Speaker Basics -- Part One
Pundits might dispute the point, but a major advance in
audio quality was the ditching of the notion that anybody could build a
loudspeaker. Most of us, if we're old enough, put together at least one pair of speakers
in our audio adolescence, usually from rough diagrams in hobbyist publications, and there
was a thriving business in garage-built devices using wine casks or 19th-century commodes
as enclosures.
It is true that most speakers are fairly simple devices
technically, and relatively easy to make. But the simplicity is deceiving: How a speaker
sounds is a result of a delicate balance between its physical and electrical elements, a
balance that is almost never achieved when a speaker is just thrown together. Even with
professionally designed models, tonal differences tend to be much greater than with other
components. But the sonic character of a speaker determines the character of the whole
audio system, so it is the most important link in the chain
The heart of a speaker is a form of electric motor that
converts electricity into physical motion. Unlike conventional motors, which are
rotational, a speaker employs a linear motor; it moves back and forth in step with a
varying electrical current. In most speakers, the active element is a coil of wire -- a voice
coil -- suspended in a strong, permanent magnetic field. As an amplifier feeds
electricity to this coil, a magnetic field is created around it, which changes in polarity
and intensity as the amplified signal changes. This varying field reacts with the
permanent field, causing the voice coil to move in and out in a motional replica of the
audio signal produced by the amplifier. The greater the current, the farther the coil will
move. Speakers using this technique -- by far the majority -- are called dynamic
speakers.
The purpose of a speaker is to create air-pressure
differences -- sounds -- that are as close as possible to the original recorded material.
To accomplish this, the voice coil is attached to a diaphragm, which compresses and
rarefies the air in front of it as the voice coil moves it forward and backward. In many
cases, the diaphragm is a shallow cone made of fibrous material, stiff paper,
plastic, or even metal. Most low-frequency speakers and virtually all full-range speakers
use cones, as do some high-frequency units. For treble speakers, however, it is more
common for the diaphragm to be in the shape of a convex dome. A few dynamic
speakers employ flat panels, driven by a single coil or by several.
An alternative design, the electrostatic speaker,
also uses a flat panel, but the diaphragm is not attached to the usual coil-and-magnet
structure. Instead, the amplifier feeds power to a thin metallic coating on the rear of
the diaphragm, causing variations in static charge. Parallel to the diaphragm is a plate
carrying a fixed charge; the interaction of the two electrostatic charges causes the
diaphragm to move. Because the plates must be very close together, the diaphragm can't
move very far, so it must be large in order to move enough air to provide adequate
listening levels.
The demands placed on a speaker are different at different
frequencies. While it is possible to make a speaker that will handle all frequencies with
reasonable ease, it is now usually considered preferable to divide the audio spectrum into
sections, each one reproduced by its own speaker, or driver. Such devices, along
with their enclosures and associated electronics, if any, are really speaker systems.
In such systems, a small driver called a tweeter reproduces
the highest frequencies, while a much larger woofer reproduces the lowest.
Tweeters typically have diameters of an inch or less, as this widens the angle over which
the sound radiates; woofers are much larger because they have to move a much greater
volume of air. If only the two drivers are employed, the speaker is known as a two-way
system. In many cases, particularly as you ascend the price scale, a third midrange
driver handles the frequencies between the other two, in a three-way configuration.
To direct the appropriate signal to the various drivers, a crossover
network is employed. In most cases, this is built into the speaker enclosure, simply
routing the signal from the amplifier to each driver; in this case it is a passive
crossover. In some more ambitious speaker designs, however, the division of the signal
takes place before the amplifier, a process known as biamplification or triamplification,
depending on the number of drivers concerned. For this, an external electronic or active
crossover is used. However the division of the signal is achieved, the drivers and
crossover must be carefully matched to create a smooth transition from one part of the
spectrum to another.
On its own, every speaker is a dipole: It radiates
as much energy to the rear as to the front. The two waveforms are opposite in polarity, or
out of phase with respect to each other, and if allowed to intermingle would
interfere with each other, canceling out some frequencies. At high frequencies this is
rarely a problem, as the signals are relatively directional and thus radiate away from
each other. Also, the higher frequencies have little energy and short wavelengths, and so
are easily contained or absorbed. In the bass, however, things are very different: Low
frequencies are virtually omnidirectional -- they radiate equally in all directions. They
also have very long wavelengths (almost ten meters -- more than 30 feet -- at 20Hz). Thus,
the front and back waves a woofer produces will cancel each other unless they can be kept
apart.
One solution is to use a baffle: a large board with
an opening into which the speaker is tightly fitted. The rear wave must travel around the
ends of the baffle before it can meet the front wave; by the time it has done so, its
level will have dropped enough that the interference is minimal. The most effective is an infinite
baffle, which prevents the rear and front waves from ever meeting. Mounting a speaker
in a wall is one way to accomplish this, although not always the most practical.
