.
AUDIO NOISE
AND AC SYSTEMS
THE TWO-PHASE SOLUTION TO A DIFFICULT PROBLEM
.
-
By Martin Glasband
-
.
-
OVER THE PAST 30 YEARS,
the art of sound recording and production has advanced perhaps as much
as any of the modern technologies. However, these advances have outdistanced
electrical standards to which we have grown accustomed in the electrical
power industry.
-
.
-
Although the consoles, transports
and digital electronics are more refined, where AC power is concerned,
we are still relying largely on the same technology used in the vacuum
tube years. It seems that no one has given enough thought to tailoring
AC power to meet the needs of sophisticated audio equipment. Some would
argue that isolation transformers with Faraday shields address this need.
But in fact, the root of the problem goes much deeper. This may sound stunning,
but power distribution systems in widespread use today are not fully compatible
with audio electronics. There are serious shortcomings in conventional
electrical engineering methods that apply to audio systems. Use of any
of today's standard 120-volt single-phase AC systems mean potential problems
for audio equipment.
-
.
-
Since the development of the
JBL "T" circuit in the mid-1950s, balanced audio signal circuits have become
the industry standard. Audio power supplies are not truly designed to operate
in an unbalanced manner; commonly they are "force-fed" (unbalanced) single-phase
AC anyway. Why? Because that's just the way it's done; because that's the
electrical industry standard. Amazingly, most audio engineers and manufacturers,
studio designers and electricians, share the same blind spot. These standards
have remained unchanged (and unchallenged) for a very long time.
-
.
-
Ideally, the correct type of
AC power for audio gear in the United States is 120-volt, two-phase (balanced)
power. Commercial use of two-phase (equi-potential) AC isn't new. In fact,
it is probably the first type of AC power ever put into widespread use
in this country (Chicago in the early 1900s). Nevertheless, its commercial
use is all but dead these days. Properly grounded 120-volt two-phase wiring
systems aren't mentioned at all in the National Electrical Code (NEC).
However, this unfortunate fact didn't deter a few innovative engineers
and studio execs from employing a 120-volt equi-potential AC system at
the Zoo Studios, Studio City, CA.
-
.
-
Soon after its opening, a studio
musician plugged in his Marshall amp and declared it was broken because
there was no hum. Much to his surprise, his amp worked just fine. He mentioned
to the studio engineer that this was the first studio he'd ever worked
in where his amp simply made no noise at all. It should also be noted that
it was not necessary to drive a single ground rod. That's pretty clean
AC.
-
.
-
EXPLAINING THE BASICS
-
.
-
The root cause of most studio
noise problems has little to do with grounding techniques. Accordingly,
the balance of this text addresses the source of studio noise, unbalanced
AC, and of course the remedy. First let's examine some typical AC systems:
-
.
-
Figure
1 is the most common type of 120-volt AC system. The transformer outputs
furnish both 120 Volts and 240 Volts to the rest of the electrical system.
Figure
2 and Figure 3
are 120/240 delta and 120/208 wye 3-phase systems. In both of these 3-phase
systems, single-phase 120-volt power can be applied to branch circuit wiring.
Figure
4 is an example of a system more commonly found in recording studios.
It furnishes 120-volts to all branch circuits.
-
.
-
The common thread in all of
these systems is a grounded neutral conductor. Here lies the sole cause
of almost all AC-induced noise in audio installations.
-
.
-
Figure
5 is a low-pass filter (EMI/RFI filter) that can be found in nearly
every power supply in audio equipment. Note that current flow exists between
the 120-volt side of the AC line and the chassis ground. Simply put, here
is the source of most AC induced audio noise. As more pieces of audio equipment
are turned on, noise in the grounding system increases. This appears as
hum in most high-gain and high-impedance equipment. There have been numerous
remedies used to eliminate this noise. The common belief is that an improved
earth ground will lessen the problem. This is true to a slight extent,
but simple logistics make desired results all but impossible. The theory
assumes that a good earth ground will somehow "vacuum" the system clean
of AC interference. However, the transmission of electrical interference
within a studio grounding system, being a measurable and even predictable
phenomenon is of course subject to Ohm's Law.
-
.
-
Commonly, the chassis-to-chassis
impedance in an electrical grounding system is fractional. (For a more
accurate estimation of ground circuit conductivity and impedance, refer
to Table 8 in Chapter 9 in the NEC, "Conductor Properties.") Even under
the most optimum conditions, substantially greater conductivity to true
earth ground is virtually impossible to achieve. Even if it were possible,
the low conductivity to earth ground would at best result in only a very
slight improvement in system noise levels. A good example of this is a
real case where a studio owner decided to have a well drilling company
sink a 60 ft. x 3 in. copper shaft into the earth for his studio's grounding
electrode. Unfortunately, the midi-racks and effects returns still had
a lot of noise.
-
.
-
A somewhat more effective approach
is to break the equipment grounding connections. Lifting the capacitors
from the chassis or using AC ground lifters are ways this is done, however
electrical safety is severely compromised. Star grounding is at best a
crude approach to the problem. The system's impedance to earth is still
far greater than chassis-to-chassis impedance. Noise continues to be a
problem to a greater than desirable extent.
-
.
-
The only cure to this dilemma
is to adapt the power system to the gear being utilized, and eliminate
the noise at its source. It seems like a simple enough task. Unfortunately
though, the electrical code lacks direction in this area. Before we tackle
that one, let's look at why a balanced (two-phase) 120-volt system works.
If one is going to attempt to amend the NEC, it's a good idea to have a
clear reason for doing so.
-
.
-
HOW IT'S DONE
-
.
-
Figure
6 is a 120-volt, two-phase grounded system applied to a typical EMI/RFI
filter. The phase to ground voltage on each side of the AC line is about
60 Volts. Each line of the iso-transformer output is 180 degrees out of
phase to the other, therefor a 120-volt potential between the lines is
present. When this voltage is applied to an EMI/RFI filter, the current
passed through the capacitors is nulled at the common chassis ground by
its inversely phased counterpart. Obviously, this noise cancellation effect
on the ground is beneficial for audio circuits.
-
.
-
Single-phase systems cause noise
flare-ups in other related applications. P.A. systems are particularly
vulnerable because of the high-impedance and high-gain equipment used.
Mobile recording units can also be adversely affected. Radio production
and broadcasting facilities, and even video editing and telecine installations,
are subject to unbalanced AC induced interference. Here, too, an equi-potential
AC system would greatly help matters.
-
.
-
At the Zoo Studios, there was
some initial apprehension when the 120-volt, two-phase system was first
fired up. Just how would the audio gear respond to this type of power?
.Optimistic
predictions proved to be correct. There were no examples of any equipment
running afoul on two-phase power. Equipment power supplies operate normally
with equi-potential power. Line-to-line voltage is critical, but line-to-ground
voltage is irrelevant as far as power supply operation is concerned.
-
.
-
The only real problem occurred
when a new electrical inspector showed up and began quizzing the contractors
about the unorthodox wiring system. Fortunately his concerns about electrical
safety had been anticipated and appropriate safety measures had been taken.
After some discussion with the inspector (and his boss), it was agreed
that the system was safe. Though not covered in any section of the NEC,
the "spirit" of the code had been maintained. In other words, it was deemed
to be safe and passed inspection.
-
.
-
Short of changing an entire
electrical system, there are some procedures that can be followed in the
interim to improve noise problems in many studios. Referring once again
to Figure 4 (the
more common studio system), if the grounded side of the output on the transformer
is lifted, a more balanced AC system will result. This is called an ungrounded
120-volt split-phase system. It has some drawbacks, but its use is recognized
in the NEC for application in hospital anesthetizing rooms. System monitoring
equipment is required as well as other conditional installation methods.
You will also need special permission from the local code authority to
use it in a studio.
-
.
-
One drawback is the lack of
system grounding protection. Should there be a short-to-ground, the entire
system would "tip over" (with one side grounded) and revert to a common
single-phase-type system with a neutral wire. Furthermore, every chassis
power supply in the studio would then have a neutral wire in its power
supply and have a chassis that is referenced to ground through a shorted
out piece of gear -- not a very nice picture. Worse perhaps would be an
ungrounded chassis that became shorted; this could create a serious shock
hazard.
-
.
-
The simple solution to potential
safety questions is the use of two-pole GFCI circuit breakers, or at least
a two-pole GFCI main breaker on the whole system. Additionally, when there
isn't a grounded center tap on the transformer, a typically uneven magnetic
flux dispersion across the output windings will result in unequal phase-to-ground
voltages at the outputs. This difference may vary as more gear is turned
on. A difference of 10-15 Volts from line-to-ground between the outputs
is not uncommon. Nevertheless, noise levels will be reduced by l0dB to
12dB.
-
.
-
When a grounded center tap system
is used, equal output windings on both sides of the center-tap ground reference
balances the output voltage. The need for costly GFCI circuit breakers
is also eliminated. This is because of a much higher ground fault current
potential at the transformer line outputs. Additionally, AC induced noise
is rendered inaudible.
-
.
-
At the time of this writing,
a course of action is being undertaken to include in the NEC provisions
for 120-volt two-phase electrical systems for special use applications
(audio/video installations). In the mean time, local authorities will need
special attention until new standards are adopted. Each project needs to
be handled on a case by case basis. More than likely, local electrical
inspectors will want to know why a two-phase system is needed. They will
also require a number of safety measures to be employed in the design and
wiring methods used. Some of the basic safety requirements are as follows:
The
isolation transformer used must have a center-tap terminal on the output
windings (refer to Figure
6) that is appropriately grounded. This will ensure sufficient fault
current to trip a circuit breaker in the event of a short circuit somewhere
in the system.
Double-pole
circuit breakers are to be used for all AC branch circuit wiring. Both
line conductors (unlike a normal circuit with a neutral wire) are ungrounded.
Therefore, both require over-current protection. The breakers selected
should include a thermal overload mechanism.
Always
use orange iso-ground receptacles to identify balanced AC system outlets.
Though they are designated as neutral-wired devices, there is no NEMA receptacle
configuration that can be used as a practical alternative. This is part
of the code change process currently underway. Nevertheless, it is important
to identify (using orange receptacles) these special system outlets. Additional
labeling may also be required. Other use outside of studio equipment should
be discouraged.
.
Inspectors may have other considerations,
but it is doubtful that anyone will want to stand in the way of progress.
The concepts discussed here are new to many and as stated, there is little
if any knowledge of these engineering methods to be found throughout the
electrical industry. Local code variances and national code changes will
take time. Electrical equipment needs to be standardized and textbooks
need to be rewritten. To quote one NEMA code representative, "This is a
multi-headed beast." Perhaps a few letters to the NFPA (National Fire Protection
Association) in Quincy, MA, will expedite these changes, Though it may
be more than a simple job getting a 120-volt, two-phase system legally
installed, the rewards are well worth the effort. The savings in downtime
and freedom from noise problems will most assuredly justify the installation
and the investment.
.
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.
.
Technical
Articles Index.....Equi=Tech
Index