|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
74.37.227.139
In Reply to: Not quite true. posted by kurt s on April 9, 2007 at 12:22:34:
"It is not necessary to have one side grounded in a twisted pair to get benefit from twisting the wires, but it improves the situation."Hum rejection in twisted pairs generally depends on both wires having equal impedance to ground (i.e. 'balanced.') Neither is connected directly to ground. Having the same impedance, both pick up equal hum and noise with the same phase; the noise is common mode. A differential amplifier rejects the common mode noise. This is how twisted pairs are used in telephone systems, computer networks, instrumentation, pro audio, ...
Typically, the signal (music) is applied to both wires out of phase (differential mode) but it does not need to be; the music might be on one wire only (referenced to ground) so long as the other wire has the same impedance to ground it will pick up identical noise and the differential amp will 'subtract' it from the music signal.
This fact can be used if you happen to have a differential input amp but a SE output source. If you build an extra resistor network into the source that is equal to its output impedance you can use twisted pair to reduce hum and noise pickup in the interconnect.
What does any of this have to do with Jarthel's original question? I don't know. :)
Follow Ups:
> > "It is not necessary to have one side grounded in a twisted pair to get benefit from twisting the wires, but it improves the situation."
Hum rejection in twisted pairs generally depends on both wires having equal impedance to ground (i.e. 'balanced.') < <you are making the case for best use of twisted pair wires, not why you twist the wires in general. I was trying to explain why we would want to twist them, and it helps in almost all cases.
Try an unbalanced approach to twisted versus non-twisted: the leads of an oscilloscope, for example. Make a big loop from the center conductor probe to the outer conductor shield, and you have a loop antenna picking up hum. Then twist them together while keeping the input shorted, and the hum nulls from the scope trace.
You could also, if it weren't for the static electricity problem, take the single-ended phono output (not differential) and feed it to the single-ended primary of a step-up transformer, all floating in the air, and not get the kind of hum you'd get if you ran two wires untwisted. In practice, you need to bleed off the static electricity from this floating arrangement because that causes other problems with noise.
But for best hum and noise rejection, you provided the best case scenario for twisted pairs, almost. For even better performance, add a grounded shield around the twisted pair. After all, in your circuit suggestion you have the established ground potential available, so why waste it? Just a little extra cost, perhaps.
Again, what was the original question? :-)
Good points, Kurt. It's just that I think that the 'real' advantage of twisted pairs are often misunderstood. We see them everywhere... like so many other people I rip apart cat5 for the individual wires. In fact, there is very little shielding action going on when you twist a ground wire around a signal wire ...I do want to make one comment:
"You could also, if it weren't for the static electricity problem, take the single-ended phono output (not differential) and feed it to the single-ended primary of a step-up transformer, all floating in the air, and not get the kind of hum you'd get if you ran two wires untwisted."
Ideally, a transformer is a differential device provided you don't ground one end of the primary (interwinding capacitance ruins that to some degree.) So, if you're talking about taking the output of a cartridge (which is balanced, not SE, BTW) then you're making my argument for me. Thanks. :)
> > Ideally, a transformer is a differential device provided you don't ground one end of the primary (interwinding capacitance ruins that to some degree.) So, if you're talking about taking the output of a cartridge (which is balanced, not SE, BTW) then you're making my argument for me. Thanks. :) < <This is in a rather ill-defined area. To call a floating two-wire unreferenced generator as "differential" is not clear either, and neither is calling it single-ended. Well defined differential requires that ground be centered, like with a center-tapped output or input of a transformer and opposite phased output lines. Without the center tap it is not clear that this is still differential. One could reference either side of the ouput of a phono cartridge, say the minus side, as the reference, and then call the positive output single-ended with respect to that reference, even though it is floating. But if it remains floating, it is not actually referenced, and so it is really not single-ended nor differential.
So I will withdraw my statement that a phono cartridge output is single-ended. It all depends on how you hook it up. If you ground one side it becomes an unbalanced single-ended output. If you ground it at it's "centered" point, then it becomes differential. But in practice you need to ground it in the system one way or the other to prevent electrostatic buildup problems, so I would call a phono cartridge generator output defined by the circuit it operates in. You can also use it in a hybrid of these two, referenced partially to one side or the other, unbalanced, but with a partial differential setup.
A floating undefined two wire generator by itself is neither differential nor single-ended, in summary. How you reference it to ground in the circuit will define its true nature.
Kurt
This post is made possible by the generous support of people like you and our sponsors: