Professional Documents
Culture Documents
Qex 2016 11
Qex 2016 11
APRS (The Automatic Packet Reporting System) is a registered American trademark of WB4APR (Mr. Bob Bruninga).
D-Star is a digital radio protocol developed by JARL (Japan Amateur Radio League).
Publisher
American Radio Relay League
Technical Illustrator
Advertising Information Contact:
Janet L. Rocco, W1JLR
Business Services
5 Determination of Soil Electrical Characteristics
Using a Low Dipole
Rudy Severns, N6LF
860-594-0203 – Direct
800-243-7768 – ARRL
860-594-4285 – Fax
F-Region Propagation and the Equatorial Ionospheric
Circulation Department
Cathy Stepina, QEX Circulation
Offices
9 Anomaly
Jim Kennedy, K6MIO/KH7
2) document advanced technical work in the Amateur In This and Future Issues
Radio field, and
Our QEX authors describe propagation measurements and modes, filter characteristics, and
3) support efforts to advance the state of the tuning an L-network. Put your favorite topic, or innovative measurement, or technical journey
Amateur Radio art. on paper. Share it on these pages with fellow readers. Just follow the details on the www.arrl.
org/qex-author-guide web page, and contact us at qex@arrl.org. We value your feedback,
All correspondence concerning QEX should be
addressed to the American Radio Relay League, comments and opinions about these pages.
225 Main Street, Newington, CT 06111 USA. In this issue, Joe Taylor, K1JT, describes and tests his new EMEcho software to predict and
Envelopes containing manuscripts and letters for
publication in QEX should be marked Editor, QEX. measure Doppler shift, frequency spread, and polarization of EME signals at 144 and 432 MHz
using amateur equipment. Gary Cobb, G3TMG compares Zolotarev low pass-band character-
Both theoretical and practical technical articles are istics with the classical Chebyshev designs and shows a universal table of values that can be
welcomed. Manuscripts should be submitted in word- used in filters for Amateur Radio bands. Ed Callaway, N4II, investigates mechanisms respon-
processor format, if possible. We can redraw any
figures as long as their content is clear. sible for gray line propagation on the low bands. Jim Kennedy, K6MIO/KH6, describes unique
Photos should be glossy, color or black-and-white ionization pattern forms in the ionosphere above the Earth’s geomagnetic dip equator, that
prints of at least the size they are to appear in provide for various of F-region propagation at 6 m. Charles MacCluer, W8MQW, describes a
QEX or high-resolution digital images (300 dots per
inch or higher at the printed size). Further non-iterative two-step process for matching an L-network. Rudy Severns, N6LF, explores mea-
information for authors can be found on the Web at surements of soils characteristics in the 80 m band. Flavio Egano, IK3XTV, followed long term
www.arrl.org/qex/ or by e-mail to qex@arrl.org. observations that suggest long path echoes might propagate with low attenuation by iono-
spheric ducts.
Any opinions expressed in QEX are those of
the authors, not necessarily those of the Editor or the Please continue to support QEX, and help it remain a strong technical publication.
League. While we strive to ensure all material
is technically correct, authors are expected to 73,
defend their own assertions. Products mentioned Kazimierz “Kai” Siwiak, KE4PT
are included for your information only; no
endorsement is implied. Readers are cautioned to
verify the availability of products before sending
money to vendors.
How to Tune an
L-network Matchbox
W8MQW describes a non-iterative two-step matching for an L-network.
A very common matchbox design is the This is mathematically possible since the
L-network of Figure 1. It is certainly the L
left hand side of [EQ 2A], as an increasing
most common design among autotuners. function of XC, takes on each value between Z
Manually tuning such matchboxes is an 0 and R exactly once. As an aside, this shows
iterative process, where an initial guess is that an L-network has at most one possible C 5!ȍ
made for the L and C , then by watching matching solution.
SWR, refinements are made alternately to Step 2. Adjust the inductor L to bring the
L and C until SWR has been lowered to an reactive part of Z resulting from Step 1 to QX1609-MacCluer01
acceptable level. Some autotuners employ zero. That is, choose XL so that
simple bisection searches for minimal Figure 1 — A common L-network matchbox
SWR, or at the other extreme, some use R2 X C . that transforms R to Z.
XL [2B]
sophisticated frequency sampling equation R 2 X C2
solving predictor/correctors. But tuning an L
network need not be an iterative process — it But is This Doable?
1:n
can in theory be done in two simple steps if toroid
The second step is easy once the first is In Out
a second parameter in addition to SWR is completed. Merely adjust the inductor for
observed during tuning. zero return power, i.e., SWR = 1. But is there 9I±9U
a method to steer the tuning of the capacitor n
Two-step Matching to accomplish Step 1? Can we, or a CPU, ȍ
Let us first develop these two steps for the observe some voltage that we may null to
special case of matching a pure resistance achieve Step 1?
R greater than 50 :. The impedance Z of Think of the left end of the network QX1609-MacCluer02
Figure 1 presented to the transmitter is, by of Figure 1 as the beginning of a 50 :
transmission line of length zero. If Vf and Figure 2 — Forward and reflected currents
the standard arithmetic of series and parallel subtract in this current sampler.
impedances, Vr are the forward and reflected voltages on
this line, then because the value of Z is the
RX C2 R2 X C .
Z jX j quotient of the net voltage across it divided to achieve Step 1 is the dc phase voltage
R 2 X C2 R 2 X C2
L
by the net current through it, and because from a phase detector that compares Vr to Vf
[1] the reflected current travels in the reverse – Vr . This dc voltage will be zero only at a
direction1, 90-degree phase shift. The reflected voltage
So the obvious approach is to proceed in Vr is of course obtainable from the reflected
two simple non-iterative steps. V f Vr 100Vr .
