[sdiy] Is everything digital?

[Magnus Danielson]

From: Aaron Lanterman <lanterma at ece.gatech.edu>
Subject: Re: [sdiy] Is everything digital?
Date: Sat, 14 May 2005 12:39:34 -0400 (EDT)
Message-ID: <Pine.GSO.4.60.0505141237380.13090 at yamsrv1.ece.gatech.edu>

> 
> On Sat, 14 May 2005, Glen wrote:
> 
> > Scott, you are confusing digital with binary. It's entirely possible to
> > have digital electronics which operate with more than two states. This is
> > somewhat equivalent to counting in different number bases. Of course,
> > binary is by far the most common form of digital electronics, but I have
> > seen trinary circuitry. (No, I'm not referring to the "high impedance"
> 
> One of my computer engineering profs at Wash U told us that a while back 
> someone figured out a computer could work most efficiently (for some 
> measure of efficiency I can't remember) if it worked in base "e". Of 
> course e = 2.7 something isn't an integer, which makes things a bit 
> tricky. They figured 3 was closer to "e" than 2, hence interest in trinary 
> circuits for while, but they were eventually determined to be more trouble 
> than they were worth.

The signal to noise requirement for a trinary system is harder. Also, binary
can make use of distinct situation as "have" and "have not" where as trinary
system must also have a stable middle state and that can be hard to acheive as
well as you can not make use of all the other tricks normally being used.
No, trinary systems is much more difficult than binary.

Konrad Zuse made the first "pseudo-logarithmic" computers i.e. binary
floating point. All done in mechanics (Z1) and later in relays (Z2 to Z5).

Cheers,
Magnus