Introduction
Q.
You don't make loudspeaker systems anymore?!?
Q.
What is the best material for loudspeaker cones?
Q.
What are the characteristics of various cone materials?
Q.
What is the best material for tweeter domes?
Q.
Why aren't the L/1 and L/2 floor-standing towers?
Q.
What is a "shielded" speaker and when do I need one?
Q:
Is an open-air/free-air/top-mount high frequency driver configuration
the best for all situations?
Q.
What is the relationship between Biro Technology and Audio by Van
Alsine, Inc.?
Introduction
One of the aims of Biro Technology as a company is
quite simply to make the whole of the audio world a better place. In
our opinion, a major obstacle keeping audio from moving forward and
better assuming its proper role in the universe -- as a servant to
truer and deeper musical experiences -- is the fact that it is treated
by a preponderance of the industry as essentially a marketing-oriented
activity rather than as an engineering or music-oriented one. What this
means is that the less you, the audio consumer, know about how things
really work, the easier it will be for the marketing people to go to
work on you and make you buy (literally) what they want you to. The
flip-side of this is that the more you know, the
more easily you'll be able to spot the exaggeration, hyperbole, or
outright lies in a marketing effort, and the better decisions you'll be
able to make. Ultimately, this means that the deserving products and
companies will get rewarded, and the ones who survive by exploiting
your ignorance, neuroses, and/or insecurities feel the pressure to get
real.
It was with this in mind that we decided to publish this FAQ. From the
outset we must apologize for being unable to break things down into
bite-sized factoids that you can easily tuck away in your brain. The
truth is, loudspeaker drivers and systems are deceptively complicated
things. And while we have tried to simplify things as much as possible,
our primary concern has been to not give so little information as to
create a bigger problem than the one we are trying to solve. Especially
true in audio, a little knowledge can be a dangerous thing.
[back]
Q.
You don't make loudspeakers systems anymore?!?
A. Sort of. Please see this.
[back]
Q.
What is the best material for loudspeaker cones?
A. (Short) There is no best material for
loudspeaker cones.
A. (Long) As anyone shopping for speakers is
undoubtedly aware, loudspeaker cones are made from a variety of
materials, each one being claimed to have some property or another that
makes it better than all the rest. Unfortunately, in spite of what the
ad copy writers would like you to believe, there is no single "best"
cone material for loudspeaker applications. Different cone materials
have different mechanical and acoustical properties that result in
various performance tradeoffs, making them better or worse suited for
various situations. However, almost always the choice of material
involves some kind of compromise. The basic material parameters that
affect the acoustic performance of a cone material are its density,
stiffness, and internal lossiness (i.e., the internal damping). Very
loosely speaking, the stiffer and lighter a cone material is, the wider
the bandwidth of the cone will be. The more lossy it is, the smoother
the response. Unfortunately, the above parameters are typically
interactive, and it is very difficult to optimize all three parameters
simultaneously. To find out why, we need to understand a little better
what happens in a speaker cone when it is making music.
At low and very low frequencies, a loudspeaker cone moves essentially
as a homogeneous unit, and there is only one cone parameter that has
significant impact on the performance of the driver: the total cone
mass (which is itself a function of the cone material's density and the
total amount of material used).[1] All other parameters being equal,
the greater the cone mass, the lower the frequency of the fundamental
resonance of the driver, the less damped the resonance, and the less
sensitive the driver will be. However, other driver component
parameters -- such as the suspension compliance and lossiness -- also
affect the resonance, damping, and sensitivity of the driver. All these
variables must be considered when performing a design analysis to get
you to the desired result. Fortunately, the mathematics describing the
low-frequency behavior of loudspeakers is not terribly complicated, and
so the modeling of loudspeaker drivers at low frequencies is a fairly
straightforward task.
At higher frequencies (where the wavelength of the sound wave becomes
comparable to the radius of the cone) the cone ceases to move as a
homogeneous unit, and our low frequency model breaks down. At these
frequencies, you'd do best to think of the sound wave as starting at
the base of the cone (at the voice-coil former/cone joint) and
propagating outward towards the edge of the cone. When the wave hits
the edge of the cone, it is reflected back toward the base of the cone
(towards the voice-coil former). When the wave hits the voice-coil
former, it is again reflected back towards the edge of the cone, and
the whole process starts again. This process is similar to sound waves
traveling in a room, hitting a wall, reflecting back, etc. In both
situations, significant standing waves result, and these standing waves
can produce really, really huge peaks and dips in the response of a
driver unless steps are taken to counteract it. (For example, the first
standing wave resonance for a good 6-1/2" driver typically falls in the
upper midrange, well within the range where it would be contributing
significantly to the output of a two-way system.)
