webform at b4h.net webform at b4h.net
Mon Nov 27 10:37:25 EST 2017

CQ Worldwide DX Contest, CW

Call: K1LT
Operator(s): K1LT
Station: K1LT

Class: SOAB(A) HP
QTH: Ohio EM89ps
Operating Time (hrs): 37 3/4
Radios: SO2R

 Band  QSOs  Zones  Countries
  160:   82    15       41
   80:  274    24       92
   40:  294    28      104
   20:  739    35      121
   15:  310    24       95
   10:   26    10       15
Total: 1725   136      468  Total Score = 2,882,288

Club: Mad River Radio Club


The last few years I have used the week leading up to the CW weekend
of the CQ World Wide contest as 'project week'.  I take a few days off
from work and combined with the Thanksgiving holidays I have a week I
can devote to some personal project.  In previous years the project
usually implemented some new station feature to make the contest
better.  This year the project was to continue my new phased array
software defined radio receiving system (still looking for a succinct
name) and make as much progress as possible with the expectation that
the new receiver would not be finished.

The basic hardware is a doubly balanced tayloe detector attached to a
nifty TI pro-audio chip.  The audio chip is a 2-channel 24-bit analog
to digital converter with sports a 120 db dynamic range spec.  The
digital side of the chip attaches to a Zedboard (a single board
computer) that shows off the Xilinx Zynq chip which combines 2 ARM9
cores (modern microprocessors) and a field programmable gate array
(FPGA, programmable digital hardware) along with memory, USB,
Ethernet, and all the other modern computing amenities.

The hardware and computer at this point can receive a 160 meter signal
and produce a data stream that can obtained via Ethernet and analyzed
on another computer.  Analysis would be better easier if an SDR
receiver and GUI were running on the Zedboard, so the project week
activity was to enable audio and video on the Zedboard.  The audio and
video controllers are implemented in the FPGA with outboard 'codecs'
to convert a hardware data stream into audio and HDMI video.  Note
that the audio codec is just an ADC and a DAC in a box while the video
codec is a very fast parallel to serial converter with a scrambler to
make Hollywood happy.

Even though the Zedboard is about 5 years old, the marriage of
hardware and software is in a tremendous state of flux.  After much
Googling and hacking, I was able to get a Linux kernel to talk to the
FPGA based audio controller and produce sound.  That took the first
day.  Although sound only would be sufficient for most of my purposes,
video running on the Zed showing a band-scope would be really cool.
After 2 more days of Googling and hacking, I determined that video was
not going to work without considerably more effort.

Thursday I decided to work on my current station rather than my future
station.  The first quick improvement to come to mind would allow me
to swap the roles of my two tribanders.  One tribander is connected
via the automated antenna switching to the transmitter while the other
connects via switching to the 2nd receiver in the K3.  One of the
beams can be switched to the transmitter of the second radio.  With
the poke of a keyboard key, I can swap the second tribander to the
transmitter, but then I have no antenna for the second receiver.

A stopgap solution would be to add a transfer switch between the
tribanders and the rest of the antenna switching.  Then the roles of
the tribanders could be swapped independently of any other switching.
It just so happens I have a transfer relay in the closest, but I don't
know how to control it.

So I figured out the operation of the Transco motorized transfer
switch (part no. 1460-30-95) that K8ND gave me some time ago.  The RF
side is a double-pole double-throw switch configured as a transfer
switch (inputs A and B connect to either outputs X and Y or inversely
Y and X).  The control side has 3 pins: a common and 2 hot pins.  The
common pin is the pin just to the left of the key when the connector
is viewed with the key up.  Apply 115 volts AC or DC between one of
the hot pins and common and the motor spins one way until a limit
switch opens the circuit.  Apply voltage between the other hot pin and
common and the motor spins the opposite direction until a limit switch
opens.  The motor spins for less than a second to make the transition.
The limit switches have latches that prevent closure until the
engagement moves significantly away from the switch.  In my particular
unit, one of the limit switches does not reliably recover from the
latching mechanism which prevents further operation after either one
or two transfers.