Z 50 50 port of a tandem coupler, while the difference
Step 1. Adjust the capacitor to bring the V f 50 Vr 50 V f Vr voltage Vf – Vr is obtainable from a current
real part of Z to 50 :. sampler, such as in Figure 2.
That is, choose XC so that [3]
2
RX Thus Z – 50 is purely reactive exactly Matching reactive loads
C
50 . when the two voltages Vr and Vf – Vr are in
R X C2
2
[2A] quadrature.
Almost certainly, a matchbox will be
asked to match loads with both resistive and
So the crucial voltage to null in order reactive parts. Thinking of such a load in its
But even with the minor interaction of matchbox, whether it is a balanced network
the L and C controls, convergence to match preceded by a balun or an unbalanced
is much quicker because of the added network followed by a balun.
information available from the phase voltage. It would be easy to add an outboard
If phase voltage is negative, decrease C; if current sampler and phase detector plus
phase voltage is positive, increase C. Thus meter to existing manual L-network tuners to Figure 7 — A phase detector employing a
MiniCircuits SBL-1.
the 50 : real part is matched immediately. achieve expedited two-step tuning.
Then proceed in the usual way to a one-to- Displaying Vr (horizontal input) against
one match by adjusting L. Vf – Vr (vertical input) as an oscilloscope
Lissajous diagram is an exceptionally
efficient aid in finding a match — one adjusts Chuck MacCluer, W8MQW, was first
Some Final Thoughts licensed in 1952 with the novice call sign
C to rotate the ellipse vertical, then L to
To save space, the current sampler WN8MQW, and progressed to Amateur Extra
shrink the ellipse to a vertical line. license. He has been very active in EME on
of Figures 2 and 6 could possibly be
An outboard current sampler/phase 432 and 1296 MHz, and recently is active on
incorporated into the tandem coupler of
detector/meter might also speed T-match SSB, CW, and digital modes on 160, 80, and 40
Figure 5 by adding its toroid as a second
tuning. meters. Chuck received a PhD in Mathematics
toroid on the through line. I did not try this.
The above matchbox construction details
Both meter A and B can be any 100 PA in 1966 from the University of Michigan and
were merely sketched. Instead the thrust of is now Professor Emeritus of Mathematics
or smaller ammeters. The meter A must be
this note is to reveal that tuning an L network at Michigan State University. Chuck is a Life
a center-zero meter since the phase detector
need not be a tedious iterative process. It can member of ARRL, IEEE, SIAM, and ASHRAE.
reports both positive and negative voltages.
in theory be done in two steps by carefully
Select by trial and error the current limiting
observing two simple-to-measure voltages. Notes
resistors, marked ‘*’, for your particular
meters. As a starting point try 1 k:. 1
The intermediate quotient of [EQ 3] is the
The fortuitous levels of the sampled Acknowledgements core computation that underlies the Smith
voltages require no active devices. This This project grew from a challenge Chart. ARRL item no. 0413, available from
your ARRL dealer, or from the ARRL Store,
permitted all signals to be piped about with from Gary Adamowicz, WA1OXT, who is Telephone toll-free in the US 888-277-5289,
my favorite coax, RG402 semi-rigid coax developing a competing approach. I thank or 860-594-0355, fax 860-594-0303; www.
with SMA connectors, lending a microwave Mike Blake, K9JRI, and Chuck Hawley, arrl.org/shop/; pubsales@arrl.org.
2
look to the construction. KE9UW, for their many helpful comments L. Phipps, N8LP, “The LP-100 Wattmeter”,
QEX, Jan/Feb 2006.
This algorithm is valid for any L network and suggestions.
Rick Karlquist, N6RK, asked on the antenna and for a couple of skin depths plotted directly on the graph yielding a good
top-band reflector about placing a dipole on down into the soil. The ARRL Antenna Book estimate of the average values of V and Er for
the ground surface to derive soil electrical discussion was pretty limited so I decided to the soil over which the antenna is installed.
characteristics — conductivity (V) and expand on it using antenna modeling software As a practical matter the reference antenna
relative dielectric constant (Er) — from combined with a spreadsheet. needs to be something easy and inexpensive
impedance measurements of the dipole. If you have a program that accurately to build. For that purpose a low dipole works
A short discussion of this technique has models the soil-antenna interaction (such well, and details of a suggested design are
appeared in the last few editions of The ARRL as NEC4) then you can use the antenna of given in Example 1. From a practical point
Antenna Book.1 For some years I’ve used your choice at whatever frequency you are of view it is necessary to have a predefined
the ground probe approach2 to measure soil interested in, see Example 2. Most amateurs antenna for each band. In this article I’ve
characteristics so I hadn’t paid much attention, don’t have this software but the technique chosen 80 m for demonstration purposes.
but in some situations this method may have can still be used. With some prompting
advantages over the soil probes and is worth from Rick, N6RK, I realized that if the What frequency, lengths and heights?
considering. The probe approach gives the antenna dimensions — length, height, wire The height above ground z and test antenna
values for a small volume of soil around the size, etc. — and measurement frequency length L will depend on the frequency of
probe, down to a depth of 3 ft or so. If you are predefined then it is possible to create interest. At what frequency within the band
want to map the properties of a large area a universal chart with contours showing should we make the measurement or do we
you need to make multiple measurements at values of Ri and Xi for a wide range of soils. need to measure across the band? Figures
different locations. The low-dipole approach If the antenna is fabricated as specified, 1 and 2 show examples of actual measured
on the other hand intrinsically averages the and impedance is measured at the specified values for V and Er at my home site using
properties of a much larger area below the frequency, the measured impedance can be soil probes.
