AW: Serge=QUALITY

From Haible_Juergen#Tel2743
Sent Tue, Apr 25th 1995, 02:51

>If you detect a certain bias towards Serge in my posts, you're right.  On
>my recent visit to CA, I was lucky enough to have breakfast w/ Mr.Rex
>Probe, the meticulous maker of Serge Synthesizers, and fellow Analogue
>Heaven listmember.  He was most generous w/ information about his products,
                                ===========================================

Now, I'll take this mail as a reason for another attempt to ask my question.
Some months ago I asked something like "What kind of high quality components
do they use to end up at these steep prices?".
Now, as You might have seen from my comments about the Cheapo-Doepfer,
this was never meant as a flame, but it was (and still is) a question of 
pure
interest. I am building and further develloping my own non-commercial
just-for-personal-fun synthesizer (of course *always* waiting for a whealthy
sponsor to take it onto a commercial basis (;->), but not seriously), and
I really like to know about these HIGH QUALITY components Serge uses.
I have some drawings of old Serge modules which are said to be very
different from later ones. *These* were very cheap indeed and have lead to
a lot of confusion in various publications. (I think it was the German 
"Keys"
mag that spoke of a "very unrelyable and useless Serge modular standing
around somewhere and collecting dust" in a recent issue, but "Keys"
and analogue is a different story: They also said "Forget the Prophet5, here
comes the Nord Lead" in a recent editorial - shiver!)

Now, Rex or Knox or someone else, please tell me as much as You can
(without violating proprietary issues of course) what components You use.
Especially VCA's and exponentiators would be of great interest,
'cause data are quite comparable in that case.

Another example:
If You build things like (linear) frequency shifters, how good is the
accuracy of Your Hilbert-transformation filters? I know some publications
of such filters which have a few degrees phase error over speach frequency
range (3kHz). They can do that with 1% tolerance components.
OTAH, if You want to make an analogue Hilbert-transforation of 16 kHz
signals with low phase tolerance, 1% components wouldn't do.
So please tell me for that example: How close tolerated do You select
the components here?

This is just out of curiosity. Sometimes I just want to know where my own 
home
built stuff stands on the quality scale (;->). So please give me a 
reference,
a glance at the state-of-the-art!

Best Regards,

JH.

(Sorry for the public-only mail; I'd like to have made a CC to Serge at
least, but I cannot find the adress after I have changed the job)