Topband: "Magnetic" receiving loop - update

[Eric Scace K3NA]

    Back in late December I posted some questions about loop 
antennas.  I got many interesting answers and opinions, and 
later Tom W8JI posted some more information about loops on 
his website.

    Several asked me to report back on my experiences.  For 
the benefit of the archives, here is a summary of my 
practical work to date.

    The original intent of this project was to construct a 
loop antenna that I could use for DFing a local QRN source 
on 160m, using a battery-powered SWL receiver (Yaesu 
FRG-100) as a detector.

    The FRG-100 is a fairly light MF/HF receiver that runs 
off 12 Vdc and has an illuminated S-meter (needle, not bar 
graph), and a reasonably robust design.  The S-meter needle 
will be useful in detecting a null in signal strength while 
DFing.  While this may not be the best performing, nor most 
compact, solution for a detector, it has the advantage that 
I already own it ;-)

    The ARRL RFI book and ARRL Antenna Book both suggested 
that a loop for 160m not exceed about 5 ft in diameter. 
 From a practical standpoint, a 5 ft diameter loop is about 
as bit as I can stick in the back of my pickup truck while 
driving around the neighborhood taking DF and signal 
strength measurements.

    I started out by building a "shielded loop": 2 turns of 
RG-6.  The shield was opened for about 2-3cm at the top of 
the loop, directly opposite the feedpoint at the bottom.  An 
octagon of 3/4" plastic electrical conduit contained the 
coax.  The feedpoint was in a 4" x 4" x 2" electrical 
utility box at the bottom.  The feedline connects to the 
utility box with an F-connector.  The feedline center 
conductor connects to the center conductor of the first RG6 
loop.  At the opposite end of that turn, a toggle switch 
connects the center conductor to the shield of the feedline 
(i.e., one turn loop) or to the center conductor of the 2nd 
turn of coax (continuing in the same direction around).  The 
center conductor end of the second turn attached to the 
feedline shield via the metal utility box.

    All the parts came from Home Despot except the 
F-connector and switch, which are available from Radio Schlock.

    This is the core of the basic design in the books.

    This simple design was tested with an Orion receiver on 
160m for general performance.  (I didn't use the FRG-100 at 
this stage in order to eliminate one more variable from the 
testing scenarios.  I know how my QRN sounds on the Orion.) 
  The loop was pretty deaf... but, on local AM stations, the 
loops showed very sharp nulls off the ends -- easily >40 dB. 
  The nulls were symmetric: rotating the loop 180° produced 
a null of the same depth.  So the system seems well balanced.

    But on 160m I could not hear the local QRN, which runs 
S-7 to 8 on the NE beverage antenna.  In fact, I couldn't 
hear any real change in the noise floor when I rotated the 
loop.  To cut a long detective story short, I discovered 
that the noise I was hearing on the Orion was actually 
leakage from the unselected beverage antenna.  The Orion 
antenna switch matrix has about 40 dB of isolation (~7 
S-units)... just enough to leak an S-8 noisy signal from one 
port to another and make it appear to be just twitching the 
S-meter above the receiver's noise floor - hi!

    Lesson #1: disconnect ALL the other antennas from the 
receiver!

    The next step was to find a variable capacitor to attach 
from the feedline center conductor to the feedline shield. 
A capacitor decade switchbox filled in for the task of 
figuring out what level of capacitance was suitable.  (The 
books suggested 400 pF, but point out it varies with the 
characteristics of the coax.)  My switchbox is pretty 
coarse, in 100 pF steps.  But the signals on 1.8 MHz peaked 
up with about 1400 pF using one turn.  Using two turns, the 
required capacitance was about 1/2 the one-turn value -- 
much in line with expectations set by reading W8JI's 
website.  (The second turn does not make the loop more 
sensitive; it just changes the inductance to be resonated by 
the capacitor.)

    Signal strengths were still very low, however... not 
strong enough for good DFing.

    W8JI's website points out that the use of shielded turns 
is really immaterial to the behavior of a loop.  At the 
bottom of the page Tom shows several unshielded designs.  So 
I decided to try a simpler design and make a comparison.

    Design #2: a single turn of ordinary hookup wire (still 
in my 5 ft diameter plastic octagon) with the resonating 
capacitor.

    This design has much better sensitivity.  The single 
turn resonates with 800 pF at 1.8 MHz, decreasing to 600 pF 
at 2.0 MHz.  The sensitivity to adjustment is fairly sharp; 
a change of 100 pF on the switchbox will drop signals by 
several S-units.  My absolute signal strength measurements 
may be lower than could be obtained by a continuously 
variable capacitor.  (I now have one on order from Antique 
Radio Supply.)

    Attaching a DX Engineering pre-amp brings the local QRN 
up to S-4 (compared to S-7/8 on the beverage).  If I add the 
Orion pre-amp, the signals come up another 2 S-units.  At 
this point we have a workable antenna.

    Last night I did some general listening tests to compare 
the ability of the loop to discern signals vs the beverages. 
  While this may not be so pertinent to DFing QRN, it could 
be of interest to those looking for low band receiving 
antennas in small spaces.

    W1AW, a high angle signal here on the southern 
Massachusetts coast.
    Loop with DX Eng preamp: S-7
    4-square: S9+30 dB
    west  beverage: S9+20 dB

    KG4SB:
    south beverage:  general noise floor at the time = S3. 
Signal = S6.
    loop with DX Eng pre-amp:  general noise floor = S2. 
Signal = S3/4.

    European signals:  In general the signal-to-noise ratio 
was lower on the loop by a couple of S-units, compared to 
the NE beverage (single wire about 420 ft long).  But I was 
pleasantly surprised at how often signals could be 
discovered by tuning around the band while listening to the 
loop.  (And, in accordance with Lesson #1, all the other 
antennas were disconnected from the radio during these 
tests.)  While conditions seemed generally good last night, 
the performance of this simple single-turn loop is better 
than what I would have expected for such a small antenna at 
ground level.  (But it wasn't hearing 6O0N well enough for 
copy; he was reliably copyable on the NE beverage but just 
above the local noise level.)

NEXT STEPS:
    1.  Check the loop for symmetry of pattern.  I plan to 
do this with some local AM stations, and to use an AEA 
analyzer as a weak signal generator a few hundred feet away 
on 160m.
    2.  Add a series capacitor (to match the impedance to 
the feedline) and low-C isolation toroid balun, as suggested 
on Tom's website.  (See the last figure on his page on 
loops.)  As presently built, the loop is balanced but the 
coax is not.  I may also slip on a string of ferrite beads 
to eliminate shield currents on the feedline if it appears 
pattern symmetry is being disturbed.  I don't expect either 
of these steps to materially increase the signal to noise ratio.

    I'll provide further results after these steps.  I hope 
this is helpful to others.

-- Eric K3NA