QX1611-Severns01 QX1611-Severns02
0.015 100
90
7 March 2015
80 monopole probe
7 March 2015
OWL Probe 70 5 March 2015
Conductivity (S/m)
monopole probe
0.010 60
Er
5 March 2015 50
Monopole probe
40
7 March 2015
19" monopole probe 30 7 March 2015 OWL
probe
0.005 20
1.0 1.5 2.0 2.5 3.0 3.5 4.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Frequency (MHz) Frequency (MHz)
128
90 free space 127 0.1/60
80 126
70 125
0.03/20 124
60 0.001/5
50 123 0.03/20
0.005/13
122
40 0.1/60
121 #17 Aluminum wire dipole
30 120 resonant at 3.7 MHz
20 119 SDUDPHWHU ı(U
10 Ideal 118
0 117
1 2 3 4 5 7 10 20 30 40 50 70 100 1 2 3 4 5 6 7 8 9 10
Height (z) [feet] Height (z) [feet]
QX1611-Severns05
standard electric fence hardware is well
25 ı ı suited for this kind of field measurement,
ı 36 in corresponds to a standard insulated
Er=5
20 ı electric fence post — a practical detail passed
15
ı to me by N6RK.
ı For a given height and resonant frequency,
ı Er=10 ı
10
ı
the resonant length will depend on the values
Er=15 for the ground constants as shown in Figure
5
4. For calculations at 3.7 MHz with z=36 in,
Er=20
0 ı L=125 ft is a reasonable compromise.
Er=25
í
;L ȍ
Er=30
A universal graph for 80 m
í If we have a physical description of the
Er=35 antenna in terms of height above ground
í Er=40
z, length L, wire size, etc., we can model
Er=45
í the antenna at a single frequency f using a
wide range of values for V and Er. This will
Er=50
í Er=60 #17 Aluminum wire dipole give us values for the feed-point impedance
f = 3.7 MHz,
Er=70 Zi=Ri+jXi at a given frequency for each pair
í z = 36", L=125'
Er=80 of V and Er values. Using a spreadsheet we
í can then graph Ri versus Xi — which are the
60 70 80 90 100 110 115 quantities we can actually measure on a test
5L ȍ antenna — as functions of V and Er, with Ri
on the x-axis and Xi on the y-axis, where V
Figure 5 — Xi versus Ri for 0.001<V<0.01 and 5<Er<80. and Er are parameters defining the contours.
After measuring the feed-point impedance at
Over the 80 m band (3.5-4.0 MHz), f we can plot the measured Ri and Xi pair as a
begin to change fairly rapidly. Many of the point on the graph. I used EZNEC pro 3 with
conductivity is 0.011<V<0.0.012 S/m and instruments used to measure impedance
relative permittivity is 41<Er<43. This is a NEC4.2 and an Excel® spreadsheet software,
don’t handle very well impedances less than AutoEZ 4, to automate the calculations and
pretty small range and a measurement near 10 : or greater than a few hundred ohms.
mid-band, say 3.7 MHz, should be more graph them. From earlier work I did on
The impedance values are smaller close to verifying the accuracy of NEC4 for wires
than accurate enough. Remember, we are not series resonance.
trying for 1% accuracy, r20% will do just The next question is “how high”? Figure
close to ground I found that the fitting at the
fine. The modest change of values shown feed point has a shunt capacitance of about
3 shows the effect of various soils (typical 6 pF. This has been added to the model.
over the 80 m band is typical of most soils. V and Er pairs) at a range of heights when
Other bands are much narrower in percentage With L=125 ft, z=36 in and f=3.7 MHz
the antenna is tuned to resonance at each we graph Xi versus Ri as functions of V and
of center frequency so the changes are even point. For heights between 1 and 10 ft the
smaller. A single frequency measurement is Er (Figures 5 and 6). The dashed contours
contours are well separated, promising represent 5<Er<80 and the solid contours
adequate for each band. reasonable resolution for variations in V and
Strictly speaking, the test antenna does represent 0.001<V<0.01 S/m (Figure 5), and
Er. However, at greater heights the contours 0.01<V<0.03 S/m (Figure 6). This range
not have to be resonant but there are practical begin to tighten up making resolution a
measurement advantages to not being too of values should cover most common soils
problem. It looks like any height z between 1 that amateurs are likely to encounter. If this
far from resonance. As you move away and 10 ft should work. I chose 36 in because
from resonance the values for Ri and Xi will doesn’t work for your site then you can use the
it’s a very convenient working height. Since
ı Er=25
í
Er=30
ı
Er=35
í
ı Er=40
;L ȍ
ı Er=45
í
Er=50
ı
ı
Er=60
í ı
Er=70
í
Er=80
í
40 45 50 55 60 65 70 75 80
5L ȍ
Figure 6 — Xi versus Ri for 0.01<V<0.03 and 5<Er<80. Figure 7 — Center connector and feed-point
support. [Rudy Severns, N6LF, photo.]
procedure described in Example 2 to generate 31 cores taped together to form a binocular
your own graph using NEC4 software. core. The winding is six turns of RG174/U way to go. In this case the trial values will be
Note that I’ve cut Figure 6 off for V 50 : mini-coax. 0.001<V<.0.01 S/m and 1<Er<50. Running
greater than 0.03 S/m. As the conductivity
the model repeatedly, we can determine Zi for
increases the scale compresses rapidly. In fact
Example 2 a matrix of V and Er values. A spreadsheet,
if we push V all the way to infinity (perfectly
If NEC4 based software is available then sample included in the QEXfiles, is a good
conducting soil) Zi converges to a single
you can create your own charts using your way to keep track of results.5
point at Zi=4.2-j76.5 :. Most amateurs are
choice of antenna, as follows. We assume a Using the spreadsheet we can graph
not blessed with soil of this high conductivity
horizontal center-fed dipole made with #17 a more restricted set shown in Figure 10.
so this limitation is not that serious. For
AWG aluminum wire at a height z of 36 in. The measured value of Zi for the antenna at
higher conductivity soils ground probe
After tuning to resonance at 3.5 MHz the 3.5 MHz is 80.26+j0 :. A dot with a label has
measurements are probably a better method.
length L is 131.11 ft. The measured feed-point been placed at that value on the graph. We see
impedance Zi at 3.5 MHz is 80.26+j0 :. From our matrix of values has bracketed this value
Example 1 this we can determine the values for V and Er nicely. The V=0.005 S/m line passes right
Figures 7 and 8 are photos of the at 3.5 MHz. First create the NEC4 model through Zi. Also, Zi lies between the Er=10
mechanical arrangements for typical test using #17 AWG aluminum wire 131.11 ft and Er=15 lines, right around Er=13. We
antenna using standard #17 AWG aluminum long and 36 in above ground. Since we do could repeat the process for multiple values of
electric fence wire and hardware widely not know the values for V or Er, we’ll run Er around 13 to refine the answer further, but
available in hardware and farm stores. The the model repeatedly with a range of possible from a practical point of view we’re already
electric fence wire is suspended at 36 in on values for V and Er. If we’re too far off in our close enough. With V=0.005 S/m and Er=13,
fiberglass (F/G) wands, with yellow plastic choice of values the process should point the we have average soil.
wire clips that slide up/down the wands for
height adjustment. The wands were spaced 10
to 20 ft apart and the wire is anchored at the
ends to steel fence posts 6 to 10 ft away from
the ends of the wire. Multiple support points
and significant wire tension kept the droop
to less than 0.25 in. High quality insulators
and non-conducting Dacron line were used
at the wire ends. Figure 7 shows the Budwig
center connecter and the common mode choke
(balun) at the feed-point. The center connector
and choke introduce approximately 6 pF of
shunt capacitance across the feed point, which
must be added to the model. The steel fence
post at the midpoint shown in Figure 8 was
replaced with the F/G wand shown in Figure 7.
The measured impedance of the common
mode choke is shown in Figure 9. The choke
comprises two Fair-Rite 2631665702 type Figure 8 — Test antenna supported with F/G wands. [Rudy Severns, N6LF, photo.]
0
ı and multiple measurements are required to
Er=15 cover a large area. Each has advantages and
80 m wire dipole
ı f = 3.5 MHz, limitations but both will work.
z = 36", L=131.11'
í Acknowledgements
Er=20 I would like to acknowledge the help that
Rick Karlquist, N6RK, provided by reading
my early drafts and asking many questions
that needed some thought to answer. My
í thanks to George Cutsogeorge, W2VJN, for
60 70 80 80 90 100 110 120
5L ȍ
relaying Ricks original top-band query.
Figure 10 — Graph of Ri versus Xi for a range of V and Er values at 3.5 MHz. Rudy Severns, N6LF, was first licensed as
WN7AWG in 1954. He is a retired electrical
engineer, an IEEE Fellow and ARRL Life
QX1611-Severns11
Member.
16
15
14 Notes
13 1
Pages 3-31 to 3-33 in, The ARRL Antenna
12 Z = 36" Z = 3" Book, 22nd edition, 2011. Available from
11 L = 131.11' L=111.22' your ARRL dealer or the ARRL Bookstore,
10 ARRL item no. 6948. Telephone 860-594-
9 0355, or toll-free in the US 888-277-5289;
Er
8 www.arrl.org/shop; pubsales@arrl.org.
2
7 R. Severns, N6LF, “Measurement of Soil
6 Electrical Parameters at HF”, QEX Nov/
5 80 m wire dipole Dec 2006, pp 3-9. Available at www.anten-
f r = 3.5 MHz nasbyn6lf.com.
4
values for E r DQGı 3
3 Several versions of EZNEC antenna model-
for resonance
2 ing software are available from developer
1 Roy Lewallen, W7EL, at www.eznec.com.
4
0.0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 AutoEZ for EZNEC, see www.ac6la.com.
5
Conductivity (S/m) See www.arrl.org/qexfiles.
Figure 11 — Values for V and Er that result in the resonant lengths shown at 3.5 MHz.
Figure 4 — The wavy line near the geographic equator is the geomagnetic dip equator. The
geomagnetic field center is displaced from the geographic center toward the Pacific side,
leaving the odd “bump” over the Atlantic. [G.Projector map and overlays.]
Figure 2 — Es produces mirror-like
reflections. Note also that the smaller the
angle of attack the higher the MUF for a
given electron density Ne.
Figure 6 — This electron density map, for 0315 UTC on April 2, 2013 at Figure 7 — The dying daytime electron fountain is shocked into a
85°W (Geomagnetic Equator 5°S), shows both EIA pools have MUFs series of standing waves of alternating high and low ionization. The
of about 57.5 MHz at about 390 km. Compare this with Figure 5. [USU- stacked layers are buoyant and float upward.
GAIM ionosphere.]
Figure 14 — Both skew paths start as south-lane TEP and then are
redirected upon entering a later lane (circles). The path from CE then
does a 2F2 hop in the north lane to BA. The FK8 path crosses the Figure 15 — Various transpolar long paths are shown between KH6
north lane and runs into it again, and is redirected to EA8. [G.Projector and the Med and A45; KH8 and 5H3; E51 and ZS; and BV and EA8.
map and overlays.] [G.Projector map and overlays.]
