Handbook for Sound Engineers

1204 Chapter 32

(32-10)

Most listeners describe 10 dB level decreases or
increases as halving or doubling loudness, respectively,
and 2 dB or 3 dB changes as just noticeable. Under
laboratory conditions, well-trained listeners can usually
identify level changes of 1 dB or less. The dynamic
range of an electronic system is the ratio of its
maximum undistorted signal output to its residual noise
output or noise floor. Up to 120 dB of dynamic range
may be required in high-end audiophile sound systems
installed in typical homes.^39

32.6.2 Troubleshooting

Under certain conditions, many systems will be accept-
ably noise-free in spite of poor grounding and inter-
facing techniques. People often get away with doing the
wrong things! But, notwithstanding anecdotal evidence
to the contrary, logic and physics will ultimately rule.
Troubleshooting noise problems can be a frus-
trating, time-consuming experience but the method
described in Section 32.6.2.2 can relieve the pain. It
requires no electronic instruments and is very simple to
perform. Even the underlying theory is not difficult. The
tests will reveal not only what the coupling mechanism
is but also where it is.

32.6.2.1 Observations, Clues, and Diagrams

A significant part of troubleshooting involves how you
think about the problem. First, don’t assume anything!
For example, don’t fall into the trap of thinking, just
because you’ve done something a particular way many
times before, it simply can’t be the problem. Remember,
even things that can’t go wrong, do! Resist the tempta-
tion to engage in guesswork or use a shotgun approach.
If you change more than one thing at a time, you may
never know what actually fixed the problem.
Second, ask questions and gather clues! If you have
enough clues, many problems will reveal themselves
before you start testing. Be sure to write everything
down—imperfect recall can waste a lot of time! Trou-
bleshooting guru Bob Pease^40 suggests these basic
questions:

1. Did it ever work right?
   2. What are the symptoms that tell you it’s not
   working right?
   3. When did it start working badly or stop working?
   4. What other symptoms showed up just before, just
   after, or at the same time as the failure?

Operation of the equipment controls, and some

elementary logic, can provide very valuable clues. For

example, if a noise is unaffected by the setting of a gain

control or selector, logic dictates that it must be entering

the signal path after that control. If the noise can be

eliminated by turning the gain down or selecting

another input, it must be entering the signal path before

that control.

Third, sketch a block diagram of the system.

Fig. 32-40 is an example diagram of a simple home

theater system. Show all interconnecting cables and indi-

cate approximate length. Mark any balanced inputs or

outputs. Generally, stereo pairs can be indicated with a

single line. Note any device that is grounded via a

three-prong ac plug. Note any other ground connections

such as equipment racks, cable TV connections, etc.

32.6.2.2 The Ground Dummy Procedure

An easily constructed adapter or ground dummy is the

key element in this procedure. By temporarily placing

the dummy at strategic locations in the interfaces,

precise information about the nature and location of the

problem is revealed. The tests can specifically identify:

1. Common-impedance coupling in unbalanced cables.


2. Shield current-induced coupling in balanced
   cables.


3. Magnetic or electrostatic pickup of nearby mag-
   netic or electrostatic fields.


4. Common-impedance coupling (the pin 1 problem)
   inside defective devices.


5. Inadequate CMRR of the balanced input.

The ground dummy can be made from standard

connector wired as shown in Figs. 32-41 and 32-42.

Since a dummy does not pass signal, mark it clearly to

help prevent it, being accidentally left in a system.

Each signal interface is tested in four steps. As a

general rule, always start at the inputs to the power

amplifiers and work backward toward the signal

sources. Be very careful when performing the tests not

to damage loudspeakers or ears! The surest way to

avoid possible damage is to turn off the power ampli-

fier(s) before reconfiguring cables for each test step.

dB 20

E 1

E 2

= log----- -

dB 20

I 1

I 2

= log----