AUDIO NOISE AND AC SYSTEMS REVISITED

RETHINKING STUDIO ELECTRICAL SYSTEMS.

By Martin Glasband

“WE’VE ALREADY ADDRESSED THE PROBLEMS peculiar to hospitals and their sensitive equipment. I guess it’s about time we turned our attention to addressing the problems in recording studios,” said one highly placed electrical inspector.

“Intriguing,” said another inspector.

“This is so simple!” laughed one elated audio engineer. “Why didn’t anyone think of this before?”

These are some of the typical responses to the June 1991 article in R-E-P, “Audio Noise and AC Systems.” The solution to audio noise proposed therein has been well-received by engineers on both sides of the “Silicon fence” dividing electrical and audio/ video industry personnel.

The subject of equipotential 120-volt AC systems for use in audio/video facilities has become the focus of discussion in many circles. For almost three years at the “Zoo Studios” in Studio City, CA, the system’s compatibility with audio equipment continues to be proven. Other audio/video facilities using equipotential AC are growing in number and at least one major TV network is currently planning to convert its sound stages and production facility’s power system to resolve noise problems.

120-volt symmetrical AC systems have been cropping up in unexpected places. One of the largest videotape duplication companies in the country has installed an equipotential system for its video racks and has reduced its noise floor 75% (in an unbalanced system). This applies in the audio and video industries, plus the communications and data processing industries. Even in high-tech “clean” environments, such as microprocessor manufacturing and R & D facilities, symmetrical AC power enables technicians to measure fractions of micro-volts with cleaner operating test equipment.

In all of these areas, a 60/120-volt, single-phase, 3-wire AC system (symmetrical or balanced AC system) outperforms all other 120-volt, 120/240-volt or 120/208-volt systems. It is the only class of 120-volt AC system that directly addresses the common problem of AC induced EMI. Symmetrical AC is uniquely transparent and causes no noise in electronic circuits. The theory behind it is simple. Unfortunately, however, other more common and less effective classes of AC systems prevail throughout high-tech industry. How is it that such a simple system remains largely unheard of and in the background?

HISTORICAL EVIDENCE

Down through the years, AC systems in recording studios, as elsewhere, have occupied a back seat in priority and understanding. Electrical engineers for years have been preoccupied with handling grounding system noise problems, which few realize are often problems of their own creation.

Often the task of wiring studio AC has been given to an electrician who has little understanding of the equipment for which he is supplying power. And then, to this person is entrusted the matter of dealing with EMI. So he installs orange isolated grounding receptacles, an isolation transformer and proceeds to attempt to clean up the grounding system by sinking it. But sinking a grounding system doesn’t work. So it’s on to the next plunge.

It’s too bad that many studios have opted for taking the high priced gamble of purchasing expensive engineering services and costly noise filtering/suppression equipment, only to come up somewhat short of their expectations. (The filtering equipment probably would have operated more cleanly with an equipotential system.) It’s not that the correct technology has never existed; it’s that the technology has never been correctly applied.

Old school electrical engineering practices, products and systems fall short of what is needed in the audio/video industry. Much of the National Electrical Code (NEC) sections dealing with audio/video and related electronic wiring systems (Articles 518, 520, 530, & 640) were written long ago and have little to do with the needs of today’s modern installations.

The time also has come for studio engineers to rethink their priorities involving electrical wiring. The simple mistake of taking for granted AC wiring in studios has led to a morass of unnecessary grounding fixes and has retarded the development of proper electrical systems for studios. The price for this mistake has been many years of studio noise problems, millions of dollars and countless man-hours wasted. Realizing that conventional studio electrical systems need to be re-examined is the first step in correcting the problem. Seeing that, one can now study the situation. First, a brief review of material from the June 1991 article:

Let’s go right to the source of noise, EMI induced by conventional AC systems. Figure 1 is a 120-volt isolation transformer plugged into a typical RF filter, common to most every piece of audio and video equipment. Note the current flow through the grounded capacitors. Here is where noise problems originate. As more gear is turned on, the voltages present throughout the grounding system (signal reference grid) are raised relative to the impedance of the grounding system. This may be only a matter of millivolts, but in high gain and high impedance audio equipment, it can be very audible.

MYSTERIES EXPLAINED

Theoretically, lowering the grounding system’s impedance should reduce the voltage present, but in reality, Ohm’s law demonstrates that significant improvement is unobtainable. Copper building wire, for example, 12-gauge wire used in branch circuit wiring, has only about 0.l ohms resistance for every 50 feet. Larger sizes have of course, even less, but the difference is insignificant. Regardless of the grounding conductor size and length, Ohm’s law, there will still be an unacceptable voltage present in the grounding system.

Furthermore, the grounding electrode (ground rod, water pipe, etc.) presents an even greater obstacle. Many have had the experience of driving one or more ground rods with huge copper wires to supplement studio grounding systems and have learned something about futility.

