An Overview of Audio System Grounding & Interfacing 9/4/2012

An Overview of Audio System

Grounding and Interfacing
by
Bill Whitlock, President
Jensen Transformers, Inc.
Life Fellow, Audio Engineering Society
Life Senior Member, Institute of Electrical & Electronic Engineers

Products mentioned in this educational material are used only as examples and no endorsement is intended. Product
trademarks are the property of their respective owners. Permission for copying and distribution of this document, or excerpts
thereof, is hereby granted, provided no charge of any kind is made for access and proper attribution is made. Entire content
© 2012 Bill Whitlock, Jensen Transformers, Inc.
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 1

An audio system is rarely a plug-and-play affair. Noises often plague the

system even if everything is done according to conventional wisdom. Did the
electrician screw up? Is the AC power dirty? Is the equipment to blame? Are
the cables poorly shielded? This presentation will discuss the "secret" forces
that drive ground loops, how they contaminate our signals, and new tools to
make troubleshooting faster and easier. Among the topics will be equipment
design errors that can make life miserable for installers and users, how
manufacturers manipulate "specs" to deceive, how new test standards and
instruments can make manufacturers honest, why expensive cables probably
won't help, and things your electrician can do to reduce or eliminate problems.

Bill Whitlock 1
An Overview of Audio System Grounding & Interfacing 9/4/2012

About Bill Whitlock

Bill began designing analog electronics for console-maker Quad-Eight in 1972 and subsequently
held chief engineer positions at Laserium® and Capitol Records/EMI prior to joining Jensen
Transformers in 1989. He’s become a widely-recognized expert on power, grounding, and signal
interfacing through his seminars and lectures at trade shows and universities, including MIT in
2007. He’s a member of InfoComm Academy’s adjunct faculty, a CEDIA certified instructor, and
NSCA students voted him Technical Instructor of the Year in 2009 and 2010.

His landmark paper on balanced interfaces appears in the June 1995 AES Journal, which has
since become the best selling issue ever printed. Other writing includes columns for S&VC and
Live Sound magazines, three chapters for Glen Ballou’s Handbook for Sound Engineers, and
dozens of magazine articles and Jensen application notes.

His four patents include the InGenius® balanced input IC made by THAT Corporation and the
ExactPower® waveform-correcting AC voltage regulator. He’s a Life Fellow of the Audio
Engineering Society and a Life Senior Member of the Institute of Electrical and Electronic
Engineers. He currently does product development and customer tech support for Jensen but is
available for consulting and lecturing as time permits.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 2

I firmly believe that the technical concepts in this class are best taught using analogies
and intuition rather than complex mathematics. In my opinion, the audio business,
especially the audiophile portion, is simply awash in bullshit and bad advice!
Therefore, the remedial part of my task is to debunk myths perpetuated as “tradition”
among practitioners and as unintentional but often self-serving misinformation from
manufacturers. You’ll find I loathe the lies, half-truths, and distortions of
marketing ... particularly when they’re used to exploit ignorance or
desperation.

Bill Whitlock 2
An Overview of Audio System Grounding & Interfacing 9/4/2012

Thanks to Our Sponsors!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 3

Bill Whitlock 3
An Overview of Audio System Grounding & Interfacing 9/4/2012

Dedicated to Neil Muncy, Nov 1938 – Aug 2012

I first met Neil in 1994 at a local AES event in Los Angeles.
He seemed to identify me right away as some one who
shared his passion and he quickly persuaded me, in spite of
my intense fear of public speaking, to "get out there and tell
those folks what you know about balanced interfaces."
In retrospect, he was one of the most influential people in my career
since he essentially kick-started my writing and lecturing. We had many
lively discussions about the grounding and interfacing that spurred both
of us to dig deeper and deeper. He was always kind, generous, and
respectful, even when we disagreed. I have so much to thank him for. I
feel very lucky to have spent several days in his Toronto home in
November 2011, enjoying conversation, movies, and a party where I
served as his robotic hands in the kitchen. He was a fine, warmhearted
man that I'll always miss.
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 4

Bill Whitlock 4
An Overview of Audio System Grounding & Interfacing 9/4/2012

Myth and Misinformation

• Do cables really “pick up” noise from the air like a
radio?
• Equipment manufacturers often don’t know ground
loops from FROOT LOOPS ... and it’s what they don’t
tell you that can plague your systems!
• Basic rules of physics are routinely overlooked, ignored,
or forgotten
• Overheard at a cocktail party:
“What do you do for a living?”
“I design and install sound systems.”
“What's so hard about that ... you just plug the stuff together, right?”

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 5

My 20 years of customer technical support work at Jensen has given me

reason to call numerous equipment makers for technical information on their
products. Sometimes, “they don’t know a ground loop from a Froot Loop” is an
understatement! A lecture to EE students at MIT in 2007 also confirmed that
typical engineers graduate with very little understanding of the vagaries of
grounding systems and interfaces!

Bill Whitlock 5
An Overview of Audio System Grounding & Interfacing 9/4/2012

Physics Police
Rule!

Courtesy of Coilcraft

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 6

The immutable laws of physics rule everything electric and magnetic ... period!

Bill Whitlock 6
An Overview of Audio System Grounding & Interfacing 9/4/2012

What We’ll Explain

• What problem noise is and how it’s measured

• Why premises AC power must be wired per Code
• How subtle changes to premises AC power wiring can make
dramatic differences in system noise
• How noise actually couples into signal paths
• How troubleshooting can pinpoint problems
• How to safely cure problems without compromising audio/video
performance
• Why some so-called "cures" have little or no effect

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 7

Much of the information taught in this course is not even mentioned in modern college
courses in Electrical Engineering. My lectures at MIT a few years ago reinforced this
observation. Perhaps it’s why noise issues have become so common in all kinds of
electronic systems! Ultimately, the issues are always analog – but, sadly, analog
expertise seems much less “glamorous” than digital to today’s students.

Bill Whitlock 7
An Overview of Audio System Grounding & Interfacing 9/4/2012

Topic Outline
• 3 – Introduction
• Basic Circuit Theory, Terminology, Laws of Physics, and Myths
• 19 – AC Power and Grounding
• Safety v Earth Ground, Magnetic Coupling, Leakage Current, and Voltage Drops
• 42 – Signal Interfaces and Noise Coupling
• Balanced v Unbalanced, Equipment Issues, Cable Properties and “Floobydust”
• 99 – Troubleshooting
• Basic Approach, Using Clues, “Dummy,” “Hummer,” and Clamp-On Meter Tests
• 122 – Solutions
• Isolators for Audio-Video-CATV-DBS-Data, Transitions, and Power Treatments
• 212 – Suggested Reading

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 8

Bill Whitlock 8
An Overview of Audio System Grounding & Interfacing 9/4/2012

Noise - Think Outside the Box

• In its broadest definition, noise is any undesired signal
• Analog signals accumulate noise as they flow through system
equipment and cables
• Once noise is added, it's essentially impossible to remove it without
altering or degrading the original signal
• Therefore, noise must be prevented along the entire signal path
• Signal INTERFACES are generally the danger zone, rather than
the equipment itself

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 9

I’m not quite as pessimistic as the quote here, but one can’t be too diligent if truly
professional performance is the goal.

Bill Whitlock 9
An Overview of Audio System Grounding & Interfacing 9/4/2012

Perspective on Noise Measurements

• Our ears perceive 10 dB reductions as “half as loud” and 2 or 3 dB
reductions as “just noticeable”
• Dynamic range is the ratio, generally expressed in dB, of maximum
undistorted signal to residual “noise floor” levels
• In audio, a dynamic range up to 120 dB may be required for high-
performance systems in typical homes
• In video, a signal-to-noise ratio of 50 dB is a threshold beyond
which even expert viewers perceive no improvement
• Audio hum or buzz (stationary noise), is much more noticeable and
irritating than the hiss (random noise) inherent in all electronics
• Likewise, the video hum bar is a “stationary” noise
• Excess random noise is a gain structure issue (not discussed in this class)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 10

Here, the term “stationary” refers to the spectrum of the noise. Hum and buzz have
spectral content that consist of one or more frequencies that do not move – they’re
stationary. The spectrum of hiss or white noise consists of countless frequencies that
are “random” in nature. An excellent 1988 AES paper by Louis Fielder of Dolby
Laboratories is the source of the 120 dB figure. Our ear-brain combination is very
adept at separating repetitive noises like buzz from other familiar sounds that may be
substantially louder in terms of sound pressure level (SPL).

Bill Whitlock 10
An Overview of Audio System Grounding & Interfacing 9/4/2012

Just Say “Know” to Noise

• Sadly, applying a few simple rules or installing a “magic” device just
anywhere in the signal path can’t guarantee a noise-free system
• Learning how interfaces and grounding systems work makes finding
and fixing problems logical, if not always easy
• Dumb luck often results in acceptably quiet systems despite
haphazard wiring and poorly-designed gear!
• You may find temporary relief, but the problem is still there ...

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 11

The question is not whether you can coax the system to work right today, it’s whether
you’ve followed sensible practices that will allow it to continue to work right when
something new is added to the system ... or the premises!

Bill Whitlock 11
An Overview of Audio System Grounding & Interfacing 9/4/2012

Basic Circuits and Ohm’s Law

• Current will only flow in a complete circuit (or circle)

• Represented by the letter I in equations and measured in Amperes
• Voltage is the electro-motive force, EMF, that “pushes” the current
• Represented by the letter E in equations and measured in Volts
• Resistance is the opposition to current flow (like pinching the hose)
• Represented by the letter R in equations and measured in Ohms
• Ohm’s Law tells us that E = I x R (and re-arranged equivalents)
• CURRENT ALWAYS FLOWS BACK TO THE VOLTAGE SOURCE
THAT PUSHED IT!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 12

I hope that, for most of you, this is nothing but a few brief reminders of the Basic
Circuit Theory 101 class that you took some time ago!

Bill Whitlock 12
An Overview of Audio System Grounding & Interfacing 9/4/2012

Impedance
• The total apparent resistance of a circuit that includes capacitance
and/or inductance
• Represented by Z in equations and measured in Ohms
• It’s the functional equivalent of resistance for AC circuits
• Unlike resistance, impedance may change with frequency
• The impedance of a capacitor decreases with increasing frequency,
becoming an open circuit at DC and a short circuit at very high frequency
• Capacitors are measured in Farads and store energy in electric fields
• The impedance of an inductor increases with increasing frequency,
becoming a short circuit at DC and an open circuit at very high frequency
• Inductors are measured in Henries and store energy in magnetic fields

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 13

Basic inductive and capacitive reactance principles: with increasing frequency, the
impedance of a capacitor decreases while the impedance of an inductor increases.

Bill Whitlock 13
An Overview of Audio System Grounding & Interfacing 9/4/2012

Impedance of Wires
• Even STRAIGHT WIRES have RESISTANCE and INDUCTANCE
• Resistance, significant only at DC and low frequencies, is directly
proportional to length and inversely proportional to diameter
squared. It’s 0.015 S for our 10-foot #12 AWG example
• Inductance is directly proportional to length, but essentially
unaffected by diameter or gauge
• 4.8 µH for our 10-foot example (straight wire)
• Increases substantially at bends or loops
• Typically dominates impedance at frequencies over a few hundred Hz
• Inductance is a result of the magnetic field that surrounds any
conductor that carries current

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 14

A wire is modeled as a resistance and an inductance in series. At low frequencies its

impedance is dominated by resistance but, above a transition frequency, the inductive
reactance dominates. For heavy-gauge wire, this transition occurs at a few hundred
Hz. For lighter-gauge wire, the transition is commonly in the 2 kHz to 10 kHz region.
When we discuss ground wiring for lightning protection, we will again mention the
increased inductance (and impedance) at bends in the wire.

Bill Whitlock 14
An Overview of Audio System Grounding & Interfacing 9/4/2012

Example of Wire Impedance vs Frequency

Inductance Causes Impedance to

Rise with Frequency

½” Solid Rod

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 15

At higher frequencies, where inductance dominates, using thicker wire has virtually no
effect on impedance. At 1 MHz, replacing the #12 wire with a ½” solid copper rod has
little effect (as noted in the plot).

Bill Whitlock 15
An Overview of Audio System Grounding & Interfacing 9/4/2012

AC Magnetic Fields and Magnetic Induction

Fluctuating field Field induces AC voltage in

surrounds every wire any nearby conductor
carrying AC current (transformer principle)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 16

The principle of magnetic induction per Michael Faraday. He is one of my heroes

because, although he was scorned by the “math snobs” of his day for not publishing
complex mathematical expressions of his findings, he made truly fundamental
discoveries about the relationship between electricity and magnetism.

Bill Whitlock 16
An Overview of Audio System Grounding & Interfacing 9/4/2012

What Does “Ground” Mean

• Also known as “Earth” in the rest of the world
• Utility Power: an actual electrical connection to SOIL
• Electronics: a common return path for various circuits, whether or
not actually connected to soil
• A FANTASY invented by engineers to simplify their work
• The “uni-potential” fantasy assumes all ground symbols in
schematics are at exactly the same voltage
• Truth: Real-world conductors have resistance, causing small voltage drops
• Truth: Ground circuits most often serve, either intentionally or accidentally,
more than one purpose
• Meaning has become vague, ambiguous, and often quite fanciful

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 17

The use of ground symbols in schematic diagrams (actually just a convenience for
avoiding more lines in the drawing) lulls us into thinking that they’re all at the same
potential. That’s the essence of the fantasy ... but far from the truth. Until room-
temperature super-conductors become a common reality, “grounds” are connected by
wires, PCB traces, or sheets of metal – all of which have both resistance and
inductance. So much for the fantasy!

Bill Whitlock 17
An Overview of Audio System Grounding & Interfacing 9/4/2012

Misguided Strategies
• Reduce unwanted ground voltage differences by “shorting them out”
with massive wires or bus-bars
• Reduce noise experimentally by finding a “better” or “quieter”
ground
• Skillfully route noise to an earth ground, where it disappears
forever!
• Is an earth ground for electronic systems really necessary? Think
about aircraft electronics ...

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 18

Random experimentation with grounding connections is the worst possible way to

solve the problems. Not only is little learned in the process, but the experiment can be
lengthy ... usually ending only when someone says “I can live with that.” The idea
that “mother earth” will simply absorb noise is often called the “sump theory” of
grounding. And when was the last time you saw an airplane dragging around a ground
wire?

Bill Whitlock 18
An Overview of Audio System Grounding & Interfacing 9/4/2012

Common Myths
• Earth ground is the absolute zero-volt reference
• Fact: Many unintentional currents flow in soil and create voltage drops just
as in any other resistance – soil is a relatively poor electrical conductor
• Ground wires have zero impedance
• Fact: Wires have impedance and cannot make multiple points in a system
have an identical “zero-volt reference”
• Noise (voltage) exists on a single wire or at a single point
• Fact: Voltages can exist only between two points
• All voltages are relative or differential
• Voltmeters have two probes ...
• Always ask “Voltage with respect to what?”

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 19

Note that voltmeters invariably have two probes! All voltage readings are differential
... even between two “ground” points. And you don’t put one of the probes in your
pocket when you take readings.

Bill Whitlock 19
An Overview of Audio System Grounding & Interfacing 9/4/2012

Connections Creating Soil Currents

• Grounds at utility power
poles and towers
• Buried metal pipes
(water, gas, waste)
• Ground electrodes at
buildings
• Ground is tied to neutral
at each location shown
• Voltage drop of
neutral wires applied
to soil
• Forms complex network
of soil currents
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 20

Gas and water pipes are required by National Electrical Code® (hereinafter referred
to as NEC) to be connected to electrical power feeding the building ... making the soil
and/or the pipes a pathway for currents that flow from building to building. We’ll show
a slide about that toward the end of the class.

Bill Whitlock 20
An Overview of Audio System Grounding & Interfacing 9/4/2012

AC Power
Distribution

NEXT SLIDE

Courtesy of ONCOR
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 21

Gas and water pipes are required by NEC to be connected to electrical power feeding
the building ... making the soil and/or the pipes a pathway for currents that flow from
building to building. We’ll show a slide about that toward the end of the class.

Bill Whitlock 21
An Overview of Audio System Grounding & Interfacing 9/4/2012

Delivery of Utility AC Power

4,800 V
“Feeder”

Transformer 120 VAC

240 V “Split-
“Split-Single-
Single-Phase”

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 22

Normal residential utility power in North America is properly referred to as split-

single-phase power.

Bill Whitlock 22
An Overview of Audio System Grounding & Interfacing 9/4/2012

The 3-wire System

• NEC requires 120-volt AC premises power distribution using a 3-
wire system (since about 1960)
• LINE (black) and NEUTRAL (white) are intended to carry LOAD
current, typically up to 15 or 20 A in branch circuits
• The “safety” GROUND normally carries no current
• NEUTRAL and GROUND are bonded only at the main panel

NEC PROHIBITS NEUTRAL TO GROUND

CONNECTIONS ANYWHERE ELSE
NEUTRAL BUS
N-G BOND LINK

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 23

When new sub-panels are installed, a frequent electrician mistake is connecting the N-
G bond. This will be discussed later ...

Bill Whitlock 23
An Overview of Audio System Grounding & Interfacing 9/4/2012

Normal Current in Branch Circuit

“LOAD”
“SOURCE”

* For clarity, meter and other “phase” not shown

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 24

None of the load current flows in either safety ground or earth ground. All
load current flows back to its source, which is the utility company’s
transformer! Remember, current always flows back to the voltage source,
whether through an intentional or unintentional path. Electrons don’t care
... they don’t read schematics!

Bill Whitlock 24
An Overview of Audio System Grounding & Interfacing 9/4/2012

Grounding for Life Safety

• Exposed conductive parts (including signal connectors) can
become “energized” at 120 volts if the equipment develops
certain internal defects
• Insulation is used in power transformers, switches, motors and
other internal parts to keep the electricity where it belongs
• Insulation can and does fail, effectively connecting “live” power
to exposed metal – this is called a FAULT
• Without equipment safety grounding, people can be seriously
shocked or electrocuted!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 25

It’s extremely important to understand the intent of the safety ground

system!

