Wired for Sound: Steve Lampen
I'll Take Differential for $5, Alex
I always like $5 terms. You can use these terms to browbeat unruly customers (or general managers). Just use a few of them and they're sure to leave you alone and realize that you know your job!
As we continue our discussion of basic wire and cable terminology, begun in the Nov. 3 issue, the first of our $5 terms is the word differential. Fig. 1 shows a microphone attached through a twisted pair to an amplified-speaker (yes, yes, there's a mic preamp in there too. Sheesh!)
Now if we graph the signal flow, this twisted pair is a circuit and the two signals on the two wires are equal but opposite polarity. If we stick the microphone in a piano and hit the key "middle A," that is defined as a note that vibrates the string 440 times a second (440 Hertz).
This 440 Hz means the arrows on our pair of wires will change direction 440 times each second. And the amplified speaker will move in and out 440 times a second. We will hear that note. And, if the microphone, cable and amp-speaker are "perfect," it will sound as if your ear is inside the piano, exactly where the microphone is.
Because the signal on the two wires is always moving in opposite directions, it is called a differential signal. It also means that, if you could measure the signal on each wire at a specific place along the pair, at an instant in time, and if you could mathematically add them together, they would equal zero. Or maybe I should say they are supposed to equal zero. The more "balanced" the pair the closer we get to 100 percent noise rejection.
Noise
Almost all professional audio signals are differential signals, carried on balanced line twisted pairs. Most computers these days also work on balanced-line twisted pairs (Category 5 or 5e or 6, for instance). And the reason is noise. Differential pairs reject noise.
The secret is in the source and destination devices. They have a magic part that rejects noise called a transformer. In Fig. 2 you can see it at the end of the twisted pair.
Many devices these days don't have a transformer. They have a circuit that pretends to be a transformer, something called "active balancing." The question of which is better remains controversial. The active balaced circuit has been getting better and cheaper, like most silicon-chip devices. Wire-and-iron trsnformers have been getting better too, although there's not as much room for improvement. The real problem is cost. A good transformer is expensive.
Both devices are designed to get rid of noise.
I suppose we should define just what noise is. Noise is any undesired signal flowing down the pair. It could be a signal from an adjacent pair in a multipair cable (crosstalk) or from a pair in an adjacent cable (alien crosstalk). Noise could come from sources outside such as RF transmitters, motors, engines (spark plugs), ballasts for fluorescent lighting, SCR lighting dimmers, almost all electronic devices. Even the sun is a great source of noise, which is why this noise goes away at night. All of these are electromagnetic interference (EMI), and the higher-frequency noise we call radio frequency interference (RFI).
This electromagnetic noise, regardless of the source, hits the two wires in our balance line, shown as big arrows in Fig. 2. The insulation on the wires does nothing to stop this.
As some readers might remember from Electronics 101, any time a changing magnetic field is intersected by a wire, a voltage is induced on the wire. I've shown that "induced noise" in Fig. 2 with two little arrows on the two wires of our twisted pair. Since the same noise source hits both wires, the induced noise also is the same, and moving in the same direction.
The key difference between the signal on the pair and the noise on the pair is that signal is differential and moves in opposite directions on the two wires, while noise moves in the same direction (called "common"). Here we use another $5 term: "common mode noise." The two noise signals travel in the same direction until they get to the transformer, or active balanced circuit, where they meet each other and they cancel out.
More to come!
Steve Lampen is Technology Specialist, Multimedia Products for
Belden Electronics Division. Steve holds an FCC Lifetime General
License (formerly a First Class FCC License) and is an SBE Certified
Radio Broadcast Engineer. On the data side he is a BICSI Registered
Communication Distribution Designer. His latest book, "The
Audio-Video Cable Installer's Pocket Guide" is published by
McGraw-Hill. You can reach him at shlampen@aol.com
|
