By Kirk A.

Kleinschmidt, NTØZ

A Balanced, Everyday
Approach to All-Band Bliss
Feed lines, antenna tuners, baluns, RFI, computer noise and all-band
antennas—now there’s a snake pit of potential conflict. After years of
experimentation, the author has found the path to multiband nirvana.

F
or most of us, the Holy Grail of tal loops until I read Dave Fischer, W7FB’s and available skyhooks could contain a
ham radio is antenna performance. life-changing article, “The Loop Skywire,” larger antenna (which should work better,
It’s often the key element in deter- in the November 1985 issue of QST (back especially on the low bands). It wasn’t the
mining ham radio success and operating when Dave signed WØMHS). Shown in 272 feet required for an 80-meter loop, but
enjoyment. You can get by with a second- Figure 1, the loop is simply a full-wave- something in between. That’s how I dis-
rate transceiver, a deep gravelly voice and length of wire cut for the lowest band of covered another rule of thumb for build-
even a severe lack of good looks and interest. Feed it with coax or ladder line ing modern-day horizontal loops: Put up
charm, but if you have an underperforming (feed lines are discussed in detail later). the largest possible horizontal loop your
antenna, ham radio isn’t nearly the fun it I had previously used quad loops and situation allows and forget about pruning
could be. dipoles (G5RV-type and conventional). it to resonance. The antenna tuner, which
You’ve undoubtedly looked with long- Sure, they worked—and they still do— you’ll need anyway for multiband opera-
ing, as I have, at magazine pictures that but I had always thought something was tion, will take care of things.
show fabulous contest “superstation” an- lacking, and I had a secret hunch about Strung about 45 feet above the ground,
tenna farms, the few fortunate hams who the Loop Skywire. I knew from past experience that my loop
have 350-foot towers festooned with big Now, Fischer’s First Rule of horizon- would be an excellent performer. Fed
Yagis (or a smattering of 200-footers), the tal loop construction is to enclose as much with 50 feet of RG-8 and my trusty old
Northern California ham who has a dozen area as possible within the confines of the antenna tuner, the loop worked very well
full-size rhombics strung between the loop. A circular loop is ideal, but a square on 40 meters and up. As before, however,
tops of giant Sequoias, and so on. loop is much more practical and doesn’t 80 and 160 were adequate, but just barely.
After imagining an antenna system that suffer performance-wise. My lot could RFI was a big headache on the higher
qualifies as a navigation hazard, most of us only accommodate a triangular “loop,” bands. In fact, 15 meters was almost im-
will scale things down to the real matter at which is pretty much the geometric limit possible. Even at 5 W power levels, ev-
hand—how to put up an affordable, easy- of what you can get away with. If you ery key-down zapped the TV screen and
to-build multiband antenna that works great make the loop any more elongated or con- all of the stereo and computer audio cir-
and fits on an average-size lot. We’re back stricted, it loses its “loop-like” qualities. cuits in my office/shack. I had converted
to the search for the Holy Grail. Although my most recent loop could the garage to an office, and my shack is
have been cut for 40 meters, my property on the west wall. Just outside the wall is
Horizontal Loop Antennas the base of my tower, which supports the
What, you were expecting dipoles? feed point of the horizontal loop. Al-
Although dipole antennas in all of
Electronic Article
though the antenna itself is at least 50 feet
their various shapes and configurations References from all of the PCs and consumer elec-
perform well, in my experience the best If you’re an ARRL Member, point tronics, the feed line was in close prox-
all-around multiband antenna is the hori- your Web browser to the links below imity. The crud generated by three or four
zontal loop. It’s efficient, omnidirectional to access Adobe PDF versions of the computers was also a headache and wiped
over most “real ground,” it’s quiet, it op- articles I’ve referenced in the test. out large chunks of the bands.
