[Fwd: Soundcards]

[Andre' Kesteloot]

vernall wrote:

> James Moritz wrote:
> snip snip
> > Since soundcards and software like Spectrogram have many other
> > uses to radio amateurs apart from reading QRSS, it would be
> > interesting to know if any data exists about the dynamic range &
> > spurious responses of the soundcard/software combination - does
> > anyone have any information?
>
> Also bear in mind the performance of the audio stages in the radio
> receiver feeding the sound card.  Some time back I was testing IF
> filters in a receiver, viewing demodulated output for feeding wideband
> noise to the RF input, and the sound card "spectrum analyser" gave a
> false indication compared to doing the test at RF.  On investigating,
> the explanation was that harmomic distortion in the audio stages was
> where the "apparent funny response" was arising.  As the sound card was
> intended for "hi fi stereo" it was better performing for distortion and
> intermod than my receiver audio lineup.  Also there is unlikely to be no
> hum from the receiver PSU being impressed on the audio output.  I came
> to a conclusion that the sound card was good at indicating filter
> PASSBANDS, but suspect for indicating steep skirts and stop bands.
> Audio noise and intermod can easily fill in deep notches on an IF
> response as indicated on the PC display.
>
> I have never been happy about the "lines" on the very low frequency end
> of a spectrogram display with no audio fed to a sound card (irrespective
> of whether the audio is open, shorted or 600 ohms shielded load).  This
> is some sort of low frequency switching breakthrough in the sound card
> itself, and varies between specific cards (and probably with sampling
> rate on the card?).  As my PC was under warranty I did not fiddle about
> with it, but in due course I may try added shielding and filtering of
> power supply rails.  Some PCs have the "sound card" implemented on the
> motherboard, so fiddling would be more difficult with those.
>
> Sound card application programmes are generally good, but be cautious in
> the interpretation of audio domain displays when attempting to use it
> for RF testing.
>
> 73, Bob ZL2CA