AW: output level of Buchla 265

[Haible Juergen]

	> does anybody know the (nominal or measured) output level
	> of the 3 noise outputs of Buchla 265 ?
	>Thanks,
	>JH.

And so I've innocently asked this question before I was reading the latest
digest ...
I wouldn't, if I'd read the digest before:

>From: Mike Peake <peake at pacificnet.net>
>If some manufacturer ever did copy the SOU, they'd have to ask Don
>for permission first or they'd end up getting some serious flame from
>the list. It would be like Don Martin all over again. Most unwelcome.

So just to clarify, not *everything* is commercial nowadays, and
I was thinking about hobby / diy stuff in the first place, and of somebody
programming the structure on his favorite (real or virtual) modular.

The direction of my initial posting was twofold, discouraging people
who think they can emulate the behaviour of a vactrol-based Buchla
VCF, and pointing out that there are other modules (like the 265)
that could be easily reproduced with garden variety components
or patched up from standard modules despite the use of "exotic" circuits 
in the original.

As for my own activities, I've designed (on paper) and simulated a cheap, 
OTA-based, version of the 265 some months ago, but I had forgotten
about it until this thread appeared on AH. Last night I started to build
it on veroboard. What I intended to do is the same as with many other
diy projects of mine: Build it for fun and my own pleasure, publish the
(new) circuit to share it among the diy community with a copyright remark
for my circuit, and with proper reference to the original design of Don
Buchla.

>No offense meant to you Juergen.

Yes, and *please* do not offend to other list members either !!

>(Initials) PS: I sent you the SOU schemo to prove that Buchla modules
>were far superior to your notion that slow random was only filtered
>noise. Let it remain an academic lesson in module design, not an
>invitation to piracy.
[...]
>I have no idea or concern what ><synth1 at airmail.net> writes.
>Faster Pussycat, Killfile Killfile.

This is not what I like to read in a thread about technological questions,
and while I'm not personally offended, I think the tone of that is quite
offending in general. I mean, *any* sensitive ear would be offended.

There has been no talk between Paul and myself about the SOU. He
has a noise module already - it's just a fun project for me. 

After discussions like that I think about that idea of publishing
and giving credit once again. You know what will happen: There
are other manufacturers around who don't have the high moral
standards of PS and who would cannibalize a simplified circuit
for their own products without giving credit to Don Buchla
nor to the modern re-engineering.
You know I've been quiet about the behaviour of a certain 
other manufacturer for a long time now. But hey, if you want
to bash someone for shameless piracy, then bash the one
who deserves it (you know who I mean), and not PS
who - for all I know - pays royalities and checks the legal
status of the origin as well.

So be it - when I have finished that single unit of a modern
implementation of the 265, I'll better not post it.
Or maybe I *should* post it, so you can see *who* will have
it announced as another addition to his ever growing module
line soon. 
Think about it - you're beating the wrong people.

============================================
technical mode on:
============================================

For all who are in the no-business / no-flames mode,
and who have followed up to that point, here's some
technical information (as I see it) - maybe it's of some
use for somebody:

You can mostly emulate the basic behaviour of the SOU with standard
modules (MOTM or not), and I can confirm most of what PS
described (He obviously described a later model - the 266,
to answer Kevin's question).

I see the following points that are important for the uniqueness
of the 265 SOU:

(1) Fluctuating voltages that are generated by a S&H with VC clock,
and a tracking subaudio filter. You raise the clock rate, the filter
will open. Original has 2pole Sallen & Key filter with low Q factor,
so any VCF that will go to subaudio range, or two VCLAG processors
in exponential mode will do, even one VCLAG would approximate
the effect. Important: The VCLAG must be in *exponential* mode;
a linear VCLAG would produce ramps with sharp corners, and not a
smooth fluctuating voltage.
If you have a MOTM system, set the response to exponential, and
make sure that the LAG time is *decreased* as the clock frequency
is increased. If you have two LAG processors, connect them in series,
the emulation should be closer.

(2) The probablity knob of the "normal" S&H.
Just make a mix (crossfade) of the S&H output and your noise source,
and feed that mix into the S&H input. The original circuit does nothing
else.
My guess is (but I don't know) that the EMS random module (forgot
the exact name) does a similar trick, so I don't know who "invented" it.
In any case, it's the easiest thing to emulate with an ordinary S&H and
a simple potentiometer.

(3) The 3 differently coloured noise outputs.
Every manufacturer apparently has a different blend of noise colours.
In case of the 265, the brightest noise is produced by transistor zener,
a discrete transistor amplifier, and an opamp stage with the non-ideal
opamp behaviour (limited bandwidth) being used as a feature.
Don't substitute it with a fast audio type or you'll get vast oscillations
and a very different coulour of noise. (I'll probably substitute it, but
that
means a non-trivial redesign of that stage)
The other 2 coloured noise outputs are derived from that by linear
filtering.
The amplitude of the zener noise is quite critical, too. Maybe that's the
reason why Buchla later switched to a digital noise source (probably with
the
side effects PS described).

The reason why I asked about the noise level in the original is related
to tis question as well. The way I have it on the breadboard now is
*nice*, but rather "dark" in comparison with the EMS VCS-3 noise for
instance.
There is some clipping in the darkest output - gives a nice "thunder" effect
on its own, but I doubt that (or: I'm curious if) that's the idea of the
original.
I think I rather have to reduce the gain of the discrete preamp.

Long mail that was, two quite different topics included.
I love technical discussion, 
and I hate it when people are flamed who deserve better.

JH.