a possibility at the “gray line” of my SS and SS). At 2325Z, to my delight I heard VK9CZ
Introduction VK9CZ SR, which occurred within three calling CQ on 3507.5 kHz. There was no
From 30 March to 13 April 2013, Chris minutes of each other. In the past I had heard pile-up, and he was not working split. To my
Tran, GM3WOJ, and Keith Kerr, GM4YXI, stories of enhanced propagation under such amazement I worked him on the first call, as
operated as VK9CZ from the Cocos conditions from grizzled low-band veterans, N4II. He called CQ again and, still with no
(Keeling) Islands, the DXCC entity farthest and I was curious to find out if I could hear pile-up, I worked him again, this time using
from my location in south Florida. When this VK9CZ on 80 m at all. the club call sign N4BRF. VK9CZ called CQ
DXpedition was announced in late 2012, I Due to CC&R restrictions at my home, I again and again, until he finally faded at 2345
determined that I wanted to work VK9CZ chose to operate from nearby club stations. Z — 5 minutes after my SS, and 8 minutes
on 80 m. The best 80 m station available at the time after his SR. The next 80 m opportunity was
The first step was to identify the period was at the Boca Raton Amateur Radio on 5 April, but a large thunderstorm sat over
of common darkness between us — if, in Association, N4BRF. It offered a SteppIR the club station, keeping me off the air.
fact, one existed. A check of the sunrise (SR) vertical, with 60 radials, in a quiet location On 7 April, wanting to hear more, I began
and sunset (SS) times at both locations for on the edge of the Loxahatchee National monitoring 80 m CW at 2315Z, 25 minutes
7 April, midway through the DXpedition, Wildlife Refuge, and a 500 W transmit before SS. At 2330Z, I heard VK9CZ again
revealed the following: power amplifier. The station did not have a calling CQ on 3507.5 kHz. There still was no
VK9CZ SS = 1132Z; N4II SR = 1106Z dedicated receive antenna. true pile-up. He worked several stations in an
VK9CZ SR = 2337Z; N4II SS = 2340Z. My first opportunity was on 3 April. Not orderly, workmanlike fashion before fading
There was no period of mutual darkness, knowing what to expect, I began monitoring at 2350Z, 10 minutes after my SS, and 13
but I thought that the low bands might still be 80 m CW at 2300Z (40 minutes before my minutes after his SR.
N4BRF
VK9CZ
QX1611-Callaway01
Figure 1 — Into daylight? Some reports indicated that the received VK9CZ signal peaked to the SSW. [DX Atlas]
QX1611-Callaway03
25
CA:21 FL:21
20
Southern states clearly
favored, even accounting for
differences in population
15
SS QSOs
KY QSO made at SR? SR QSOs
TX:11 (in US)
Very interesting, if true;
K9LA emailed the op, 10
but got no response
0
AK WA OR CA NV AZ NM CO TX AL TN KY GA OH FL WV NC VA PA MD DE NY NJ MA ME
West East
Figure 3 — VK9CZ 80 m QSOs made by state, ordered West-to-East. Southern states were favored in both SS and SR openings.
QX1611-Callaway04
115
100
Eastern openings much shorter
than western openings
86
72
57 Minutes
Eastern QSOs were much closer
to the VK9CZ antipode meridian
than western QSOs 43
28
FL: 24 min
14
ME: 6 min
00
AK WA OR CA NV AZ NM CO TX AL TN KY GA OH FL WV NC VA PA MD DE NY NJ MA ME
West East
Figure 4 — Duration of VK9CZ 80 m opening by state, ordered West-to-East. The eastern openings were of much shorter duration than the
western openings. Note the short duration of the Maine opening.
450 At the
At midnight terminator
400
350
Altitude (km)
300
250 F Layer
200
150
Better E-F duct E-F duct
100
E Layer
50
0
1×109 1×1010 1×1011
Electron per Cubic Meter
QX1611-Callaway06
Figure 6 — Propagation along the terminator is unlikely. Figure 7 — The duct between the E and F layers of the ionosphere. Note
The horizontal ionization gradient along the terminator the more pronounced duct at midnight, compared to the duct at the
would refract a signal away from the terminator, into the terminator. [Adapted from Robert R. Brown, NM7M, “On the SSW Path
dark ionosphere. [DX Atlas] and 160-Meter Propagation,” QEX, Nov/Dec 2000, pp. 3-9, Figure 1.]
0
<- greatCircle(N4BRF,
Day Night
4 dpN4150, n=360)
VK9CZ VK9CZ
antipode [Line 7] lines(gcN4150,
205°
lwd=2,lty=’dashed’,
–50 Terminator N4BRF col=’black’)
antipode
2 3 Line 2 locates N4BRF on the map by
longitude and latitude. Line 5 defines a
Auroral IonizaƟon destination point 15 million meters away
from N4BRF, along the great circle heading
–150 –100 –50 0 50 100 150 of 150 degrees. Line 6 defines the Great
Longitude Circle containing the locations of N4BRF
and the destination point, and Line 7 draws
Figure 11 — The path through the dark ionosphere, with the ionization of the southern polar the line on the map.
oval as the skewing element between N4BRF and VK9CZ. The ionization skews the signal Using R, I made a map that had Great
towards the equator, i.e., to the north, from one Great Circle to the other. Dashed line: N4BRF Circles leaving N4BRF to the SSE (I used
Great Circle. Dotted line: VK9CZ Great Circle. Solid line: terminator.
Coaxial Cable
Assemblies
These pre-made low-loss cable
assemblies are available with
DX Engineering’s patented new PL-259 bhi Noise Cancellation Products
connectors, featuring the best qualities of both crimp-on and bhi Noise Cancellation Products harness cutting-edge digital
solder-on connectors. Use the free online Custom Cable Builder at signal processing (DSP) technology to enhance the audio quality
DXEngineering.com to build assemblies made to your exact specs. of incoming radio signals. By removing the noise, the improved
audio makes it much easier to discern distant and weak
DX Engineering’s coaxial cable is also available by the foot or contacts. DX Engineering now carries bhi Noise Cancellation
in bulk spools. Products, plus bhi accessories like switches, cables, adapters
and extension speakers.
