Newsgroups: rec.arts.theatre.stagecraft
Subject: Re: DMX optoisolator
References:    <[email protected]>
"Frank Wood"  writes:

> E. Lee Dickinson wrote:
> > "Christian Marg"  wrote in message
> > news:[email protected]...
> > > Hello...
> > >
> > > E. Lee Dickinson wrote:
> > >> As DMX is balanced, is there any real reason I should connect Pin1?
> >
> > I've just read an article on "differential signals," and reread the few I
> > have on "balanced signals" and I don't understand the difference.
> > Differential, according to your post, requires a ground reference.
> >
> > But I fail to see why it's needed.  Can't the inverting and non-inverting
> > signals reference off eachother?  That's what they do in, for repeating
> > example, Krystof Nys's design.  Data+ and Data- drive the input
> > optoisolator, and ground appears unconnected.
> 
> This is a very tricky subject. The easy bit first. Balanced signals,
> such as are used in high quality audio installations, have NO earth
> reference. Absolutely none at all, apart from the cable capacitance.
> This means that they are totally immune to common-mode interference.

You are rigth that there are balanced signals that have NO earth 
reference, apart from the cable capacitance. 

But there are also ba�anced systems where the signal output 
could be wired to earth. The signal output could be a center tapped 
output transformer the center tap grounded. Or two operational 
amplifiers driving anti-phase signal to two siognal wires through 
the suitable resistor (for example 120 ohm resistor). Both outputs 
have exactly the same output impedance. This system is balanced as well. 
The balancing is defined on the couplign impedance to earth from 
both signal wires. If they are same on both wires on wire pair then 
the system can be considered balanced. The impedance can be close 
to infinite (on your floating inputs and outputs case) or something smaller 
(50-120 ohms on many differential outputs on many modern mixers, 
usually around 10 kohms on many electronically balanced line inputs 
on mixers). 

None of the real-life systems I know are totally immune to common-mode 
interference, but they are very immune to the common mode interference. 
Well built balanced audio system is pretty immune to the common 
mode interfarence, attenuation of common mode can be in 80-90 dB 
range on good circuits based on transformers. 
 
> Differential signals need something to be different from. Think of a
> centre-tapped transformer. Leave the centre-tap alone, and you have a
> balanced output: ground it, amd you have a differential output.

I wouldn't say exatly this...

Differential signals are just defined to be signals that are 
tranfered as difference of voltage levels on two signal wires. 
It is the voltage difference of the signals on two wires that 
carries the signal, the common mode voltage on both of those 
wires is irrelevant and is canceled on the differential receiver 
(in real world system this common mode canceling is more or less 
ideal). 




-- 
Tomi Engdahl (http://www.iki.fi/then/)
Take a look at my electronics web links and documents at 
http://www.epanorama.net/