Bill Whitlock 25
An Overview of Audio System Grounding & Interfacing 9/4/2012

Fault Current Trips Breaker

• Grounding wire in a power cord connects equipment’s exposed
metal to safety ground via 3-prong plug
• Outlet safety ground returns via green wire or metallic conduit to
neutral at main panel
• Low-impedance connection causes high fault current, which trips
breaker quickly!
• Typical fault currents range from 150 A to over 1,000 A
• Safety grounding must return to NEUTRAL – THE EARTH
GROUND CONNECTION IS IRRELEVANT!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 26

The impedance of earth ground is far too high to cause enough circuit current to trip a
circuit breaker.

Bill Whitlock 26
An Overview of Audio System Grounding & Interfacing 9/4/2012

FAULT Current Trips Circuit Breaker

Fault currents range

from 150 A to 1,000 A
NOT INVOLVED Trip times range from
under 10 ms to 2.5 s

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 27

The quoted range of fault currents and trip times is based on a UL study (circa 2003-
2005) of over 1,000 residential 15A and 20A outlets.

Bill Whitlock 27
An Overview of Audio System Grounding & Interfacing 9/4/2012

Earth Ground is for LIGHTNING

• Power lines are frequent targets of lightning

• Before NEC required grounding of power lines, they often guided
lightning strikes directly into buildings!
• Power distribution lines are grounded at intervals
• Ground rod impedance is low enough to control lightning
• NEC requires protection of every line that penetrates a
building

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 28

Lightning involves voltages in the hundreds of millions and currents ranging from
5,000 to 150,000 amps (an average strike is about 20,000 amps), briefly unleashing
incredible amounts of power. Neighborhood distribution power lines are typically
grounded at least every 1,500 feet.

Bill Whitlock 28
An Overview of Audio System Grounding & Interfacing 9/4/2012

Earth Ground Rod is Useless for Fault Currents

Circuit Breaker Will NOT Trip!

YIKES!

HIGH IMPEDANCE

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 29

Its connection to earth is not what makes “safety ground” safe ... its the
connection to neutral. The impedance of the soil between the ground rods
could be 25 ohms, causing less than 5 A to flow through the circuit breaker.
The chassis of the defective equipment would remain at nearly 120 volts – a
lethal shock hazard. A few years ago, the author attended a product training
session for a maker of power conditioners where the “engineer” actually
advocated this technique to provide “quiet ground” to an audio listening room!
He was subsequently barred from instructing at that trade show – after I
reminded them of their legal liability.

Bill Whitlock 29
An Overview of Audio System Grounding & Interfacing 9/4/2012

NEC Grounding Rules

• During lightning events, thousands of volts can be induced between
separate rods
• Can cause serious damage to any equipment that bridges them
• Telephone lines per NEC Article 800
• Telco NIU protector and its ground connection must be as close as
practicable to point of entry
• Use insulated, listed wire no smaller than #14 AWG
• Bond to main power grounding electrode if within 20 feet
• If not, use separate 5-foot ground rod and bond to utility power ground
with #6 AWG or larger wire
• Wiring should be as straight as possible, using gentle curves rather than
sharp bends, and physically protected (PVC conduit is preferred)
• CATV has similar, but not identical, rules (only summarized here)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 30

Soil has relatively high impedance, allowing very large voltages to develop during even
a modest lightning strike to one of them. PVC conduit is preferred because steel would
increase the inductance of the wire within, raising its impedance. The #6 bond wire
helps protect against power lines that fall onto telephone lines (due to an auto
accident, for example) by completing a path back to the utility power neutral. Sharp
bends in the wire raise its inductance, therefore making it less likely to contain the
strike energy ... it may jump out of the wire at the bend and strike something (or
someone) nearby.

Bill Whitlock 30
An Overview of Audio System Grounding & Interfacing 9/4/2012

The “Conduit Transformer”

• This finally explains what drives 99% of all ground loops!
• Load current in line and neutral produces opposing magnetic fields
since instantaneous current flow is in opposite directions
• Imperfect cancellation magnetically induces voltage over the
length of the nearby safety ground conductor
• Strongly affected by geometry and proximity of wires
• Highest voltages with randomly positioned wires in conduit
• Lower voltages with uniform geometry of Romex®
• Voltage is directly proportional to load current, wire length, and
rate of change in current or ∆I/∆t
• Mechanism favors high-frequency harmonics of 60 Hz
• For constant current in L and N, induced voltage rises at 6 dB/octave

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 31

The favoring of higher harmonics of 60 Hz is why we usually hear “buzz” more often
than the more fundamental 60 Hz “hum” and why light dimmers have such a horrible
reputation for “causing” noise problems.

Bill Whitlock 31
An Overview of Audio System Grounding & Interfacing 9/4/2012

Magnetic Fields in Romex®

Current in L and N are equal but flow in opposite directions

L ● ●N

Safety Ground “Sees”

Zero Magnetic Field at
Exact Midpoint
(cross-section view)
Instantaneous L and N currents are
flowing into page and out of page

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 32

There is a magnetic “null zone” exactly midway between line conductors

Bill Whitlock 32
An Overview of Audio System Grounding & Interfacing 9/4/2012

The “Conduit Transformer”

Wires randomly positioned in conduit
produce the worst possible results!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 33

The “elephant” that’s been in the room for years but, until recently,
unrecognized as THE major driving force behind ground loops! This author,
and my co-author Jamie Fox, presented an AES paper on this subject in November
2010.

Bill Whitlock 33
An Overview of Audio System Grounding & Interfacing 9/4/2012

How the Tests Were Performed

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 34

Excerpt from “Ground Loops: The Rest of the Story” by Bill Whitlock and Jamie Fox, a
paper presented 5 Nov 2010 at the AES 129th convention in San Francisco. Available
from AES as preprint #8234.

Bill Whitlock 34
An Overview of Audio System Grounding & Interfacing 9/4/2012

Results

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 35

Excerpt from AES paper Ground Loops: The Rest of the Story by Bill Whitlock and
Jamie Fox, presented 5 Nov 2010 at the 129th AES convention in San Francisco.
Available from AES as preprint #8234.

Bill Whitlock 35
An Overview of Audio System Grounding & Interfacing 9/4/2012

AC Power is Not Just 60 Hz

• Many loads, especially “brute force” electronic power supplies,
draw current in pulses only at the peaks of the voltage during each
cycle
• “Power Factor Corrected” or PFC power supplies alleviate this problem
• 120 VAC waveform distortion is typically 2% to 6% THD
• High harmonics are created by sudden current changes
• Typical low-cost light dimmers have a current risetime of about 5 µs,
producing strong harmonics to at least 70 kHz!
• The conduit transformer favors those harmonics and couples them into
safety ground wiring

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 36

And the most notorious generator of fast current changes is ... drum roll .....
LIGHT DIMMERS!

Bill Whitlock 36
An Overview of Audio System Grounding & Interfacing 9/4/2012

High-Frequency Noise Sources

• Phase-control light dimmers
• Fluorescent lamps
• Electronic power supplies
• Arcing switches, relays, or motors
• Power-line corona discharge Current at 50% brightness

• AM radio power line pickup

In the frequency domain, this

risetime creates very strong
harmonics up to 70 kHz.

Expanded at 5 µs/div

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 37

Now you can see why a light dimmer is at its worst (noise wise) when set for 50%
brightness – the line voltage is at maximum when it “snaps” on (taking only about 5
us). This was a Lutron “Skylark” dimmer operating with its full rated load of six 100 W
incandescent bulbs. Scale in upper oscilloscope image is 4 A/div.

Bill Whitlock 37
An Overview of Audio System Grounding & Interfacing 9/4/2012

Typical Power-Line Spectrum

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 38

An actual spectrogram of power-line AC leakage current through 1 nF of parasitic

(coupling) capacitance. Measured at a former location of Jensen Transformers in a
light industrial area of Van Nuys, CA. No light dimmers were operating, but most
lighting was conventional fluorescent and several pieces of office equipment and test
instruments were running – manufacturing was in another building at the time.

Bill Whitlock 38
An Overview of Audio System Grounding & Interfacing 9/4/2012

Parasitic Capacitances
• Exist in all real equipment connected to AC power
• Transformer winding to winding (not shown in schematics)
• Internal EMI filters
• Capacitances cause leakage current to flow from power line to
chassis in real-world equipment
• UL established safety limits for “listed” equipment
• 3.5 mA for devices with 3-prong plugs
• 5 mA for devices with 3-prong plugs and a special warning label
• 0.75 mA for devices with 2-prong plugs

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 39

5 mA is a shock that’s very unpleasant, but not life-threatening. 0.75 mA, as

mentioned earlier, results in just a tingling sensation.

Bill Whitlock 39
An Overview of Audio System Grounding & Interfacing 9/4/2012

Equipment with 2-prong AC Plugs

Leakage current flows to any ground and/or

between devices to eventually return to neutral

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 40

When capacitors have AC voltages impressed across them, they must constantly
charge and discharge to the varying voltage. This is the origin of 60 Hz “leakage”
current.

Bill Whitlock 40
An Overview of Audio System Grounding & Interfacing 9/4/2012

About 2-prong Plugs

• UL listing demands extraordinary measures:
• Insulation “creepage” distances must be large
• One-shot thermal cutoffs must be embedded in transformers and motors that
might overheat
• Equipment must remain safe in spite of component failure, overload, and
mechanical abuse
• Chassis voltage with respect to ground may approach 120 V but
available current is harmlessly low
• But it’s enough to drive a high-impedance voltmeter to readings that frighten
those who don’t understand!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 41

Even external “wall-wart” transformers must be thermally protected and well insulated
to get a UL listing.

Bill Whitlock 41
An Overview of Audio System Grounding & Interfacing 9/4/2012

How AC Wiring Affects A/V Systems

• Hum is rarely caused by “improper” grounding
• Properly installed, fully NEC-compliant premises wiring normally
creates small voltage differences between outlet safety grounds
• Generally, safety ground voltage differences:
• Are lowest (milli-volts) between nearby outlets on the same branch circuit
• Are highest (often over 1 volt) between distant outlets on different branch
circuits
• Over 3 volts indicate a serious wiring error that should be
investigated immediately!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 42

To meet the NEC guideline of no more than 5% voltage drop (6 volts for a
120-volt circuit) at the farthest outlet on a branch circuit under full load, the
drop per conductor would be 3 volts or less. Therefore, 3 volts is the maximum
voltage difference that should ever exist between neutral (N) and safety ground (G) at
an outlet. Even if neutral and safety ground are swapped somewhere on the branch
circuit, this places an upper limit of 3 volts on the voltage seen between any two
safety grounds.

Bill Whitlock 42
An Overview of Audio System Grounding & Interfacing 9/4/2012

The Facts of Life

• Voltage differences will always exist between grounds of outlets

• Lowest between nearby outlets on the same branch circuit
• Highest between distant outlets on different branch circuits
• Over 3 volts between any outlets indicates a serious wiring error!
• Leakage currents will always flow in signal cables connected to
equipment powered via 2-prong AC plugs
• This occurs in every properly wired, fully Code-compliant premises

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 43

Leakage currents are an issue only with unbalanced interconnections. Voltage

differences, due almost entirely to the conduit transformer effect, are disaster to
unbalanced interconnections and the driving force behind problems in balanced
interconnections.

Bill Whitlock 43
An Overview of Audio System Grounding & Interfacing 9/4/2012

Ground Loop Links Aggressor and Victim

Signal Cable

VICTIM

AGGRESSOR
Safety Ground Wiring
Ground Voltage Difference

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 44

Bill Whitlock 44
An Overview of Audio System Grounding & Interfacing 9/4/2012

SIGNAL INTERFACES & NOISE COUPLING

• An interface consists of a line driver (device output), the line or
cable itself, and a line receiver (device input)
• Two conductors are always required to complete a signal circuit
• The purpose of modern interfaces is to transfer signal voltage
• Before amplifiers existed, interfaces like 1900 telephone systems were
designed to transfer maximum power ― but those days are long gone!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 45

Bill Whitlock 45
An Overview of Audio System Grounding & Interfacing 9/4/2012

Unbalanced vs Balanced
• Status depends ONLY on IMPEDANCES, with respect to ground,
of the two signal conductors!
• In unbalanced interfaces, one impedance is zero (grounded) while
the other is significantly higher
• In balanced interfaces, impedances are equal
• Since they’re all in parallel, this requires that driver, line, and receiver each
have equal impedances to ground

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 46

This is, without a doubt, the single most misunderstood concept about
interfaces!

Bill Whitlock 46
An Overview of Audio System Grounding & Interfacing 9/4/2012

Unbalanced vs Balanced

Grounded

Unbalanced

Balanced

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 47

Bill Whitlock 47
An Overview of Audio System Grounding & Interfacing 9/4/2012

Line Drivers
• Every driver has an internal impedance called output impedance
or “output source impedance,” shown as Zo
• Zo is often confused with load impedance
• Load impedance is external to the equipment and generally consists of cable
capacitance in parallel with the input impedance of the receiving input
• Sadly, manufacturers often omit Zo and specify only a minimum load
impedance … which tends to perpetuate the confusion
• Low Zo is desirable but it can’t reach 0 Ω in real-world equipment
• IEC standard 61938 (2010) specifies
• Zo ≤ 2.2 kΩ for “general-purpose consumer” outputs
• Zo ≤ 1 kΩ for “general-purpose professional” outputs
• Zo ≤ 50 Ω for “broadcast and similar line amplifier” outputs

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 48

Many, many people confuse “output impedance” with “load impedance” –

and manufacturers are no help as they continue to confuse customers by
omitting source impedance!

Bill Whitlock 48
An Overview of Audio System Grounding & Interfacing 9/4/2012

Line Receivers
• Every receiver has an internal impedance called input
impedance, shown as Zi
• High Zi is desirable but it can’t reach infinity in real-world equipment
• IEC standard 61938 (2010) specifies
• Zi ≥ 22 kΩ for “general-purpose consumer” inputs
• Zi ≥ 10 kΩ for all professional and broadcast line amplifier inputs
• A legacy term, relating to 600 Ω telephone lines, for high-impedance
(≥ 10 kΩ) inputs is “bridging” since many of them can be driven
simultaneously by a low-impedance (≤ 600 Ω) source

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 49

For balanced inputs, Zi is the “differential” input impedance … i.e., between

the two lines with no reference to ground

Bill Whitlock 49
An Overview of Audio System Grounding & Interfacing 9/4/2012

An Interface is a Series Circuit

• Driver Zo and receiver Zi form a series circuit when connected
• Since voltage drop is to impedance in a series circuit, it is often
called a voltage divider
• Maximum signal voltage across Zi, the receiver input, is developed
when Zi is at least 10 times Zo
• Typical audio interfaces transfer 90% to 99.95% of the available
signal voltage
• 50 Ω to 1 kΩ is a typical range for Zo
• 10 kS to 100 kS is typical range for Zi
• Because video interfaces must use terminated true transmission
lines, where Zo = Zi = 75 Ω typically, they transfer 50% of signal
voltage

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 50

Making input impedances much, much higher than output impedances is sometimes
called “voltage matching”, but I think the term is somewhat misleading.

Bill Whitlock 50
An Overview of Audio System Grounding & Interfacing 9/4/2012

Voltage Dividers in Interfaces

Unbalanced

Balanced

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 51

Arrows indicate the signal current path.

Bill Whitlock 51
An Overview of Audio System Grounding & Interfacing 9/4/2012

Impedance “Matching”
• It’s a myth that audio output and input impedances, Zo and Zi,
should have the same or matching values!
• Apparently based on use of 600 Ω “matching” by telephone companies
• If impedances are matched, maximum power is transferred
• Impedance matching wastes half the signal voltage and forces the line
driver to drive an unnecessarily heavy load
• As shown earlier, all modern audio interfaces, including power
amplifier to speaker, are designed to transfer voltage, not power!
• However, depending on the highest frequency it must pass and its
physical length, a cable may behave as a true transmission line
that does require the same impedance at each end.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 52

As we showed earlier, maximum voltage is transferred by using low source

impedances and high load (or input) impedances.

Bill Whitlock 52
An Overview of Audio System Grounding & Interfacing 9/4/2012

Termination of Transmission Lines

• Transmission line effects become significant when the cable’s
physical length becomes about 10% of an electrical wavelength at
the highest signal frequency of interest
• A wavelength is the distance traveled by the signal during one cycle
• Signals travel in cable at about 70% to 90% of the speed of light
• Characteristic impedance is the impedance of an infinite length of cable
• Signals will reflect in the cable, causing video ghosts or data errors, unless
terminated in its characteristic impedance at each end
• Video, RF, and data cables over a few inches long need terminations
• For analog audio cables, termination is unnecessary!
• 10% of a wavelength at 20 kHz in cable with a propagation velocity of 70%
is 3,442 feet … if your cable is longer than this, consider terminations

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 53

Characteristic impedance of a particular cable depends on its physical dimensions

(wire size and spacing) and the insulating material (air, foam, plastic, etc.). Typical
coaxial cables come in 50 Ω and 75 Ω varieties and typical twisted-pair (CAT-5 for
example) cables range from 100 Ω to 130 Ω. Calling a 6-foot audio cable a
“transmission line” is bullshit ...

Bill Whitlock 53
An Overview of Audio System Grounding & Interfacing 9/4/2012

Why 600 Ohms?

• First US telephone systems used
existing telegraph lines
• Miles long ... they are true audio
transmission lines
• Wire size/spacing gives them a
600 Ω characteristic impedance The 600 Ω “standard” does
• Became Bell standard for system NOT apply to modern audio
transformers, filters, etc. systems!
• Equipment “migrated” to early The term “600 Ω“ is often
radio and recording systems … misappropriated to mean
simply “line-level.”
• Remains “standard” in some
communications systems

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 54

And old legendary practices fade away very, very slowly!

Bill Whitlock 54
An Overview of Audio System Grounding & Interfacing 9/4/2012

UNBALANCED Audio Interfaces

• INHERENTLY SUSCEPTIBLE to noise coupling!
• It’s truly ironic that they still dominate consumer audio after 50
years, even though the dynamic range of readily-available program
material has steadily increased!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 55

When vinyl records and FM radio were the “quality” audio media, 70 dB of dynamic
range was enough. But today, any consumer can access CD or digital broadcast media
with 100 dB or more dynamic range.

