[sdiy] Voltage references in VCO

[Gene Stopp]

I think there's a place for both ultra-stable and less-than-perfect.

A few months ago I was approached by three different people to fix their
Minimoogs. Through some kind of confluence of karma, I had them all at the
same time in my garage along with my two. One very old s/n (metal panel,
clear wheels, all transistor), two middle s/n's (3046 VCO's, white wheels)
and two late s/n's (ua726 VCO's, ribbed white wheels). After they were all
cleaned and tweaked, I set them all up in a row for a couple reasons: a) it
looked SOOO COOOL! and b) I decided to compare their sounds. Kind of a
synthesis myth-busters. I used headphones so I could hear the nuances and
besides, Minimoogs have great headphone amps.

All of these machines have a very slight warble in VCO pitch, compared to
something like a test tone generator. A truly pure perfect tone, one that
would be an infinitely thin lone spike on a spectral plot, would gradually
desensitize your ear and fade into the background. These Minimoogs varied
from one to the next on the amount of pitch warble they had. One of the
middle s/n's was the worst, where a high note (like 2-3 khz) sounded like it
was being modulated by another VCO at a very small mod depth, as if the
depth pot were cracked open just the tiniest perceptable amount. It was
barely even noticable without headphones, and if you joined in the other
VCO's and set the filter up for a nice sweep, it just sounded like a fat old
Minimoog. Part of the charm.

The purest tone of all five was the ancient all-transistor Mini. I would
move the headphones over to it and it had this kind of silky pureness where
the other ones were a little ruder. This of course was also the least able
to hold its tune over time, with the 726-based Minis having the best
long-term stability.

The thing I'm getting at is that pure tones have their own character. I once
was using a big bench generator (HP3325 I think) and I was building modules
and testing them with headphones. This generator made very pure tones,
laboratory precise. I was testing a 4-pole lowpass (EN quad 3080 integrator)
with a sawtooth from the 3325. I remember as I turned the resonance up a
little and swept the filter, I could pick out the harmonics of the sawtooth
one at a time and isolate them as pure tones. It was quite a bit different
than doing the same thing on a commercial synthesizer. On a normal synth you
start high and sweep the filter down, you get "weeeeaaaaaaaoooooooouuuuuw"
but with this frequency synthesizer through a filter I got
puretone-puretone-puretone-puretone all the way down the harmonic series. I
suspect that jitter in the oscillator pitch "smears" the tone of the filter
resonance peak on most synthesizers.

So if you could have multiple perfect oscillators, and you detuned them
slightly, would the cancelling and reinforcing harmonics dance around with a
brighter sound than we're used to? Would you hear hetrodyning things that
are smeared and lost in regular synthesizers? Anybody got two HP3325's?
Maybe a perfect VCO could be useful for sparkle and shimmer, and
not-so-perfect VCO's are good for that good old thick sound.

- Gene


-----Original Message-----
From: owner-synth-diy at dropmix.xs4all.nl
[mailto:owner-synth-diy at dropmix.xs4all.nl]On Behalf Of René Schmitz
Sent: Friday, April 14, 2006 6:49 AM
To: Richard Wentk
Cc: 'Synth DIY List'
Subject: Re: [sdiy] Voltage references in VCO


Hi Richard,

Richard Wentk wrote:

> Ultra-stable VCOs are not good for music making.
> 
> Slight exaggeration for effect, but as a principle - not entirely untrue.

Exactly.

> For musicality you want something with a little life in it, not a test 
> tone generator. Too much precision and the sound can easily become
sterile.

I said that before. I want something which has enough life, but still 
doesn't wander all over the place. So it might be noticeable but has to 
be not annoying.

Cheers,
  René