1 balun in a coaxial line and getting a beautiful transfer of power.  Either would work, but only at specific frequencies that may or may not even be in the ham bands.  I wish it was that easy!  Coax is much easier to work with.

But I've learned not to fear open wire feedline.  And the stuff has some wonderful advantages, not the least of which is that is very low loss and, in effect, makes the Great Satan "SWR" almost a non-factor.  Ladder line, with a good tuner (balanced or with a hefty enough balanced input to do the job), is a good way to go, despite the need to keep it away from metal, trees, and the ground.  Putting a remote auto-tuner at the antenna feedpoint is a good way to go as well, but you have the problem of supporting it, getting voltage to it, etc..

If you only intend to use the antenna on one band and bands that are resonant harmonics, cut it for one wavelength for the frequency closest to where you intend to operate and feed with coax.  But you are only using a fraction of the capability of that nice, long piece of wire if you do that.

There are several methods of getting ladder line/window line into the shack.  I run the two sides of window line to the center conductors of two pieces of RG-8 that I tape together.  I tie the shields of the coax together at the station end and solder to the station ground.  The two center conductors go right into the balanced output of the tuner.  I tried to keep the coax as short as possible, but that enables me to get the run through a hole in the brick wall, past a bunch of other coax, alongside some air conditoner ducts and telephone lines, and past a lot of other metallic stuff that open-wire feedline doesn't need to be near.

There are plenty of sites where you can learn more about the care and feeding of open wire feedline.  Just do a Google search for "ladder line" or similar.

Good luck, and don't be afraid to ask questions.  None of us are born knowing the relative merits of coax vs. ladder line or how to feed a horizontal loop!  Well, maybe W4RNL was, but not most of us!


Don N4KC


Posts: 17352

« Reply #5 on: January 19, 2007, 06:29:13 PM »

While I also recommend ladder line or twinlead for feeding, I've had
good results matching a horizontal loop using the autotuner in my
TS-450.  I did this by designing the loop near bottom of the 80m band
(so LONGER, not shorter) which moved the resonance down closer to
the bands on 40/20/15/10m.  Since I like to work CW mostly this
wasn't a problem, though I could still use it across the band with a higher
SWR.  Something about 70 feet on a side (total length 280 feet) looks
like a good choice.  (This resonates below 3.6 MHz.)

Then I used a 4 : 1 balun at the feedpoint.  This gives a higher SWR on
80m where the impedance is around 90 ohms at resonance (depending
on height above ground) but the SWR is better than 1.5 : 1 on parts of
40 and 20m (and at 10.8 MHz) and under 2 : 1 across all of 15m.
I figure the cable losses are higher on the higher bands, so I want to
minimize the SWR there.  If your tuner won't match the loop on the
high end of 75m then I'd try the approach in one of the recent issues
of QST where the author put a T connector in the feedline in the shack
and plugged in a coax stub to shift the impedance to give a better
match on the upper part of the band.  (Or you could use a tuner with
a "bypass" position and leave it pre-tuned for the upper end of the
band - switch it in when you want to operate there and the internal
tuner should be able to find a match.)  The SWR on 80m is under 5 : 1
up to 3.85 MHz, so it should be only the top of the band (where the
SWR is over 8 : 1) where you have any problems).

Posts: 220

« Reply #6 on: January 19, 2007, 08:18:23 PM »

Here's a good article

I feed mine with RG-59 and NO balun like the article says. Works excellant on 10-80m with low swr. If your fussy about keeping swr to a minimum then use an outboard tuner, although my IC-736 tunes it up fine. It sees more service than the Mosley Tribander on my tower. Try it, you'll like it.

73, Doug

Posts: 2429

« Reply #7 on: January 20, 2007, 10:10:16 AM »

Where do these guys get this stuff?  The author of that document makes this statement:

"The SWR of either of these loops with approximately 100 ft feedline is rarely over 3:1."  

He uses coax to directly feed a horizontal loop and even says you can use zip cord for a feedline.  Taking his "star performer" loop, 68 feet on a side, at 40 ft height, fed in the corner, with 100 feet of RG8X, here is what EZNEC and TLW say about it.

The numbers are frequency in MHz, SWR at the antenna, SWR at the station end of the line, and feedline loss in dB.

Freq _ SWR Ant _ SWR Stn _ Loss dB
3.5 ____ 9.9 _____ 7 _______1.7
4.0 ____ 13.5 ____ 8.4 _____ 2.1
7 ______32.5 ___ 11.2 _____ 4.8
10.1 ___51.7 ____ 11.1 ____ 6.7
14 _____25 _____ 8.2 _____ 4.8
18.1 ____8 ______4.6______2.5
21 ____ 25.9 ____ 7.5 _____ 5.9
24.9 ___17.2 ____ 6.1 _____ 4.5
28 ____ 37.9 ____ 7.2 _____ 7.3
29 ____ 5.8 _____ 3.4 _____ 2.3

If you used zip cord for a feedline, you might get lower SWR because the feedline loss will be huge.  

Also if you do not use a choke on the feedline, the coax shield will have high common mode currents.  If you attach the other end to a lossy ground with the right length of feedline, you may be able to lower the SWR by dissipating more power into the ground.

He also says:
" ... open wire (feedline) does not appear to make the  loop perform any better or matching it any easier."

Since feedline loss with open wire feedline is negligible, all you have to do is look at the feedline losses in the table above to see what the improvement will be.

He uses this antenna on 160 as a vertical without a ground radial system.  EZNEC says the gain should be somewhere around -4.3 dBi.  If you put in a good radial system it should work OK.

Jerry, K4SAV

Posts: 351


« Reply #8 on: January 20, 2007, 05:32:45 PM »

Oh, Doug/KE4VBG, you must be living a charmed life!  If you are using a loop on multiple bands and feeding it with RG-59, and not using any sort of tuning device other than what's inside your radio, and you are seeing low SWR, then I'd like to invite you to accompany me on my next visit to the casino.  Hopefully some of that magic will rub off!

As K4SAV points out in his EZNEC modeling, the impedances and resulting SWR are excessive at best and completely prohibitive in most cases without some serious matching going on.  I think any number of us loopy hams can verify his results from practical experience.

Yes, your loop will present a 75-ohm load at certain frequencies, and the thing might even work as you describe at that frequency and harmonic multiples (one of the advantages of a loop as opposed to a dipole that is only resonant at odd multiples).  But those multiples will mostly fall on frequencies where you probably don't want to have a signal.  Say you cut the antenna for 3.75, the middle of the 80/75 meter band.  The loop would also be resonant and present lowest SWR at 7.5 mhz, 11.25, 13.8, 18.75, 22.5...well, you get the picture.  But with a tuner to get a good match and low-loss feedline (such as ladder line) to feed it with so SWR is really no factor, you would still get excellent results.  Otherwise, sorry.  I'm a skeptic.

Note that there was a followup QST article to the one you linked to.  It was in the April 2002 issue, written by NT0Z.  He takes you through the many problems he had with RFI on his loop when he tried to feed it with coax (he had to use an outboard tuner to get a match in the first place when he tried to use the antenna on multiple bands, by the way).  He finally went to ladder line and a balanced tuner in order to make the antenna work.

If you can tell me how you do it with RG-59 and no tuner, I'm all ears!


Don N4KC

Posts: 220

« Reply #9 on: January 20, 2007, 07:38:35 PM »

Well I don't know much about Fancy antenna modeling software, But, I've had this loop up for about 8 years. It's 272ft. long(a square shape) and is about 40ft up supported and fed at one corner at my tower. About 75ft. of RG-59 down the tower to the shack. It is directly over my pool, maybe this helps? Have worked stations on 10,15,18,20,40,and 80m with great success. Most of the time it tunes up fine with just the radio's AT. I do use the Vectronics tuner some on 75 and 80m. I have worked the world with it with 559-599 signal reports. I worked 3Y0X on Peter Island on 20m SSB (look it up)as well as 4Z4UN,RA6AFB,HL5BLF and others. I made a lot of these contacts on 18 and 10m when it was open a few years back. Plus a lot of 40m CW. I have a wall full of cards from everywhere around the world. So, maybe the antenna don't Know that it's not supposed to work. Hi Hi

All I know is that it works, and I can't wait for 2010 for the bands to open back up again.