Bill Whitlock 55
An Overview of Audio System Grounding & Interfacing 9/4/2012

Irony of the Unbalanced Interface

• “RCA” or “phono” connector appeared about 1948
• Home audio evolved into separate “components” in the 1960s and
used ever longer cables to interconnect them
• Unbalanced interfaces remain the norm in consumer gear, home
theater, and even in outrageously expensive “audiophile” gear
• Balanced interfaces, hallmarks of truly professional systems, avoid
the inherent noise problem with unbalanced interfaces!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 56

When talking audio, the words “unbalanced” and “professional” should

never appear in the same sentence! And this author is still trying to figure out
exactly what “semi-professional” and “musical-instrument grade” mean!

Bill Whitlock 56
An Overview of Audio System Grounding & Interfacing 9/4/2012

The “Good Ol’ Days”

• In 1960, most home theater
systems were self-contained
• RCA cables were all internal
and very short
• A single AC cord, with no safety
ground, powered it all
• Life was good!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 57

I graduated high school in 1962 and spent my last two years working part-time in a
TV repair shop (Webb’s City in St. Petersburg, FL). I worked on many of these TV,
AM-FM radio, and phonograph “combo” units. Invariably, they contained at least two
separate chassis inside – interconnected with numerous very short RCA cables.

Bill Whitlock 57
An Overview of Audio System Grounding & Interfacing 9/4/2012

“Component” Systems Changed Everything …

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 58

And consumers were largely “on their own” to solve hum and buzz problems ... most
manufacturers, as today, just pretended the problems didn’t exist and that
it’s a fanciful “plug-and-play” world.

Bill Whitlock 58
An Overview of Audio System Grounding & Interfacing 9/4/2012

… to Unbalanced Nightmares!

Courtesy CHT Systems, Inc.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 59

This is just the video portion of a huge AV system in a 10,000 square-foot residence.
It’s only reasonable to EXPECT ground loop noise issues in any system covering such
large physical distances with unbalanced interconnections.

Bill Whitlock 59
An Overview of Audio System Grounding & Interfacing 9/4/2012

The BIG Problem with Unbalanced

• When two devices, each having its own power cord, are
connected by an audio cable, a small power-line current will flow
in the audio cable … a fact of life
• Essentially all this current flows in the grounded conductor, typically
the “shield” of audio cables
• Since this conductor has resistance, a noise voltage appears over its
length … per Ohm’s law
• The noise voltage is directly added to the signal seen at the
receiver … just like two batteries in series

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 60

Maybe if we have room-temperature super-conducting wire someday ...

Bill Whitlock 60
An Overview of Audio System Grounding & Interfacing 9/4/2012

Common-Impedance Coupling

Two currents, power-line and signal, flow in the

same impedance, which allows them to couple!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 61

Technically, this coupling mechanism is called “common-impedance coupling” because

an impedance (resistance) is “common” (shared) to two circuits. One circuit is
between device A output and device B input (the signal circuit) and the other is
between the AC power connections at each device. In this example, since there are no
safety ground connections, the current in the second circuit is due to “leakage
current” (which flows from power-line to chassis in each device). The “common
impedance” is, of course, the resistance of the grounded conductor (usually, but not
always, the shield) in the interconnect cable.

Bill Whitlock 61
An Overview of Audio System Grounding & Interfacing 9/4/2012

Leakage Current Effect

< 0.75 mA

Neither device is grounded

120 V 0V

Generally OK if interconnect cable is under 10 feet long

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 62

The “parasitic capacitance” in each device causes a small AC current to flow. Just as
with a battery, current must flow into a capacitor to increase the voltage across it – or
“charge” it. Conversely, current must flow out of a capacitor to decrease the voltage
across it – or “discharge” it. Since these capacitors connect to the 120 VAC 60 Hz AC
power, the voltage across them is constantly changing and a predictable amount of AC
current will flow in the process. For equipment with UL listing and a 2-prong AC power
connection, this current can be no higher than 0.75 mA – just enough to cause a
slight tingling sensation if it flows in your body. Although harmlessly low, this current
can cause enough voltage drop over the length of the signal interconnect cable to add
significant “buzz” to the audio. While it’s remotely possible that leakage current could
cause a “hum bar” issue in video, this author has never seen or heard of such a case.

Bill Whitlock 62
An Overview of Audio System Grounding & Interfacing 9/4/2012

Leakage Current Effect - A Calculated Example

• A 25-foot cable (foil shield, #26 AWG drain wire) has an end-to-end
shield resistance of 1 Ω
• Measured leakage current between the ungrounded devices is 316
µA (well under the UL limit of 750 µA)
• From Ohm’s law, noise voltage E = I x R = 316 µA x 1 Ω = 316 µV
• Consumer −10 dBV reference level = 316 mV
• Signal to Noise ratio = 20 x log (316 mV⁄316 µV) = 60 dB
• This is 35 dB worse than an audio CD!
• Same length of Belden #8241F cable, with its shield resistance of
only 0.065 Ω, makes S⁄N 84 dB, an improvement of 24 dB!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 63

From a noise perspective, shield resistance is the most important parameter

of all. How many times have you seen it specified on a data sheet or in
advertising hype? I rest my case about clueless manufacturers!

Bill Whitlock 63
An Overview of Audio System Grounding & Interfacing 9/4/2012

It Gets Far, Far Worse!

• The ugly truth: unbalanced interfaces simply can’t deliver on
the “plug-and-play” promise
• Only the smallest of systems in the most benign imaginable electrical
environments could possibly live up to the promise
• Turn the volume to maximum and listen closely ― all you should hear is hiss
• If unbalanced interfaces exist between two grounded points in a
system, hum and buzz can easily become larger than the signal!
• Ground voltage differences between outlets is effectively applied across the
ends of the audio cables and directly added to the signal
• Other “alien” ground connections, such as CATV or DBS, can create
comparable voltage differences

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 64

No hum/buzz at all, or hum/buzz that’s barely audible in the hiss, should be

the goal for a serious listening system.

Bill Whitlock 64
An Overview of Audio System Grounding & Interfacing 9/4/2012

Voltage Difference Effect

Ground Voltage Difference

Even a few millivolts here can be noise disaster!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 65

Remember in our previous example with ungrounded (2-prong) power cords, the
voltage across the cable ends was only 316 µV. Imagine what a rather common
30 mV of ground voltage difference between outlets would do to S/N ratio!
Actually, the resulting S/N would be only about 20 dB … unbearable to most listeners.
Even installing a “cheater” (3-to-2 prong adapter) or cutting off a safety
ground pin from one of the power cords would still leave as much as 3.5 mA
flowing in the interface cable … that’s 10 times the current of the previous 2-
prong example, so a serious noise problem would still exist … as well as a
much more serious safety hazard!

Bill Whitlock 65
An Overview of Audio System Grounding & Interfacing 9/4/2012

THE VIDEO HUM BAR

“A dark, horizontal bar in a TV picture caused

by hum interference in the (analog) video
signal.”

• In NTSC video, the slow upward movement is the “tell-tale”

signature of coupling from 60 Hz AC
• Movement caused by the 0.06 Hz frequency difference between the 59.94 Hz
NTSC field rate and 60.00 Hz AC power
• Takes 16 seconds for each of 2 bars to move from bottom to top of screen
• The pair of disturbances (120 Hz rate) occurs because power-line load
current induces the voltage in the “conduit transformer” and it has a positive
and negative peak in each cycle
• In component color interfaces, the bar is often a distinctive color

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 66

The symptoms for digital video interfaces such as DVI or HDMI are completely
different.

Bill Whitlock 66
An Overview of Audio System Grounding & Interfacing 9/4/2012

Coupling in Video Cables

• Common-impedance coupling – again

• This time, only half of ground voltage difference adds to signal
because of 75 Ω terminations at the driver and receiver

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 67

Just as in the unbalanced audio interface, tiny voltage drops appear over the length of
the cable’s shield. This voltage adds itself to the signal seen at the receive end of the
cable.

Bill Whitlock 67
An Overview of Audio System Grounding & Interfacing 9/4/2012

What’s Visible
• Standard analog video signal is 1 V peak-to-peak
• About 700 mV produces visible black to white range
• About 7 mV peak-to-peak hum bar interference is visible
under worst-case conditions
• Equals 14 mV peak-to-peak across cable ends
• “Pedestal clamping,” a technique used in some video input
circuits, can increase tolerance to hum bars

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 68

Pedestal clamping is also sometimes called “DC Restoration”.

Bill Whitlock 68
An Overview of Audio System Grounding & Interfacing 9/4/2012

The BALANCED Interface

• Inherently immune to noise coupling – that’s why professionals use
it!
• Telephone companies popularized it and use it to this day
• But its true nature is widely misunderstood:
“Each conductor is always equal in voltage but opposite in
polarity to the other. The circuit that receives this signal in the
mixer is called a differential amplifier and this opposing polarity
of the conductors is essential for its operation.”
This WRONG explanation, like many others in print, doesn’t
even mention the defining property of a balanced
interface!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 69

This misconception has led to lots of bad equipment design ... and to the old
IEC test method

Bill Whitlock 69
An Overview of Audio System Grounding & Interfacing 9/4/2012

It’s All About Impedance!

“Only the common-mode impedance balance of the driver, line, and

receiver play a role in noise or interference rejection. This noise or
interference rejection property is independent of the presence of a
desired differential signal. Therefore, it can make no difference
whether the desired signal exists entirely on one line, as a greater
voltage on one line than the other, or as equal voltages on both of
them. Symmetry of the desired signal has advantages, but they
concern headroom and crosstalk, not noise or interference rejection.”
IEC Standard 60268-3 (2000)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 70

I finally convinced them ... now they “get it”

Bill Whitlock 70
An Overview of Audio System Grounding & Interfacing 9/4/2012

An Accurate Definition
“A balanced circuit is a two-conductor circuit in which both conductors
and all circuits connected to them have the same impedance with
respect to ground and to all other conductors. The purpose of
balancing is to make the noise pickup equal in both conductors, in
which case it will be a common-mode signal which can be made to
cancel out in the load.” Henry W. Ott, Distinguished Member of the Technical Staff, AT&T Bell Labs

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 71

Henry Ott and Ralph Morrison are two physicists whose work I truly admire!

Bill Whitlock 71
An Overview of Audio System Grounding & Interfacing 9/4/2012

Balanced Interface Circuit

The desired signal is “differential-mode” or DM

Ground noise and interference creates “common-mode” or CM

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 72

Drivers can have different topologies, but this one, where the outputs are virtually
grounded by the driver, is the most common ... and easiest to understand.

Bill Whitlock 72
An Overview of Audio System Grounding & Interfacing 9/4/2012

About Common-Mode
• A voltage that appears identically on both lines or input
terminals is called Common-Mode or CM voltage
• Three main sources of common-mode voltage:
• Voltage difference between driver and receiver grounds
• Voltage induced in cable by a magnetic field
• Voltage induced in cable by an electric field
• A balanced line receiver can reject any interference that results in
identical voltages at its two inputs
• “Identical” ± 0.1% = 60 dB rejection, ± 0.01% = 80 dB rejection, etc.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 73

Extreme matching is required to achieve useful noise rejection.

Bill Whitlock 73
An Overview of Audio System Grounding & Interfacing 9/4/2012

Common-Mode Rejection Ratio or CMRR

• Ideal receiver would respond only to differential voltage – with no
response whatsoever to common-mode voltage
• Real receiver is either an “active” circuit or a transformer
• Limited common-mode rejection
• The ratio of DM (signal) gain to CM (noise) gain is called Common-Mode
Rejection Ratio, or CMRR
• Commonly expressed in dB
• Not a measure of signal level − it’s wrong to express it as dBu, dBV, or dBm!
• Is a positive number (not negative) for any useful receiver

• Higher CMRR figures mean better noise rejection

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 74

CMRR is the most useful, and widely used, measure of noise rejection in a
balanced interface.

Bill Whitlock 74
An Overview of Audio System Grounding & Interfacing 9/4/2012

A Balanced Interface is a Wheatstone Bridge

• CM impedances form the bridge
GROUND
• If not precisely “nulled,” the bridge
converts ground noise Vcm to signal
• A “null” requires precise ratio match of
the + and − voltage divider arms
BALANCED LINE
• Most sensitive to component tolerances
when all arms have the same impedance
• Least sensitive when upper and lower arms
have widely differing impedances – the
wider the better!
• Although low driver and high receiver CM
GROUND impedances are the norm, the Zcm of
most receivers is not nearly high
enough!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 75

A Wheatstone bridge uses a generator and a detector. The resistance in one

of the arms is adjusted for a “null” in the detector. In the case of the
balanced interface, the “generator” is the ground voltage difference and the
“detector” is the balanced line receiver itself.

Bill Whitlock 75
An Overview of Audio System Grounding & Interfacing 9/4/2012

“Reality” vs “Advertised” CMRR Specs

• Zcm tolerance of real-world drivers typically determined by ± 5%
series resistors and ± 20% (or worse) coupling capacitors
• Typical Zcm imbalances are about ± 10 Ω at 60 Hz
• Typical real-world “active balanced” line receivers have Zcm ranging from
10 kΩ to 50 kΩ
• This low receiver Zcm makes real-world INTERFACE CMRR
exquisitely sensitive to normal driver Zcm imbalances!
• CMRR of the popular SSM-2141 suffers a 25 dB drop with only a 1 Ω
imbalance in driver Zcm
• Noise rejection or CMRR of an interface may be much less
than the CMRR number advertised for the line receiver,
depending on exactly what equipment is driving it!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 76

In fact, most electrolytic capacitors have even looser tolerances, commonly +80%, -
20%. To make matters worse, they also change capacitance with age! However, the
larger their capacitance, the less imbalancing effect their tolerance and aging will
have.

Bill Whitlock 76
An Overview of Audio System Grounding & Interfacing 9/4/2012

Bad Tests Lead to Bogus “Specs”

• Only a complete interface can have CMRR because the driver,
cable, and receiver critically interact!
• Traditional CMRR tests completely ignore “real-world” signal source
impedance imbalances that can drastically reduce CMRR
• Most commonly, the receiver inputs are shorted together, simulating
a “bogus” source with zero impedance and zero imbalance
• Similarly, the former IEC test source had real impedance but it was “tweaked”
for zero imbalance – to laboratory perfection!
• Real-world signal sources tested had impedance imbalances ranging
from 0.3 Ω to 30 Ω, so my new test specifies a 10 Ω imbalance
• My new real-world test method was adopted and published August 2000 in
IEC Standard 60268-3, Sound System Equipment - Part 3: Amplifiers

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 77

In 1999, the IEC issued a call for comments about the existing CMRR test.
They recognized that its results had little correlation with results in real-
world audio systems ... duh!

Bill Whitlock 77
An Overview of Audio System Grounding & Interfacing 9/4/2012

Tests for CMRR – Old vs New

Establishes 0 dB
reference for
common-mode gain

Old – Tweaked Until Perfect

New – Imbalanced ±10 Ω

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 78

The old test used a very clever method make sure the source impedances to
the two DUT inputs were perfectly matched. Unfortunately, this perfection
was its undoing!

Bill Whitlock 78
An Overview of Audio System Grounding & Interfacing 9/4/2012

Audio Precision Model APx525

First Commercial Analyzer to Incorporate the New IEC Test

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 79

Bill Whitlock 79
An Overview of Audio System Grounding & Interfacing 9/4/2012

How Much CMRR is Enough?

• Professional reference signal level is +4 dBu = 1.23 V
• A “hostile” 1.23 V ground voltage difference (common-mode
voltage) might exist in a system
• In this case only, SNR (signal-to-noise ratio) = Interface CMRR
• Total dynamic range = SNR + “headroom”
• If headroom is 15 dB, CMRR of 80 dB would result in 95 dB of total
dynamic range, which is comparable to an audio CD
• A “less hostile” 123 mV ground voltage difference would increase dynamic
range by 20 dB
• 80 dB of “real-world” CMRR is generally acceptable, but 50
dB is rarely enough for a commercial sound system

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 80

According to my esteemed colleague Neil Muncy, “50 dB is NOT enough

CMRR” in most systems.

Bill Whitlock 80
An Overview of Audio System Grounding & Interfacing 9/4/2012

The “Pin 1 Problem” − A Designed-In Defect

• Common-impedance coupling that occurs inside equipment, turning
their shield connections into very-low-impedance signal inputs!
• Dubbed the “pin 1 problem” (XLR pin 1 = shield) by Neil Muncy in
his 1994 AES paper
• This defect has been inadvertently designed into an alarming
number of products … and, sadly, it continues
• It allows shield current (power-line noise) to flow in wires
or PCB traces shared by internal amplifier circuitry
• This problem can exist at I/O ports, whether analog or
digital, in any piece of equipment

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 81

Neil Muncy’s comprehensive paper about this was published in the June 1995 AES
Journal. Since wires and PCB traces have small but significant resistances, the current
flow is converted to small voltages that couple into the signal path ... adding noise to
the output signal.

Bill Whitlock 81
An Overview of Audio System Grounding & Interfacing 9/4/2012

Good Designs Avoid “Pin 1 Problem”

“Hybrid”
Ground

Non-Metallic Metal Metal

Housing Chassis Chassis

Shield currents “want” to flow back to safety ground –

giving them a separate, direct route avoids the problem!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 82

Hybrid grounding is a technique that allows the shield to effectively be grounded at

high frequencies (RF) but prevents problems associated with low-frequency (60 Hz
and audio) current flow in shields. This topic will be mentioned later ...

Bill Whitlock 82
An Overview of Audio System Grounding & Interfacing 9/4/2012

CAUTION, “Pin 1 Problem” Inside!

Shield current flows in signal reference “ground” inside equipment

Power-line noise, coupled by the power supply, flows in signal reference ground in
two of the boxes. This design error produces so-called “sensitive” equipment!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 83

The tiny voltage drops created along the “SIG REF” conductors are amplified by the
circuitry, resulting in noise at the output. So-called “sensitive” equipment
contains a related design error that causes power line noise (coupled
through the power supply) to flow through signal reference ground on its
way to safety ground (to return to its source). I think a more appropriate
name for such equipment is “power-line primadonna”.

Bill Whitlock 83
An Overview of Audio System Grounding & Interfacing 9/4/2012

CABLE ISSUES
• “A cable is a source of potential trouble connecting two other
sources of potential trouble.”
• Electric Field Shielding
• Provided by copper braid or aluminum foil
• Resistance critical for unbalanced
• Construction critical for balanced
• Capacitance
• Can reduces audio bandwidth in long runs
• Magnetic Field Shielding
• Provided by coaxial construction for unbalanced
• Provided by twisting of inner signal pair for balanced
• Copper and aluminum have no effect on LF magnetic fields

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 84

Strange that cable manufacturers still don’t “get it” ... they seem to think
“floobydust” is more important.

Bill Whitlock 84
An Overview of Audio System Grounding & Interfacing 9/4/2012

Electric Field Shielding

• Electric fields exist when voltage between two points is high
• Capacitance Cc is formed by air space between conductors
• AC voltage causes AC current flow through Cc
• Current flows in signal line without shield, creating a noise voltage
• Grounded shields divert noise current harmlessly to ground
• Without shielding, noise current may couple unequally to the two
signal conductors of a balanced interface (creating noise signal)
• Twisting improves match by averaging physical distances to the field
• Braided shields with ≥ 85% coverage are usually entirely adequate
• An issue with high I/O impedances in vintage vacuum-tube gear but
trivial in modern systems
• Some “audiophile” cables have no shielding at all — wires are woven
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 85

Capacitance exists between any two conductive objects, whether intentional or not.
Closer physical spacing and increased surface area both increase capacitance.

Bill Whitlock 85
An Overview of Audio System Grounding & Interfacing 9/4/2012

Electric Field Coupling and Shielding

Current couples
to signal

Current diverted
to ground

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 86

Shielding is a diversionary tactic, which stops coupling to one conductor by enclosing it

in another.

Bill Whitlock 86
An Overview of Audio System Grounding & Interfacing 9/4/2012

Balanced Lines – Ground at Driver or Receiver?

95 Ω

105 Ω BAD
3.22nF 3.38nF

95 Ω

105 Ω
GOOD
3.22nF 3.38nF

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 87

Unless you’re familiar with filters, this may not make much sense to you ...

Bill Whitlock 87
An Overview of Audio System Grounding & Interfacing 9/4/2012

“Good” vs “Bad” Balanced Shield Connections

• Commercial cables have capacitances from each signal conductor to
shield are typically mismatched by 4% to 6%
• It’s because the two wires have different colors … honest!
• Common-mode output impedances of drivers can also have 5% mismatch
• A shield grounded only at the receiver forms pair of low-pass filters
for common-mode noise … bad
• The two mis-tracking filters will convert a portion of common-mode (noise)
to differential-mode (signal), degrading CMRR at high audio frequencies
• A shield grounded only at the driver eliminates both filters … good!
• Because all filter elements move together at the driver ground voltage

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 88

Wires of different colors must be made on different machines – the machines that
extrude the liquid insulation onto the wire as it is pulled through a die. Die diameter
has a tolerance on the order of 0.001”, according to Belden. Therefore, two wires of
different color but otherwise identical, can have insulation thickness differences.
Calculations of capacitance, using the dielectric constant of the PVC insulation,
generally confirm capacitance differences (to the shield) in the vicinity of 5%.

Bill Whitlock 88
An Overview of Audio System Grounding & Interfacing 9/4/2012

Shield Connections and Crosstalk

• Driver signal asymmetry and cable capacitance mismatch cause
signal current flow in the shield at high audio frequencies
• Shields grounded only at the receiver force this audio current to return to
the driver via an undefined path – often resulting in crosstalk, distortion,
or oscillation problems as it flows in sensitive circuitry!
• Shields grounded only at the driver allows this audio current to return
directly to where it came from – the line driver and avoid problems
• Always ground the driver end of a balanced cable, whether or
not the receiver end is grounded
• It’s OK to ground both ends – experience has proven it but, for
the ultimate attainable interface CMRR, ground only at the driver
end

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 89

I remember a prototype console checkout at Quad-Eight (when I worked there around

1973) where the internal wiring grounded shields only at the receive end. As the fader
was pushed up, with no signal input, suddenly the output VU meters slammed into the
pegs. The entire signal chain was breaking into oscillation at 300 kHz!

Bill Whitlock 89
An Overview of Audio System Grounding & Interfacing 9/4/2012

Immunity to Magnetic Fields

• Unbalanced cables of coaxial construction are essentially immune to
AC magnetic coupling
• Centerlines of the two conductors are coincident, so there is no “loop area”
• Balanced cables may have mismatched voltages induced into the two
signal conductors, the difference becoming noise added to the signal
• TWISTING improves matching by averaging physical distances of each wire
to the external field source (just as with electric fields)
• Effective magnetic shielding for 60 Hz fields requires enclosing
conductors in steel conduit or special magnetic alloys
• The most effective prevention comes from
• increasing the distance between magnetic offender and victim
• avoiding open loops in the two signal conductors

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 90

Remember that mated XLR barrels and “punch-down block” open up

twisting, thus creating vulnerable “loops” for magnetic field pickup. Of
course, this is especially important for mic-level cabling.

Bill Whitlock 90
An Overview of Audio System Grounding & Interfacing 9/4/2012

Magnetic Fields from Power Cords

Current in L and N are equal but of opposite polarity

L N

Strong magnetic fields exist only very near the conductors

(cross-section view)
Instantaneous L and N currents are
flowing into and out of this page

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 91

For power cords of normal construction, either twisted (round) or parallel (flat),
magnetic field cancellation is near-perfect unless you get very close to it.

Bill Whitlock 91
An Overview of Audio System Grounding & Interfacing 9/4/2012

Magnetic Fields from Power Cords

Far from cord, magnetic field cancellation is nearly complete

At distance of 10 times the conductor spacing,

magnetic field is about 1% of close-in value

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 92

At a distance of 2” or 3”, magnetic fields from power cords are negligibly

weak.

Bill Whitlock 92
An Overview of Audio System Grounding & Interfacing 9/4/2012

Nullifying Magnetic Coupling

Wires exactly 90° to each In a “hairpin loop,” equal

other have zero coupling induced voltages series
oppose and cancel ...
twisting further improves

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 93

It’s the physics of magnetic fields ... we’ll see later that this principle is used in the
“dummy tests” to determine if magnetic fields are inducing noise into a signal cable
(we short one end of the cable as shown with the “hairpin loop” and listen to the
voltage at the other end).

Bill Whitlock 93
An Overview of Audio System Grounding & Interfacing 9/4/2012

Shield Current Induced Noise (SCIN)

• Current flow in cable shield creates a magnetic field extremely
close to the twisted pair inside
• Physical imperfections in real cables result in unequal induced
voltages, which adds the difference to the signal
• Dubbed SCIN in 1994 AES paper by Neil Muncy
• Best cables use braided or counter-wrapped spiral shields to create a very
uniform magnetic field around the twisted-pair
• WORST cables use a DRAIN WIRE, which “hogs” the shield current and
distorts the magnetic field (whether it has a braid or foil shield)
• For low-level signals, cables with drain wire should either be
avoided or connected so that no current flows in the shield (i.e.,
grounded at one end only)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 94

Pronounced “S” “C” “I” “N”, not “skin.” The “skin-effect” is quite different and
we won’t discuss in this class. In addition to Neil’s original paper, Jim Brown and I
have also written AES papers on the subject.

Bill Whitlock 94
An Overview of Audio System Grounding & Interfacing 9/4/2012

“Hybrid” Shield Grounding

• Resolves the audio frequency vs RF conflicts about shield
grounding for balanced audio cables
• At audio frequencies, grounding shields only at the driver produces
the best possible interface CMRR
• But, at radio frequencies, grounding shields at the receiver often
reduces or eliminates RF interference
• “Hybrid” grounding uses a capacitor to ground the receiver end of
the cable at RF frequencies only
• AES Standard on Interconnections, AES48, issued in 2005 (after years of
agonizing debate) includes this method
• The capacitor must be constructed and connected in a way that makes it
effective up to at least 1 GHz … and that’s not trivial!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 95

The hybrid grounding technique generally uses a capacitor value in the 10

nF to 100 nF (that’s 0.01 µF to 0.1 µF) range. To be effective at high
frequencies (over a few MHz), the capacitor leads must be extremely short!

Bill Whitlock 95
An Overview of Audio System Grounding & Interfacing 9/4/2012

The Neutrik EMC Connector

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 96

The first commercial connector to embrace the “hybrid” grounding idea. The
ferrite bead functions to impede the ingress of RF to the inside of a shielded
equipment enclosure, while the capacitor ring passes RF currents to the outside of
the equipment enclosure. The capacitive pathway remains low-impedance to
frequencies as high as 1 GHz. Thank you Neutrik!

Bill Whitlock 96
An Overview of Audio System Grounding & Interfacing 9/4/2012

High Capacitance Cables

• Some “exotic” audiophile cables have very high capacitance

• Cable C and driver Zo form a low-pass filter
• For typical consumer Zo = 1 kΩ and Cc = 50 pF/ft
• 20 kHz response down 0.5 dB @ 50’, 1.5 dB @ 100’, and 4 dB @ 200’

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 97

It can sure take the “sizzle” out of the top end ...

Bill Whitlock 97
An Overview of Audio System Grounding & Interfacing 9/4/2012

High-Frequency Cable Behavior

Over 40 kHz, shielded

cables themselves
eliminate ground noise!

• At audio frequencies, Rs is responsible for common-Z coupling

• Impedance of shield rises above 4 kHz (typical) due to Ls
• At 10 times this frequency, essentially all voltage across Ls is
magnetically induced into Lc, canceling receiver common-mode noise
• Voltage in Lc subtracts S to S (common-mode) voltage from signal
• Shielded cables inherently cancel CM noise above ≈ 40 kHz
• Weigh this against claims made for power-line filters ...
• This does NOT apply to UTP cables such as CAT-5 types

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 98

Power-line filters in “conditioners” only begin to work at frequencies over about 40

kHz … but these frequencies do not cause either common-impedance coupling in
unbalanced cables or common-mode voltages at balanced inputs, provided that cables
with normal coaxial shielding is used in either case. Power-line filtering, unless the
filter starts working at a few hundred Hz (which would make it the size of a
refrigerator), are useless in the reduction of hum and buzz ... duh! Those who
choose UTP over STP, especially in audio and video work, should consider
what they’re giving up in terms of noise rejection in the interface!

Bill Whitlock 98
An Overview of Audio System Grounding & Interfacing 9/4/2012

Physics from Another Universe?

The Emperor’s New Audio Cables?
• Double-blind tests prove that when
audible differences among cables
actually do exist, they’re entirely
explainable
• Marketing hype often invokes
transmission line theory and implies
that nano-second response is
somehow important
• Real physics reminds us that audio
cables only begin to exhibit subtle
transmission line effects when some
3,500 feet long

“High-end" audio is awash in pseudo-science, anti-science, and mysticism.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 99

Tom Nousaine has published excellent information about ABX “double-blind” testing
and some very interesting results. See www.nousaine.com.

Bill Whitlock 99
An Overview of Audio System Grounding & Interfacing 9/4/2012

Physics from Another Universe?

• Exotic cables, even if double shielded, made
of 100% pure un-obtainium, and hand
woven by virgins, will have no significant
effect on hum and buzz!
• Truly high-performance unbalanced audio cables
combine very low shield resistance, low
capacitance, and reliable connectors
• I like Belden 8241F (even if they call it video
cable)
• It combines very low shield resistance, low capacitance,
high flexibility, and availability in lots of pretty colors!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 100

Make your own “designer cables” that actually do something worthwhile!

Bill Whitlock 100

An Overview of Audio System Grounding & Interfacing 9/4/2012

TROUBLESHOOTING
Without method, it can be both frustrating and time-consuming!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 101

The silent scream “says” it all ...

Bill Whitlock 101

An Overview of Audio System Grounding & Interfacing 9/4/2012

“Think” … Like a Detective

• DON’T start by changing things!
• Ask questions: “Did it ever work right?”
• Equipment controls can provide valuable clues
• Volume Control – if turned down, does noise
disappear or remain unchanged?
• Source Selector – do all sources have the noise or
just one?
• Keep written notes, memory is less reliable
• Problems that go away by themselves are also
very likely to unexpectedly reappear by
themselves ...
• “Gentlemen, let the finger-pointing begin!”
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 102

Effects of equipment controls are extremely valuable but nearly always

overlooked – I know “firsthand” from experience in helping thousands of customers
troubleshoot audio systems! For example, if a hum disappears when the volume is
turned down, logically it must be entering the signal path before the volume control.
Conversely, if it is unaffected by the volume control, logically it must be entering the
signal path after the volume control. Similarly, source selector switches can identify
which source is affected or if all are affected. These are very important clues when
troubleshooting!

Bill Whitlock 102

An Overview of Audio System Grounding & Interfacing 9/4/2012

Draw a “One-Liner”
• Show ALL signal cables and indicate their approximate length
• ANY cable can be a pathway for a ground loop!
• Show data and control cables, even though not part of the signal path
• Mark any balanced inputs or outputs
• Show stereo pairs, RGB, etc. as single lines
• Note all equipment grounded by its 3-prong AC power plug
• Note any other ground connections such as CATV feeds, DBS dish
feeds, or rack mounting

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 103

Jensen offers free help with troubleshooting but sometimes a caller whines about the
documentation I ask for. I’ve been known to ask them “If you can’t take the
trouble to sketch up the system, why should I take the trouble to try to
figure it out with so little information?” Once I have the sketch and know what
the symptoms are, I most often ask them to perform a couple of simple tests (like
temporarily lifting a ground with a 3-to-2 prong adapter or unplugging that USB
cable to a PC) and then tell them where and what the problem most likely is … and
sometimes they end up not needing to buy anything from us, which is OK with me. I
believe they’ll remember where they got real help, not hype, the next time they need
assistance.

Bill Whitlock 103

An Overview of Audio System Grounding & Interfacing 9/4/2012

A Sample “One-Liner”

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 104

All equipment is assumed to have 2-prong (ungrounded) AC power unless noted

otherwise. It’s not necessary to show speakers unless they’re the powered type
(amplifier built-in).

Bill Whitlock 104

An Overview of Audio System Grounding & Interfacing 9/4/2012

Work Backwards!
• Unless clues suggest otherwise, always begin at audio power
amplifier or video monitor inputs and test interfaces in
sequence back toward signal sources
• REMOVE “ground lift” devices before troubleshooting!
• These illegal “cheats” generally disguise the real problem
• Although having a 3-prong power cord and 3-prong outlets, some
“audiophile” power strips internally disconnect the safety ground!
• Look for a UL label on every AC power component in the system.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 105

Bill Whitlock 105

An Overview of Audio System Grounding & Interfacing 9/4/2012

Testing: Unbalanced Audio Interfaces

• The “Dummy Test” can locate the exact noise coupling point
• It also reveals the exact nature of the problem
• Common-impedance coupling in cables – by far, the most “common”
• Common-impedance coupling inside equipment – the “pin 1 problem”
design defect often exists at unbalanced interfaces, too
• Magnetic or electric field pickup by cables

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 106

The “dummies” are not people, they’re devices!

Bill Whitlock 106

An Overview of Audio System Grounding & Interfacing 9/4/2012

The Unbalanced Audio “Dummy”

ISO-MAX® Model TA-R1 “Dummy”

Remember, the “dummy” DOES NOT pass signal!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 107

Jensen offers a pair of TA-R1 dummy adapters (as pictured) along with a
printed troubleshooting guide for $10.

Bill Whitlock 107

An Overview of Audio System Grounding & Interfacing 9/4/2012

STEP 1

Unplug the existing cable from Box B and replace with dummy
• Noise — problem is internal to Box B (or downstream)
• Quiet — go to step 2

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 108

If clues don’t suggest otherwise, Box B would be the power amplifier that drives the
speakers and we’d test the first line-level interface as shown in the following steps.

Bill Whitlock 108

An Overview of Audio System Grounding & Interfacing 9/4/2012

STEP 2

Leave the dummy in Box B and plug the existing cable into the
dummy
• Noise — Box B has a “pin 1 problem.” Confirm with the “hummer” test
• Quiet — go to step 3

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 109

The “hummer” test is described later in this section, right after we cover
balanced interfaces.

Bill Whitlock 109

An Overview of Audio System Grounding & Interfacing 9/4/2012

STEP 3

Remove the dummy and plug the existing cable back into Box B.
Unplug other end of existing cable from Box A and plug it into the
dummy. The dummy must not touch anything conductive!
• Noise — a magnetic or electric field is inducing noise into the cable. Check
shield continuity or re-route the cable to avoid the field.
• Quiet — go to step 4

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 110

Bill Whitlock 110

An Overview of Audio System Grounding & Interfacing 9/4/2012

STEP 4

Leaving the dummy on the existing cable, plug the dummy into Box A.
• Noise — common-impedance coupling in the cable. Install an isolator!
• Quiet — noise must be coming from output of Box A. Perform the 4-step
test sequence on next upstream interface.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 111

In some cases, replacing the cable with another having much lower shield
resistance may provide enough noise reduction to be acceptable.

Bill Whitlock 111

An Overview of Audio System Grounding & Interfacing 9/4/2012

Testing: Analog Video Interfaces

• Because video monitors typically revert to a “blue screen” when the
signal is absent, video doesn’t lend itself to “dummy” testing
• Instead, a sensitive clamp-on ammeter is used to identify the route of
the ground loop and measure the loop current
• For every video cable that’s part of the loop, either a measured or
estimated shield resistance is used to calculate noise coupling voltage
according to Ohm’s law, E = I x R
• Hum bars become visible when the coupling voltage reaches or
exceeds a threshold of about 5 mV rms (roughly 15 mV peak-to-peak)
• A cable with end-to-end shield resistance of 0.5 Ω is found to have 20 mA rms
flowing in it, making the coupling voltage 20 mA x 0.5 Ω = 10 mV
• A component video RGB bundle has 50 mA flowing in it and the paralleled
shield resistance is 0.5 Ω, making the coupling voltage 25 mV

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 112

Bill Whitlock 112

An Overview of Audio System Grounding & Interfacing 9/4/2012

Clamp-On Ammeter Locates Ground Loops

• Senses current magnetically
Video
• Currents that flow bi-directionally Cable
in the cable (like signal currents)
are ignored and produce no
reading
• Currents that flow in one direction
and return elsewhere (like ground
loop currents) are measured
High-sensitivity models most often called
“Leakage Current Meters”
A reading resolution of 1 mA or
AEMC model 565
better is highly recommended

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 113

The AEMC 565 sells for about $270 (it’s a versatile multimeter, too). Signal
currents are ignored because they flow in opposite directions (center conductor and
shield) in the video cable, making their magnetic fields cancel. Ground loop current
passes one-way only through the cable and is therefore detectable because of the
magnetic field it creates around the cable. It can be clamped around signal cables as
well as power cords to identify the route and magnitude of ground loops. This method
applies to audio, video, data, and signal lines of all kinds … even laboratory
instruments.

Bill Whitlock 113

An Overview of Audio System Grounding & Interfacing 9/4/2012

Current vs Cable Type and Length

Cable Type Diameter 50’ 20’ 10’

RG-11 0.285” 80 mA 200 mA 400 mA

RG-59 0.242” 35 mA 90 mA 180 mA

Precision 0.233” 26 mA 65 mA 130 mA

Miniature 0.125” 17 mA 43 mA 86 mA

Sub-Miniature 0.097” 9 mA 22 mA 43 mA

Current listed typically produces a barely visible hum bar

based on threshold coupling voltage drop of 15 mV peak-to-peak

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 114

In general, the smaller the diameter of the cable, the higher the resistance
of its shield conductor – which makes it more likely to become a hum-bar
problem. Why is the industry so obsessed with tiny cables anyway?

Bill Whitlock 114

An Overview of Audio System Grounding & Interfacing 9/4/2012

Testing: Balanced Audio Interfaces

• Same basic concept as unbalanced “dummy” tests
• Again, tests pinpoint location of problem interfaces
• Coupling mechanisms are different:
• Shield-current-induced-noise or SCIN in cables
• Magnetic or electric field coupling in cables
• “Pin 1 problems” in poorly-designed equipment
• Inadequate Common-Mode Rejection in equipment inputs

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 115

Unfortunately, this test will not find Pin 1 problems at equipment outputs.

Bill Whitlock 115

An Overview of Audio System Grounding & Interfacing 9/4/2012

Building a Balanced Audio “Dummy”

Prototype made with housing, connectors and

switch of a CalRad #35-465 in-line mic switch

Closing S1 creates the 10 Ω

imbalance for IEC-like CMRR test

Remember, this device DOES NOT pass signal!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 116

SHORTCUT: Modify a commercial “microphone kill switch” that normally

uses the switch to short pins 2 and 3.

Bill Whitlock 116

An Overview of Audio System Grounding & Interfacing 9/4/2012

STEP 1

Unplug existing cable from input of Box B and plug in just the dummy.
(switch remains in “NORM” position until Step 5)
• Noise — internal problem with Box B (or downstream).
• Quiet — go to step 2.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 117

Bill Whitlock 117

An Overview of Audio System Grounding & Interfacing 9/4/2012

STEP 2

Leave the dummy at input of Box B and plug existing cable into
dummy.
• Noise — Box B has a “pin 1 problem” … confirm with the Hummer Test.
• Quiet — go to step 3

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 118

Bill Whitlock 118

An Overview of Audio System Grounding & Interfacing 9/4/2012

STEP 3

No Contact with Metal

Remove dummy and plug existing cable back into Box B. Unplug
cable from Box A and plug into dummy.
• Noise — magnetic or electric field induction in cable. Check shield
continuity or re-route cable to avoid the field.
• Quiet — go to step 4.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 119

Bill Whitlock 119

An Overview of Audio System Grounding & Interfacing 9/4/2012

STEP 4

Leaving the dummy on the cable, plug the dummy into Box A.
• Noise — Shield Current Induced Noise or SCIN. Replace cable or
take steps to reduce shield current, possibly lifting shield at receive end.
• Quiet — Go to step 5.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 120

“Replace cable” would mean replacing a cable that uses a drain wire shield
connection with a cable that does not.

Bill Whitlock 120

An Overview of Audio System Grounding & Interfacing 9/4/2012

STEP 5

With same setup, move the dummy switch from “NORM” to “CMRR.”
• Noise — Poor input CMRR at Box B. Replace Box B or add an input
transformer isolator to increase CMRR.
• Quiet — Noise must be coming from output of Box A. Use the “Hummer
Test” to determine if Box A has a “pin 1 problem” at its output. If not,
perform this 5-step test sequence on the next upstream interface.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 121

This CMRR test is a simplified version of the IEC test that uses the existing system
common-mode voltage rather than a signal generator. In some ways that makes it a
real-world “in-situ” test.

Bill Whitlock 121

An Overview of Audio System Grounding & Interfacing 9/4/2012

“Pin 1” Testing with the “Hummer”

Pin 1

Chassis or other Pin 1

80 mA to 100 mA of Rectified 60 Hz Flows in Suspect Shield Connections

Concept by John Windt, June 1995 AES Journal

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 122

This can be easily made by modifying a standard “automotive test light”, adding the
transformer and diode. The diode is used to add harmonic content to the 60 Hz
current, making it easier to hear. Thus far, I have not created a simple test
to reveal the “sensitive” equipment or “power-line primadonna” syndrome.

Bill Whitlock 122

An Overview of Audio System Grounding & Interfacing 9/4/2012

Using the “Hummer”

• Monitor one output and disconnect all other I/O cables
• For reference, listen to output without the hummer connected
• Connect clip to metal chassis and touch the probe to shield contact of
each I/O connector
• If no metal chassis, connect clip to shield of another I/O connector
• A “clean” design will have no output buzz or change in noise floor
• Other test paths include safety ground to I/O shields and input
shields to output shields
• Safety ground is accessible at the AC outlet and testing the path from there to
chassis could reveal a problem due to a poorly chosen internal connection

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 123

Bill Whitlock 123

An Overview of Audio System Grounding & Interfacing 9/4/2012

DON’T BET YOUR LIFE!

• NEVER defeat safety grounding to
solve a noise problem!
• Adapter is intended to provide a safety
ground when a 3-prong plug must be
used with a 2-prong receptacle
• Outlet cover-plate screw provides
connection to safety ground via outlet
OR EVEN WORSE
saddle and metallic box/conduit
• You’ll be held legally liable if someone is
shocked or electrocuted because you
installed a “cheater” … the judge simply
won’t care about the buzz it cured!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 124

This is not only illegal and dangerous, but it may get you sued right out of business. It
can, and does, happen. Especially devious are those who clip off the safety ground
prong so that nothing seems to be wrong to a casual observer!

Bill Whitlock 124

An Overview of Audio System Grounding & Interfacing 9/4/2012

With Utter Disregard for Safety!

“One of the coolest new innovations ...

PS engineers invented a better way: a
screw-in ground pin. Instead of using a
cheezy cheater plug and ruining all the
benefits of the xStream's machined
connectors, you simply unscrew the
ground pin and voila! No more buzz.”
www.psaudio.com

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 125

A supreme example of commerce without conscience!

Bill Whitlock 125

An Overview of Audio System Grounding & Interfacing 9/4/2012

And Downright Dangerous Advice!

“The safety benefit of having even one component earthed extends to
all the other components in the system because they are all connected
together by the interconnects. So if one component faults, the
dangerous currents would travel through the interconnects to
the earthed component, and drain safely away through its AC cord.”
and “But the cheaters are hardly a high-end solution. Some audiophiles
deal with this by cutting the third prong off the plug so they can
plug it straight into the outlet without needing a cheater. That's
what I do.” from www.altavistaaudio.com/hum.html May 2012

Can you imagine a typical RCA cable carrying a 150 A

fault current for the several seconds it might take to
trip the AC power circuit breaker?

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 126

This is just one example of bad advice that’s rampant on the internet ...
audiophiles are particularly vulnerable victims. The 150 A fault current, with a
corresponding breaker trip time of about 2.5 seconds, comes from a UL study of
over 1,000 typical residences.

Bill Whitlock 126

An Overview of Audio System Grounding & Interfacing 9/4/2012

Consider Signal Cables

• They will distribute lethal voltage from a failed but “ground-lifted”

piece of equipment throughout an entire system, or
• Fault current may find its way to safety ground through a signal
cable, causing it to melt or burst into flames!
• DEFEATING SAFETY GROUNDING IS PROHIBITED BY CODE
• NEC 2008 Articles 250.6(D) and 250.114 specifically prohibit
removal of safety grounding for plug-and-cord-connected
equipment to solve “noise” issues
• Signal cables are NOT safety ground wiring!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 127

A judge in the liability trial won’t give a rat’s ass how your “trick” cleared up the hum
problem! Even if the customer lives, he can claim he can’t sleep or his sex life is
ruined since he got that nasty shock ... he’ll sue and win!

Bill Whitlock 127

An Overview of Audio System Grounding & Interfacing 9/4/2012

The DANGER is REAL

• Typical Year in USA
• 10 ELECTROCUTED
• 2,000 FIRES causing
• 20 DEAD
• 100 INJURED
• $30 M property lost

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 128

Note the “AC-DC” album in his hand! This is a commissioned cartoon.

Bill Whitlock 128

An Overview of Audio System Grounding & Interfacing 9/4/2012

Shock and Electrocution

• Dry skin has high resistance – little current flows
• Increased moisture, contact area, and pressure
result in more current
• Current determines the severity of shock
• Under 1 mA is simply an unpleasant “tingling”
• About 10 mA causes involuntary muscle contraction,
the “death grip,” or suffocation if through chest
• 50 to 100 mA through the chest may induce
ventricular fibrillation, causing subsequent brain death
• ALWAYS HAVE A VERY HEALTHY RESPECT
FOR ELECTRICITY!
• To help prevent lethal chest currents,
always keep the “other” hand in your back
pocket!
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 129

Statistically, high voltages are less dangerous than 120 volts. The slightest contact
with high voltages causes severe muscle contractions that usually throw the victim
clear of the danger. He’ll suffer skin burns where he made contact, but he’ll likely live
to tell about it. At 120 volts, many victims are found dead but still gripping the
energized devices or contacts that killed them. Note that UL leakage current limits are
less than 5 mA in any circumstance, well below the involuntary muscle contraction
threshold.

Bill Whitlock 129

An Overview of Audio System Grounding & Interfacing 9/4/2012

SOLUTIONS – Breaking the Loop

• NEVER defeat safety grounding!
• Ground loops may include many cables and pieces of equipment
• In general, problems become worse with distance
• Coupling is worse in longer signal cables
• Ground voltage differences in safety ground are proportional to length
• Ground isolators in signal path are preferred solution
• Determining correct location is critically important
• Sometimes we have more than one location that works
• Ungrounded equipment can pose a problem for isolators

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 130

In many large systems, placement of a single (or single pair) of isolators in the signal
path solves a myriad of noise issues related to a single ground loop. The key, as in
real estate, is location, location, location!

Bill Whitlock 130

An Overview of Audio System Grounding & Interfacing 9/4/2012

Ground Loop in Simple Example System

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 131

The two different ground points create a ground loop

Bill Whitlock 131

An Overview of Audio System Grounding & Interfacing 9/4/2012

ISOLATORS for Unbalanced Audio

• Ground isolators solve the inherent problem in unbalanced
interfaces
• An isolator is a differential-responding device with high common-mode
rejection (discussed later in balanced interfaces)
• Simply put, an isolator stops noise current flow in the cable shield
• An isolator is NOT A FILTER
• Filters are devices that pass certain frequencies while rejecting others
• If a filter could be “taught” what spectral “footprint” to remove for every
conceivable hum or buzz, it could be inserted anywhere in the signal path!
• To eliminate noise, an isolator must be installed at the interface
where the noise coupling actually occurs …
• Location, Location, Location …

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 132

Such filters have been attempted in the digital domain. While moderately successful at
removing a steady hum, they don’t do well with more harmonically complex buzz. It’s
essentially impossible to pass high-quality music through such a filter without altering
the sound in a very audible way.

Bill Whitlock 132

An Overview of Audio System Grounding & Interfacing 9/4/2012

Two Basic Types of Audio Transformers

• An ideal transformer would completely stop ground loop current
… at all frequencies
• In reality, capacitance between windings allows some current to
flow “around” the transformer at high audio frequencies
• Two basic types differ greatly in capacitance:
• OUTPUT types have HIGH capacitance
• Typically 10,000 pF or more
• Noise reduction about 40 dB worse than input type
• The only type made by nearly all Jensen competitors …
• INPUT types have VERY LOW capacitance
• Typically 100 pF (1% of output type) due to internal Faraday shielding
• Large capacitive load on output will reduce HF bandwidth
• Use long cable on input side, keep output cable under 3 feet

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 133

Jensen makes both types of transformers, and each is well-suited for a particular
application. Output transformers are simply a poor choice if maximum noise
reduction is the goal.

Bill Whitlock 133

An Overview of Audio System Grounding & Interfacing 9/4/2012

Isolators: “Friends with Benefits”

• Benefits of INPUT type transformers
• Noise rejection over 80 dB, whether the
signal source is balanced or unbalanced
• Filter out ultra-sonic and RF interference,
reducing “spectral contamination”
• Solve “Pin 1 problems”
• No power required
• Simple, passive, robust, reliable, and
virtually immune to transient surges “Simplicity is the
ultimate sophistication”

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 134

Transformers have no concept of “ground” – they “don’t care” whether a

signal source is balanced or unbalanced.

Bill Whitlock 134

An Overview of Audio System Grounding & Interfacing 9/4/2012

Transformer Isolators for Unbalanced Audio

“Isolation is under-appreciated but astoundingly useful.”

Insulated Barrier

The electrical path between the boxes via the signal cable is broken
and the signal is magnetically coupled across an insulated barrier

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 135

The insulated barrier inside the transformer, which can easily withstand several
hundred volts, makes it impossible for ground loop current to flow through the audio
cables.

Bill Whitlock 135

An Overview of Audio System Grounding & Interfacing 9/4/2012

Noise Rejection vs Transformer Type - Unbalanced

None

Output

Input

60 Hz 3 kHz

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 136

Input type gives 30 dB better noise rejection at all audio frequencies.

Bill Whitlock 136

An Overview of Audio System Grounding & Interfacing 9/4/2012

Commercial Ground Loop Isolators - Unbalanced

• Overwhelming majority use unshielded output transformers
• Many units use tiny $2 transformers designed for telephone
couplers, where 300 Hz to 4 kHz “flat” response is enough
• Device below typical of many low-cost “ground loop isolators”
• Data lists details like DC resistance but doesn’t even mention noise rejection
… which is the main reason isolators exist!

Ground LP Isolator (270-054) from http://support.radioshack.com

Specifications, ELECTRICAL DATA:
Impedance Ratio: Z1:Z2...............................1000 Ohms:600 Ohms
Frequency Response:...........................300 Hz to 4000 Hz +/- 3 dB
DC Resistance @25 degrees C:
Z1:....................................................200 Ohms +/- 20%
Z2:....................................................150 Ohms +/- 20%
Radio Shack 270-054 at $20 Insulation:.........................500 VDC apply to primary and secondary
more then 100 Meg Ohms Min.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 137

The Radio Shack unit is popular with car stereo installers (but not in my car)

Bill Whitlock 137

An Overview of Audio System Grounding & Interfacing 9/4/2012

Specifications often “Deceive by Omission”

• When inappropriately-designed (usually cheap) transformers are
used in high-quality audio systems, the result is:
• Loss of deep bass
• Distorted bass, especially during loud passages
• Poor high-frequency transient response and imaging
• But their so-called “specifications” are often just a feature list that
omits the “dirty little secrets” of their poor performance!
• Audio transformers intended for professional use should,
above all, have complete and unambiguous specifications!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 138

Bill Whitlock 138

An Overview of Audio System Grounding & Interfacing 9/4/2012

Beware of “Sketchy-Specs”
• Noise Rejection or CMRR
• Meaningless without stated source impedance and frequency
• Maximum Signal Level and Distortion
• One strongly affects the other
• Both also depend on driving source impedance and frequency
• Any P.O.S. transformer will have low distortion at any signal level
when it’s driven from a very low or zero-impedance source and at a
midrange frequency like 1 kHz!
• Ask how much signal it will handle at 20 Hz or 30 Hz, while keeping
distortion under 1% THD, when driven from a real-world source –
THAT’S WHAT REALLY MATTERS!
• Frequency Response
• Again, since source and load impedances strongly affect response, specs are
meaningless unless both are stated
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 139

Maximum signal level and distortion are the transformer specifications most
deceptively advertised and listed by our competitors. Many won’t even
mention CMRR, let alone disclose the test conditions. After all, NOISE
REJECTION is the top reason for using a transformer at interfaces in the
first place!!

Bill Whitlock 139

An Overview of Audio System Grounding & Interfacing 9/4/2012

Typical “Sketchy-Specs” Information*

• The entirety of published electrical data for a stereo “Hum Killer”
offered for “professional” use:
• TYPICAL PERFORMANCE:
• Inputs (2): −10 dBV nominal, 10 kΩ
• Outputs (2): −10 dBV nominal, 10 kΩ
• Frequency Response: 20 Hz to 20 kHz (+/− 3 dB)
• THD+N: <0.5% (30 Hz to 20 kHz); 0.05% @ 1 kHz
• Crosstalk: <60 dB (20 Hz to 20 kHz; 100 Ω source, 10 kΩ load);
<70 dB @ 1 kHz
• Only crosstalk could possibly be verified (test conditions are stated) –
the rest is meaningless and most likely hiding the ugly truth!

* Sketchy: wanting in completeness, clearness, or substance; slight, superficial. (merriam-webster.com/dictionary)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 140

Data from http://www.rdlnet.com/pdf/Data_Sheets/av-hk1.pdf on 15 May 2012.

Bill Whitlock 140

An Overview of Audio System Grounding & Interfacing 9/4/2012

Honest, Complete Specifications

• Isolator for unbalanced audio interfaces
• Faraday-shielded INPUT transformers
• Out cable ≤ 3 ft for rated HF bandwidth
• It has real specifications:
Jensen Transformers

ISO-MAX® Model CI-2RR Condensed Specifications

(see www.jensen-transformers.com/datashts/ci2rr.pdf for complete version)

Test conditions: 600 Ω source, 47 kΩ load, 1.23 V rms level unless noted
Frequency Response, ref 1 kHz: − 0.15 dB at 20 Hz, − 0.7 dB at 20 kHz
Bandwidth, − 3 dB, ref 1 kHz: 0.25 Hz to 50 kHz
Total Harmonic Distortion: 0.04% typ at 20 Hz, <0.001% typ at 1 kHz
Max Input Level, THD < 1%: 7 V rms at 20 Hz, 11 V rms at 30 Hz
Noise Rejection (CMRR): 95 dB typ at 60 Hz, 85 dB typ at 3 kHz
Phase Distortion, 20 Hz ~ 20 kHz: ±2º max deviation from linear phase
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 141

Notice that the ability to handle large signals is very frequency dependent.
Who but Jensen would dare mention phase distortion? If you want to know
more about “deviation from linear phase” or DLP, just ask us for a copy of
the paper by the late Deane Jensen. It explains why “phase shift” is not
necessarily phase distortion.

Bill Whitlock 141

An Overview of Audio System Grounding & Interfacing 9/4/2012

“Active” Amplifier-Based Isolators

• Many powered or “active” devices claim
to cure ground loop problems in
balanced or unbalanced interfaces, but RDL

• ...
This unit touts an “instrumentation” input stage, but its low CM input
Z, 5 kΩ per leg, cause its noise rejection to be exquisitely sensitive to
the slightest Z imbalance in “balanced” signal sources
• With unbalanced sources, their entire output impedance, Zo, typically
200 Ω to 2.2 kΩ, becomes Z imbalance
• Even if properly connected to an unbalanced source, over that same
Zo range, CMRR will range from 28 dB to as little as 7 dB
• However, the manufacturer’s recommended hook-up for unbalanced
sources results in zero noise rejection (more on this later)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 142

We’ll talk about “diff-amps” in the section on balanced interface line receivers.

Bill Whitlock 142

An Overview of Audio System Grounding & Interfacing 9/4/2012

CMRR vs Input Zcm and Source Imbalance

Input Zcm
(per leg) Typical
”Active”
=
5k
Ω Range
Zcm

kΩ
50
=
m
Zc
InGenius®
Ω
10
M Input
=
Zc
m Transformer
MΩ
50
=
m
Zc

BAL UNBAL 2k

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 143

This graph shows the fundamental relationship that limits CMRR for any
combination of receiver common-mode input impedance (per leg) and
driver common-mode output impedance imbalance. These are calculated
results that assume that all system components are ideal. The performance of actual
circuits will generally be very close to these CMRR numbers.

Bill Whitlock 143

An Overview of Audio System Grounding & Interfacing 9/4/2012

Floating Equipment vs Isolators

• The chassis of devices with 2-prong AC plugs can "float" at 50 VAC
or more (relative to safety ground) if no ground path exists
• This often happens when an isolator is installed at an interface
• No shock hazard because available current is very low
• This “float” voltage is an extreme challenge for an isolator – the
common-mode voltage can be 100 times larger than the signal
• This unavoidably degrades its noise rejection by about 40 dB!
• An added equipment ground solves the problem
• Replace the original 2-prong AC plug with a new 3-prong type
• Add a wire connecting the ground pin of the new AC plug to the chassis of
the ungrounded equipment
• Recommended wire: #18 or #20 AWG, stranded, with green (or green w/yellow
stripe) insulation. Wrap it around original 2-wire cord in a long spiral and connect
with a lug to a rear-panel screw-head.
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 144

Since an isolator has no electrical connection between its input and output, the ground
path previously provided by the signal cable vanishes when the isolator is installed in
the signal path. Remember that CMRR is a ratio comparing an isolators response to
a normal signal to its response to the same voltage applied as a common-mode (or
ground voltage difference) voltage. If the ground voltage difference is now 100 times
that of the signal (40 dB larger), the noise rejection of the isolator will be 40 dB
worse! It explains why an occasional customer will call to tell us that the buzz got
worse when the isolator was installed!

Bill Whitlock 144

An Overview of Audio System Grounding & Interfacing 9/4/2012

Grounding “Floating” Devices

New 3-prong plug

Added ground shown in green removes “floating voltage” from the PREAMP and POWER AMPL,
allowing rated noise rejection in isolators shown. If the TV had a CATV isolator at its input, there
would be no need to replace its 2-prong plug and ground it or to use the TV to PREAMP isolator.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 145

Without an added ground, the isolators allow the preamp and power amp to
“float” with respect to the TV or sub-woofer, placing a huge noise-rejection
burden on the isolators.

Bill Whitlock 145

An Overview of Audio System Grounding & Interfacing 9/4/2012

Tips to Avoid Noise in Unbalanced Interfaces

• Keep cables short and do not coil excess
• Use cables with heavy gauge shields
• Maintain good connections
• Do not add unnecessary grounds
• Never, ever disconnect a safety ground!
• Use ground isolators at problem interfaces
• Bundle all signal cables between same two boxes
• Bundle AC power cords separately
• Separate signal and power bundles by at least 6 inches

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 146

Longer cables have more shield resistance end-to-end, which produces more
common-impedance coupling. Coiling excess cable invites magnetic induction from
any nearby motor or transformer (better to fold and bundle cables that are too long).
For long cable runs, use cable with lots of heavy copper braid as shielding to keep
resistance down. Bonded-pair stereo or RGB video cables are a good idea because
they reduce magnetic pickup area just as bundling all signal cables between same
boxes does. Although AC power cords actually radiate very little magnetic field, when
many are bundled together the magnetic field tends to drop even further due to
averaging. If signal and power bundles (or individual cables) must cross, arrange for
them to cross at 90° angles. Gently wiggling connectors will often reveal intermittent
problems. Adding grounds not required for safety will almost always make
noise problems worse! The only exception is a ground added to equipment having
a 2-prong AC plug to limit common-mode voltage as discussed earlier. Ground
isolators at signal interfaces are the “silver bullet” solution for ground-loop noise ...
because they directly solve the fundamental coupling problems.

Bill Whitlock 146

An Overview of Audio System Grounding & Interfacing 9/4/2012

Unbalanced to Balanced Transitions

• Also called “Consumer to Pro” conversion
• Signal reference levels are different
• Consumer ref = −10 dBV (or −7.8 dBu) = 0.316 V rms
• Professional ref = +4 dBu (or +1.8 dBV) = 1.228 V rms
• Requires voltage gain of about 4, or 12 dB
• 0 dBu = 0.775 V rms
• 0 dBV = 1.000 V rms
• 0 dBV = +2.2 dBu

• Why not use a 1:4 step-up transformer you ask?

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 147

Bill Whitlock 147

An Overview of Audio System Grounding & Interfacing 9/4/2012

“Consumer to Pro” Converter?

• Uses a 1:4 step-up transformer
• A 1:4 transformer reflects (1:4)² or a
1:16 impedance ratio
• For balanced Zin from 10 kΩ to 40 kΩ,
the consumer output sees loads from
625 Ω to 2.5 kΩ while IEC standards
specify 10 kΩ rated load
• Overloading degrades headroom,
Rane Corp. distortion, and frequency response
• Actual gain isn’t the expected 12 dB,
NOT A GOOD SOLUTION but only 3 to 8 dB
• 12 dB of gain “reach” is normally
available for balanced inputs, making
external gain generally unnecessary

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 148

The Rane engineering staff is fully aware of my comments about this

product. They’ve defended it by saying that they keep selling them! And they’re not
the only one – Ebtech also sells a similar product and my criticism would apply to it as
well. The problem is theoretically insurmountable. If gain is actually required, it
should come from an amplifier of some sort.

Bill Whitlock 148

An Overview of Audio System Grounding & Interfacing 9/4/2012

Close, But No Cigar … or Noise Rejection!

RDL

• Just as with a simple RCA to XLR adapter,

all noise-rejection benefit of its balanced
input is lost
• In this case, an easy change to this wiring
could add 7 dB to 28 dB of rejection! from “Installation/Operation” page for RDL
Model STA-1M with “Instrumentation
Input to Isolate Ground Loops”

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 149

Bill Whitlock 149

An Overview of Audio System Grounding & Interfacing 9/4/2012

Again … Close, But No Cigar!

• Internal wiring of this “Audio Format
Converter” renders its internal audio
transformer useless
• Input shield is tied to output pin 1,
which allows noise current to flow in
the incoming cable shield … exactly
how noise couples to the signal!
• No useful isolation 2

• No voltage gain 3
1
• No output voltage symmetry
• No Faraday shield, bifilar winding
instead (precluding symmetry)
Internal Schematic

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 150

Because the input and output grounds are tied, no “isolation” or common-mode
rejection is possible. Because the transformer ratio is 1:1, there can be neither
impedance matching nor voltage gain. Because the transformer has no Faraday shield
(it is bifilar wound instead), it can’t create voltage symmetry. If it doesn’t perform one
of these useful functions, what value does it have?

Bill Whitlock 150

An Overview of Audio System Grounding & Interfacing 9/4/2012

Simple but Smart Transition or “Converter”

• Transition and noise rejection are the issues, not gain
• “Adapters” and most “adapter cables” throw away the noise
reduction benefit of the balanced input
• An RCA to XLR adapter at a balanced input converts the entire interface
to a noise-prone unbalanced one!
• A 3-conductor (shielded twisted-pair) hookup allows the balanced
input to reject ground noise
• If the balanced input uses an input transformer or the InGenius® IC, noise
rejection can be over 90 dB!
• Even if the balanced input is only a “garden variety” diff-amp or
instrumentation-amp, rejection is still typically 30 dB

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 151

Bill Whitlock 151

An Overview of Audio System Grounding & Interfacing 9/4/2012

2 Conductors ... or 3?

Unbalanced cable + adapter = 0 dB rejection

Balanced (STP) cable = ≥ 30 dB rejection

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 152

The mere existence of these adapters makes people think it’s the right thing
to do ... it’s definitely NOT! And makers of “adapter cables” generally don’t
GET IT either … continuing to make cables with single-conductor shielded
cable with an adapter essentially built-in at the XLR end … DUMB!

Bill Whitlock 152

An Overview of Audio System Grounding & Interfacing 9/4/2012

Noise Rejection in Consumer to Pro Interface

2-C cable + adapter

3-C cable

3-C cable + output transformer

3-C cable + input transformer

60 Hz 3 kHz

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 153

Input transformer adds 70 dB of hum and 35 dB of buzz rejection. Standard

(output) transformer improves hum by about 20 dB but does essentially
nothing for buzz.

Bill Whitlock 153

An Overview of Audio System Grounding & Interfacing 9/4/2012

Should I “Balance” an Unbalanced Line?

Unbal to Bal Bal to Unbal

• NO – two converters on same line completely unnecessary

• This myth is based on the superficial idea that signal symmetry somehow
has anything to do with noise rejection and wrongly assumes that cheap
“balun” transformers create such symmetry … they do not!
• An input-transformer-based isolator at the receive end will have
much better noise rejection than a pair of “baluns”
• Creates a high-rejection, source-impedance tolerant balanced input at the
receive end of line
• The line itself, whether coaxial or twisted-pair, is not more vulnerable to
noise because “it’s not balanced” at the send end!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 154

Buying two “baluns” to fix noise on an unbalanced line is a waste of money!

And you don’t need new cable either!

Bill Whitlock 154

An Overview of Audio System Grounding & Interfacing 9/4/2012

ISOLATORS for Balanced Audio

• Solve the two most common problems
with balanced interfaces:
• Equipment “Pin 1 problems”
• Low CMRR in the real world
Jensen Transformers
• This isolator solves both
• DIP switches (on bottom) can reconfigure
shield connections to solve Pin 1 problems
• Uses Faraday-shielded INPUT transformers
boost CMRR, typically by 40 to 60 dB

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 155

The 2-channel model PI-2XX is shown.

Bill Whitlock 155

An Overview of Audio System Grounding & Interfacing 9/4/2012

Getting High CMRR in the Real World

• Real-world interface CMRR will likely be much less than
advertised for the balanced input alone!
• The “marketing” CMRR figure is VERY MISLEADING
• INPUT transformers have a BIG advantage!
• Zcm of a real input transformer is about 50 MΩ at 60 Hz – 1,000 times that
of ordinary “active” inputs – making them very source tolerant
• Interface CMRR is essentially unaffected by 500 Ω imbalances and little
affected up to 2 kΩ, allowing sources to be either balanced or unbalanced

the author’s license plate

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 156

Real-world audio equipment is not made with the precision of laboratory

test instruments, so conventional tests for CMRR are nearly meaningless.

Bill Whitlock 156

An Overview of Audio System Grounding & Interfacing 9/4/2012

Boosting Rejection of Typical Balanced Input

Marketing “90 dB”

becomes real 65 dB
using new IEC test

+ output transformer

+ input transformer

60 Hz 3 kHz

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 157

The input transformer, which has an internal Faraday shield, does about 35 dB better
than a commodity output type. The output type improves hum rejection by
about 20 dB but does nearly nothing for buzz.

Bill Whitlock 157

An Overview of Audio System Grounding & Interfacing 9/4/2012

Commercial Ground Loop Isolators – Balanced

Ebtech

• Again, Beware of “sketchy-specs” that hide the ugly truth

• Overwhelming majority of isolators use output transformers
• As shown in the graph, such units do nothing for “buzz” and little for “hum”
• In fairness, they have one possible advantage – they can be installed
without regard for output cable length, making it a “no-brainer”
• Many isolators simply can’t handle professional signal levels …

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 158

Is there a hidden message in the fact that “buzz” isn’t mentioned?

Bill Whitlock 158

An Overview of Audio System Grounding & Interfacing 9/4/2012

J e n s e n T ra n s f o rm e rs , I n c 0 6 /1 1 /0 9 1 5 :2 2 :1 0

+3
“Pro” Isolators by Sescom and Jensen – Frequency Response

+ 2.25

+ 1.5

+ 0.75 Sescom IL-19

d
B +0
r

Jensen PI-XX
-0 . 7 5

-1 . 5

-2 . 2 5

-3
20 50 100 200 500 1k 2k 5k 10k 20k 50k 100k
Hz

Test Conditions: Audio Precision System 1

Sweep Tra c e C o lo r L i n e S tyl e T h ic k D a ta Axi s Com m ent

1 1 Red S o lid 5 An l r .L e ve l A L e ft J e n s e n P I-XX

2
Source Impedance = 100 Ω balanced
1 B lu e S o lid 5 An l r .L e ve l A L e ft S e s c o m IL - 1 9

Load Impedance = 20 kΩ balanced

T e s t C o n d i ti o n s :
1 0 0 O h m S o u r c e Im p e d a n c e
2 0 k O h m L o a d Im p e d a n c e

Signal level = − 10 dBu = 245 mV rms

Al l te s t i n b a l a n c e d m o d e

c o m p a ri s o n te s t m a g s w e e p te s t i n p u t.a t1

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 159

Both the Jensen ISO-MAX® model PI-XX and the Sescom model IL-19 are
isolators with XLR input/output connectors and intended for “professional”
applications. The Jensen model uses the exact same (model JT-11P-1HPC)
transformer as the model CI-2RR “consumer” isolator shown in the previous
slide. The gradual drop in its high-frequency response is intentionally designed
to give it a second-order Bessel rolloff, which accounts for its impeccable time-
domain response … seen as square-wave response without overshoot or
ringing. It is standard measurement practice to measure frequency response at
a low signal level to avoid confusing the results with overload – magnetic
saturation in this case.

On the other hand, the Sescom IL-19 has a huge frequency response peak in
the vicinity of 100 kHz. This will not only result in square-wave overshoot and
ringing but such peaks can also cause or aggravate “spectral contamination”
distortion in downstream amplifiers. Ask us for a copy of the 1988 AES paper
“Spectral Contamination Measurement” by Gary Sokolich and the late Deane
Jensen. Also note that the IL-19 frequency response falls sharply at about 30
Hz. It would no doubt be flat at a lower signal level, but at only − 10 dBu
(equals 245 mV rms) the 30 Hz signal has magnetically saturated the
transformer core. This magnetic saturation manifests itself as high THD at 30
Hz in the next slide.

Bill Whitlock 159

An Overview of Audio System Grounding & Interfacing 9/4/2012

J e n s e n T ra n s f o rm e rs , I n c 0 6 /1 1 /0 9 1 5 :2 8 :5 0

10
“Pro” Isolators by Sescom and Jensen – Distortion vs Level
5

0.5

0.2

Sescom IL-19
% 0.1

0.05

0.02

0.01

0.005 Jensen PI-XX

0.002

0.001

+4 dBu
-3 0 -2 5 -2 0 -1 5 -1 0 -5 +0 +5 + 10 + 15 + 20
dB u

Test Conditions: Audio Precision System 1

Sw eep Tra c e C o lo r L i n e S tyl e T h ic k D a ta Axi s C om m ent

Source Impedance = 100 Ω balanced

1 1 Red S o lid 5 An l r.T H D + N R a ti o L e ft J e n s e n P I-XX
2 1 B lu e S o lid 5 An l r.T H D + N R a ti o L e ft S e s c o m IL - 1 9

Load Impedance = 20 kΩ balanced

T e s t C o n d i ti o n s :
1 0 0 O h m S o u rc e Im p e d a n c e

Signal frequency = 30 Hz
2 0 k O h m L o a d Im p e d a n c e
Al l te s t i n b a l a n c e d m o d e

c o m p a r i s o n te s t 3 0 h z th d s w e e p te s t i n p u t.a t1

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 160

High distortion at low signal levels is a property of inexpensive core materials

like steel. Reaching over 10% distortion at the nominal “pro” operating level of
+4 dBu (shown by purple line) indicates that the IL-19 transformer core is far
too small. Ad copy for the IL-19 makes a big deal of “headroom,” but
at bass frequencies, there is no headroom at all! Although the Jensen is
larger, and a bit more expensive, it has some 20 dB of real headroom and
remains sonically transparent in even the most demanding audiophile systems!
In general, it’s impossible to make a high-performance audio
transformer that will fit into an XLR barrel.

Bill Whitlock 160

An Overview of Audio System Grounding & Interfacing 9/4/2012

Isolators: “Friends with Benefits”

• Benefits of INPUT type transformers
• Noise rejection over 80 dB, whether the
signal source is balanced or unbalanced
• Filter out ultra-sonic and RF interference,
reducing “spectral contamination”
• Solve “Pin 1 problems”
• No power required
• Simple, passive, robust, reliable, and
virtually immune to transient surges “Simplicity is the
ultimate sophistication”

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 161

Transformers have no concept of “ground” – they “don’t care” whether a

signal source is balanced or unbalanced.

Bill Whitlock 161

An Overview of Audio System Grounding & Interfacing 9/4/2012

The Balanced Interface Checklist

• Twisting of all balanced line pairs
• Provides immunity to magnetic fields
• Especially important in low level microphone circuits
• Terminal blocks and XLR connectors vulnerable to magnetic fields
• “Star-Quad” magnetic pickup ≈ 40 dB less than STP
• Provides immunity to electric fields for unshielded pairs
• Be careful with cable shield grounding
• ALWAYS at DRIVER
• OK at BOTH
• NEVER at RECEIVER ONLY

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 162

Bill Whitlock 162

An Overview of Audio System Grounding & Interfacing 9/4/2012

Balanced to Unbalanced Audio

• Also called “Pro to Consumer” or “+4 to −10” converter
• Signal level difference is a legitimate concern
• Consumer inputs are easily over-driven by pro levels
• Interface requires voltage loss of 12 dB
• If pro output level is reduced, metering and S/N are degraded
• Resistor “pad” can be used but provides no isolation
• Variety of balanced output circuits make direct connections risky!
• Some will misbehave or be damaged if either output line is grounded
• “Servo-balanced” or transformer outputs must have one line grounded to
drive an unbalanced input
• Grounding one line at the driver reduces the entire interface to unbalanced, making
it vulnerable to ground noise
• “Servo-balanced” outputs (SSM2142 and clones) may become unstable if one line is
grounded at receive end of cable, which is necessary to reject noise

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 163

Making this transition with a CABLE ALONE is indeed RISKY BUSINESS! The
12 dB voltage loss IS a big deal. Recall that +4 dBu is “pro” signal reference level and
that −10 dBV is “consumer” signal reference level.

Bill Whitlock 163

An Overview of Audio System Grounding & Interfacing 9/4/2012

“Universal” Pro to Consumer Converter

• Interface works with any output stage
• Step-down 4:1 input transformer attenuates signal 12 dB
• Superior noise rejection

Jensen Transformers

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 164

The 2-channel model PC-2XR is shown. It’s available with various other input and
output connectors – or it can be permanently wired using the detachable screw-
terminal strip (on its rear).

Bill Whitlock 164

An Overview of Audio System Grounding & Interfacing 9/4/2012

Noise Rejection in Pro to Consumer Converters

No transformer

Output transformer

Input transformer

60 Hz 3 kHz

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 165

Again, the INPUT transformer performs some 50 dB better than an output

transformer.

Bill Whitlock 165

An Overview of Audio System Grounding & Interfacing 9/4/2012

Solutions for Analog Video

• Objective is to reduce voltage drop over the length of the video cable
• Reduce ground voltage difference (AC power work may be expensive)
• Reduce cable length (may not be possible)
• Replace cable with lower shield-resistance type
• Use an isolator in the signal path to reduce ground loop current
• There are 3 types of isolators, each with its own tradeoffs
• Baluns for video over UTP generally have no isolation

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 166

Bill Whitlock 166

An Overview of Audio System Grounding & Interfacing 9/4/2012

1 – Video Isolation Transformer

• Magnetically couples signal across an electrically-insulated barrier
• Signal bandwidth 10 Hz to 10 MHz is near limits of technology
• Extremely high common-mode impedance at 60 Hz
• Effectively stops ground loop current and has very high CMRR
• Tolerates extreme ground voltage differences
• 300 V-rms typical, limited only by internal insulation
• Suited for composite or S-video, even when ground voltage
difference is very high and/or cables are very long
• Wide application in surveillance and security TV
• Lowest cost

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 167

Security and surveillance video is a primary market for these units.

Bill Whitlock 167

An Overview of Audio System Grounding & Interfacing 9/4/2012

Video Isolation Transformer

LOOP CURRENT ELIMINATED

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 168

True transformers will tolerate the highest ground voltage differences of the 3
solutions. The Jensen VB-1BB will withstand 300 V between input and output,
but this is obviously a safety hazard and not recommended.

Bill Whitlock 168

An Overview of Audio System Grounding & Interfacing 9/4/2012

Composite Video Isolation Transformers

Jensen Transformers

RDL

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 169

There are also models available for S-VHS format. These transformers find their
main application in security systems.

Bill Whitlock 169

An Overview of Audio System Grounding & Interfacing 9/4/2012

2 – Video Common-Mode Choke

• Called “hum eliminator,” “hum suppressor,” “hum-bucker,” “ground
loop inhibitor”, and (mistakenly) a “transformer”
• Fundamentally different from a transformer
• Signal not converted to magnetic field, only ground voltage
difference produces magnetic field in its core
• Maximum ground voltage difference limited by core saturation
• Most will tolerate up to 2 V rms, but some less than 200 mV rms
• “Windings” actually a length of miniature coaxial cable
• LF response extends to 0 Hz (DC)
• HF response determined by coax, typically >400 MHz
• Suited for high-resolution component or computer video

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 170

Bill Whitlock 170

An Overview of Audio System Grounding & Interfacing 9/4/2012

Video Common-Mode Choke

CURRENT TYPICALLY REDUCED TO < 1%

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 171

Bill Whitlock 171

An Overview of Audio System Grounding & Interfacing 9/4/2012

Video Common-Mode Chokes

3 and 5-Channel Units for RGB, RGBHV, etc.

Jensen Transformers

“Breakout cables”
allow use with
Allen Avionics computer video
signals

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 172

The 3-channel Jensen VBH-3RR and the 5-channel VBH-5BB are shown here.
The channels on these units are completely independent electrically. Single-channel
units are also available and are often used on SDI video interfaces.

Bill Whitlock 172

An Overview of Audio System Grounding & Interfacing 9/4/2012

3 – Video Isolation Amplifier

• Called “hum-bucking” or “differential-input” amplifiers
• Use wideband diff-amp to provide CM voltage rejection
• CMRR is affected by accuracy of the 75 Ω source and cable
resistances (a function of their length)
• “Trim” adjustment necessary for best rejection
• Require power and use semiconductors, making them potentially vulnerable
to transient voltage damage
• Bandwidths typically range up to 200 MHz
• Often have useful features like multiple outputs and adjustable gain

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 173

Bill Whitlock 173

An Overview of Audio System Grounding & Interfacing 9/4/2012

Video Isolation Amplifier

CURRENT SIGNIFICANTLY REDUCED

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 174

Bill Whitlock 174

An Overview of Audio System Grounding & Interfacing 9/4/2012

Video Isolation Amplifier

Video Accessory Corp.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 175

The unit pictured is VAC part number 100-0-007.

Bill Whitlock 175

An Overview of Audio System Grounding & Interfacing 9/4/2012

How Much Overall System Bandwidth?

Source Format Resolution @ FPS BW
VHS NTSC 240 L @ 30 4 MHz

Broadcast NTSC 330 L @ 30 6 MHz

Broadcast, DVD SDTV 480i 720x480 @ 30 15 MHz

Prog DVD EDTV 480p 720x480 @ 60 30 MHz

Computer XGA 1024x768 @ 60 71 MHz

HDTV, DVD 720p 1280x720 @ 60 83 MHz

HDTV, DVD 1080i 1920x1080 @ 30 93 MHz

Computer SXGA 1280x1024 @ 60 118 MHz

Courtesy Peter Putman

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 176

Peter Putman is a video expert who writes for SVC and other magazines.

Bill Whitlock 176

An Overview of Audio System Grounding & Interfacing 9/4/2012

Video Isolator Considerations

• System bandwidth required
• Length and type of existing cables
• Voltage difference across the interface
• Remember (for all types of devices):
• Noise rejection will decrease with cable length
• Ground voltage difference may exceed isolator capability
• Cables themselves limit overall bandwidth
• Critical specs may be conspicuously absent or unqualified as to test
conditions – beware the “sketchy-spec” syndrome!
• Baluns for video over UTP generally provide NO isolation!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 177

Not surprisingly, marketing obscures the truth about many of these

products, too.

Bill Whitlock 177

An Overview of Audio System Grounding & Interfacing 9/4/2012

Cable TV Grounding
• Requirements of NEC Article 820
• Coaxial cables entering a building must have their outer sheaths grounded as
close as practicable to their point of entry
• Grounding wire must be as short as possible, insulated, and have current-
carrying capacity equal to the shield of the coaxial cable – no smaller than
#14 AWG, but larger than #6 AWG is not required
• Bond to main power grounding electrode if within 20 feet
• If not, an 8-foot ground rod must be used and bonded to main power
grounding electrode with #6 AWG or larger wire
• A metallic water pipe, if within 5 feet of where it enters the building,
may be used as a grounding electrode
• All such lightning ground wiring should be a straight as possible, using gentle
curves rather than sharp bends
• PVC conduit is preferred for physical protection of this wiring

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 178

Same comments apply regarding #6 bond wire and protection of wiring.

Bill Whitlock 178

An Overview of Audio System Grounding & Interfacing 9/4/2012

Satellite Dish Grounding

• Requirements of NEC Article 810
• Outdoor masts or antenna supports must be grounded
• Signal cables must use listed “antenna discharge unit” (grounding block)
located as near as practicable to the point of entrance, inside or outside
• Grounding wires to be no smaller than #10 AWG copper or #17 AWG copper-
clad steel, need not be insulated, and may be outside or inside the building
• Bond to main power grounding electrode if within 20 feet
• An 8-foot ground rod may be used but it must be bonded to the main
power ground electrode with #6 AWG or larger
• An interior metallic water pipe, if within 5 feet of where it enters the
building, may be used as a grounding electrode
• Ground wiring should use gentle curves, not sharp bends
• PVC conduit is preferred for physical protection of this wiring

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 179

The NEC requirements are often modified by local “authorities having jurisdiction” or
AHJs.

Bill Whitlock 179

An Overview of Audio System Grounding & Interfacing 9/4/2012

CATV Isolator Breaks Loop

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 180

The VRD-1FF is shown.

Bill Whitlock 180

An Overview of Audio System Grounding & Interfacing 9/4/2012

CATV/Cable Modem Isolators

• Also used on TV/FM/MATV antenna cables
• Must be downstream of lightning ground connection
• Splitters connect all shields, enabling ground loops between
systems
• Prevent loops with an isolator at each destination
• Beware of “sketchy-spec” isolators – some are truly awful!
• Two 75 Ω to 300 Ω matching transformers back-to-back is particularly bad
• Isolators will not work between DBS dish and receiver!
• They block DC power to dish

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 181

A popular DIY isolator uses two back-to-back 300Ω to 75Ω converters ...
and has really horrible frequency response!

Bill Whitlock 181

An Overview of Audio System Grounding & Interfacing 9/4/2012

CATV Isolator Specifications

Frequency Response:
5 MHz – 1,300 MHz ± 0.5 dB

Standing Wave Ratio

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 182

VSWR and Reflection Coefficient are measures of 75Ω impedance accuracy.

VSWR under 1.2 is considered excellent.

Bill Whitlock 182

An Overview of Audio System Grounding & Interfacing 9/4/2012

Satellite System Isolation

• Satellite dish must be grounded per NEC section 810
• “Isolated Grounding Block” or IGB allows dish and downleads to
remain safely ungrounded under normal conditions
• If DBS receiver has 3-prong AC plug, that’s a separate issue that
may require isolators at receiver outputs – in which case no IGB is
needed

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 183

Bill Whitlock 183

An Overview of Audio System Grounding & Interfacing 9/4/2012

The Isolated Grounding Block

Ground circuit remains open until

voltage reaches 90 volts, which ionizes
the gas-discharge tube that can sustain
18,000 A surge. It then automatically
re-opens, ready for the next “event.”

To Ground Connection

NOTHING can fully protect from direct strikes that can reach 150,000 A

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 184

The GI-2FF is a new model this year!

Bill Whitlock 184

An Overview of Audio System Grounding & Interfacing 9/4/2012

Ground Loops and Data Interfaces

• RS-232 digital interface is unbalanced

• Since it’s digital, noise symptoms are usually deemed “unexplainable”
• Since it’s unbalanced, ground noise voltage directly adds to the signals
• RS-422, RS-485, and USB are balanced but their common-mode
voltage range is quite limited
• Shielded cables, grounded at both ends can help reduce common-mode
voltage, especially at high frequencies
• Ethernet uses integral isolation transformers … making it very robust
• Many data line isolators are optically-coupled and can withstand
ground voltage differences up to 2 kV

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 185

Unbalanced interfaces are “trouble waiting to happen” no matter where

they are!

Bill Whitlock 185

An Overview of Audio System Grounding & Interfacing 9/4/2012

Ground Isolators for Data Interfaces

RS-232 Isolator USB 2.0 Isolator

see www.bb-elec.com see www.electronics-shop.dk
also, for similar products
see www.telebyteusa.com

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 186

Jensen is currently exploring the feasibility of an HDMI isolator ...

Bill Whitlock 186

An Overview of Audio System Grounding & Interfacing 9/4/2012

Recap: AC Power & Unbalanced Interfaces

• Safety grounding must NEVER be compromised!
• Earth ground has essentially nothing to do with noise
• There are always voltage differences between “grounds”
• Magnetic coupling in premises power wiring is the major source of
these voltage differences
• Unbalanced interfaces are a potential noise disaster!
• Noise in unbalanced interfaces is not about shielding
• “Ground loops” are always created when signal cables interconnect
AC-powered equipment
• Ground isolators cure the “unbalanced” problem

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 187

Bill Whitlock 187

An Overview of Audio System Grounding & Interfacing 9/4/2012

POWER-LINE ISSUES and TREATMENTS

“Today’s residential systems contractors face unprecedented challenges
where high resolution, trouble-free operation is required. From inducing
AC ground loops, video hum bars, static bursts, damage from AC line
surges and variable audio and video performance, comprehensive control
and conditioning of AC power is no longer an option.“

Similar assertions are made by so many, so often that

most customers accept them as truth. This “power-line
paranoia” exploits the seductively-appealing notion that
“dirty power” is to blame for all system noise problems!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 188

“Power Conditioning” has been a high-growth, and high-profit, business for nearly 20
years. Most of the claims made for such “conditioning” are never
substantiated in any meaningful or scientific way. Many of the claimed
benefits, as well as carefully contrived demonstrations, truly test the limits of
credibility in an engineering sense ... “deeper colors” ... “better soundstage imaging”
... WHAT?

Bill Whitlock 188

An Overview of Audio System Grounding & Interfacing 9/4/2012

“Mesh” vs “Star” Controversy

• “Mesh” grounding
• Attempts to create a uniform potential over large
areas using a network of ground wires that, to some
degree, approximates a solid ground plane at high
frequencies. Ground loops exist but are ignored.
• Long advocated by the IT community since 60
Hz is “out of band” for most data and signal-to-noise
requirements are very low. As clock frequencies have
risen, this method has gradually fallen out of favor.
• “Star” (single-point) grounding
• Prevents coupling of noise by avoiding shared
conductors wherever possible
• Long advocated by the analog community since
60 Hz is “in band” for audio and video and signal-to-
noise must be very high. This method routes noisy
currents in conductors separate from sensitive ones.
Varous circuit returns are joined at a single point.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 189

The IT community generally persists in claiming that everyone should ground systems
the way they do – failing to appreciate the reasons for “star” grounding in audio
equipment. The PANI system was developed by telephone companies primarily to deal
with lightning and fault currents arriving at switching stations via subscriber lines. The
PANI system should NEVER be used as a substitute for the premises wiring safety
ground system. NEC requires that the safety ground conductor be inside the same
conduit as the line and neutral conductors ... always!

Bill Whitlock 189

An Overview of Audio System Grounding & Interfacing 9/4/2012

Buildings with Metallic Conduit

Normal saddle-grounded or
“SG” outlets connect safety
ground to conduit when the
J-box is installed.

“SG” Outlet

Leviton

VDIFF is magnetically induced

into safety ground wires as
current flows in other wires
in the conduit. VDIFF is what
drives ground loop currents
through signal cables.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 190

A good way to effectively eliminate this induced voltage is tight twisting of

each L-N conductor pair in the conduit.

Bill Whitlock 190

An Overview of Audio System Grounding & Interfacing 9/4/2012

“Technical” or “Isolated” Grounding

• Conduit touching ANY separately-
“IG”
grounded metal causes new noise currents Outlet
in safety ground system
• Isolated-ground outlets do not connect safety
ground to their mounting saddle but only to the
green wire Leviton

• IG outlets reduce extraneous, and often

intermittent, ground noise problems
• Covered by NEC Article 250-74
• Does not apply to premises wired with Romex®
and/or plastic J-boxes
• It’s already an isolated ground system!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 191

It remains controversial, especially among electrical contractors, but even

among audio experts. I believe there are many situations where it can be of
great benefit – which ones depends on many factors.

Bill Whitlock 191

An Overview of Audio System Grounding & Interfacing 9/4/2012

Hidden Connections

... add complications

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 192

If conduit is part of the safety ground system, these contacts allow other
power-line currents to flow in them – with resulting voltage differences.
THIS IS WHY ISOLATED GROUNDING IS POPULAR IN LARGE COMMERCIAL
BUILDINGS WHERE UNINTENDED, AND OFTEN INTERMITTENT, CONTACTS
EXIST BETWEEN CONDUIT AND OTHER GROUNDED STRUCTURES. Contact
with building steel, HVAC structures, gas pipes, etc. can cause intermittent
problems.

Bill Whitlock 192

An Overview of Audio System Grounding & Interfacing 9/4/2012

And Lots of Magnetic Fields

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 193

Magnetic fields can induce voltages in nearby cables. They can also cause serious
hum in electric guitars in the vicinity! Note that VDROP is created as high
current (100 A or more) flows in the resistance of the neutral wire
connecting the two utility poles. When neutral is tied to plumbing (as NEC
requires) in each building, another route for current flow is created through
the plumbing in each building and the water pipe in the street. In a
consulting job, I found 13 A flowing in a water pipe under the stage in a
rehearsal hall. Their complaint was that no one could play an electric guitar
(very sensitive magnetic field detector) on that end of the stage. Is it any
wonder? The current was generating a magnetic field with a strength of
about 70 mG at waist level on that end of the stage. During a trip to the
basement, my clamp-on ammeter measured the 13 A in the pipe. The cure
was to install an insulated coupler in the water line right at the building side
of the meter. That kept the premises plumbing properly grounded while
stopping the current flow into the street!

