The war of noise

There's nothing more irritating than a buzz, tizz, or hiss permeating the room when we're critically listening. How distracting! You close your eyes to hear deep into the symphony's soft passage and visualize the musicians at work. "Hummmmmmm." The illusion is suddenly shattered by an unnatural hum or electronic sound. Unless we rely upon an integrated or receiver, our separates must be interconnected through cables. As discussed in yesterday's post the one music carrying conductor in a single ended cable is protected from radiated noise by a grounded outer "chassis" of braided wire, called a shield. This is effective for most types of radiated noises. But not all. When we discuss radiated noises we use the acronym EMI (Electro Magnetic Interference). The cable's shielding takes care of one—electro—but not the other—magnetic. Transformers generate magnetic fields. These invisible lines of energy penetrate just about everything, including the outer copper braid of a cable. Once inside the wire, hum is the end result. There are a few shielding remedies but none very practical when it comes to wires. This is where our friend, the balanced cable, comes into its own. Balanced cables are no more impervious to magnetic interference than single ended cables. Yet, if you interconnect your equipment with balanced cables, you eliminate both types of noise: electro and magnetic. How is that possible if the cable itself is no better than single ended? The answer lies in the amplifier's input. Where the magic happens. Recall the construction differences between a single ended cable vs. a balanced. They are the same, save that the balanced has gained a second conductor. That second conductor is key to removing radiated noise, both electrical and magnetic. When a magnetic field cuts through the cable's outer shield, it equally pollutes both signal conductors within the balanced cable. This means that both signal wires now have the identical magnetic noise. Connect the two wires into a difference input (a balanced input) and whatever is common to the two wires (the noise) is ignored. Only differences are amplified. I'll show you how that's done tomorrow.