Evidently I'm not too Lucky cause I haven't hit the Lotto yet. Hi Hi

73, Doug


Posts: 220

« Reply #10 on: January 20, 2007, 07:43:55 PM »

Oh yea, this antenna is featured in the 1994 ARRL Antenna Handbook 17th Edition. Thats where I found it.

Posts: 18

« Reply #11 on: January 21, 2007, 05:37:47 AM »

If you want to make this an all-bander go with balanced line. Use real balanced line, not the  300 , or 450 ohm line.  Check out the reviews on W7FG balanced line on eham , it would do the job. You can also home-brew your own line. I made mine using thin wall pvc conduit cut 7 inches long , drilled i inch from the end for 6 inch spacing, using 14awg stranded copper. The impediance is in the range of 625 ohms. I use a Palstar AT1KM Tuner .My dipole is 176 ft long and tunes flat on all of the bands I work including 160. I would have made it longer, but dont have the space. I run only balanced line, and use coax only for jumpers between the rigs. As far as the tuner built into the rig, it probally wont do the job. The tuner in my K-wood TS570 is useless, a big waste. Balanced line is far superior to coax, hands down.

Posts: 322


« Reply #12 on: January 21, 2007, 08:22:07 AM »

Plain and Simple, for many years I have used the 80 Meter Horz loop, and used a 4:1 High Current Balun at a corner Fed with RG8 Coax with absolutely no problems, of course I do use a Nye Viking Tuner with it to work other Bands but again...always works GREAT.

Posts: 17352

« Reply #13 on: January 21, 2007, 08:59:21 AM »

We have here a classic case of the difference between "what is optimum"
and "what works".  (And I've certainly been accused of muddying that
distinction in the past!)

Feeding a loop directly with coax cable certainly will work, in that some
power will be radiated.  On the lowest band the loop impedance is
close to 90 ohms at typical heights, so a quarter wavelength of 75 ohm
coax can give a very good match, and any length will give an SWR of
better than 2 : 1 at resonance.

The more wavelengths of wire in a loop, the higher the feedpoint
impedance.  On 40m it is around 250 ohms at resonance; on 15m it
is around 400 ohms.  (These presume that the loop length is set so
that the resonances fall on these bands, and also ignore any possible
radiation from the feedline if you aren't using a balun.)  This is why
I chose a 4 : 1 balun because on most bands it gives an SWR of 2 : 1
or less.

But if you feed it directly with 75 ohm coax you'll still get power to the
antenna - just perhaps not as much due to the line losses.  As long as
your tuner will match the impedance at the rig end of the cable
efficiently, the system loss shouldn't exceed 10dB, and probably will
be less on many bands.  You'll still radiate more power than I do with
my QRP rig to a more efficient antenna, and I've worked DX with that
as well.  You'll still make lots of contacts and have fun.

So how much difference does it make having, say, 6dB of loss in your
antenna?  Or running 25 watts instead of 100 watts - the difference
is the same at the other end.  When conditions are good and signals
are strong, the answer is "not much".  Think about the signals you
typically listen to - if they are 2 or 3 S-units above the noise level and
they drop by 1 S-unit, you can still copy them - but perhaps not quite
as well.  If you are listening for very faint DX signals down in the noise
6dB can make a big difference.  So, having worked QRP for many years,
my experience is that the day-to-day (or minute-by-minute) variations
in propagation are far more than the 6dB difference in radiated power.
That drop in power simply reduces the period of time over which
signals are strong enough for the other station to hear you.  Perhaps
an opening will last 3 hours instead of 4 because conditions are just
too marginal at the beginning and the end.  With good operating style
you can still work most of the same stations, you may just have to work
harder at it, or listen to the station working others for an hour before
propagation picks up enough for your signal to get through.

So it depends a lot on your personal operating preferences.  Often it
is better to put up a less efficient antenna and get on the air rather than
waiting for the parts to make it "perfect".  I've certainly been known to
feed loops directly with coax, or with zip cord or twisted pair when
that was what I had handy.  Yes, it works and I made contacts.  No, it
may not be the most efficient option in many cases - but sometimes
all I had was a mess of wire salvaged from blasting a rock pit, and that
got me on the air.

