Message-ID: <001101c3f91c$82ba8c10$687a37c0@w2ksn>
From: "Stewart Nelson" <sn@scgroup.com>
To: "LowFER reflector" <lowfer@qth.net>,  rsgb_lf_group@blacksheep.org
Date: Sun, 22 Feb 2004 00:18:52 -0800
MIME-Version: 1.0
Subject: LF: feedback requested for new WOLF
Content-Type: text/plain; charset=iso-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
Sender: <owner-rsgb_lf_group@blacksheep.org>
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org

Hi all,

I am working on a new version of WOLF, which will feature frequency
and message timing synchronized to UTC, as well as real-time
transmission and reception via a sound card.  The raw data rate is
slowed to one bit per second, to reduce bandwidth requirements and
to improve tolerance to propagation delay uncertainties. 

There are many different ways to achieve synchronization, and I hope
to eventually implement a variety of transmission and reception
options.  However, adding a new method entails construction of a
suitable setup for testing, as well as substantial coding effort.  So
that the first releases will be usable by many operators, I'd like to
hear which of these modes would work at your station, with hardware
that you already have, or would be willing to build.  Also, please
suggest any other methods that you feel would be effective for your
station.

Reception modes:

R1. This would use LORAN "QRM" present in the received signal for both
    frequency and timing lock.  The obvious advantage is that no
    special hardware is needed.  However, it requires that the LORAN
    sidebands be strong enough to show visible lines when receiving
    QRSS30.  In addition, it is necessary to use a wide (SSB) filter,
    so it won't work if there is a strong interfering carrier that
    demands a narrow filter.  If the distance to your closest LORAN
    station is such that groundwave isn't always dominant, then this
    method will not be as accurate as one using GPS.

R2. The 1 pps signal from a GPS receiver is coupled into the LF
    receiver antenna input, and used for both frequency and timing
    lock.  This also requires use of an SSB filter.

R3. Your receiver LO is locked to GPS by external hardware means.  For
    example, you have an amateur transceiver that derives all
    frequencies from a single reference, and you feed that from your
    Z3801A.  Or, your Z3801A feeds a DDS or divider that provides the
    LO signal to your homebrew LF receiver.  Then, you feed a 1 pps
    signal into the other sound card channel, providing a sample rate
    and time reference.

Transmission modes:

T1. You generate a precise carrier frequency by any suitable means,
    feeding one input of an XOR gate.  WOLF would key a tone on or
    off, corresponding to the phase desired.  You would rectify and
    filter the tone signal, feeding the result to the other XOR input.
    The gate output then feeds the transmit output chain.  A 1 pps
    signal is used as a timing reference only.

Methods T2-T5 below all involve the use of an upconversion mixer,
combining the sound card output at e.g. 18 kHz with an LO to generate
the desired Tx frequency.  The sound card frequency would be
adjustable in fine steps for tuning, so the LO could be a simple fixed
divider from a suitable reference.  Sound card output would be
available on both channels in quadrature, to allow use of a simple
image reject mixer if desired.  It may also be possible to perform the
upconversion with an SSB transceiver, modified to allow LF
transmission.

T2. The Tx LO is precise.  A 1 pps signal provides a sample rate and
    timing reference.

T3. The Tx LO (in addition to driving the mixer) is divided by 10 and
    the result fed to one sound card input, with 1 pps on the other.
    WOLF would servo the audio output to compensate for LO errors.
    This setup needs manual switching or additional hardware to change
    from Tx to Rx.

T4. The Tx LO is generated by a DDS or divider driven from your sound
    card's time base oscillator.  A 1 pps signal provides sample rate
    and timing reference, and allows WOLF to compensate for LO
    errors.

T5. The Tx LO, or the final carrier output, is monitored with an LF
    receiver that is also fed with a 1 pps signal.  One needs to
    ensure that the Tx signal does not swamp the 1 pps.

In methods T6-T7 below, WOLF generates a submultiple of the desired
output, e.g. 18.5 kHz for Tx on 185 kHz.  There are various ways
to multiply the result.  A 1 pps signal is used to keep the output on
track.  These methods are likely to produce highly impure spectra, so
they are probably unsuitable for high power operation.

T6. The carrier is generated by a PLL.  For example, a 185 kHz LC VCO
    is divided by 10 and combined with the sound card output in a
    mixer.  The result is filtered and used to control the VCO.

T7. A frequency multiplier, or series of frequency multipliers,
    generates the desired carrier.

Except for T5, one could replace the 1 pps signal with an LF receiver
tuned to ~100 kHz.  This would permit transmission without a GPS
receiver, although it requires that your transmitter not interfere
with LORAN reception on 100 kHz.  In the case of T3, T4, T6, or T7, a
precise oscillator is not needed.  If you have a reasonably precise
frequency source, e.g. rubidium, but no GPS receiver, then you could
use LORAN as a timing reference with T1 or T2.  Please refer to the
LORAN locked transmit modes as T1A, etc.

As usual, I've left out lots of details, so any questions or comments
are welcome.

73,

Stewart KK7KA