Commonly, if one can achieve 25 ohms or less above true earth ground, one is doing quite well. Shaving millivolts off a signal reference grid (grounding system) through grounding techniques is truly a logistical nightmare.

But what if RF filters could be made to operate more cleanly? By design, they are most efficient at doing what they are intended, suppressing stray radio frequencies from entering or exiting equipment chassis. Notice in Figure 1 that they are balanced. However, the voltage supplied to them is unbalanced. Currently, the electrical industry recognizes only one basic type of single-phase, 120-volt circuit. To ensure system safety, one of the two 120-volt supply conductors is always grounded. However, this means that one side of the RF filter has potential relative to the grounding reference; the other is neutral. Consequently, AC leaks through the capacitors unchecked directly into the grounding system.

What would happen if the AC supply was balanced? Figure 2 is an example of such a 120-volt application with an RF filter. The potential to ground on each side of the AC line is 60 Volts with each side of the circuit 180 deg. out of phase to the other. Thus, a 120-volt supply is maintained.

As a result of this configuration, trace currents flowing 180 deg. out of phase through capacitors on opposite sides of the filter null at the common chassis ground connection. Voltage formerly present in the AC grounding system is thereby eliminated. The difference is dramatic. So pronounced is the effect that conventionally accepted studio grounding methods are exposed as essentially worthless in countering low-frequency noise problems. Harassing EMI ceases. When a symmetrical AC supply is used, the noise floor of any audio or video facility is greatly reduced.

ASSEMBLY REQUIRED

The key component in this type of system is the transformer. In electrical terms, this particular transformer is called a 120/60-volt, single-phase (center-tapped) isolation transformer. No primary voltage has been indicated here because applications vary among cases. However, all common primary voltages are available.

Current improvements in transformer specifications, such as lower impedance and reactance and better shielding, all contribute to the “purity” of an equipotential system. But even a run-of-the-mill version can provide a dramatic (42db in one measured case) improvement in noise levels. Of course, the size of the transformer would depend on system requirements. 1 kVA to 50 kVA (8.3A to 400A) is the general range. By design, this kind of transformer is the only type that delivers a balanced 120-volt supply.

Similar isolation transformers providing a 120 Volt supply have either one grounded line that unbalances the power or two ungrounded “floating” lines that are unstable and potentially dangerous. One should never use these transformers. On the other hand, a center ground transformer has a stable output voltage, and, for safety’s sake, an adequate system fault current capacity that enables fuses and breakers to operate properly.

Symmetrical AC wiring has also been shown to reduce EMI in equipment where RF filters are not used. Two prong AC plugs are generally an indication of this design. In such equipment, magnetic/capacitive coupling in poorly wound power supply transformers can create hum as do RF filters. This is commonly true where interference from power supplies originates in audio grounded chassis.

In the case of symmetrical power, because the audio ground exists roughly at a mid-point between primary line potentials in the power supply, a greater measure of nulling occurs in transformer coupling (except for poorly made transformers), as is similarly the case with capacitors in RF filters. Furthermore, the highest potential to ground is only half that of conventional AC systems. Once again, an equipotential supply demonstrates inherent compatibility with audio equipment.

On paper, symmetrical AC looks like a balanced audio circuit. It is quite possible that audio and AC grounding systems will no longer be considered separate (but connected) entities. A properly designed symmetrical AC system is actually a part of the audio electronics. There is no reason why fully balanced electrical/electronic systems should not be treated as a whole entity. The concept of virtual integration could be the technology of tomorrow.

NOT ON THE BOOKS

Currently, there are no provisions in the NEC for this class of system. There isn’t even and approved receptacle device that one can use to plug in one’s equipment. However, guidelines have been developed that make it possible for studios to convert to this system to the satisfaction of most electrical inspectors.

In the February 1992 issue of EC&M (R-E-P’s sister Intertec publication), there is an article titled “Applying a 120-volt System with 60 Volts to Ground,” by Fred Hartwell. EC&M is among the leading and most respected of the electrical industry’s trade journals. It is widely read by electrical engineers and inspectors.

This article makes reference to the June 1991 R-E-P article and is generally a discussion of a 120-volt equipotential AC system from the viewpoint of an electrical inspector. The article may be of some assistance in the short term for those who may need to provide an authoritative reference for an electrical inspector.

A possible code restriction could place symmetrical AC in highly controlled or authorized personnel use areas only. It would seem impractical to restrict symmetrical AC outlets to these locations. The average studio has large rooms where musicians and lesser trained studio personnel do their work. Portable symmetrical power supplies are also applicable outdoors and in a variety of remote locations. Use of the proposed NEMA 5-15E receptacles (see sidebar) in conjunction with more accurate GFCI (ground fault circuit interrupter) devices would ensure routinely accurate GFCI performance and greater safety, especially where symmetrical AC is used outdoors.