Active Magnetic
Loop MF/HF
Receiving Antenna
The well-known
RF-PRO-1B is now
manufactured and
sold exclusively by
DX Engineering. This
original Pixel loop
design incorporates
a high performance
preamplifier for
excellent broadband
reception from 50 kHz
Footswitches to 30 MHz. Installation
Go hands-free and speed up your operating. These new PTT is easy, because it’s
footswitches from DX Engineering blend performance and value. only 38 inches in diameter. When rotated, this antenna provides
You have three options: a budget friendly plastic model, the deep nulls for effective reduction of directional noise and
stalwart cast iron version, and a new style with an extra-wide interfering signals. Ideal for Amateurs, SWLs and AM DXers, this
pedal—perfect for enthusiastic stomps. loop also offers up to 30 dB rejection of electrostatic field noise.
Lattitude
340°
and physics of the ionosphere — I know Night DU6
0 Day
of no measurements taken to confirm DU6
that propagation actually occurs via this antipode
mechanism. 20°
PY3
–50
The TOFU Project Terminator
One way to confirm this propagation Auroral IonizaƟon
mechanism is to use time-domain techniques
to determine the path delay and, therefore,
its path length. The problem of gray line –150 –100 –50 0 50 100 150
propagation seems particularly amenable Longitude
to this type of analysis, since it is relatively
predictable in both time and location, and
Figure 17 — The path through the dark ionosphere, with the ionization of the northern polar
there are specific candidate paths having oval as the skewing element between PY3 (southern Brazil) and DU6 (central Philippines). The
predictable path delays. The Time-Of- ionization skews the signal towards the equator, i.e., to the south, from one Great Circle to the
Flight Unit (TOFU) project is an attempt to other. Dashed line: PY3 Great Circle. Dotted line: DU6 Great Circle. Solid line: terminator.
use time-domain techniques to answer this
question.3
The concept of TOFU is simple. Both Those interested in working on the TOFU polarization of signals received via gray
ends of the link are synchronized in time, project — a project still at the concept stage line propagation. This could be done, for
via GPS. At a known time, the transmitting — should contact the author. example, using Waller flag antennas having
station sends a predetermined pseudo- rotatable polarization.
random (PR) sequence (e.g., 255 bits in
length) of ones and zeros. The receiving Conclusion and Future Work Ed Callaway, N4II, received the BS degree
station stores what it receives in a time- This investigation of gray line propagation in mathematics and the MSEE degree from
stamped file. The received data can be post- started with a surprising 80 m QSO, and the University of Florida in 1979 and 1983,
processed (with, e.g., Matlab) by passing it is still going strong, three years later. The respectively, the M.B.A. degree from Nova
through a sliding correlator and the time of tentative conclusion is that polar refraction (now Nova-Southeastern) University in 1987,
maximum correlation — the “correlation of low-band signals in the E-F duct is most and the PhD degree in computer engineering
peak” — determined. Since both stations are likely responsible for the phenomenon, from Florida Atlantic University in 2002. He
time-synchronized, once the delays through although this has yet to be confirmed holds 58 issued US patents, and is the author of
the transmitter and receiver are removed, a experimentally. several books, book chapters, and papers. Dr.
One corollary to this hypothesis is that Callaway is a Registered Professional Engineer
simple calibration, the difference between the
relatively low takeoff angles would be best and was named an IEEE Fellow in 2010. After
time of the correlation peak and the time that 25 years with Motorola, in 2009 Dr. Callaway
the PR sequence was sent is the measured to participate in gray line propagation,
since they are needed to inject a signal into co-founded Sunrise Micro Devices, a leading
path delay. This delay can then be compared provider of low-voltage, low-power wireless
against those delays predicted by the various the E-F duct at SR and SS. Interestingly, if
long-range, low-band propagation generally IP to the semiconductor industry, where he
candidate propagation mechanisms. was CTO. In 2015 the company was sold to
At present the US has a 300 baud limit occurs via the E-F duct at times other
ARM, Ltd., where he is now Director of Radio
on digital signaling at HF, which means than SR and SS, antennas with higher Standards. Dr. Callaway holds an Amateur
the minimum duration of bits in the PR angles of radiation also may be useful, so Extra Class license and is a Life Member of the
sequence is 3.333 ms. Given the signal-to- that signals may be injected into — and ARRL. He is an avid DXer, and is fascinated by
noise ratios common on the low bands and detected coming out of — random E-layer ionospheric propagation of all types.
the limited symbol transition times due to the inhomogeneities. Perhaps the main value
limited signal bandwidth, but accounting for of gray line propagation is that it provides a
oversampling at the receiver, one may hope predictable time and location for an entrance Notes
to achieve a measured path delay accuracy of into the E-F duct — SR and SS.
1
Experimental confirmation of the polar shop.spacew.com.
better than plus or minus 1 ms, leading to a 2
https://www.r-project.org/.
path length measurement uncertainty of plus refraction hypothesis is a subject of future 3
Carl Luetzelschwab, K9LA, first suggested
or minus 300 km. work. Also of interest is a study of the the use of time-domain techniques to me.
$0RUH(I¿FLHQW/RZSDVV)LOWHU
G3TMG compares Zolotarev quasi-elliptic low pass-band characteristics
with the classical Chebyshev design in terms of the expected performance for
lumped-element circuits, and shows a universal table of values that can be
used in filters for Amateur Radio bands.
6 ±
6
±
±
±
66 G%
66 G%
±
6
±
±
±
6
±
±
±
± ±
)UHTXHQF\ 0+] )UHTXHQF\ 0+]
4;&REE 4;&REE
Figure 2 — A loss-less 7th order Elliptic response. Figure 3 — A 5th order response from OptLowpass program, tweaked
and with losses.
Amplitude
to the circuit produced by OptLowpass at 2
the cost of one more component — now 8.
1
The 3rd harmonic band is clearly worse at 1
63 dB, as we might have expected with no
coincident zero, but the general rejection is 0
better, monotonically improving beyond the Chebyshev
3rd harmonic frequency. Taking a leaf out of –1
–1
Jim Tonne’s book, the rejection could still
further be improved by providing a matched –2
region only in the top half of the low-pass
pass-band. How might this be done and what –3
would be the cost?
It turns out that Jim has discovered, –4
intentionally or otherwise, the characteristic –1.5 –1.0 –0.5 0 0.5 1.0 1.5
that has become known as the Zolotarev Frequency (radians)
function. This function is similar to the
QX1609-Cobb05
ubiquitous Chebyshev characteristic which
has a y-valued unit amplitude cyclic behavior
between x-axis values of -1 and 1 radian. Figure 5 — A 7th order Chebyshev and Zolotarev low-pass function.
In the Zolotarev case, which has an extra
parameter, the first cycle about the origin can
have a magnitude greater than 1. Figure 5
compares the two functions.
This issue was researched more than 40
years ago by Ralph Levy5 who determined 4;&REE
&S)
a method of incorporating this characteristic
into low-pass filter designs for, in principle,
any order. The mathematics therein is not = ȍ /Q+ /Q+ /Q+ = ȍ
trivial but the basic outcome is that even 3RUW 3RUW
ordered functions are analytic and can be 3 3
generated by modifying the root locations &S) &S) &S) &S)
of a conventional Chebyshev polynomial
function in a predetermined geometric
fashion. For these to be realizable however,
it is necessary for the source and load to be
unequally valued, the ratio depends on the
bandwidth compression factor chosen. Figure 6 — Basic circuit configuration for analysis. Element values are in Table 2.
Odd-ordered functions — more
appropriate for low-pass filters — are
unfortunately not so easily dealt with,
same electrical circuit. Second, a couple into the Zolotarev domain using a freely
requiring some rather knotty mathematical
of practical examples will demonstrate the chosen fractional bandwidth compression
techniques. Nevertheless, these are of greater
achievable performance using typical and factor providing a new polynomial function.
value since the resulting network source
appropriate inductor construction techniques. Prototype lumped element component
and load are always equal. Because of this,
values are then extracted from the equivalent
a simpler design methodology for any odd
admittance polynomial and scaled in both
ordered function has been recently developed Approximation and Synthesis of 7th
Order ZWAZ Filter frequency and impedance so as to produce
and fully described previously in QEX.6 The
To create a 7th order ZWAZ filter, we the necessary component values.
mathematics will not be described here but
first generate a 7th order CWAZ polynomial Two filters using the common network
the Zolotarev With Added Zero (ZWAZ)
filter function whose characteristic pole configuration (Figure 6), have been selected
will first be compared to that of the CWAZ,
and zero singularities are then mapped for comparison, one with Chebyshev and the
by way of example, using essentially the
other with Zolotarev pass-band characteristics
±
tolerance parts. Some allowance was made
for stray or parasitic capacitance.
In the first instance, a design chosen for ±
the original 70 MHz band of interest was
realized using the typical VHF technique ±
of shielded air-cored solenoid inductors
with a designed Q of 200 (Figures 10a and ±
10b). The measured result for transmission
loss, 0.28 dB, is close to the target function,
0.19 dB, although the cut-off frequency ±
is in error by approximately -1.5 MHz
due to component tolerances and tuning ±
inaccuracies. Good stop-band lobe definition
with a peak attenuation ~71 dB is achieved )UHTXHQF\ 0+]
with quite simple mechanical screening
4;&REEE
arrangements.
(B)
In the second instance, a design chosen
for 60 m operation uses toroidal inductors Figure 9 — (A) Insertion loss for ZWAZ (solid) and CWAZ (dashes). (B) — ZWAZ (solid) and
with a designed Q of ~200. In this case, CWAZ (dashes) shows rejection at 160 MHz for the ZWAZ is 70.7 dB, and for CWAZ is 58.2 dB.
±
Conclusion
Two low-pass filter types (CWAZ and ±
ZWAZ) have been synthesized and compared
as an evaluation of a preferred output filter ±
for power amplifiers in general. The ZWAZ
filter is demonstrated to be significantly better,
by design and realization, in terms of both )UHTXHQF\ 0+]
harmonic and general stop-band attenuation, 4;&REEE
as compared to the CWAZ filter. In fact, the
(B)
transition rate and the ultimate stop-band
attenuation beyond the 3rd harmonic frequency
all exceed that of the ideal Elliptic filter. Figure 10 — (A) A 4 m band Zolotarev low-pass filter test piece construction. (B) — The
The two test samples generally show 4 m band Zolotarev low-pass filter S21 test result – measured (solid), predicted (dashed).
[Gary Cobb, G3TMG, photo]
good agreement with the target design
(A)
800-522-2253
This Number May Not
Save Your Life...
±
But it could make it a lot easier!
±
Especially when it comes to
ordering non-standard connectors.
±
RF/MICROWAVE CONNECTORS,
CABLES AND ASSEMBLIES
± • Specials our specialty. Virtually any SMA, N,
TNC, HN, LC, RP, BNC, SMB, or SMC
delivered in 2-4 weeks.
± • Cross reference library to all major
6 G%
manufacturers.
• Experts in supplying “hard to get” RF
± connectors.
• Our adapters can satisfy virtually any
combination of requirements between series.
± • Extensive inventory of passive RF/Microwave
components including attenuators,
terminations and dividers.
± • No minimum order.
±
±
)UHTXHQF\ 0+]
NEMAL ELECTRONICS INTERNATIONAL, INC.