erates well on all HF frequencies above • The Loop Skywire:
www.arrl.org/members-only/tis/ RFI
its design frequency (and even those be- info/pdf/8511020.pdf
low, as we’ll see later), it fits on most lots • A Balanced Balanced Antenna Despite the RFI and computer noise,
and, contrary to tradition and many offi- Tuner: www.arrl.org/members- I was pleased with the performance of the
cial-looking radiation pattern plots that only/tis/info/pdf/9002028.pdf big loop.
can be found in various antenna manu- • The Lure of the Ladder Line: I tried all of the standard RFI-fighting
als—can be an outstanding antenna for www.arrl.org/tis/info/pdf/ techniques. I made sure I had a good, short,
domestic and DX contacts alike. 9312070.pdf low-impedance connection to an earth
As a kid I made my share of vertical • Home-brew Your Own Induc- ground at my operating position. I installed
tors: www.arrl.org/members-only/
loops/quads from bamboo poles and cop- tis/info/pdf/0107066.pdf
ferrite cores and clamp-on chokes on
per wire, but I didn’t think about horizon- seemingly endless numbers of audio and
April 2002 47
video cables. I installed ac line filters and Lure of the Ladder Line,” from the De- lines isn’t always as easy as simply swap-
connected some sensitive gear to the ac cember 1993 issue of QST. It’s also avail- ping one cable type for another. Coaxial
mains via an uninterruptible power sup- able to ARRL members in PDF format cables are unbalanced (which can make
ply, complete with RFI filtering, surge sup- from ARRLWeb (www.arrl.org/). them vulnerable to common-mode RFI),
pression and line conditioning. Steve tells us how he dramatically and conventional antenna tuners are de-
I picked up a copy of The ARRL RFI improved the performance of his multi- signed to handle them, so when they’re
Book, which is handy to have even if band dipole, formerly fed with coax and matching coaxial cable loads, everything
you’re not besieged. I read up on com- a tuner, by replacing the feed line with works out fine as long as the impedance
mon-mode interference, front-end over- 450-Ω ladder line (see Figure 2). There at the tuner end of the feed line isn’t too
load, nonlinear rectification and even it was in black and white: A key piece of extreme.
interference that’s re-radiated by other antenna system wisdom that I’d been Ladder-line and open-wire line are
affected components. lacking for years. intended for balanced operation where
I disconnected dozens of speaker In that article I discovered that the loss equivalent currents flow through each of
wires, audio and video leads, and coaxial figures commonly attributed to coaxial the two wires that make up the feed line.
TV cables. Is the RFI getting in through cables of various sizes and compositions To accommodate balanced feed lines,
this wire or is it through that one? It was are only accurate under low-SWR condi- conventional tuners almost always use a
test, test, test—back and forth. tions. As shown in Table 1, when the balun transformer at the tuner output to
In the end, all of that poking and prod- SWR on the coaxial feed line between make the transition from balanced line to
ding made the RFI situation quite a bit your antenna tuner and your antenna’s an unbalanced tuner network.
better on all of the bands except 15 feed point is high—as it usually is when In a perfect world, this works pretty
meters, which was still unusable. feeding antennas on frequencies at which well and users can conveniently feed un-
Still, life was pretty good on 40, 30, they’re not resonant—the signal losses in balanced and balanced loads without need-
20 and 10 meters, and I was working lots the coax can be staggering. ing separate tuners. The problems arise
of juicy DX and enjoying solid stateside No wonder my antenna worked great when we consider how difficult it is to
ragchews. at the design frequency and on all higher make a single balun that works well over
HF bands, but fell off on 80 and 160 a wide range of frequencies and power lev-
Climbing the Ladder Line of meters. On those bands, even a 40-meter els. In short, it’s easy to build a tuner-out-
Success full-wave loop is physically small and the put balun that works well on a single band
At this point I was feeding my loop high SWR on the line between the tuner or on a few adjacent bands, but when it
through a standard antenna tuner and a and the antenna wasted most of the power.
50-foot length of coax. I knew about open- But look at the loss figures for the lad-
wire line, TV twinlead and ladder line, der line. They are a lot better, especially
but I’d never used the stuff to feed an an- on the lower bands.
tenna. Then I got wind of Steve Ford,
WB8IMY’s, excellent and eye-opening Baluns and Conventional Antenna
introduction to multiband antennas fed Tuners
with 450-Ω ladder line. The article is “The Unfortunately, using open-wire feed

Figure 2—A dipole formerly fed with coax

Table 1
Loss Comparisons for Belden 8214
Coaxial Cable and 450-Ohm Ladder
Line
Cable length: 50 feet.
Antenna: 66-foot dipole at 30 feet.
Calculated by Dean Straw, N6BV,
Senior Assistant Technical Editor.
Freq Loss (dB)
(MHz) 8214 Ladder Line
1.9 26.9 8.82
3.8 13.7 1.37
7.15 0.19 0.07
10.14 2.85 0.07
14.27 5.30 0.15
18.14 6.96 0.31
21.40 0.78 0.12
Figure 1—Put up the largest horizontal loop your lot can support while keeping the 24.90 3.94 0.13
“loop” as square as possible. Don’t worry about perfect symmetry. For size reference, 28.50 5.69 0.18
values are shown for 80 and 40 meters. See the text for additional information.

48 April 2002
comes to a single dc-to-daylight tuner-out-
put balun, things often don’t work so well.
Also, because of where the balun re-
sides in the tuner/antenna circuit, output-
style baluns are often subject to extreme
RF voltages, which can cause arcing,
sparking, burning and other undesirable
behavior. Output-style baluns also tend
to have a tough time staying electrically
balanced over a wide frequency range.
All of these factors combine to decrease
the tuner’s efficiency and increase power
losses in the tuner and balun that are in
addition to the losses of the feed line. Of
course, all antenna tuners have some loss, Figure 3—A simplified schematic of the balanced tuner as originally presented by Rich
and that loss tends to vary by frequency Measures, AG6K. The home-brew coaxial balun is on the input side of the network,
and load impedance. Matching extreme which consists of two roller inductors that are adjusted in sync and a single variable
impedances usually means more loss. capacitor. The capacitor, usually on the output side of the coils, can be moved to the
Steve experienced this when feeding input side to match some lower-impedance loads. Some builders use fixed, tapped
inductors to save money. See the text and the References sidebar for more information.
his 40-meter dipole with ladder line on
80 and 160 meters (extreme impedances
for a 40-meter dipole). At anything more in that fashion, the SG-231 matches the ing about how they’d built balanced an-
than just a few watts, his tuner would arc, antenna to the 50-Ω impedance of the tenna tuners to feed their big horizontal
snap, sizzle and pop. To operate on those coax that runs from your rig to the tuner/ loops with ladder line. I knew I’d feel
bands he could run QRP or switch to an feed point. Because the SWR on the cable right at home chatting with these guys.
expensive megapower tuner. between the tuner and the radio is low They were spread out from the Carolinas
After a quick trip to the Twin Cities (matched), SWR losses are minimal and to Arizona, and they were booming in on
to fetch some ladder line, I dropped the essentially of no concern. their own horizontal loop antennas. The
feed point, made the switch to ladder line The big problem for me was, I couldn’t guy from Carolina was running a kW and
and hooked everything up to my time- use the autocoupler at the lake if it was he sounded like a shortwave broadcast
tested—conventional—antenna tuner, hanging 50 feet up in the air, precariously station.
making sure I connected the jumper wire attached to my loop. And, even though The gist of our conversation centered
that brought the output balun into the cir- it’s designed for such service, I couldn’t around building a simplified version of
cuit. get myself to mount the coupler outside, the balanced antenna tuner detailed in, “A
With high hopes, I keyed the rig and, exposed to the elements and those nasty Balanced Balanced Antenna Tuner,” by
as usual, the impressive audio thumps and Minnesota winters. Rich Measures, AG6K, in the February
TV screen blackouts began. I still felt the So, I snooped around on the Internet and 1990 issue of QST (available to ARRL
lure of the ladder line, and I was now saw that some ops were using their SGC members in PDF format from ARRLWeb.
