FOCAL Sopra N° 2 Carrara White Sopra_WhitePaper.pdf

SOPRA

SOPRA

OUR CRUSADE ...

Since his founding of Focal,
Jacques Mahul has launched
ambitious projects just to push
the limits of the loudspeaker:
the Audiom 15A in 1981 ; the
TC120 FC for magnetic power;
the first K2 sandwich cones in
1986 creating a low-mass, high­stiffness diaphragm. Conceived
originally as prototypes these
projects can sometimes seem
approximate but have given birth
to numerous products that have
made Focal a leader, envied in
the field of loudspeaker science.
The basic aim is very simple:
creating the maximum magnetic
force driving minimum moving
mass for an expressive sound.
This has also not changed in
35 years and this consistency
of purpose has allowed us to
make very significant progress:

multi-ferrites/Power Flower,
EM and IAL in the first respect,
and W, Flax and Beryllium
diaphragms in the second. So
many milestones that have built
the reputation of the Brand.

But this quest for the Holy
Grail in the field of the sound
transducer is not always kept
pace with the expectations
of markets always wanting
more convenience and
miniaturisation.

Also, in parallel, to provide the
means to continue ‘crusade’ we
had to be pragmatic by:

• Passing our technology down

to more affordable products (the
JMlab series in the 80’ is a good
example);

• Applying our expertise to

other areas such as car audio
and pro audio, where we have
been active for 25 and 10 years
respectively;

• More recently, expanding

our expertise by exploring
new promising fields such as
headphones and sound bar.

This has been in order to
maintain a company of critical
mass, able still to explore the
limits, and so continue to exist
and grow in an industry that has
seen many changes over the
last 30 years. So has Focal built
and consolidated.

OUR TECHNOLOGICAL
ADVANCES IN 2014

MATERIALS AND DIAPHRAGMS

The work of recent years
that led to the development
of our Flax sandwich clearly
demonstrated that our W
sandwich W (with the mastery
that we have for almost 20
years as well as its many
improvements) remains
unsurpassed for making cone
woofers or midrange drivers.
For tweeters, beryllium likewise
remains the ultimate material,
combining all the best qualities
(lightness, stiffness, damping).
The only way forward was to
work more closely on the driver
suspension. We had already
begun this work on the woofer­midrange of the Diablo Utopia
by developing a new spider
to "soft clip" mechanically,
and by improving the join
to the cone surround when
developing Utopia III in 2008.
Since then, the simulation
tools at our disposal have

improved, allowing us to
analyze the nonlinearities
that cause distortion, mainly
in the midrange, where the
1-1.5kHz area remains very
critical. Whereas in Utopia III,
with improvements wrought
by the third-generation W
sandwich and laser cutting,
we could hand-build the drive
unit to limit these distortions,
this is not possible in our
premium ranges for obvious
cost reasons.

The problem is well known:
there is a sudden change in
mechanical impedance where
the wave passes from the cone
to the surround, which itself
radiates sound and has a
"boomerang" effect, causing
deformation of the cone.
Many attempts to resolve
this have been considered
by our competitors but none

is satisfactory, since all are
detrimental to cone travel and
cause dynamic compression. A
committed research project of
more than two years duration
has allowed us to develop
a remarkable solution by
addressing the problem at
source, without limitation of
cone movement, using a "TMD"
suspension for which a patent
was filed in the third quarter
of 2014.

TMD SUSPENSION OR

OUR MIDRANGE OBSESSION...

In continuation of the work done
for the woofer-midrange driver
of the Diablo Utopia, and with
hindsight and experience gained
through the manufacture
of thousands of drive units
revealing the high criticality of
the cone-suspension assembly,
we realized that the multitude
of parameters involved made
it illusory to suppose we could
develop a solution empirically.
We had instead to develop a
computer simulation model
that represents this complex
mechanical connection well
enough, first to correlate the
results observed from many
prototypes, and second to
develop a non-empirical,
reliable solution. Today,
computing power is enabling

the finite element method to
calculate the dynamic response
of complex objects and thus to
predict the behaviour of virtual
prototypes. And consequently
to devise accurate and relevant
solutions.

Standard measures such as
frequency response, distortion,
even laser interferometry,
highlight the problem but as a
snapshot, a still photograph.
What we need is a ‘movie’ to
reveal the motion, understand
the phenomenon as a whole
and thus devise effective
solutions. ‘Patch’ correction
of static defects linearizing
the response curve can
be devastating to dynamic
behaviour. The midrange is

undoubtedly the most complex
area and the most demanding
musically; it is a key element of
the sound signature of Focal,
and spoiling its resolution is
simply not an option!

With a numerical model that
correlated closely with classical
acoustic measurements on
numerous prototypes (over a
hundred have been tested!), we
could at last consider a remedy.
The computer model is used to
accurately assess the addition
of mass or stiffness to the
suspension, the added masses
acting as a dynamic vibration
absorber. The technique, which
is well known, is termed a
"tuned mass" or "harmonic"
damper.

