Unbalancing audio with scrap transformers.

[headcrab]

I've been running sound backstage through CAT5E for a while now, basically lifting the ground and using the pair. This never really worked because I needed a console at the other end to receive the signal. But then, while digging through my boxes of junk I found a transformer which I had pulled from a switching power supply. I used that to unbalance the signal to feed into a TV. And it worked surprisingly well.
Unfortunately I can't say what impedance the transformer is, since it's not an audio transformer, but my point is that nearly any transformer will work for this application. I even wound an isolation transformer, to see if it would work, and while it did, the SMPS transformer was a thousand times better.
Though, were I doing this for a PA or recording application I would buy a transformer, but since it's only for actors I'm not inclined to spend anything.

 

[gcpsoundlight]

Possably that would work, but using Cat5 is asking for trouble! you've got 4 (i am assuming) balanced lines without ground. I can see so many interferance problems.

 

[BigBC68]

If I remember correctly, you can get away with twisted pair for balanced audio without a ground by sending the +/- down each wire of one pair, and the properties of the twisted pair should cancel any interference. I haven't personally used it, but I've heard of it being done. I don't think I would trust it for long runs though.

 

[museav]

Actually, lifting the shield at one end, usually the receiving or input end, is very common for balanced lines. Much of the noise cancellation and EMI resistance comes from the twisted pair configuration along with the balanced circuit and a differential input. The impedance balancing and twisted pair configuration help ensure that any induced noise is induced equally in both the + and - conductors which then maximizes the noise reduction provided by the differential input. You might also find this interesting, http://www.audiosystemsgroup.com/SCIN-2.pdf. The net result is that the best way to address the shield can depend upon the system power grounding, the cable configuration, the type of shielding, how the audio device addresses the signal ground and the RF environment.

 

[FMEng]

Every telephone line on the planet is an un-shielded, twisted, audio pair. It is possible to send high quality, balanced audio over thousands of feet of un-shield, twisted pairs without degradation. I have done it many times.

The key is that it has to be well balanced, and the receiving end has to have very good common mode rejection, and electrostatic isolation. Good transformers offer both. And good transformers cost significant bucks.

For very long lines, more than a few hundred feet, the capacitive characteristics of the cable start to become a factor. When that happens, the current capabilities of the driving stage matters. The typical op-amps won't do it alone without using a transformer to raise the impedance seen by the output stage.

For really, really long lines, a mile or more, then you have to get into using special equalization techniques to combat the high frequency roll-off that builds up. But even that is possible. Broadcasters used to send studio quality audio over telephone cables for miles, even coast to coast in the days before satellites.

For meatball stuff like backstage feeds, or feeds to overflow rooms, it is amazing what can be done on the cheap. Radio Shack sells their 273-1374 1:1 audio transformer for $4 (in stores). It won't work as well as a $75 Jensen in a demanding application, but for cheap it can solve many audio over distance problems and it sounds pretty decent if it isn't driven with too much level. It is worthy to be in any audio tech's parts box.