I worked in a high-end audio repair shop for a few years, and we had
one of these. It was big, expensive, and the only real purpose I ever
saw it accomplish was to allay the fears of paranoid customers who were
afraid that their equipment had mysteriously fallen out of design
specs.
We had other tools for aligning FM RF's, IF's, detectors and such (are
such concepts dinosaurs nowadays?). AF power section bias adjustments
we performed with current measurements.
I would view software distortion analyzers skeptically. First, in order to
get the sound into your computer, it has to be sampled by the sound card.
Sampling, by its nature,can't yield a 100% true representation of the
original waveform. The software has to interpolate the waveform between
samples. The higher the sample rate, the more accurate the interpolated
waveform will be, but most sound cards are limited to 44 KHz. That may be
good enough to listen to, but I would view it to be inadequate when trying
to do distortion analysis on a high frequency audio signal.
The software only has the samples of the original to do analysis on,
therefore, it can't give very accurate results over a wide range of
frequencies.
If I absolutely needed a distortion analyzer, I would shop Ebay or a hamfest
for an HP 334 or similar. Those instruments are superbly constructed, and
don't suffer from the sampling problem. You can't go wrong with one of
those.
--
Dave M
MasonDG44 at comcast dot net (Just substitute the appropriate characters in
the address)
Never take a laxative and a sleeping pill at the same time!!
Leonard
"powerdoc" <dkin...@mindspring.com> wrote in message
news:1138809307.7...@g14g2000cwa.googlegroups.com...
What do you hope to see that you won't see on a 'scope?
Amounts like 0.05% on a sine wave.
RE: doing it in software. FWIW (probably nuthin' at this point), I
agree with previous posters. By the time you get a sine
wave through a PC, it's probably already got 2,3 or more % distortion
on it.
I remember the analyzer we had, it may well have been an HP. I think,
on its highest setting, it was something like 0.1% full scale, so if
you trusted the linearity of the meter you could measure down to 0.01%
At those levels, everything was very sensitive. All connectors had to
be clean and tight or you'd get erratic/false readings. You could tap
a shielded cable with your finger and make the meter jump.
Good luck!