From: "Klaus von der Heide" <v.d.heide@on-line.de>
To: rsgb_lf_group@blacksheep.org
Date: Fri, 02 Oct 2009 13:00:51 +0200
MIME-Version: 1.0
Message-ID: <4AC5F983.16585.906600@v.d.heide.on-line.de>
In-reply-to: <22E5FF108AB94135883D0A402EC9B85C@JimPC>
References: <22E5FF108AB94135883D0A402EC9B85C@JimPC>
Content-description: Mail message body
Subject: Re: LF: RX and Sound Card
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org




Dear Members of the Group,


since years I use a very simple method of measuring 
the output sample rate of the soundcard:

(1) Choose an LF frequency normal such as HBG, 
    MSF or DCF77

(2) Generate a sinewave on the soundcard output 
    with a harmonic very near to the frequency 
    normal. Use maximum volme.
    Example: For DCF77 on 77500 Hz generate 
    (77500 +- 1.5)/19 = 4078.868 or 4079.026 Hz 
    to get a signal 1.5 Hz apart from the normal 
    (if the samplerate has the nominal value).

(3) Use a diode or transistor to distort the sound 
    output and couple it appropriately to your LF 
    antenna such that the harmonic has similar 
    amplitude as the frequency normal in your 
    receiver.

(4) Change output frequency such that the frequency 
    normal lies symmetrically between both harmonics 
    (i.e. use both frequencies of the example in 
    sequence).

(5) Sample the SSB receiver output, compute the 
    spectrum, and identify the peaks. Usually the 
    generated peaks are far from what one expects. 
    So the audio frequencies have to be improved 
    iteratively. 

The accuracy only is limited by the time interval 
used for the measurements. With intervals of 10 
minutes the accuracy is about 0.001 Hz. That is 
enough to measure the drift of the samplerates.
 
I have some JPGs with measured spectra which will 
be sent on request.


73 de Klaus, DJ5HG