making easy contacts on 80 and 160 tuners to feed balanced lines in a conven- The tuner uses a balanced L network
meters—electrically goofy feed line and tional sense (tuner in the shack, feeding a instead of the conventional pi network
all. I was gaining ground, but because of multiband antenna via 450-Ω ladder line). employed by almost every commercial
the RFI, etc, operating wasn’t convenient. I quickly connected the SG-231 to my antenna tuner in service today (see Fig-
rig and to my ladder line, fired it up and ure 3). The balanced L network can di-
Autocoupler to the Rescue watched the magic happen with my own rectly feed 450-Ω balanced lines while
The next remedy I tried didn’t cure all eyes. The autocoupler easily matched maintaining a high degree of electrical
ills, but it became a piece of “can’t do everything from 160 through 6 meters balance on each leg of the ladder line. It’s
without” gear. In preparation for portable with the exception of a small chunk of this balance, I learned, that’s critical in
operation at the lake, I acquired an SGC 80 meters where the impedance at the keeping the ladder line from radiating RF.
SG-231 autocoupler, a computerized, au- shack end of the ladder line was probably And instead of being at the tuner out-
tomatic supertuner that can match practi- weird. put, the balun is placed at the tuner’s in-
cally any load from 160 through 6 meters I enjoyed using this arrangement for put, where baluns really do work over
in the blink of an eye. several years. It matched my loop, which wide frequency ranges and where RF
The SG-231 has no controls and is is probably resonant at 5 MHz or so, in a voltages and RF losses are minimal.
designed to be mounted outdoors, even snap, and RFI on 80 and 160 meters was So what’s the catch? Well, although
in harsh environments. You supply the dc pretty much eliminated. The 231 added the balanced L network could hardly be
power to operate the innards and some tremendous convenience to the loop’s simpler, Rich’s version requires two
RF from your rig and the ’231 does the superb performance. matched roller inductors, which are dif-
rest. Simply key the mike on whatever ficult to find and somewhat expensive,
frequency and the autocoupler matches The Final Balancing Act and a turns-counting dial.
the load in a jiffy, remembering the tun- Clearly, the horizontal loop, the lad- Still, it was clear from reading Rich’s
ing solution so when you return to a der line and the autocoupler were work- article that using a balanced tuner with
nearby frequency, the tuner matches the ing famously. I probably would have put the balun on the input was the proper way
load in about a quarter of a second—fast. up with the RFI/noise problems for a good of doing things for antennas fed with lad-
This amazing piece of hardware, long while if I hadn’t run across a dis- der line. So, I decided to build one and
called an autocoupler by the manufac- cussion during one of my late-night jaunts see for myself.
turer, is designed to be mounted at the on 75 meters. Thankfully, I had two brand-new roller
feed point of the antenna. When mounted The guys in this roundtable were talk- inductors on hand. I’d purchased them
April 2002 49
from MFJ several years ago when I was
planning to build a pair of conventional
antenna tuners. I also had a turns-count-
ing dial I’d purchased almost 20 years
earlier. The capacitor was a junkbox
transmitting unit. The only parts I had to
purchase were the sprockets (¼-inch shaft
size) and the toothed belt required to turn
each roller inductor in sync. These cost
only a few dollars from McMaster-Carr
(www.mcmaster.com).
Building this tuner seemed like car-
pentry, with a little radio thrown in for
good Measures (pun intended). The tuner
is built on a piece of plywood with a
wooden (or other non-conducting) front
panel because the capacitor and inductor
shafts are hot with RF. Building on metal
would have required stand-off insulators
and a lot of tedious wrangling. With wood
it’s a no brainer.