Bill Whitlock 193

An Overview of Audio System Grounding & Interfacing 9/4/2012

Recommended Instruments

Ideal SureTest® AEMC AlphaLab TriField®

Circuit Analyzer Clamp-On Multimeter EMF Meter
Model 164 Model 565 www.trifield.com
(about $230) (about $270) (about $170)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 194

The Sure-Test analyzer can measure line voltage, test for full-load voltage drop (on
either 15A or 20A rated branch circuits), test for ground resistance, and discover
dangerous outlet wiring errors. The Tri-Field meter will indicate AC magnetic fields
(with 3 milli-gauss and 100 milli-gauss full-scale ranges), AC electric fields, and has a
broadband RF detector as well (suitable for testing leakage for microwave ovens or
finding hidden “bug” transmitters). Being able to measure low-level magnetic fields
can be especially handy in cases where the complaint is that “my electric guitar hums”
– which is very often due to ambient magnetic fields.

Bill Whitlock 194

An Overview of Audio System Grounding & Interfacing 9/4/2012

“Power Conditioning” Can Include

• Short-term over-voltage protection (mostly MOV-based)
• “Transients” and “spikes”
• Long-term over/under-voltage protection (voltage regulation)
• “Swells” and “surges”
• “Sags” and “brown-outs”
• Isolation transformers (control “common-mode” noise)
• “Balanced” power (control leakage current noise)
• Filters (control high-frequency noise)
• Other “bizarre” stuff (presumably to invoke magic)
• “Tice clock”
• Exotic cords, plugs, outlets, and wire

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 195

It’s dangerous to generalize, but some of these functions are “band-aids” for poorly-
designed equipment and some are legitimate concerns.

Bill Whitlock 195

An Overview of Audio System Grounding & Interfacing 9/4/2012

Facts to Keep in Mind ...

• Ground voltage differences are the driving force behind nearly all
audio and video noise problems
• They are generated by magnetic induction in the premises
wiring itself, the “conduit transformer”
• They approach zero between system devices powered from
closely-spaced AC outlets
• Many benefits attributed to “power conditioning” are
actually due to the tight cluster of output outlets!
• The coupling mechanism in signal cables that causes hum
and buzz ceases to operate beyond about 30 kHz
• Power-line “common-mode” (neutral to ground) noise is, by
definition, zero at the N-G bond

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 196

Common-mode noise can only be generated within a premises, between the N-G bond
and the load in question. For system noise reduction, a simple outlet strip often
works as well as, or even better, than a “power conditioner.” Filtering in most
power conditioners only begins to work at frequencies above 30 kHz – at that and
higher frequencies, noise no longer couples in cables like it does at lower (audio)
frequencies.

Bill Whitlock 196

An Overview of Audio System Grounding & Interfacing 9/4/2012

Power Isolation Transformers

• Ground loops unaffected since input and output grounds are tied
• Applies to all cord-connected filters and isolation transformers
• UL 1950 and IEC 950 prohibit “isolation” of safety ground
• Electrostatic/Faraday shielding can add noise current to safety ground
• Touted high-frequency noise reduction specs are unrealistic
• Measurements made in lab on low-Z ground plane
• Real-world grounding is via high-Z wires or conduit

• Only a permanently-installed, hard-wired transformer, configured as a

“separately-derived system” per Code, can establish a new N-G
bond, which has many noise-reducing benefits
• See Middle Atlantic Products white paper (see Recommended Reading at end)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 197

See www.middleatlantic.com to download the white paper.

Bill Whitlock 197

An Overview of Audio System Grounding & Interfacing 9/4/2012

Power-Line Noise Spectrum vs Coupling

Noise Coupled in Noise Cancelled in
Shielded Cables Shielded Cables

40kHz
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 198

This plot is an actual spectrum analysis of leakage current in a 3 nF capacitance

connected to an AC outlet at former Jensen headquarters in Van Nuys, California. It
represents a more-or-less typical common-mode + normal mode noise spectrum. The
graph is split into sections below and above about 40 kHz. Below 40 kHz, magnetic
coupling in shielded cables is minimal so that a) common-impedance coupling occurs
in unbalanced interconnect cables, and b) end-to-end voltage difference is presented
to the line receiver as common-mode voltage in balanced interfaces. Above 40 kHz,
magnetic coupling in shielded cables is maximum, allowing the end-to-end voltage
difference to be fully coupled to the cable’s inner conductor(s), which subtracts it from
the signal(s) seen at the receiver, resulting in zero noise coupling in unbalanced
interfaces and zero common-mode voltage in balanced interfaces. It should be clear
that any power-line filter that only begins to attenuate at 40 kHz is a case of
“too little, too late.”

Bill Whitlock 198

An Overview of Audio System Grounding & Interfacing 9/4/2012

Power-Line Filters Useless at Audio Frequencies!

• Virtually all AC power-line filters,
whether very simple or complex, have
trivial attenuation below 30~50 kHz
• Shielded cables (balanced or
unbalanced) nullify the effects of
shield current above 10~30 kHz
• A power-line filter that might help
would be as big as a refrigerator
• Serious attenuation would need to begin at
a few hundred Hz, not tens of thousands!
• Filters are most effective when
located at the noise source

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 199

Bill Whitlock 199

An Overview of Audio System Grounding & Interfacing 9/4/2012

Power Isolation Transformer

Cord-connected “point of use” type shown

Safety ground can’t legally be interrupted, so ground loops will remain.

The Faraday shield (also mistakenly called an “electrostatic shield”) prevents
capacitive coupling of high-frequency noise to the output but it couples the noise
into the local safety ground system. This explains reports that such transformers
often make noise problems worse.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 200

Plug-and-Cord connected isolation transformers are generally of very little

benefit. Further, they cannot legally disconnect safety ground.

Bill Whitlock 200

An Overview of Audio System Grounding & Interfacing 9/4/2012

So-Called “Balanced Power”

• Properly called SYMMETRICAL power
• Has very seductive intuitive appeal
• NOT similar to balanced audio lines in any way!
• Uses transformer having 120 V center-tapped secondary
• Both line and neutral output blades are energized at 60 V
• Although advertising often implies endorsement, NEC seriously restricts
its use – because it’s potentially dangerous!
• ONLY FOR PROFESSIONAL USE
• NOT to be used with lighting equipment, especially screw-base bulbs
• MUST have GFCI at outputs
• Only technical function is to reduce leakage currents
• Leakage currents are trivial system noise sources
• Reported noise reduction generally less than 10 dB
• Any real benefit likely due to its clustered outlets

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 201

This is an example of “marketing gone wild” if ever there was one!

Bill Whitlock 201

An Overview of Audio System Grounding & Interfacing 9/4/2012

Symmetrical AC Power

Total leakage current to safety ground could be cancelled

IF parasitic capacitance pairs, shown as C1/C2 and C3/C4,
were precision matched. This is not likely to ever happen!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 202

The theory is pretty simple: if symmetrical (equal but opposite instantaneous polarity)
AC voltages are applied to equal-valued capacitors, say C1 and C2 above, the two
current flows would also be equal but opposite and cancel each other … no net
current into safety ground. But leakage currents flowing in safety ground
wiring are not a significant source of ground voltage differences ... as so
many apparently believe.

Bill Whitlock 202

An Overview of Audio System Grounding & Interfacing 9/4/2012

The “Ground Lift” Feature

Legally Disconnects Safety Ground … BUT

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 203

Because this unit has the required GFCI on its output outlets, it is legally
allowed to interrupt safety ground. Paradoxically, the same can’t be done
when a premises AC outlet is replaced with a GFCI type. NEC says that, if a
safety ground is available to the outlet, it MUST be connected … even if the
outlet is a GFCI protected one. Although this unit claims to eliminate ground
loops via the “LIFT” switch, there’s a catch ...

Bill Whitlock 203

An Overview of Audio System Grounding & Interfacing 9/4/2012

“Nuisance Trips” May Occur

Now, leakage currents must flow in signal

cables to return to ground/neutral! The GFCI
may trip if current exceeds 4 mA.

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 204

This is especially likely if several devices are plugged into the “lifted” outlets
... individual leakage currents accumulate!

Bill Whitlock 204

An Overview of Audio System Grounding & Interfacing 9/4/2012

GFCI for Safety

• Ground-fault circuit interrupter or GFCI does not
require a ground connection to function properly
• Senses difference between line and neutral currents
• Current NOT returning on neutral is presumed to be
Leviton
flowing through a person
• Required to trip between 4 and 7 mA
• NEC allows upgrade of 2-prong AC outlets (in pre-
1960 homes) with GFCI outlets
• Ground must be connected if available
• If no ground is available, outlet must be marked “No
Equipment Ground” and may be used without grounding

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 205

Often these GFCI units are the best solution for an older home with 2-prong
outlets.

Bill Whitlock 205

An Overview of Audio System Grounding & Interfacing 9/4/2012

Noise IS Sometimes the Electrician’s Fault

• In new construction, some AV contractors claim that up to 25% of AC
outlets are mis-wired.
• Neutral-Ground swaps at outlets are common but cannot be detected by
ordinary low-cost outlet testers.
• Multiple Neutral-Ground bonds is also common when sub-panels are part
of premises power distribution
• NEC allows only one and it must be at the main disconnect panel
• Many boxes arrive with the N-G jumper in place but it must be removed if the
box is used as a sub-panel
• Both cause extraordinarily large voltage drops in the safety ground wiring
• Load current, up to 20 A, normally flows in Neutral wires
• Leakage current, up to perhaps 0.1 A, normally flows in Ground wires

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 206

NEC allows a 5% voltage drop at an outlet under full load. This says that the voltage
at a 120 V outlet under full-rated load could drop by 6 V to 114 V. This drop will, of
course, be highest at the outlet farthest from the panel. Since wire gauge must be the
same for L and N, it means that 3 V will be dropped on each (which makes the wires
warm – that’s 60 watts of heat for each on a 20 A circuit). But if the current returns
on safety ground rather than neutral, we have up to 3 V of voltage difference in the
safety ground system. This is some 100 times higher than voltage differences in a
properly-wired system. But, because everything except your AV system works as it
should, electricians often dismiss us by saying “It’s not my problem. Look – everything
works!”

Bill Whitlock 206

An Overview of Audio System Grounding & Interfacing 9/4/2012

The N-G Swap, a Real Problem THIS

WON’T
AEMC
FIND IT

Load current now flows in safety ground wiring. Its voltage

drop, up to 3 volts, appears as a ground difference voltage to
the system – creating truly severe noise problems! The swap is
not detectable with any outlet tester, but a clamp-on ammeter
can easily find it!
Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 207

This wiring error is normally not a safety hazard but definitely a big headache
regarding system noise. The simple outlet tester can’t find a N-G swap because both N
and G are grounded at the breaker box. The simple tester can only verify that there’s
120 V between the hot blade and the other two. If the other two are reversed, it has
no way to know! Again, it’s all about where the current is flowing – and it should NOT
be in the safety ground!! In general, any current over 100 mA in a safety
ground wire indicates a problem.

Bill Whitlock 207

An Overview of Audio System Grounding & Interfacing 9/4/2012

Another: Multiple or “Bootleg” N-G Bonds

• NEC prohibits more than one N-G bond
• Must be at “main disconnect” power panel
• Others allowed ONLY at “separately-derived system” transformers
• Extra bonds most often found at sub-panels
• New sub-panels often arrive with N-G bond jumper in place
• Electrician forgets to remove jumper when panel is installed
• Accidental connections
• Neutral wire pinched and shorted to metallic J-box
• Staple shorts N to G in Romex
• Electric dryers (240 V) in old homes may tie N to plumbing
• Testing should be done only by a qualified electrician!

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 208

There are various methods for finding “stray” N-G bonds. An ohmmeter is rarely
useful. Because most methods involve disconnecting the legal N-G bond and
then re-energizing circuits to trace the errant current flow, the procedure is
potentially dangerous and should be done only by (or under the direct
supervision of) an experienced, qualified professional. NEVER WORK ALONE
AND KEEP ONE HAND IN YOUR BACK POCKET!

Bill Whitlock 208

An Overview of Audio System Grounding & Interfacing 9/4/2012

IS IT SAFE?
• It’s NOT UL Listed!
• This device claims to eliminate
ground loop problems
• Uses a pair of back-to-back rectifier
Ebtech
diodes in series with safety ground
connection
• Essentially an open circuit until
ground voltage difference reaches
about ± 0.6 V (~ 400 mV rms)
• But will it survive a 100 A fault for
2.5 seconds or a 1,000 A fault for
10 ms* without opening up?

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 209

The current and time numbers noted * were taken from a survey by Underwriters
Laboratory of 1,000 homes and represent a realistic range of short-circuit current and
time taken for the circuit breaker to trip open (it depends mostly on the distance
between tested outlet and breaker panel). I’ve done a “teardown” (de-potting) of this
product. Based on what I saw (and know about the failure mechanisms of
semiconductors), I would not depend on this device to protect my life during an
equipment fault scenario. These are serious questions!

Bill Whitlock 209

An Overview of Audio System Grounding & Interfacing 9/4/2012

About Surge Suppression

• Paranoia and pseudo-science make it easy to sell surge protection
devices of all kinds
• Mindless use of conventional suppressors can actually increase
risk of equipment damage!
• By far, the most widely-used suppressors employ three MOV (metal-
oxide varistor) devices that divert (shunt) surge current into the
safety ground system
• When surges occur, they create brief but extreme voltage
differences in the safety ground system!
• Equipment interfaces are often damaged …

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 210

A case of marketing that involves selling “protection”. Do you know what a

Joule is?

Bill Whitlock 210

An Overview of Audio System Grounding & Interfacing 9/4/2012

DANGER Will Robinson, DANGER!

6,000 V

“3-Mode” MOV Suppressor

1,800 V
0 V REF

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 211

Upwards of 2,000 volts across any signal interface is very likely to release the
factory-installed smoke!

Bill Whitlock 211

An Overview of Audio System Grounding & Interfacing 9/4/2012

Think Different …
• Protect from the biggest threat, lightning-induced surges, with a
“whole house” device at main power entry panel – MOV is OK there
• Whether on a branch-circuit or in the equipment rack, I recommend
only series-mode suppressors
• They present a high impedance (inductor) to block the surge energy
• They DON’T dump surge currents into safety ground

SurgeX

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 212

Surge-X is one of several makers of series-mode surge suppressors. I have no

financial ties to the company, I just like the product.

Bill Whitlock 212

An Overview of Audio System Grounding & Interfacing 9/4/2012

Thanks for Your Attention!

Have a question? Want a copy of this presentation? See me at CEDIA
Booth 3052 or send me an e-mail: [email protected]

Focal Press
Handbook for Sound Engineers, 4th edition,
has three chapters written by Bill Whitlock

Jensen Transformers, Inc.

9304 Deering Avenue
Chatsworth, California 91311
www.jensen-transformers.com
Tel: (818) 374-5857 - Fax: (818) 374-5856

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 213

Jensen Transformers, Inc., 9304 Deering Avenue, Chatsworth, CA 91311

Phone: (818) 374-5857 - Internet: www.jensen-transformers.com - E-mail:
[email protected]
Open 9 AM to 5 PM, Monday thru Thursday, closed Fridays, visitors by appointment
only.
All content Copyright 2012, William E. Whitlock

Bill Whitlock 213

An Overview of Audio System Grounding & Interfacing 9/4/2012

Suggested Reading
1. Integrating Electronic Equipment and Power into Rack Enclosures, Middle-Atlantic Products, at
http://www.middleatlantic.com/pdf/PowerPaper.pdf and its addendum at
http://www.middleatlantic.com/pdf/Power%20Paper%20Addendum.pdf. This document was a collaborative
effort of the MAP staff and several industry experts, including Neil Muncy and myself.

Following Papers at http://www.apc.com/prod_docs/results.cfm?DocType=White%20Paper&Query_Type=10

2. White Paper #8, Inter-System Ground Noise: Causes and Effects, Rev 2, APC, Neil Rasmussen
3. White Paper #9, Common-Mode Susceptibility of Computers, Rev 1, APC, Neil Rasmussen
4. White Paper #21, Neutral Wire Facts and Mythology, Rev 1, APC, James Spitaels

Following Papers at http://www.jensen-transformers.com/apps_wp.html

5. Audio Transformers, by Bill Whitlock, from Handbook for Sound Engineers, Third Edition, Glen Ballou, Editor

For a good, yet entertaining introduction to electricity, check out The Manga Guide to Electricity by Kazuhiro Fujitaki

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 214

Bill Whitlock 214

An Overview of Audio System Grounding & Interfacing 9/4/2012

The Demonstration Setup

Generator Isolator Analyzer
+ + + +
−
Balanced Cable
− − −
GND GND GND GND

+ + Unbalanced Cable +
− − −
GND GND GND
Adapter

Conduit Wiring Simulator

(modified power strip)

Bill Whitlock, 9/4/2012 Overview of Audio System Grounding & Interfacing 215

Since the Audio Precision instrument has a common-ground for its generator and
analyzer. Therefore, the “isolator” above allows the generator outputs to reference a
different ground potential (the green wire). This re-creates a real-world scenario
where the signal source device and destination device are physically separated and
have a small ground voltage difference between them.

The “Conduit Wiring Simulator” inserts the primary of a transformer between the
neutral connections of the two outlets. The transformer’s secondary is inserted
between the ground connections of the two outlets. Any current drawn by loads on
the orange outlet flows through the transformer primary and magnetically induces a
ground voltage difference between the outlet ground connections. This is exactly what
happens in normal commercial premises wiring. It simulates a conduit run about 15
feet long (each winding of the transformer has an inductance of about 7 uH) where
wire positioning is worst case. For more details on the “conduit transformer”, see
“Ground Loops: The Rest of the Story” by Bill Whitlock and Jamie Fox, presented at
the 129th AES Convention, Nov 4-7, 2010.

Bill Whitlock 215