Connections are Vital
from Component to Component
Audio manufacturers go to considerable lengths to make each
component behaves itself, but there is little they can do to make sure the overall system
will remain free of signal degradation. And yet, the division of a stereo system into
separate units is really only a matter of convenience; it remains a system, and its
various parts must work well together. A significant contributor to this is the method of
connecting one part to another.
Wherever it is placed in an audio system, any wire is
subject to a number of ills, but within any particular component the designer is able to
prevent most problems. Once the signal leaves this component however, it snakes its way
from device to device by wires several feet long and hooked up to unpredictable input
circuits.
The ideal way to connect two audio components would be to
make sure the output impedance of one exactly matched the input impedance of whatever it
feeds. This is standard practice in professional installations, but there is no such
consistency in consumer audio, so the risk of mismatches is high. These can cause gross
distortion, so home audio equipment usually avoids the problem by means of a technical
compromise: output stages generally have low internal impedance, while input impedances
are many times higher.
This arrangement is much less efficient in terms of power,
necessitating somewhat higher amounts of amplification along the way to achieve the same
output, but audio equipment is very tolerant of upward mismatches, so this technique gives
some assurance that a signal will pass through a whole system with its waveform intact.
The amount by which the input impedance must exceed the
impedance of the output stage is not critical, as long as the difference is great enough,
which is why we can plug virtually any piece of audio equipment into any other with
reasonable assurance that they will both perform optimally. The reduced gain at each stage
is a small price to pay for this level of compatibility.
Digital sound has injected some new factors into all this,
however. For one thing, many owners of compact disc players are unwilling even to
entertain even the thought that their cables might affect the pristine digital
signal, so the use of high-tech interconnects is increasing. In addition, many amplifier
manufacturers now include a bypass function that routes the signal from the compact disc
input directly to the power amplification stage, on the theory that the fewer electronic
components there are in the signal path, the less the chance for degradation. Things like
tone controls and input switching are not allowed to interfere with the signal from the CD
player, although users do have the option of routing the signal conventionally on those
occasions when the extra controls are required.
Such techniques may pay some sonic benefits, although the
audible improvements are likely to be subtle at best, but they would apply equally to any
signal source and don't really address potential problems that are unique to digital
audio. From the photoelectric cell that actually reads the data on a CD to the
digital-to-analog converter, the information is all digital; after conversion it is
analog. Some experts feel that housing both digital and analog circuits on the same
chassis can cause problems because the digital is rich in harmonics in the radio-frequency
part of the spectrum. Analog audio is prey to radio frequency interference from external
sources at the best of times; the situation is much worse if the radio frequencies are
generated within the same component as the analog.
One solution is to separate the digital and analog
functions, housing the former in the CD player and the latter in the amplifier. Most
high-end CD players now provide at least one form of digital output for this purpose, and
amplifiers and receivers with built-in digital-to-analog conversion exist as well. There
are also several freestanding digital-to-analog converters available that can be used with
any amplifier.
The advantage of this system is that the generators of
electromechanical noise, as well as most of the interference-producing radio-frequency
circuits, are in the player, with their own chassis and power supply. The digital signal
processed there is unaffected by the other sorts of noise, and the analog section is far
enough away (and separately shielded) that it remains clean as well.
There are drawbacks to all this, however. For one thing,
the majority of compact disc players that offer digital output also have built-in D/A
converters, so taking advantage of the system usually means paying for two converters.
The main problem with the separation of the digital and
analog portions has to do with the interconnection itself. A metal (coaxial) cable used to
transmit the bit stream from the player to the amplifier may be subject to the same
problems as ordinary analog patch cords. The digital signal will ignore things like hum
and interference, to be sure, but it might be degraded by cable capacitance.
One answer to this has been to make the connection
optically. By using the player's output to drive a fast-acting light source, and then
providing the amplifier with a photoelectric device to turn this back into electricity,
the player and the amplifier may be linked by a fibre-optic cable unaffected by any of the
problems associated with the coaxial method. Most of the top CD players on the market
offer optical outputs, and a handful of amplifiers have optical inputs.
All of this might well become irrelevant, however. The
exotic new digital media, with their reliance on severe data reduction schemes and
perceptual coding may make the simple matter of digital-to-analog conversion -- wherever
it may take place -- seem as quaint as the problem of getting a clean signal from your
turntable to your phono preamp.
...Ian G. Masters
ian@mastersonaudio.com
|