The only accommodation I made in
my prototype is the use of banana posts
and jacks to let me occasionally switch
the capacitor from the output side of the
coils to the input side to match balanced
loads that are less than 50 Ω (most are
Figure 4—The parts layout of the author’s balanced tuner. At the right is an easy-to-
greater than 50 Ω). build, 160-10 meter coax balun. At the center is a pair of roller inductors and the front-
The thing went together in a couple panel tuning capacitor. The larger banana jacks allow the tuning capacitor to be
of hours, and once I synchronized the switched from the input to the output side of the coils. The smaller banana jacks are
roller inductors I replaced the autocoupler for connecting the feed line. The smaller variable capacitor was added to facilitate
with the balanced tuner. Figure 4 shows testing. See the text and References sidebar for more information).
the completed tuner in all its glory.
Ahhh. The tuning was smoooooth, was 5U1A—a DXpedition to Niger, as I (1) Horizontal loops are fabulous—if
with no sharp, hard-to-find dips. It was learned later. After re-remembering how not the best—all-band non-resonant an-
also convenient because it had only two to operate split, I jumped into the fray. It tennas. They noticeably outperform di-
controls instead of the usual three. I me- took me about 15 minutes to work through poles when used at frequencies above
thodically tuned up and down the bands the pileup, but at that relatively low power resonance and they’re easier to match
with a 5-W signal and was pleased to no- level I was more than pleased (as I was there as well (impedance wise). They’re
tice that every spot on every band tuned the next night when I worked V51AS in efficient, quiet and forgiving.
up without a hassle, except for part of 75 Namibia on the same band). (2) As long as in-shack RFI and com-
meters—the same part that the SGC After using the prototype tuner for a puter noise aren’t part of your equation
autocoupler didn’t like. while I discovered a couple of minor you can happily feed the loop with coax
The other thing I almost forgot to no- things I need to iron out. The first is the via a standard antenna tuner for use on
tice was that—as if by divine decree— funky part of 80 meters, which I hope to the band of resonance and all HF bands
there was no more RFI. Anywhere. On fix by lengthening the feed line. The sec- above the fundamental frequency.
any band. At any power output from 5 to ond involves the wide tuning range I’m (3) If you want to operate your loop
100 W. Not even the faintest of audio asking the L-network to accommodate. on bands below its fundamental fre-
thumps could be heard in any of the many Using the first capacitor I tried (a 15 quency, you’ll probably want to replace
electronic goodies just three feet away to 300 pF unit) I can easily tune the bands the coax with ladder line to minimize
from the tuner and the stub of ladder line from 160 through 20 meters, but there’s SWR losses in the feed line.
that pokes through the wall. apparently too much stray capacitance to (4) In doing so you may discover first-
But there’s more. The computer noise match things on the higher bands. When hand that most conventional tuners don’t
was reduced by a good 80%. Now I could I substitute a smaller capacitor I can tune work very well when feeding balanced
operate anywhere. The noise pickup prob- 40 through 10, but not 80 and 160. lines, and that most tuner-output baluns
lem that I’d made incremental progress on I think my eventual solution is to use the don’t keep things balanced over a wide
over the past few years was now virtually smaller capacitor and switch a fixed-value range of frequencies.
eliminated, and what remained was barely capacitor in parallel as needed, or to use If it isn’t obvious by now, I’d like to
noticeable. I rushed to get on the air. the larger unit and switch a fixed C in se- offer my heartfelt thanks to Dave Fischer,
Because I was familiar with how well ries. My short-term solution, however, is to W7FB; Steve Ford, WB8IMY; Rich Mea-
the horizontal loop works as a DX an- mount two tuning caps on the front panel sures, AG6K, and the loopy guys on 75
tenna, I wanted to really push the new and use their respective banana plugs to meters.
setup to the edge. I tuned up on 30 meters switch them in and out as necessary. Oh,
with about 1.5 W showing on the QRP the joys of breadboard construction.
wattmeter. As I tuned to the low edge of You can reach the author at 16928 Grove
the band I heard a nasty pileup. Under- The Path to Your Bliss St, Little Falls, MN 56345; kirk@
neath it all (and a few kHz down the band) So, what have I/we learned so far? cloudnet.com.
50 April 2002