[sdiy] New synth DIY site ?

[Magnus Danielson]

From: Neil Johnson <nej22 at hermes.cam.ac.uk>
Subject: Re: [sdiy] New synth DIY site ?
Date: Sun, 28 Apr 2002 21:33:16 +0100 (BST)

> Magnus,

Neil,

> > > compared to a single PIC/AVR and two DACs this is a BIG/Complex circuit.
> > You've totally missed the point. Finally we can LOOSE those f**king
> > MPUs!
> > Enjoy the simplicity of pure logic! ;O)

                                         ^
First, please note the smiley-symbol ----+

> Actually, one could view the *discrete* circuit as just a specialised
> processor, optimized down to the minimum required to achieve the desired
> functionality.
> 
> Put another way, an MPU is "pure logic" that steps through a user-defined
> state machine.  Nothing more, nothing less.

Please note that I do have a pretty good comprehention of logic in its
many fashions. I am not unfriendly with both lower and higher forms of
CPUs, as well as logic in itself. I happily hack stuff in VHDL for
FPGAs, but has also been part of the design team rather largish
ASIC. You will also find me hack C and C++ and odd times also
Assembler. Daytime I worry about bit-fiddeling in hardware and
protocol design in software, and how the hell I am going to convey
that to others.

Given that you now understand that, maybe now you can see both the
irony in my little statement as well as my childish love for the
"wrong" solutions.

> Its not really "pure logic" -- it is just a physical interpretation of
> logical rules (I get too much logic from the theory boys at the lab).
> And its not really digital either, just analogue that saturates.  DRAMs
> are better -- at least they have those nice little sense amps doing it
> properly, ie. amplifying signals and storing charge on dinky little
> capacitors.

Digital is just a simplification of analog. If you toss any digital
signal up on a scope, then you see it. As you push the limits for a
certain technology, then you learn the hard way how un-digitalness
digital signals really are. If you want to acheive low bit error rates
in "digital" transmission, you have to pay attention to zillions of
analog aspects. Oh the joy that it brings.

Few people actually understand the insight that Claude Shannon had
when he in 1948 explained the benefits of a Pulse Coded Modulation of
telephony signals. It took until the 70thies for a wider deployment to
emerge.

Cheers,
Magnus