[sdiy] Interesting f vs. t graphs of pitch instability

[Andre Majorel]

On 2010-02-20 10:37 -0800, David G. Dixon wrote:

> > Finally fixed it. It wasn't VR4, it was the CA3140 (IC5).
> > Replacing it made the wavering go away.
> 
> Andre, I recall our discussion about VCOs vs DCOs and how you
> thought the former might sound better than the latter because of
> slight frequency noise.  It would be interesting to see whether
> there is any measureable noise now that the unit is fixed.

Thought you might catch on that. :-) The graph you're looking for
is already there, it's number 3 (ms20_pitch_instab_ext_volt.flac)
in http://www.teaser.fr/~amajorel/ms20/files/waver1.png

But I suspect that any short term deviations lesser than 6 Hz tell
you more about my frequency counter than they do about the signal.
See my answer to Magnus on 2010-02-20 for the reason why.

To measure frequency jitter of that magnitude, I think you'd need
a much higher sample-rate / signal-frequency ratio and/or a much
better frequency counter.

It is interesting, however, that there is no frequency noise to
speak of on the mathematically generated sawtooth (bottom right
graph). So perhaps there is a way...

-- 
André Majorel http://www.teaser.fr/~amajorel/