This crude ASCII art diagram, best viewed with a mono-spaced (fixed
pitch) font, illustrates the operation of a transfer switch:

                          position   position
                            one        two
                           A---X      A   X
                                      |   |
                           Y---B      Y   B

If port A is a radio, port Y is an antenna, and ports X and B are the
input and output of an antenna tuner, then the transfer switch could
be used to bypass the antenna tuner.  Alternately, ports A and B could
be connected to 2 radios, and ports X and Y to 2 antennas.  Then the
transfer switch could swap the antennas between the radios.

Since the transfer switch was no good, I reassembled it and put it
back in the closest.  Some day, when I am brave, I'll disassemble the
10000 RPM to quarter-turn per second gear train to gain access to the
defective limit switch.  On to the next project.

SO2R operation on the low bands at my station works reasonably well
because all of the antennas are far apart and the receiving antennas
just don't pick up much RF.  But transmitting on 40 trashes reception
on 20, 15, and 10.  I previously made a 2nd harmonic trap coax stub,
but I never made it work.  So I dug out the stub, tweaked it with the
VNA, and tried to make it work again.  Despite the trap, transmitting
on 40 trashes a K3 receiving on 20.  After much disassembly and
reassembly of everything in the antenna switch and 40 meter vertical
looking for a noisy joint, I decided that the problem was really
fundamental overload.  So I dug out the W3NQN filter I made a couple
of years ago, and that prevents the overload when the filter is placed
between the K3 and amplifier.

The real problem is that I have only one filter and 2 radios and no
switching for filters.  Making another filter is no big deal, but the
filter must be switched along with the band the radio operates.  A big
pile of relays could make a stack of filters track the radio, but
housing the big pile of relays becomes an issue, Maybe pin diodes ...
For now, the 2nd radio is confined to either 20 or 40 meters.  On to
the next project.

Oops, no time left now.  A quick shack cleaning and a couple of
domestic chores, and now its is show time.  Since I had relatively low
motivation, I decided to operate assisted and make my goal to maximize
multipliers.  I would run only when that was pleasing without any
attempt to reach some QSO target.  On the other hand, chasing
multipliers would allow me to see the 'maximum reach' of my station.
Ideally any multipliers I miss are ones that are unavailable to me.

Maximizing multipliers instead of maximizing score changes operating
strategy.  To maximize score, one runs as much as possible on the band
mostly likely to be productive of QSOs and multipliers, hunt
multipliers as thoroughly as possible, and change bands strategically.
To maximize multipliers, one just follows the multipliers around and
runs when productive.  In other words, less work.  SO2R becomes a tool
to avoid changing bands and no longer provides a means to hunt
multipliers while running.  I configured Writelog to show me spots
that originated in Ohio so I wouldn't chase stuff I would likely not
hear.  On the other hard, there could be stations that I could hear
with a beam that a spotting receiver on an omni-directional antenna
could not.

The first night was 'normal', after accounting for the declining solar
cycle and better than last year conditions.  15 was not open at the
start here and 20 closed very shortly after the state.  40 and 80 were
fairly quiet and had strong signals, although the most distant
stations were not evident.  160 had strong signals but very few
Europeans.  It was hard to tell if they were absent or inaudible or

The basic operating strategy was to work all multipliers, prioritizing
those that might go away soon.  It was hard to in gore easy to work
stations that were not new multipliers.  When most of the 'yellow' on
the Writelog band map was gone, I would work other stations or even
try CQing.  By about 0800Z, I had worked almost everyone I heard.
There were always more stations to work, but finding them took longer
and longer.

I took a 3 hour nap.  When I was awakened by the alarm, I stumbled
into the shack and looked at the Writelog band map for 40 and 80.  No
yellow!  I went back to bed for another hour.  When I got up again,
there was a little yellow so I stayed and operated.  It takes about an
hour to get thoroughly awake and past the 'I want to sleep' mood.

Hitting the high bands presents a small dilemma: every station is a
new multiplier.  Clicking and calling is fun when the multiplier is
somewhat unique, but a crowded band full of European stations is just
bulk contacts, regardless of multiplier status.  So I CQed for a while
to work the multitudes.  I stated on 20 and my rate was terrific,
signals we strong, and the 'packet bursts' were minimal.  After about
45 minutes though, I moved to 15 so I wouldn't miss the short opening.
I ran there for about 30 minutes then reverted to multiplier pouncing.

