Jitter in vcos- Arp Odyssey related

[jean-charles maillet]

>Relating to the recent "jitter in vcos" thread, I recently made a small
>discovery on an Arp Odyssey-
>
>In the later Arp Odyssey models, the noise source uses a zener diode for a
>smoother quality of noise. The original designs used a "noise" transistor .
>What is interesting is that the transistor style noise source caused
>miniscule bumps and spikes on the power supply lines. The zeners don't and
>thus the vcos behave in a more stable manner. After modifying an older style
>unit, the pitch stability was noticeably improved.
>
>-Kevin

I wonder if overly high impurity levels in the silicon would be responsible
for spurt-like (partial) avalanche effects in a standard older device
connected as a zener, i.e. flurting with reverse breakdown would vary as
well giving it a random chaos feel in relation to time and our adjusted
senses ... natural analogs like rain or pop-corn or vynil have similar
musical functions on us ... unlike the quantization noise inherent of
digital systems.

If so, the repetition rate of the surges responsible for creating
intermittent supply dips would vary randomly.  This would vary
inconsistently from one device to the next, and therefore from one synth to
the next depending on circuit sensitivity and amplitude of manifestation.
I can see, for fiscal reasons, a push to shy away from chaotic variance at
this level.

Zener diodes have fairly low circuit impedance (in relation to loads) once
in avalanche mode (compared to a zener biased transistor), and therefore
are more resistant to loading effects from the oscillator (switch) it is
driving.  This spells out stability and overall operations come closer to
idealised targets.

Your observation affords much because it points to a zener device playing
the role of an extreme in the "single device" league (short of going the
buffered route which has huge drive and is accompanied by a totally stiff
and non-singing quality)  Whereas the Arp's transistor (?!) or VCS3's
transistors my lay further away from shore, i.e., the limit set by the
zener, where each display subtle to non-subtle large-signal and loading
effects which results in a VCO waveform that is characterised by "higher
personality", part of what makes a musical instrument stand out it seems.

Similarly, the classic Fender/Marshall Bassman/SuperLead circuits have such
types of "randomly chaotic moving parts" found in the coupling.  Through
this it  contributes to very definite tonal and feel quality which directlt
fall back on the large-signal transfer functions acting out their temporal
destiny. Anyway, this suggests that the switch over to a bone-fide zener
device, designed to operate intentionally in "controlled avalanche" without
spurtingly shorting out, would result in a more uniform sounding design.

JH's input :) confirms much of what you're saying and vice-versa, all this
could point towards a design direction that allows relatively crude designs
to sound more musical (randomly chaotic of a "certain" type) and maybe more
"natural" than other high-tech ones ... sorry for my ascii getting out of
line back there ...


can we access a gif of the circuits (pre/post) and the component list ?!

thanx
jc maillet

    "I'd like to spread mayonnaise and kao?pectate all over your body,
                and take you down to Hollywood boulevard ..."
                                                                  Fz