Most U.S. operation thus far under Experimental licenses in the 137 kHz band has been one-way,
patterned on earlier work done under Part 15 in the 160-190 kHz band. Since all of us have
ham tickets, there has been some natural pressure to start making 2-way contacts. At least one other
QSO took place prior to this date, between WD2XES and WD2XGJ, with straight-key CW over a distance of 25 miles, which
was no great challenge. A short description of that event may be found here.
WD2XDW is in Bartlesville, OK (EM26AQ), and is run by Laurence Howell, KL1X. His QTH is 1340 miles from
my WD2XES in Holden, MA (FN42ch). We completed a 2 hour, 15 minute WOLF QSO on October 29, 2005 between 0015 and 0230 UTC.
We had pre-arranged the frequency of 137.422 kHz, which is in a small gap between sidebands of the Loran-C
signals on 100 kHz in North America. The hope was that we could attract some listeners (viewers, actually)
from around the U.S. and Canada. After some pre-QSO emails, we avoided Internet chatter during the actual contact.
Long distance QSO's at this frequency are helped by weak signal techniques which lower the data rate.
CW can be sent with dots of 30 seconds or more in length, but the time to complete a 2-way contact can be
considerable, especially with signal fading. "WOLF" is a binary phase-shift keyed mode developed by Stewart Nelson, KK7KA in 2001.
It is highly error-corrected, and gives usable copy at signal levels where
CW at a 60 second per dot rate can just be copied. It sends a group of 15 characters that
are heavily encoded and error corrected into a package of 960 bits. The phase of
the transmitted BPSK signal is shifted by 180 degrees at a 10 bit/sec rate, giving one full
transmission in 96 seconds. The decoding software builds up copy over successive 96 second
intervals, enhancing the weak signal capability. The mode requires high frequency and phase stability at both ends, and
over the path, and was specifically designed for low-frequency work. Stewart's original program was command-line
based, and required the receiving operator to record and subsequently process large .wav files.
The current software was developed by Wolf Buscher, DL4YHF, and runs with a GUI screen under
Windows. Incoming audio from a sound card is processed in real-time. The WOLF GUI
program is available at DL4YHF's web site.
The QSO
The contact began at 0015 UTC on 29 October 2005 with WD2XDW sending WD2XES WD2XDW N for 15 minutes. The constraints
of the 15 character string will become obvious as you read on!
The "N" is a signal report in the following format:
N = Nothing received
T = Signal present but no copy
M = Partial copy
O = Complete copy starting at line #
R = Received or acknowleged
WD2XES replied from 0030 to 0045 with XDW XES RO 384, indicating that the signal was copied completely
starting at WOLF line 384. The line numbers are measured in seconds since the program was put into receive mode, meaning that it took
6 minutes, 24 seconds to decode Laurence's signal in this case. As you can see from the QSO screen dump at the end of this article,
successive lines tend to repeat that decoded message unless the signal falls apart.
From 0045 to 0100, WD2XDW replied with XES XDW RO 0096, indicating copy on line 96. However, a deep fade prevented
copy at WD2XES. The signal was just building at the end of the 15 minute period, but I ran out of time before having to make a response.
So from 0100 to 0115, WD2XES sent XDW XES N SORRY. Fortunately, that was the last big fade during the QSO.
WD2XDW then repeated his 0045 transmission from 0115 to 0130 with XES XDW RO 0096. I received this easily, and the rest
of the exchanges require little explanation:
0130-0145 WD2XES sent RO 384 32DEG HR.
0145-0200 WD2XDW sent RO288 51F HR VA.
0200-0215 WD2XES sent 192 TU73 WD2XES.
The contact concluded with WD2XDW sending TU2 WD2XDW EM26 from 0215 to 0230 UTC.
Equipment
WD2XES was running 200 watts output into a large vertical loop antenna
supported by trees. The transmitter PA was an unmodified Hafler P3000 audio amplifier feeding a step-up
transformer and low-pass filter. The SSB phasing exciter is homebrew, and uses a dual DDS
circuit fed by the disciplined 10 MHz output of a GPS receiver. The WOLF GUI program was running on a separate computer from that used for
receiving, with the sound card output going to the SSB exciter. I used the large transmit loop as a receiving antenna, connected directly
to an ICOM R75 receiver(no external preamp).
WD2XDW was running 1 KW into another vertical loop also supported by
trees. For modes requiring a sound card source, Laurence uses a TS-870 transceiver set to 5.137 MHz with a homebrew transverter to get the transmit
signal down to 137 kHz. An external GPS-locked source provides external L.O.'s for the 870 and the transverter. His PA is a converted DECCA
rig. On receive, he uses the TS-870 directly on 137 kHz without the transverter. A description of the WD2XDW setup may be found at the
KL1X website.
Thanks
Besides thanking each other, KL1X/WD2XDW and W1TAG/WD2XES would like to thank Stewart Nelson, KK7KA, and Wolf Buscher, DL4YHF, for their work
in developing and improving the WOLF software. We also thank a number of people who copied all or part of the QSO, including W1VD, W3NF, W4DEX,
VE2IQ, and Hartmut Wolff (who copied part of the exchange from Germany).
How It Looked
I left the WD2XES receiving setup running during my transmit periods, so here is a screen dump of what was seen during the whole QSO: