Cable Question

[Ziller]

Wow, thanks a bunch Skippy. That was a great explanation. I knew balanced was better, but not why.

I am interested in this topic at this point, because I need to run a couple of cables from my mixer to my computer, about 20' away. I figure I should use balanced, but like Popcorn, my soundcard only has unbalanced ins. Usually not a problem, except for the distance I have to run the cable. My plan is, at this point, to run 2 balanced cables (L/R) out of my mixer, 20' to my computer, and put 2 of those low z -> high z transformers on the end, then adapt it down into the 1/8" stereo input on the soundcard. From what you explained then, it must be those transformers that do the noise cancelation, before sending the signal out as unbalanced?

Thanks again for your clarification.
Ziller

 

[skippy]

You're most welcome. I'm surprised that you guys will still put up with my blatherings...

For your 20-foot run from mixer to computer, I'd try it single-ended first, and see if you find the noise floor to be acceptable. If so, you'll save quite a bit of money, and some hassles.

The reason is threefold: transformers cost money, the little inline ones are usually not my favorites in terms of quality and are not usually appropriate for line-level use, and you may or may not get decent ground isolation with them- which may leave you with a ground loop and induce hum in the system despite the fact that you're using balanced right up to the plug.

Balanced interconnect is not a panacea, and it's either confused or pissed off a lot of people over the years to do a bunch of rewiring and still have hums because of ground loops (and the "pin 1 problem", which is an excellent topic for a site search or a web search).

You're exactly right about the transformer doing the noise cancellation. The primary has its center tap grounded, and the signal hot and signal cold attached to each end of the winding. The signal cold and signal hot induce a magnetic field in the core of the transformer, effectively subtracting the cold from the hot magnetically. Which is how the common mode noise gets cancelled: it bucks itself right out of the equation.

The problem with the cheap little inline transformers (say, from Radio Shack) is that they may have fairly awful frequency response on a good day, and they are usually designed for mic levels: which means that the core will saturate and the damned thing will distort *terribly* when used at line levels, which are 40-60dB hotter. Magnetics can be a pain in the ass...

Good line-level transformers are pretty large and expensive (shoot, so are good mic-level transformers!), so I'd try the single-ended setup first. Then, if the resulting noise floor is too high due to ground loop induced noise, get back to us and we'll go to the next step in nerdhood: which may involve going balanced, or using transformers. Or it may not...