4;&REEE 12240 N.E. 14TH AVENUE
(B) NORTH MIAMI, FL 33161
TEL: 305-899-0900 • FAX: 305-895-8178
E-MAIL: INFO@NEMAL.COM
Figure 11 — (A) A 60 m band Zolotarev low-pass filter test piece construction. (B) — The 60 m BRASIL: (011) 5535-2368
band Zolotarev low-pass filter S21 test result – measured (solid), predicted (dashed). URL: WWW.NEMAL.COM
[Gary Cobb, G3TMG, photo]
1. Introduction A smooth moon would produce a the 432 MHz station at W2PU, the Princeton
For me, one of the fascinations of amateur specular reflection that preserves linear University Amateur Radio Club. The two
Earth-Moon-Earth (EME) communication polarization and reverses the sense of circular stations are configured in a similar way.
is the range of interesting physics that polarization. A rough moon (on the scale Both have four dual-polarization Yagis —
accompanies it. Motivated in part by a of one wavelength) would produce diffuse 4×2Mxp28’s at 144 MHz1 and 4×15LFA-
need to test a new software program called echoes and significant depolarization; JT’s at 432 MHz.2 Both stations use a
EMEcho, and also by a desire to see how well cross-polarized return echoes might be single low-loss feed line for transmitting
I could predict and measure the phenomena just a few dB weaker than the dominant and separate LNAs and receive feed lines
of Doppler shift, frequency spread, and polarization. At VHF and UHF frequencies for each polarization. The receivers use
polarization of EME signals using amateur the circumstances are closer to the specular dual-channel down-converters to produce
equipment, I recently made an extensive limit. Received echoes should be almost fully four baseband signals, I and Q (in-phase
series of lunar echo measurements at 144 polarized, and with linear polarization they and quadrature) for each polarization.
and 432 MHz. This paper describes how the should have a polarization angle that depends WSE converters by SM5BSZ3 were used at
measurements were made and presents a on geographic locations of the transmitter 144 MHz, and the IQ+ receiver by HB9DRI4
selection of results. and receiver and the amount of ionospheric at 432 MHz. Four-channel sound cards
Doppler shifts of EME signals are Faraday rotation. (M-Audio Delta44) digitize the I/Q signals
caused by continuous changes in the total Together with our knowledge of solar- at 96000 samples per second, and in normal
line-of-sight distance between a transmitter, system dynamics, the relevant physics EME operation all further processing takes
reflecting or scattering spots on the lunar is such that EME Doppler shifts can be place in the computer programs Linrad5
surface, and a receiver. The relevant rates calculated with high accuracy (parts in 1010, and MAP656. For this echo experiment my
of change are usually dominated by Earth or better) for any time and any terrestrial new program EMEcho was used in place
rotation, which at the equator amounts to location. Maximum Doppler spread across of MAP65. Transmitter power was about
about 460 m/s. As a consequence, two-way the full lunar disk is also predictable. Faraday 500 W at the antenna, at each station.
Doppler shifts can be as large as r 440 Hz rotation depends on latitude, moon elevation,
at 144 MHz, r 4 kHz at 1296 MHz, and time of day, solar activity, and ionospheric 2.1 EMEcho Software
r 30 kHz at 10 GHz. Different reflection “weather”; the resulting effects are generally EMEcho is a new program designed to
points on the lunar surface produce slightly not predictable in detail. For optimum make reliable tests of lunar echoes from
different Doppler shifts, so the echo of a efficiency, EME operators must know an amateur EME station. It goes beyond
monochromatic signal is spread out over a about and take account of this full range of the Echo mode available in WSJT in two
small and predictable frequency range. The phenomena, both predictable and otherwise. important ways. Doppler calculations are
full range of spread can be as large as 4 Hz at done with state-of-the-art accuracy, based
144 MHz and 300 Hz at 10 GHz. However, at on the Jet Propulsion Laboratory’s DE405
2. Equipment Setup
VHF and UHF a majority of reflected power planetary ephemeris.7 In addition, EMEcho
I used single-station echo tests to
is returned from a region near the center of takes full advantage of a dual-polarization
measure Doppler shift, frequency spread,
the lunar disk, so the observed half-power system by measuring the polarization as well
and polarization during the moon pass of
Doppler spread is always considerably less as the frequency and strength of echo signals.
January 2-3, 2015. My equipment was that
than the full limb-to-limb amount. The basic echo-testing cycle is similar
of the 144 MHz EME station at K1JT and
In Stock Now!
Semiconductors
for Manufacturing
and Servicing
Communications
Equipment
Figure 5 — Examples of contributions that might affect the accuracy of computed Doppler shifts.
• RF Modules
• Semiconductors
• Transmitter Tubes
Phone: 760-744-0700
Toll-Free: 800-737-2787
(Orders only) 800-RF PARTS
Website: www.rfparts.com
Fax: 760-744-1943
888-744-1943
Email: rfp@rfparts.com
Figure 6 — Measured polarization angles at 144 MHz (left scale) and 432 MHz (right scale).
K6PZB #2
Se Habla Español • We Export
–20
K6PZB #2B Calc.
Phone: 760-744-0700
Toll-Free: 800-737-2787
Excess Loss (dB)
–30
KK6EEW Calc. (Orders only) RF PARTS
Website: www.rfparts.com
–40 KK6EEW
Fax: 760-744-1943
–50
888-744-1943
Email: rfp@rfparts.com
–60
0 20 40 60 80 100 120 140 160 180
Altitude (m)
QX1609-LettersA
1(: The TICC is a two channel counter that can time events
with 60 SLFRVHFRQG resolution. It works with an Arduino
Mega 2560 processor board and open source software.
Think of the most precise stopwatch you've ever seen, and
you can imagine how the TICC might be used. The TICC
will be available from TAPR in early 2017 as an
assembled and tested board with Arduino processor board
and software included.
TICC
High-resolution 2-channel Counter
TA P R
PO BOX 852754 x Richardson, Texas x 75085-2754
Office: (972) 671-8277 x e-mail: taproffice@tapr.org
Internet: www.tapr.org x Non-Profit Research and Development Corporation
*Actual dealer prices may vary. Shipping and handling charges apply. Sales Tax is required for all orders
shipped to CT, VA, and Canada. Prices and product availability are subject to change without notice.
Q(; 11/2016