SAFETY CONSIDERED

It has long been the contention of – electrical inspectors that studios are operated in an unsafe manner because of various grounding compromises that are made. Without question, they are right. But what other choice is there for studio engineers? By virtue of the fact that for safety reasons, audio and electrical grounding systems are indelibly linked, problems with noise have always existed.

It seems unlikely, but there are actually few examples of 120-volt apparatuses, manufactured under UL guidelines, that could not be operated safely – (some light fixtures excepted) with symmetrical AC power. The voltage and phase are the same as conventional 120-volt power. Only the line to ground voltage is different, which should be of no consequence.

According to UL standards, power circuits in equipment must be isolated away from the ground. UL allows the use of 2-prong cord plugs on ungrounded equipment, provided that manufacturers provide a double insulated chassis. Equipment using this type of cord connection shouldn’t be a problem if the equipment chassis is not grounded to the neutral side of the AC supply. In most cases, the chassis ground reference is left floating to avoid ground loops in audio circuits.

Sometimes in older musical instruments or in vintage consumer audio/video equipment, the chassis is referenced to the neutral. This most certainly means trouble in an equipotential AC system. A chassis so referenced would become energized. Fortunately, this is not the usual case because UL regulations prohibit the use of neutrals for grounding purposes. But sometimes it happens anyway. (Check your old unbalanced audio gear and your imported video monitors!) Some equipment slips through the cracks because of confusion that exists with manufacturers over some of the nomenclature used in the electrical industry.

In a standard 120-volt single-phase wiring system, the neutral (white wire) is referred to as the grounded conductor. It’s intended for use only as a current-carrying supply conductor. Being grounded, however, does not mean that it may used as a grounding reference for equipment chassis. In any AC system, the grounding conductor (green wire) is the only correct reference for chassis grounding. The code specifically prohibits its use as a current-carrying conductor.

In every situation, prudence dictates that 2-wire AC cord connected equipment be checked out. (In Japan, nearly all equipment used and exported has a 2-wire cord.) If a neutral-to-ground chassis is found, modification of the chassis grounding and retrofitting of the equipment with a 3-wire, U-ground cord would be the proper course of action to take.

RECENT UPDATES

A number of sound safety measures have been developed since the original R-E-P article was published. Because of characteristics unique to this type of system, simple misuse can easily compromise the most basic of electrical code safeguards. For example, lamp sockets on music stand lights would remain energized even when turned off.

Most studio equipment currently uses a single-pole power switch. This means that (among other things) greater care must be taken by service personnel while working on switched-off gear. Two-pole circuit breakers and exclusive use of GFCI devices or hard-wired AC equipment connections in lieu of GFCI protection seem to have become standard safety measures upon which most electrical inspectors agree.

Use of 2-pole GFCI breakers have been recommended and some will probably work if their internal op-amps can function at about half the normal AC circuit voltage. By design, a 2-pole GFCI is connected to the neutral bus. Operating within a 120-volt symmetrical system, the input voltage on a 2-pole GFCI breaker’s op-amp power supply would therefore be only 60 Volts instead of 120 Volts.

Even though some may work, at $130 each a more practical alternative is suggested. Standard GFCI mastering devices (used with spas and jet tubs) without slots for cord plugs are a fraction of the price. They too operate at 120 Volts, but they derive their power from both sides of the line (120 Volts) on the device, which is a more compatible configuration with symmetrical power.

A DISCLAIMER

An unsettling practice continues by a few well-meaning audio technicians who attempt the electrical installation themselves. Often, this turns out to be an overzealous and uninformed approach to hazardous electrical safety matters. Beware! Electrical power is a technology requiring years of education and training. Numerous safety factors, and in many cases manufacturing specifications, must be considered in any electrical installation. Misapplication can result in damage to property, liability problems and serious injury or worse.

The National Electrical Code is a comprehensive collection of safety measures developed through many years of real world experience. It is not intended to be a bureaucratic obstacle (as viewed by some) but a collection of proven guidelines designed to protect property and human lives. A more positive approach would be a willingness to be open to work with those whose business is electrical safety.

Safety standards don’t necessarily have to be impractical standards. There are many sides to an issue. Electrical Code Panel 15 is the forum where input is needed. (Panel 15 handles motion picture studios, theaters and similar locations.)

As proposals are made for code changes, various documentation will be made available. Inquiries can be sent to the National Fire Protection Association, Electrical Code Section, 1 Batterymarch Park, Quincy, MA 02269. Full lists of technical committee panel members are available by writing to the above address.

Let’s proceed to implement this system properly and in a spirit of cooperation. The problem of AC-induced audio noise is soon to be a thing of the past. Adequate standards are almost, at long last, a reality.

Martin Glasband is an electrical engineering consultant and contractor in Selma, OR. He has designed and built electrical systems for KCET-TV, Music Animals (now The Post Complex), Baby’O Recorders, New World Pictures and the ABC Radio Network.