Fortunately, most cone materials have a degree of lossiness in them --
meaning that they are imperfect sound conductors. A portion of the wave
energy travelling through the material is converted to heat, and the
wave is gradually attenuated as it travels down the cone. Such
lossiness reduces the intensity of the standing waves by reducing the
intensity of the reflected wave energy, thereby smoothing the response
of the driver. Different cone materials vary greatly in the amount of
internal damping they have, ranging from almost none (metal) to a lot
(some plastic materials).
Another means of controlling the intensity of standing waves in a cone
is the cone surround. Typically, a speaker cone is supported around its
edge by some kind of material -- usually a rubber-like elastomer or
foam, but sometimes cloth or even accordioned paper. One function of
this surround is to allow the cone to move back and forth with relative
ease at low frequencies while providing an air-tight seal. At higher
frequencies, it can be used to absorb some of the cone's standing wave
energy. As the wave travelling out from the base of the cone hits the
surround/cone interface, a portion of the wave energy is actually
transmitted into the surround material, with the remaining energy
immediately reflecting back into the cone. Depending on how lossy the
surround material is, the portion of the wave energy transmitted into
the surround may be converted into heat (effectively damping
resonances), or it may be bounced around inside the surround and then
back into the cone (creating a more complicated series of resonances).
Synthetic rubber-like surround materials are typically formulated to
have very high internal losses, although there are a few that are
surprisingly low. Foam surrounds are typically less lossy than "rubber"
ones, although I'm betting that someone, somewhere makes a foam that is
very lossy. In either case, the amount of loss in a surround (or a cone
for that matter) may or may not be constant with frequency.
The correct amount of damping in the cone and surround depends on the
demands of the situation. Generally speaking, you want enough loss in
the combined cone/surround system to produce a smooth and
well-controlled high frequency response, unless all the standing wave
resonances occur well outside the bandwidth under which the driver will
be used.[2]
The usable bandwidth of a cone is determined largely by the frequency
of the first standing wave, and the faster the wave travels through the
cone material, the higher in frequency this will occur. The primary
mechanical properties which determine the rate of sound propagation
through the material are its stiffness, density, and thickness: stiff,
light, thick cones producing faster rates of sound propagation than
limp, dense, thin ones. Unfortunately, the general tendency is that the
lighter and stiffer a material is (yielding wider bandwidths) the less
internal loss it has -- meaning that the less damped the standing waves
and rougher the frequency response will be. In addition, the more
dissimilar the cone material is from the surround, a stiff metal cone
with a loose and lossy elastomer surround, for example, the less wave
energy will be transmitted into the surround, and the less effective it
can be at damping the standing waves. These characteristics make it
very difficult to get wide bandwidth and smooth response simultaneously
from a cone. To further complicate matters is the very annoying
phenomenon that the lossier a surround material is, the less linear it
tends to become at high excursions -- as it might experience when the
driver is reproducing large low-frequency signals. So if on top of a
wide, smooth bandwidth you also want good low frequency performance,
you are stuck with a very complicated juggling act where no one can be
completely happy, but with luck nobody will be overly let down.
While the above is certainly not an exhaustive description of every
aspect of loudspeaker cone behavior, it does hit on some of the major
ones. Indeed, one could write several books on the subject. (And we
wish someone would!)
______
[1]
Having said that, we would be amiss if we did not mention that there
might be additional mechanical properties which influence the low
frequency behavior of cones in subtle, difficult to measure ways.
Unfortunately, opinions vary widely on this subject, and there is very
little research backing up any claims one way or the other.
[2] Another option is to
use active or passive equalization to compensate for the resonances.
This turns out to be a pretty bad idea since the actual frequencies of
the resonances move around a little depending on manufacturing
variances as well as ambient conditions. A compensation circuit that
works for one driver on one day may or may not work on a different
driver or on a different day.
[back]
Q.
What are the characteristics of various cone materials?
