[sdiy] Unstability of oscillators and psychoacoustic qualities

[Magnus Danielson]

From: "jhaible" <jhaible at debitel.net>
Subject: Re: [sdiy] Unstability of oscillators and psychoacoustic qualities
Date: Sun, 22 Sep 2002 13:16:31 +0200

Hi Jürgen!

> > Now comes the question, if we have two VCOs playing the same notes, how
> will we
> > know they are distinct? If "coldness" would be due to synchronisation,
> then
> > we should at least be able to explain it in terms where some pair of
> overtones
> > is near each other, or? If one oscillator locks to another by
> synchronisation,
> > each time it synchronises we should have a jump in frequency, thus giving
> a
> > modulation. However, then "coldness" would be due to modulation. How do we
> > then handle a single VCO case where we can label it as "cold" or "warm"?
> 
> As long as the VCOs are beating there is some "warmth". If you tune them
> close enough so they will lock -> no more beating, "cold". (or "thin" ?)
> If a modulation kicks the VCOs out of lock periodically -> "warm" again.
> Different from free running, though.

Hmm... are we discussing a form of PLLing here?

I view a synched VCO as an oscillator being repeatedly and abruptly phase
modulated.

What you seems to describe looks to me like that of a PLL, which beats until
the frequency is close enought for the loop to be strong enought to
"pull it in" and actually move the center frequency of the oscillator and thus
reduce the modulations from the synching. You can form a PLL from basically
nothing once you have an oscillator as such. You, Jürgen, of all people here
should known that!

I never though about the phase locked loop aspect of beating in oscillators,
but you just put me on that track... thanks!

The question is how one best analyses this behavior in a real oscillator?

> > Does my oscillator sound "warm" or "cold" by itself?
> 
> I say "cold", as long as you listen to it "dry" on headphones. In practice,
> however, room ambience will make delayed versions of one VCO
> signal interact, and if the frequency is modulated, the phase of the
> delayed signal will wander against the phase of the direct signal,
> causing some "warmth".

A classic form of synchronous jitter analysis, delay the signal and compare
against itself! Doesn't work as well for wander at all times thought.
There are however some ultrasensitive methods for phasenoise measurements that
builds up on multiple delay-lines which is used mostly to measure phase noise
of oscillators above 1 GHz.

Video signals is still modulated to be measured with a delay of 90 lines using
an oscilloscope. A still-to-be-released wander instrument from Pendelum will be
able to measure it by other means... I guess Fluke will also sell it too.
I helped them out some... (like telling them that they should do it in the
first place).

> (You can get a remarkable amount of "warmth" from the most cold
> and digital sound source by running it thru a modulated delay line.
> BUT this needs _time_, and thus you loose "direct-ness". I think
> the big plus of analog is not the capability to be "just warm", but
> to be "warm" _and_ "direct" at the same time.)

Just a variable phase delay (not to be confused with a phase) is one method.
Phasers and flangers adds too, but they also mix up with variable notchfilters
which helps to mess the situation up a bit.

> > I may have an incorrect personal assumption that "warm" oscillators have
> some
> > sort of phase noise in them. This causes them to wander about the note
> > frequency.
> 
> Yes, of course! Phase noise and random frequency modulation are two
> aspects of the very same thing.

Indeed.

> > I also think that near equal frequency tones feel "warmer" than any
> > of those notes on their own.
> 
> Certainly.

It should be noted that this is also similar to having a single tone amplitude
modulated by half of the difference signal.

> > Do we design our oscillators with low phase noise as a matter of design
> > concern? Do we design them with _any_ phase noise concerns? Not what I
> have
> > ever heard off in this buissness.
> 
> Phase noise comes into a classic saw VCO core at two points:
> a) the comparator (or whatever reset-) threshold (affects phase directly)

I agree. Noise on any of the inputs to the comparator will cause phase
modulation.

A sync-signal is usually set to control the reset cycle, and it will those
cause a phase modulation by phase-jumping the oscillator into "0 degrees" of
phase.

I wonder if it is not the reset comparator that acts the phase detector when
the oscillator PLLs in. There is PLLs which operates by insering the reference
signal straight into the oscillator. That's how the color-burst in the start
of each TV line keeps the local crystal oscillator up in frequency and phase
lock.

> and
> b) the ordinary CV path that controls VCO frequency (d_phi = d_f * d_t)

Indeed. This is the path in for frequency modulation. This in form of linear,
CV scaled linear and exponential frequency modulation.

> I think (b) is the more important part.

Yes, but it would be interesting to see how much of the noise that could be
found to be from the comparator side.

> Some VCOs use noisy components and have that modulation built-in,
> some VCOs avoid noisy components and give the choice between
> stable and warm, by choosing external modulation.

Yes, but do people really *think* actively about phase noise when designing
oscillators for musical applications? Please present me a designer that does
this! ;O)

> We should also keep in mind that modulation does not necessarily mean
> noise only. (And the noise part does not mean unfiltered "white" noise.)

I agree. We discuss any form of modulation signal. It should also be recalled
that all noise is not Gaussian. It is wellknown that phasenoise needs better
claws of analysis than pure averaging since normal averaging will not converge.
This is where Allen deviation, Modified Allen deviation and other forms of
higher-degree analysis comes in.

> IMO the magic of a "warmth" in groups of VCOs lies in a mix of
> deterministic and random modulation.

It would be interesting to record two oscillators together and then separate
with the other one turned off (no power!) and mixed together afterwards.

Cheers,
Magnus