Welcome to a new audio column, Audio by Design. Here I plan to explore issues related to audio system design, as well as audio and acoustic test and measurements, with-I expect-a few side trips along the way.
As a brief introduction, among the many hand-knitted hats I wear, I am a systems design engineer, project manager, consultant and writer. I've worked for CBS Engineering and freelanced with top systems integrators.
Along with university degrees, I'm also a graduate of Synergetic Audio Concepts (Syn-Aud-Con), the independent systems design, audio and acoustics education company founded by Don and Carolyn Davis and now operated by Pat and Brenda Brown. I first started attending seminars while still in college and have continued my education with them ever since, participating in numerous workshops and seminars, including Grounding and Shielding.
And it's a grounding-and-shielding issue that I wish to start with here, namely the "Pin-1 Problem."
While I've seen coverage of this topic in the commercial sound realm, I don't think it's gotten as much attention in the broadcast world.
As systems designers, we have learned that for balanced line-level audio circuits (unbalanced audio is a whole different issue), we need to cut the shield at one end of an audio cable interconnecting two devices to prevent ground loops from causing hum and noise in the audio.
Depending on the client or the systems integrator, various shielding schemes have been used, including connecting the shield at the input only, at the output only, at the jackfield only, or at the input and output of the gear but not at the jackfield.
Designing a consistent grounding-and-shielding scheme for a particular application can take a great deal of engineering time and attention to detail in installation. And with one end of the shield disconnected, the shield could become a good antenna, picking up RF-induced noise, which could be carried into the audio circuitry. When this happens, more fixes are needed, like inserting a small capacitor between the cable shield and the equipment chassis.
But even with all these machinations, if a consistent design was rigorously followed, acceptable results were usually obtained. So it was probably no surprise that no one really looked deeper into the problem. Until...
...Neil Muncy, an audio and acoustical consultant based in Ontario, Canada, said wait a minute. Why can't we just take a standard twisted-pair shielded cable and simply plug a balanced output from one piece of gear directly to a balanced input of another and have everything work perfectly, without noise, hums or buzzes?
Why not indeed? Was there something fundamental going on here, a root cause?
It turns out there was. Muncy presented his findings at the 1994 Audio Engineering Society (AES) Convention in San Francisco (along with others in a grounding-and-shielding workshop). His paper, "Noise Susceptibility in Analog and Digital Signal Processing Systems," was subsequently published in the Journal of the Audio Engineering Society in June 1995, in an issue dedicated to grounding-and-shielding topics.
Muncy discovered that the ground-and-shielding problem often originated not with any particular cabling scheme, but with the audio equipment itself and, in particular, how the shields at the inputs and outputs were connected.
A common approach, which produces undesirable effects, is to connect the balanced cable shield to the signal ground inside the equipment. This means that any noise or hum induced in the cable shield would be assured to enter the audio circuitry, mixing with the audio signal itself and ultimately be present at the output. In other words, the shield connection becomes another audio input connection, albeit an unwanted one, with noise as the "signal."
Muncy called this phenomenon the "Pin-1 Problem," with Pin 1 referring to the pin on a three-pin XLR connector used for the shield. (For a tip-ring-sleeve phone plug, the shield would be on the sleeve.)
As an antidote to the Pin-1 Problem, the best engineering practice is to connect the cable shield at the equipment metal chassis directly at the cable's point of entry to the equipment.
If this is done properly on both pieces of equipment, and if the cable shields are connected at both ends, then the cable shields form an extension of the shield formed by the metallic enclosures. This serves to keep electromagnetic interference from entering the equipment (and for that matter, keeping EMI from escaping the equipment). Since the chassis is connected to earth ground via the three-prong AC power plug (as it should be in any professional installation), any outside interference is shunted to ground and never enters the audio circuitry.
(As a reminder, never lift the safety ground on the power plug. Doing so can be potentially hazardous.)
Chassis-grounding the cable shields also help protect against RF interference.
Now alert readers may say that the chassis ground must be connected to the signal ground 0V potential inside the box.
That's correct, and there are proper ways to do that, but the point here is that to achieve the best-noise immunity and gain the noise-canceling benefits of well-designed balanced circuits and shielded twisted-pair cables, the cable shield should never be allowed to connect directly to the signal-ground inside the box. In addition, the cable shield should be connected at both ends of equipment free of pin-1 problems.
In an ideal universe, with all equipment designed properly to prevent the pin-1 problem, cabling becomes much easier. Although things are improving as manufacturers recognize and correct the problem, it is still out there. So let's not throw away all the tricks of the trade we've learned all these years, (but wouldn't it be nice if we didn't have to use them anymore?).
Next time: How to test for the pin-1 problem.