A. If you haven't done so, you should probably read "What is
the best material for loudspeaker cones?" before consuming the material
below.
Paper
Paper is the traditional material for speaker cones and is widely
(though mistakenly) considered to be an outdated technology
inappropriate to high performance audio applications. Among its virtues
are that it can easily be formed into a wide variety of shapes without
overly complex or expensive tooling and its mechanical properties can
be varied over a usefully wide range. Unfortunately, untreated paper is
very sensitive to environmental conditions -- humidity in particular.
As the ambient humidity changes, the moisture content in the paper also
changes, and this leads to changes in cone mass and other parameters.
Also, while it is possible to manufacture paper to be stiff enough to
get extended frequency response, the paper itself is usually
insufficiently lossy to achieve a smooth rolloff. Finally, paper is not
the easiest material to manufacture consistently, with the possible
result that there will be wide production variances.
The first two of the above shortcomings can be greatly alleviate by the
application of various types of surface treatments including latex and
PVA-based coatings and impregnations. These coatings help to isolate
the cone from ambient environmental conditions while increasing
transmission loss, thereby smoothing out the upper range of the driver.
(Note: many coated paper cones are coated largely for aesthetic
purposes using God-only-knows what kind of materials that do
God-only-knows what to the acoustical performance of the driver.) The
potential for wide production variances can be ameliorated by tightly
controlled production processes and quality material sourcing.
Despite the seeming low-techness involved, a well engineered paper cone
can deliver a combination of bandwidth and smoothness that is at least
as good as any "higher-tech" material. Additional research into new
paper formulations, manufacturing methods, and treatments continues.
Don't be surprised if some of the newest "breakthroughs" in cone
technology are based on lowly, old-fashioned paper.
Polypropylene
Polypropylene is probably the most common plastic material used in
speaker cones. Most cones advertised as being made of polypropylene are
in fact a combination of polypropylene and a mineral or other filler
(e.g., carbon fiber and Kevlar®). These fillers can be used
both to control costs and to alter the mechanical properties of the
material. Polypropylene cones tend to be inherently well damped with
the result that they can deliver smooth, if not terribly extended,
frequency responses. They are also largely immune to changes in ambient
humidity. The material itself and the methods used in manufacturing
cones with it are such that tight tolerances are easily achieved. In
fact, polypropylene is the material of choice for may researchers
involved in finite element analysis (FEA) of drive units because it is
easy to reliably characterize.
Polypropylene acquired something of a bad reputation in its early days
due to fact that it's a difficult material to get things to bond to.
Luckily, modern adhesive technology has completely solved this problem.
However, this is not to say that polypropylene is free of problems.
While a quantitative study has never been published (not to my
knowledge at any rate), there are some who feel that drivers made with
polypropylene cones tend to exhibit an audible degree of hysteresis or
hysteresis-like behavior. (Hysteresis is a kind of nonlinearity where
the parameters of a system which should be constant vary depending on
the system's recent history.) The most common thinking is that it is
the viscoelastic creep present in all plastic materials that is
responsible. (Viscoelastic creep refers to the tendency of plastic
materials to slowly stretch when under stress. This process may or may
not be linear and typically is related to the lossiness in the
material.) One colleague whom I respect greatly feels that the joint
between the voice—coil former and the cone may be to blame.
He suggests that the heat generated by the voice-coil and dissipated by
the former may soften either the plastic cone material or the glue at
the joint -- the amount of softening depending on how much power the
coil is dissipating.
Despite these actual or imagined problems, polypropylene cones remain a
popular choice for high performance systems largely because of their
well-behaved high-frequency response and consistent performance.
Other plastics
Apart from polypropylene, there are numerous other plastic and
plastic-based materials that have appeared over the years including
TPX, HD-A, and HD-I (all manufactured by Audax), Neoflex (manufactured
by Focal), and Bextrene (which polypropylene largely replaced). All
these represent attempts at finding combinations of stiffness,
lossiness, density, and sound velocity that are somehow optimal for a
given application. They generally have the same virtues and potential
pitfalls as polypropylene.
Resin-bonded high-strength woven fibers
In this class of materials belong most carbon-fiber, fiberglass and
Kevlar® cones. These cones are made from a fabric of fibers
bonded together with an epoxy or similar resin. The fibers themselves
have a high degree of tensile strength; when embedded in an appropriate
resin, a material of considerable stiffness results. Not surprisingly,
these woven cones tend to have extended bandwidths. However, it comes
with the cost of quite a bit of roughness as the internal losses of the
basic resin-bonded material are quite low. It has been suggested the
random orientation of the fibers helps to break up standing wave
patterns on the cone, thereby smoothing the response of the driver. In
our experience, this phenomenon has at best a minor influence on the
high frequency response of the driver as every woven-cone driver we've
examined has exhibited a rather rough high-frequency response.
Attempts have been made to improve on the basic construction of simple
woven fabric cones. One manufacturer of raw driver units employs two
thin layers of Kevlar® fabric bonded together with a resin and
silica microball combination. The laminated structure is purported to
be very stiff and the core material has the potential of introducing a
controllable amount of damping. Another driver manufacturer employs a
similar sandwich structure but with a honeycomb Nomex core. While these
technologies are very exciting, they tend to be extremely costly and
suffer, to greater or lesser extent, from the same high frequency
roughness as their simpler cousins.
It is highly unlikely that a woven fabric cone will have any hysteretic
properties. (Although the surround and spider -- even the motor system
-- may still suffer from hysteresis, but that's another issue.) So,
while they may not generally be the best choice for wide-range
applications, woven fabric cones are well suited to low-frequency
applications owing to their inherent stiffness and immunity to
hysteresis. In addition, woven fabric cones typically are insensitive
to environmental conditions and are not likely to be bothered very much
if exposed to a direct heat or light source (e.g., the sun). Thus they
may be well suited to a variety of mobile or even outdoor applications
as well.
Metal
Metal is seeing something of a surge in popularity as a cone material.
Of all the materials we have discussed so far, it is the least well
damped and so suffers from extreme peakiness in the high frequency
region -- peaks of 12 dB at 5 kHz for a 6-1/2" driver being not
uncommon. However, below their first breakup mode, metal cones tend to
be very well behaved, and this is a major source of the attraction to
metal cones.
The most common materials used in metal cones are aluminum (and its
alloys) and magnesium. Given the broad range of forming and surface
treatment options possible with these materials, it is not
inconceivable that we may one day see the advent of a well-controlled
metal cone driver. However, even with the best crossover design, the
high-frequency peaks present in currently available cones make them a
poor choice for wide-range applications.
Everything else
Driver manufactures are constantly experimenting with new permutations
of basic materials and constructions in an attempt to find (at best) a
better compromise for a given application, or (at worst) a product that
merely has greater market appeal. Laminates of all sorts,
Kevlar® and paper composites, and Kevlar® and plastic
composites are but a few of the materials that have recently been made
available. As with any new technology, all claims made for or against
such new materials must be considered very, very carefully.
The Bottom Line
I hope that by now it is clear that the "best" cone material to use for
high performance audio depends on what you need to do with it and that
at best it will only be some kind of compromise. It is also important
to bear in mind that a loudspeaker driver is much, much more than the
material from which its cone is made. The profile of the cone and
distribution of material, the properties of the surround and spider at
various frequencies, the voice coil geometry and materials, the
magnetic structure, etc. all play a large role in the final performance
of the driver. What all this means, dear reader, is that you simply
cannot judge a driver by its cone material.
[back]
Q.
What is the best material for tweeter domes?
A. (Short) There is no best material for tweeter domes.
A. (Long) Coming sooner or later. But many of
the kinds of issues that are true for woofer cones (above) apply here
as well.
[back]
Q.
Why aren't the L/1 and L/2 floor-standing towers?
A. One of the recent trends in loudspeaker marketing has been
the move toward floor-standing systems that have footprints similar to
more conventional stand-mounted designs. This is chiefly due to a
perception in the market that larger boxes necessarily deliver "better
bass". By simply packaging a system in a floor-standing format the
seller can easily increase the "perceived value" of a product -- no
matter what the technical consequences.
Unfortunately, the premise that
bigger boxes necessarily deliver better bass is by no means universally
or generally true. In reality, for a given alignment (e.g.,
second-order Butterworth, fourth order Bessel, etc.), the size of the
box merely fixes an efficiency-bandwidth figure. The exact relationship
is:
n0
= k * f3 * Vb
where n0 is the system efficiency, k is a constant
determined by the system's alignment, f3 is the system's low-frequency
3dB cutoff, and Vb is the size of the box. This equation tells us the
following:
* If we increase the box volume
and keep the low-frequency cutoff the same, the system's efficiency
will increase.
