"Arny Krueger" wrote in message
...
"Serge Auckland" wrote in
message

I rarely use the sound card as it is too difficult to use
. Firstly you need an attenuator to reduce the incoming
level, as few soundcards take more than a couple of volts
of input before clipping themselves, idealy

A 5K 2 watt high quality potentiometer suffices.

you need a
millivoltmeter in parallel with the sound card input to
make sure the level stays when it should.

You mean a voltmeter across the UUT output. We often used those with the
old nulling-type analyzers for one reason or the other.

No, I mean across the sound card input, so I know what's going into it to
avoid it clipping. In practice, that may also be across the DUT (or UUT if
you prefer) output, but what's important to me when using a soundcard is
that I don't clip its input.

Then, with software, I never really know what it's measuring

That's your fault - a good experimentalist should be able to determine
that with a few real world measurements. On a bad day, analyze some
signals generated for the purpose. Generating complex tones is very easy
these days.

Agreed, if I could be bothered, but as I have two null-type meters
available, I really can't be fagged.

unless you have a suitable sound card and sample
at 192kHz, you can't measure over the 100kHz bandwidth
that most THD meters manage.

Virtually every sound card that I'd consider to be an alternative to test
equipment samples that high.

My current laptop's internal sound card offers 192k sampling, and it
actually works! However, I prefer to use my Digigram card for anything
serious as its noise performance is much better, but it samples only up to
48k.


I use the software test set
for analysing WAV files. For example, it's easier for me
to take my laptop to the hi-fi than test gear, so if I'm
making any measurements to my room or system, I will
record the DUT as a WAV and analyse it later.

Agreed - it is easy to capture data in the field, and analyze it in detail
later on.

If I were still a practicing engineer I would probably get with modernity
and have PC based test tools, but as now my engineering is for personal
pleasure only, I have a set of old-fashioned instruments that are good
enough for the purpose of hobbying. If noise/distortion etc is below what I
can measure, I go and worry about other things.

S.

--
http://audiopages.googlepages.com