If you can take the time to prune the loop length then coax feed can
work with a reasonably low SWR.  (This also depends on the matching
range of your tuner.)  If you only have one chance to put it up, then
something like twinlead or balanced line may be a better choice because
the SWR may be very high on the frequencies you want to work, and
that will keep the losses low.  (Again, presuming you have a tune that
will match it.)  This is often the case with discrete antennas, where
repeated trips up and down to trim the length will be more noticable
than the wire itself.

If you want the lowest feedline losses while still enjoying the flexibility
of coax feed, use two pieces of 75 ohm coax in parallel to make a
150 ohm balanced line.  Cut two pieces the same length, preferably
an electrical quarter wavelength on 80m.  (There are some double coax
cables available that make this particularly convenient.)  Tie the shields
together at each end and use the two center conductors as the balanced
feeders.  I like to ground the shields to the tuner case, but that isn't
entirely necessary.  In the shack connect this balanced line to the tuner
(or the rig) with a 4 : 1 balun.

Why does this work?  On 80m the impedance is around 90 ohms, and
a quarter wavelength of 150 ohm feedline will transform this to 250
ohms, for a 50 ohm SWR of 1.25 : 1 at the output of the 4 : 1 balun.
On 40m and up the impedance will be in the range of 200 to 500 ohms,
and (at least on the even harmonic bands) the feedline will be close to
a multiple of half a wavelength so the impedance won't be transformed
much.  This gives an output from the balun of 50 to 125 ohms, which
is a worst case SWR of 2.5 : 1.  Of course the SWR will be higher on the
other bands, and in other parts of some of the wider bands. You still
have the same inherent losses of the 75 ohm coax, but the additional
loss due to SWR is less.  So while the losses are still higher than they
might be using ladder line, they shouldn't be excessive, and the
impedances should be within the range of the internal auto-tuner on
most bands.

Posts: 2429

« Reply #14 on: January 21, 2007, 10:49:53 AM »

Doug, no one said you can't make contacts on an antenna system with high SWR, or on one that has high losses.  The point is that the antenna could be improved without a lot of effort.

Antennas become complex systems when there are multiple structures involved and an antenna is set up in a way to allow coupling to the other structures.  Because of no balun used, and I'm guessing that the feedline runs down the tower, and the feedline will have current on it, so those currents will be coupled into the tower.  The tower will become a radiator, as well as a loss contributor because of being coupled into lossy ground with no radial system.  The feedline at the shack is eventually coupled into lossy ground also.  All of this becomes part of the antenna system.

Not knowing what your tower looks like, what the tower ground system looks like, and where the coax lines are tied I can't do an accurate estimate of your particular installation, or even a rough estimate, however, just doing some guesswork as to what these thing might be, here are some numbers for a similar system.  You system may be considerably different because of tower resonance and coax tie points.  

Previous SWR numbers were relative to 50 ohms because of the RG8X used.  These SWR numbers are relative to 75 ohms since the coax is RG59.  Since this data is only a possible estimate, I didn't include a complete set.  

The numbers are frequency in MHz, SWR at the antenna,  SWR at the station end of the line, coax loss in dB, and near field ground loss in dB.  The near field ground loss is due to the connection with lossy earth.

Freq _ SWR Ant _ SWR Stn _ Loss dB_ Gnd dB
3.5 ____ 7.4 _____ 4.5______2.5 ____ 0.1
4.0 ____ 11.5 ____ 5.4 _____ 3.2 ____0.2
7 ______14.5 ____ 5.1 _____ 4.7 ____ 0.8
10.1 ___17.1 ____ 4.7 _____ 5.8 ____ 1.8
14 _____10.6____ 3.7 _____ 4.8 ____ 1.6

You can see that things can change a lot with seemingly minor system changes.  This system has more loss that the previous one which partially accounts for the lower SWR numbers.  The numbers also appear lower because the reference used was 75 instead of 50.  Note the two loss numbers.  These represents the possible improvements that can be made with the system just by using low loss balanced line.  Those numbers don't mean that the signal in the direction you want will improve by that amount.  It means that the total radiated power will improve by that amount.  Remember your power doubles every 3 dB .  Again, these numbers do not accurately represent your system, just a similar system.

Jerry, K4SAV
Pages: [1] 2 Next   Go Up
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!