After a couple of band changes, it was time to deploy SO2R between 20
and 15.  Oops!  SO2R doesn't work.  My automated antenna switching
apparatus seems to deny one radio or the other a high-band antenna.  I
considered troubleshooting the problem immediately, but I decided that
since my goal was not rate oriented, that I could live with the
inconvenience of band switching and solve the problem later.

After several hours of pouncing, including a very brief and weak 10
meter opening during the 16Z hour, finding interesting prey became
fairly boring.  The 19 and 20Z hours were particularly slow.  By 21Z I
noticed that I could work stuff on 40, so I spent a couple of hours
there.  At 23Z I observed that 20 meters had propagation to Japan so I
stared tuning on 20.  I very much like old fashioned tuning for Far
East stations, probably because they are relatively rare and weak and
exotic.  Nevertheless, the JAs were very strong!  I even tried CQing
and had a short run.

At the half-way point, I had 100 multipliers on 20, 88 on 40, 78 on
80, and just 77 on 15 meters and not quite 1000 QSOs.  If I could
repeat my progress on the second day, the results would be quite
reasonable.  Or at least I could try to get over 100 multipliers on
each band 80 through 15.  Goal set!

After 0030Z 20 faded.  I spent the rest of the evening rotating
through 40, 80 and 160.  The second night 80 meters again looked like
20 on the band map.  I spent much time clicking through all of those
stations.  By 0800Z I had really run out of stations to work.  By
careful hunting or dedicated CQing, I could continue to make contacts,
but I was getting very sleepy.

After another a 4-hour nap that skipped the dead time observed the
morning before, I found several Pacific and Far East stations on 40
bringing the multiplier total over 100 with a couple to spare.  80
though had only 90 multipliers.

Sunday high bands repeated the same pattern as Saturday except I spent
much more time running.  Also, I figured out how to use the 2nd VFO on
the run radio to set up a new multiplier pounce without interrupting
the run.  Then, when enough new multipliers accumulated on the other
band (20 versus 15), I would change bands, establish a new run, and
repeat the process.  This mode worked until 15 began to fade and 20
began to get noisy and and mushy (assume absorption was getting high).

The 19 and 20Z doldrums were even worse on Sunday except for a very
wimpy 10 meter opening around 1840Z.  By doldrums time, the 15 meter
multiplier count was 95 and firmly fixed.  I added 40 to the doldrums
rotation after 2000Z but there were relatively few new stations there,
except for Canadians and the occasional European.

During the doldrums, I noticed that there is another sub-window on
Writelog that shows packet spots as they come in.  By checking off
enough check-boxes on the GUI, I could get this information to show me
new multipliers rather than having to scroll the band maps.  Thus when
new African station would pop up on 20 or 15, I could get there
quickly and often beat the pileup.

At 2200Z I stated CQing on 20 hoping for Asian stations and by 2230Z I
got a steady run established.  The JAs were fairly strong and and kept
coming at a decent pace.  At this point I was using both tribanders
pointed the same direction.  I will have to think about whether
diversity reception made the JAs sound louder or if 20 meters was just
in particularly good shape, but I had one of my best JA runs ever.
Maybe a 100-foot high tribander just works better than a 60-foot
tribander.  My run included a couple of BYs, a YB, several VE7s and
numerous K6s.

Editorial: I think there should be no 0-point contacts.  Yes, it is a
DX contest, so DX contacts should be preferred over domestic contacts.
But assigning some small point value, like 0.1 points, to domestic
callers would help alleviate some of my animosity.  Some of these
run-interrupters are potentially new contesters who should be
encouraged rather than ignored or disparaged.  Alternately, just
automatically assign the domestic multiplier so that 0-point QSOs are

I also worked 1 maritime mobile station.  He gave me zone 7 and the
rules say /MM calls are also worth 0-points.  Why discourage anyone?
Why not apply the same 2 or 3 points to a guy on a boat?  Is this a
safety issue?  </Editorial>

As my 20-meter JA run was beginning to slow down, the Writelog spot
window showed C6AGU on 160.  The rate window said a new multiplier was
worth 3 minutes, and the that number was creeping up slowly as the run
declined.  My daughter was impatiently asking when the contest was
going to end (She was fondly recalling the doldrums when I had a bit
of time to talk to her.)  Also, my bladder was hurting.  Finishing the
contest with a new multiplier would be very satisfying.  Tactical
decision needed!  At 2340, A44A popped up as a multiplier on 80.
Would a second multiplier happen?