* If we increase the box volume, we can decrease the low-frequency
cutoff and maintain the same efficiency.
* If we decrease the box volume, we can keep the same low-frequency
cutoff by reducing the system efficiency. (In all cases we've kept k
the same so that the bass quality is unchanged.)
This last result -- which
probably comes as a surprise to most -- shows that a small box is
capable of the exact same extension and quality as any large box as
long as you're willing to pay for it with efficiency [1]. Furthermore,
this cost is not astronomical: halving a box's volume while keeping the
extension and bass quality constant will reduce the system efficiency
by about 3dB.
But perhaps more important than
the size vs. extension vs. efficiency issue is the annoying fact that
there are two significant factors conspiring against a system
effectively taking advantage of the additional volume a floor-standing
configuration has to offer. The first (and possibly most obvious)
factor is cost. A large, well-braced floor-standing cabinet is
appreciably more expensive to build, handle, and transport than a
comparable stand-mounted one. This extra cost must be passed on to the
customer as either an increase in the selling price of the finished
system or a reduction in the quality of the drivers, crossover
components, or even the cabinet itself. Second, a floor-standing
cabinet of the format suggested (e.g., 90cm (36") high with a footprint
matching our L/1's 26cm (10.25") by 29cm (11.5")) when left to its own
devices will possess a very strong 1/4-wave column resonance around 100
Hz. This resonance will manifest itself as a spectacularly unpleasant
coloration and will thoroughly perturb the system's low-frequency
alignment.
There is very little you can do
about the cost issue. But if you're still really committed to the
floor-standing concept, there are a few things you can do to mitigate
against the column resonance. Unfortunately, all of these solutions
drive up the cost even further, and most of them eliminate or
drastically reduce any advantage the additional box volume may have
contributed in the first place.
For example:
You can resistively damp the
airflow inside the cabinet. While this approach can eliminate the
column resonance it will also largely nullify the benefits that the
larger volume may have given you vis-a-vis low
frequency extension or system efficiency.
Or...
You can incorporate an angled
surface to break up the column. To be effective, the angled surface
must be comparable in size to the sides of the cabinet. (A simple
angled top is not sufficient!) Such a construction, whether it happens
as an additional panel inside the cabinet or as an integral part of the
cabinet exterior, is a very expensive proposition, and it robs you of a
good chunk of the extra volume.
Or...
You can block off the lower part
of the cabinet with an additional panel, but now you've lost all the
extra volume you're trying to gain.
Or...
You can divide the cabinet into
top and bottom halves and use two drivers, one in each section. Not
only does this approach increase the crossover complexity, it also
means your low-frequency driver budget now needs to buy two inexpensive
units rather than one really good one. It can, however, have the
advantage of increasing the radiating area of the system at low
frequencies.
Or...
You can increase the width and
height of the cabinet to distribute the resonant modes and then deal
with those distributed resonances with a moderate amount of resistive
damping. The cabinets of our L/1 and L/2 are currently narrow enough
that we can very precisely compensate for the acoustical influence of
the cabinet baffle on the frequency response of the system. With a
wider baffle, this would no longer be possible -- at least not without
dramatically increasing the crossover network's complexity and cost.
Of course, one of the hidden
costs in a stand-mounted design is the price of the stand itself.
Suffice it to say that you can spend anywhere from $50 on a stand that
gets the job done effectively and not too unattractively to as much as
you'd ever care to spend for a pair of custom sculpted marble pedestals
that would make even Brancusi envious. The important thing to keep in
mind is that you, the end customer, are in control of how much to spend
on utility and how much on aesthetics.
To conclude, the primary benefit
of a stand-mounted cabinet over a floor-standing cabinet of equal
footprint is that it permits the use of construction methods, drivers,
and crossover components of an appreciably higher quality for the same
end-cost. Nature being what it is, there is a downside associated with
using the smaller cabinet and that is that if you want the same bass
extension and quality as the (marginally) larger floor-standing box,
you need to sacrifice a little system efficiency. However, in this day
of relatively inexpensive amplifier watts, we don't consider this a
significant compromise. In fact, we barely consider this a compromise
at all. In our 3200 cubic foot auditioning room we are consistently
able to drive our L/1 and L/2 to very satisfying listening levels using
a high quality, yet modestly powered 85 wpc solid-state amp and an
equally high quality 35 wpc tube amp.
______
[1]There is a practical
limit on the minimum size of the box that is determined by the maximum
required sound level and the nonlinear compressibility of air. The air
inside the box cannot be compressed more than a few percent before its
nonlinearity becomes noticeable. A system reproducing low frequencies
will typically reach this point when the cabinet approaches 3 liters
(0.1 cubic feet). As a frame of reference, our L/1 has an internal
volume of just under 19 liters (0.67 cubic feet).
[back]
Q.
What is a "shielded" speaker and when do I need one?
A. Read this.
[back]
Q:
Is an open-air/free-air/top-mount high frequency driver configuration
the best for all situations?
A. (Short) No.
A. (Long) An open-air mounting technique as
used in our own L/1 loudspeaker and by others when
implemented competently can
simultaneously solve three loudspeaker design problems. The first
problem involves cabinet edge diffraction and the effect it has on
high-frequency system behavior. In short, a conventional system's
cabinet edges introduce radiation discontinuities, and these
discontinuities introduce time and frequency response aberrations. One
way to reduce, but not eliminate, these aberrations is to place the
high-frequency driver on top of the box. This isssue increasingly
becoming less of an issue at Biro Technology as our Diffraction
Optimized Positioning driver placement technology matures. Diffraction
Optimized Positioning allows us to create conventional systems whose
diffraction behavior is essentially as good as an open-air mounting
configuration.
The second issue involves the placement in time of the low and high
frequency drivers. High-frequency drivers are physically shallower than
low-frequency drivers. When the both drivers are mounted on a common
plane (as in a conventional system), the low-frequency driver's
physical and acoustical center will be behind the high-frequency
driver. This introduces a small time delay between the drivers. With an
open-air configuration, the high-frequency driver can be physically set
back, thereby bringing the acoustical centers of the two drivers back
into perfect synchronization.
The third issue follows directly from the above. When the acoustical
centers of two drivers are identical, it becomes possible to use
"textbook" crossover configurations that are optimized for smoothest
off-axis response and high out-of-band rejection. Our crossover target
function optimization technology lets us develop system-specific
"non-textbook" crossover configurations for conventional systems whose
performance is very close to that of the ideal crossover configurations
attainable with open-air configurations. However, the open-air
configurations still has a slight advantage in this respect.
The above three benefits do not come without cost. In particular,
open-air configurations have a frequency-dependent radiation pattern
that may present problems in some situations. In both open-air and
conventional systems, the low frequency driver radiates in a
forward-facing hemispherical pattern in the midrange, with the
radiation narrowing as the frequency increases. The high-frequency
driver's radiation in conventional systems is essentially
forward-facing hemispherical up to very high frequencies, making it a
reasonably close match to the low-frequency driver's radiation pattern
throughout the crossover region. In contrast, in an open-air system the
high-frequency driver's radiation pattern is mostly hemispherical but upward-facing
from below the crossover frequency up to very high frequencies.
If your listening room is reasonably absorbent, the differences in
radiation patters of the two drivers in an open-air system will be of
no consequence. However, if your listening environment is particularly
reflective, the differences in radiation patterns may lead to
subjectively greater brightness, potential hardness, and possibly
compromised imaging.
Therefore, the overall conclusion is that open-air mounted
configurations when implemented competently will be
of greatest advantage in listening environments that are relatively
absorbent. However, in acoustically reflective environments their
frequency dependent radiation pattern may introduce perceived response
and imaging artifacts that overshadow the advantages that this
configuration facilitates.
[back]
Q.
What is the relationship between Biro
Technology and Audio by Van Alsine, Inc.?
A. Biro Technology and Audio by Van Alstine, Inc. are two
inpedendent companies that have shared a close working relationship.
Audio by Van Alstine have assembled and sold Biro Technology products
under an exclusive licence for several years. At the same time, Biro
Technology founder Mithat Konar has provided technical design services
to AVA, being responsible for much of the innovation implementation at
AVA from 2001 to 2005.
[back]
