Content-Transfer-Encoding: 8bit
Message-ID: <65AECDF1F89AD411900400508BFC869F0D756A@pdw-mail-1.dera.gov.uk>
From: "Talbot Andrew" <ACTALBOT@dera.gov.uk>
To: rsgb_lf_group@blacksheep.org
Subject: LF: 2 tone
Date: Fri, 26 Jan 2001 14:39:58 -0000
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1; format=flowed
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

>As a general point for computer-generated signals, if the soundcard or
DSP could be persuaded to generate the >output at a few kHz instead of
800Hz or whatever, it would be a simple matter to mix the signal
directly up to >.136kHz, with a fairly simple bandpass filter to give
good unwanted sideband rejection. 

G4GUO and myself have been thinking along the same lines for generation
of MSK.  While it is fine generating a constant envelope waveform at
audio from a Soundcard, getting this up to RF obviously requires
frequency translation.  There are several options :

1)  Use an HF rig and transverter.   Too complex, as few operators need
such a system.  Although anyone equipped for PSK31 on LF will have this
already.

2)  Dedicated upconverter from a high 'audio' frequency as Jim suggests.
The highest audio freq available is 15 - 18kHz using 44.1k sampling
rate, which will need a decent bandpass filter at 137 to reject the
image, but of a straightforward design / construction.   WIth realistic
pot-core inductors, a three pole design to cover 135 to 138 without
having to tune with the band.

3) Use an I/Q upconverter and let the soundcard generate the two
quadrature components from left and right channels.  This will give
something like 30dB isolation immediately, at the expense of a slightly
more complicated mixer and LO generation, followed by a much relaxed
filter spec.

4) Multiplication of the soundcard output, say by 9, from 15.2kHz; an
odd number  is preferable.    It will be straightforward to modify the
MSK modulation scheme to generate a lower modulation index, but
soundcard stability becomes of paramount importance as it directly
affects RF accuracy  in ppm terms.  Only applicable to constant envelope
schemes

Therefore 2) and 3), would seem to be the optimum solutions

>I think VK2ZTO has already carried the multi-tone idea to one of it's
logical conclusions, with seperate tones for >each different
alphabetical characters - see 'FDK' on his web pages
http://www.qsl.net/vk2zto. 

The first version wasPiccolo as designed in the 1950s by the UK Foreign
Office for diplomatic comms.  32 tones corresponding to the 32 states of
the 5 bit RTTY code transmitted at 100ms per tone.  Various amateur
reincarnations appeared over the years, with the number of tones ranging
from six to 32 with assorted coding schemes.   A modern 12 tone varient
is often heard on 14MHz and has been around for the last year.   
All single tone schemes give the constant envelope advantage we want for
LF, not usually an issue at HF where linear transmitters are always
used.

Andy  G4JNT


-- 
The Information contained in this E-Mail and any subsequent correspondence
is private and is intended solely for the intended recipient(s).
For those other than the recipient any disclosure, copying, distribution, 
or any action taken or omitted to be taken in reliance on such information is
prohibited and may be unlawful.