So at 2340Z I abandoned the 20 meter run, and set up the QSO with C6
on the 160 receiving device and called when I heard a rather loud
C6AGU.  Immediately, someone said 'up'.  So I rotated the transmitter
knob 1 kHz and tried again a couple of times.  Then the C6 said down!
Twiddled the knob again and finally made the QSO.  I was glad I
allowed 20 minutes for this contact!  Since I had 10 more minutes I
called TM1A who came right back.  Still 8 minutes left!  Why not go
for the A4 on 80?

I changed bands, which was a 15-second thing by this point.  The A4
was quite readable and the pile-up was basically 3 guys at a time.  My
daughter came in again to ask for dinner so I told her I would cook as
soon as I made this contact.  When she asked how long, I said 10
minutes tops!  I was cooking her dinner by 2357Z.

Epilogue: Finishing the contest by terminating a JA run to work 3 new
multipliers was a blast.  I missed my goal on 15 by 5 multipliers and
missed on 80 by 9 multipliers.  I beat my goal on 40 by 3 and 20 by
plenty.  But I've never had this many multipliers on 80 before.
Missed zones 22, 23, 26, and 36.

DX worked: 3B9, 3V, 4O, 4X, 5B, 5H, 5X, 6W, 6Y, 7Q, 7X, 8P, 8R, 9A,
9G, 9H, 9K, 9M6, 9Y, A3, A4, A6, BV, BY, C6, CE, CE9, CM, CN, CT, CT3,
CU, CX, DL, E7, EA, EA6, EA8, EA9, EI, ER, ES, EU, F, FG, FJ, FM, FY,
G, GI, GJ, GM, GU, GW, HA, HB, HB0, HC, HH, HI, HK, HK0/a, HP, HR, I,
IS, *IT9, JA, K, KH0, KH2, KH6, KL, KP2, KP4, LA, LU, LX, LY, LZ, OA,
OE, OH, OH0, OK, OM, ON, OX, OZ, P4, PA, PJ2, PJ4, PY, PY0F, PZ, S5,
S7, SM, SP, SU, SV, SV9, TF, TG, TI, TK, TU, TZ, UA, UA2, UA9, UN, UR,
V2, V3, V4, V6, V7, VE, VK, VP5, VP9, XE, YB, YL, YN, YO, YU, YV, Z3,
ZA, ZB, ZF, ZL, ZP, and ZS for a total of 137 entities.

Worked TI7W, 6Y0W, PJ4Y, P40L, 9Y4/VE3EY, PJ2T, LT1F, and PX2A on six
bands.  I tried to get P40W on 6 bands but missed on 80.  Every time I
visited the 80-meter spot, they were gone.

Equipment: K3/100, P3, Alpha 8410; K3/10, Hardrock-50, ETO 91B
(thanks, Jeff); high bands: X7 at 101 feet and X7 at 61 feet; low
bands: full sized 40 vertical over 32 radials, full sized 80 vertical
over 60 radials, 65-160 "tee" over 75 radials.  Homebrew SO2R and
audio routing stuff.  Homebrew W3NQN filters for 20 and 40.  Ancient
Writelog 11.

I didn't put up a 40 meter beam this year as I had planned last year.
I dug the old Cushcraft Discoverer-72 out of the barn 3 times to begin
assembling it.  The first time, the missing boom to mast clamp was a
set-back.  The clamp also joins the boom segments.  The last couple of
times the antique linear loading put me off.  If I am to have a short
beam, I would rather have the higher efficiency "top" loading.  I
would much rather have a 3-element OWA design which seems like it
would not be much longer than a 2-element shorty, although a lot
heavier.  I thought the vertical worked pretty well anyway.

Posted using 3830 Score Submittal Forms at: http://www.3830scores.com/

More information about the 3830 mailing list