DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
  s=s1024; d=btopenworld.com;
  h=Received:X-Yahoo-SMTP:X-YMail-OSG:X-Yahoo-Newman-Property:Message-ID:From:To:References:In-Reply-To:Subject:Date:MIME-Version:Content-Type:Content-Transfer-Encoding:X-Priority:X-MSMail-Priority:X-Mailer:X-MimeOLE;
  b=ZP4ZVMNgVvhoWdoGLWCD0fQ3AnnM88bfURrK6beEA8QGmXDA9rXOeaJGrUBu/i+iTM8VFGXTux3p5h7Y59HgIWo/VnlLp+G5fmk8kJ5rq+8RdCM21w7Hi1Z5Et6FNE3bzBC8ylTorLppH6gbNRc04w3Nid1UX/w5cGaN/hXh7TY=  ;
Message-ID: <4E3B8A72F432464293A0DA58BA5D170B@JimPC>
From: "James Moritz" <james.moritz@btopenworld.com>
To: <rsgb_lf_group@blacksheep.org>
References: <22E5FF108AB94135883D0A402EC9B85C@JimPC> <000e01ca4340$5ba6b3a0$0900a8c0@lark>
In-Reply-To: <000e01ca4340$5ba6b3a0$0900a8c0@lark>
Date: Fri, 2 Oct 2009 20:24:47 +0100
MIME-Version: 1.0
DomainKey-Status: good (testing) 
Subject: LF: Re: Re: RX and Sound Card
Content-Type: text/plain;
	format=flowed;
	charset="iso-8859-1";
	reply-type=original
Content-Transfer-Encoding: 7bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org


Dear Steve, Alan, Klaus, LF Group,

Thanks for the comments. This calibration method is a bit fiddly in terms of 
doing the calculations, but I think it is simpler than W1TAG's Loran 
technique as far as making the frequency measurements goes. Also it gives 
you the receiver frequency error as well as the sound card sampling rate. I 
think the main difficulty with the Loran method is actually selecting and 
identifying  a suitable spectral line to measure the frequency of - 
especially in the UK where the received Loran signal is "dual rated" with 
interleaved spectral components from at least 2 different GRIs in the 
demodulated signal. A lot of promising looking peaks in the display are 
actually two peaks very close together.

As with the Loran technique, it has the advantage that, as far as hardware 
is concerned, it is "something for nothing", with no additional equipment 
required (something always appealing to amateurs ;-)). If you are prepared 
to make some simple calibration hardware, the calibration process is much 
more straightforward and potentially more accurate too. I have used a 10MHz 
xtal oscillator divided down to 1kHz (2 x 'HC390) to provide an audio 
calibration signal. The harmonics of the 1kHz signal can be compared with 
198kHz etc. using the RX and soundcard with spectrogram software - in this 
way, the calibrator frequency can easily be set to well within 1ppm. In my 
case, I used a salvaged TCXO module for the 10MHz, which is stable enough to 
provide in-band 1kHz marker pips for checking the receiver calibration too.

I don't know why the WOLF frequency calibration utility only works within 
+/- 1Hz. The trouble is, if the offset is outside that range, it gives you a 
sensible-looking value that is actually wrong. So it is important to check 
the audio frequency on the WOLF spectrum display to make sure it is within 
the right limits.

Cheers, Jim Moritz
73 de M0BMU