Most modern speaker systems deal with the rear wave by
mounting the woofer in some sort of enclosure (the other drivers are usually
mounted there as well, but the enclosure mainly affects low-frequency performance). One
common sort is a type of infinite baffle called an acoustic-suspension enclosure.
This is a sealed box that prevents the rear wave from radiating into the listening room.
The air pressure created inside the enclosure is used to spring-load a loosely suspended
woofer.
Equally popular is a type of enclosure that does allow the
rear wave to radiate, but modifies it so that it enters the listening room in phase with
the front wave, thus reinforcing it. This is known variously as a ducted, vented,
or ported enclosure, after the opening through which the rear wave emerges; more
often, it is called a bass-reflex design. A variation on this employs a passive
radiator instead of an open port. This contains an unpowered speaker cone. The
air-pressure differences within the enclosure determine its movement. In a bass-reflex
speaker, the mass of the air in the port tends to smooth the bass response; the physical
mass of a passive radiator does the same thing.
Although less common in consumer audio than in professional
applications, the horn speaker offers probably the greatest acoustic return for a
given amount of amplifier power. Other speaker types tend to convert relatively little of
their energy to air motion. By placing a driver at the narrow end of a flared horn,
however, its coupling to the air is improved and much more acoustic output is
possible. In practice, this means that small drivers, which need little power, can be used
to produce a lot of sound. The main drawback, at least when it comes to low frequencies,
is that horns have to be large to work effectively.
In operation, a speaker and the amplifier driving it form a
single circuit, so the electrical characteristics of the speaker dictate the overall
performance of the circuit. In essence, the amplifier is missing one resistor, which is
replaced in the circuit by the speaker itself, along with its crossover network and the
cables used to connect everything together.
The "value" of this composite resistor is its impedance,
like resistance measured in ohms. Impedance is unlike resistance, however, because it
varies with frequency. A speaker may have a nominal impedance of, say, 8 ohms, but
this is only an average -- actual impedance may be much lower at some frequencies. As with
any circuit, an amplifier is designed to work ideally with a particular value, or a fairly
narrow range of such values. If the impedance is too low, the current drain on the
amplifier will rise to a potentially damaging level; if it's too high, the speaker will
produce less output than it might otherwise.
Speakers vary widely in the amount of acoustic output they
can produce for a particular level of amplifier power, an attribute called sensitivity.
Sensitivity is often measured by feeding the equivalent of 1W of power from the amplifier,
then measuring the acoustic output at a distance of 1m in front of the speaker. The result
is stated in decibels of sound pressure level (dB SPL). A relatively
insensitive speaker might have an output of 85dB SPL or less, while anything in the 90s is
considered quite sensitive, and a few speakers can put out more than 100dB SPL.
Neither impedance nor sensitivity affect how a speaker will
sound, but they are important when it comes to matching speakers and amplifiers. By the
same token, a speaker's power-handling capacity -- the maximum number of watts it
can handle continuously without damage -- should be important. Speaker companies have
different ways of stating this, however, so the numbers are often less useful than they
might be.
While a good speaker will perform well all by itself, it
also has to function well as one of a pair, because a majority of us listen to our music
in two-channel stereo. One of the most desirable features is stereo imaging: the
ability to create the illusion of a firm center image and to place individual instruments
in their proper positions in the "soundstage." A number of things contribute to
imaging, not least being the nature of the listening room and the positions of the
speakers and the listeners. Also important is proper wiring between the amplifier and the
speakers: The speakers must be in phase -- the diaphragms must move in and out
simultaneously when fed a center-image or mono signal. Out-of-phase speakers will
prevent proper imaging, and usually degrade low-frequency performance as well.
The speaker itself can enhance imaging if its design allows
it to simulate a point source. In this case, the acoustic origin of all frequencies
is both physically close together and primarily aimed at the listener. A few advanced
designs enhance imaging ability further by electronic means, while others take a different
tack and deliberately reduce imaging in favor of widening the soundstage, usually by means
of reflecting large portions of the signal off nearby walls. Increasingly, imaging -- at
least with signals that really are equal in the left and right channels -- is enhanced by
means of a center-channel speaker, almost always associated with surround-sound
systems. Often these employ a number of small satellite speakers to achieve
proper directionality, and one or more subwoofers to handle the more-or-less
omnidirectional low frequencies.
For the most part, today's best speakers come very close to
producing an accurate acoustic replica of the amplifiers' electrical signal. Once the
sound is radiated into a listening room, however, all bets are off. A speaker is only one
half of an acoustical system; the other is the room itself, and it can have a
profound effect on the sound you hear.
More on that aspect of sound reproduction next month.
...Ian G. Masters
ian@mastersonaudio.com
|
