Slightly OT - balanced to unbalanced cables?

[Fraser, Colin J]

> -----Original Message-----
> From: Bill Layer [mailto:b.layer at vikingelectronics.com]
> Sent: 22 January 1999 19:40
> To: Fraser, Colin J
> Subject: RE: Slightly OT - balanced to unbalanced cables?
> 
> >Wiring pin 1 & 3 together wil give you gnd (by fixing the 
> -ve signal at
> >earth), and pin 2 will be the signal, but (IIRC) you will 
> end up with twice
> >the signal level at pin 2 and may overdrive input stages. 
> 
> Ummm.... not to step on any toes, but that doesn't sound 
> right. In a true
> differential system, if you tie 1 to 3, you neither gain nor 
> lose any signal.

This isn't a true balanced system - it's a balanced output driving a
single-ended input.

> A level change will ocurr if you are driving a balanced load, and then
> begin to drive it single-ended. In that case, you will lose 
> 6db of gain as
> the diff-stage loses sensitivity when half of it is tied to ground.

That's true.
At a balanced *input* the -signal is subtracted from the +signal. This gives
2 x signal, which is then reduced by 6db to give just the signal.
When you drive a balanced input from a single ended output, the -signal line
should be tied to ground, so subtracting -signal from +signal only gives you
1 x signal, so you lose 6db. Some gear has balanced inputs that detect if
the -signal line is connected to gnd, and compensate by not dropping the
6db.

My understanding is that if you take a balanced *output*, and tie -signal to
ground, then the +signal should still be driven such that
(+signal)-(-signal) is 2 x signal, giving you an extra 6db if you connect
+signal to a single ended input.
Is that not correct ?


Colin f