60

with absorber, optimized damping

with absorber, small damping

with absorber, no damping

without absorber

m

2

m

1

c

2

c

1

k

2

k

1

50

40

30

20

10

Transmissibility T (dB)

0

-10

Frequency Hz

120100806040200

It was used with great success
in the suspension of the
Renault R25 F1 car in 2005
– and was quickly banned
because it was considered
anti-competitive by the FIA! It is
also the basic principle of anti­seismic systems for modern
skyscrapers. (fig A)

This principle applied to a drive
unit surround consists of two
small circular rings, acting
as additional masses, which
oscillate in opposition to the

m

2

resonance frequency of the
surround. (fig B)

c

2

This device offers major
advantages. We are able to:

m

1

• Use an exponential cone

without absorber

with absorber, no damping

with absorber, small damping

with absorber, optimized damping

k

2

profile that extends the

k

1

c

1

bandwidth to more than

Thus we can combine benefits
which were previously
irreconcilable: low mass,
optimum damping and
extension of the frequency
response. (fig E)

This leads to several benefits
at the listening level: improved
transient response coupled
with a flat frequency response
and reduced distortion, of the
order of more than 50 per cent
in an area where the ear is
highly sensitive around 2kHz.
This results in more accurate
timbre, improved definition and
better stereo imaging. To clarify
the last point, the resonance of
a conventional surround blurs
the soundstage, especially
when the resonance is marked.
TMD suspension eliminates the

problem at source. (fig F)
5kHz and thus achieve better
transient response;

Fig A: The principle of the tuned harmonic damper (TMD) is shown in the graph,
above. A system comprising a single mass m1 and spring k1 has a marked
resonance (red trace). Addition of a second mass-spring system, m2/k2 as
shown in the diagram, results in two resonance peaks (pink trace in the graph).
If the anti-resonance dip is aligned with the resonance of m1/k1 we obtain the
green curve and finally, with careful application of damping, the black curve –
the resonance has virtually disappeared! Photo down shows the TMD (m2 = 660
tonnes) used in the skyscraper Taipei 101.

• Choose a very light surround,

eliminating the resonance that
is even stronger when the mass
is low;

• Damp resonance in the

direction of the sound radiation
(the radial plane) whereas
competitors’ devices damp
circumferential resonances of
the suspension. (fig C et D)

Fig B: Two moulded circular beads in the surround form our tuned harmonic damper. While
this appears to be a simple solution, over a hundred different configurations were tested
to optimize the result.

Fig E : Les meilleurs paramètres enfin conciliables : très
faible masse de l’équipage mobile pour haute définition,
profil exponentiel à réponse en fréquence et amortissement
optimal pour grande linéarité et faible distorsion.

Fig C: Effect of the harmonic damper on linearizing the frequency
response between 1.5 and 2kHz (blue trace with TMD, red trace
without, all other parameters the same).

Fig E: Optimum parameters combined: very low moving
mass for high definition, exponential profile for extended
frequency response, and optimum surround damping for
linear response and low distortion.

Fig D: Effect of the harmonic damper on nonlinear distortion, which
is halved between 1.5 and 2kHz (blue trace with TMD, red trace
without, all other parameters the same).

Fig F: Frequency response of our latest midrange driver (blue trace) compared
to the previous-generation W-cone midrange, representing prior state-of-the­art performance (red trace). Note the improved response linearity between
1 and 2kHz and the frequency extension provided by the exponential cone
profile. Improvements to the magnetic circuit also contribute (see next section).
NB: The dip at 3kHz in the blue trace is due to the tested driver not being fitted
with a dust cap.

REFINING THE MAGNETIC CIRCUITS

In recent decades our work
on the magnetic circuit
primarily comprised optimizing
the magnetic field strength
and therefore the force factor,
a key criterion in terms
of acceleration and thus
expressive musical rendering.

Also we have:

• Developed motors equipped

with advanced magnetic
materials for the Utopia III and
the Be tweeter, and field coil
woofers with our EM
technology;

• Reduced magnetic losses

by optimizing the geometry
of the pole piece, sizing of the
plate-core (T-Yoke) and flatness
of field plates, as in the multi­ferrite Power Flower midrange

unit. Further research work on a
concept speaker with its voice­coil around the circumference
of a flat diaphragm (US Patent
2013/0064413) was very
informative.

To improve performance
further is an extremely
complex problem because this
is a dynamic electromagnetic
system with multiple
interrelated factors. As the coil
moves in the gap it modulates
the magnetic field (Lenz’s law),
in different ways depending on
its position. And as the voice
coil current varies depending
on the music signal, it induces
Eddy currents in the magnetic
circuit which have the effect
of slowing the movement of
the coil.

Both these phenomena vary

according to a third factor,

which is the frequency. Ideally

we would like to eliminate

these effects but they are very

complex and only sophisticated

numerical modelling could

help us progress. This work,

undertaken over three years

of research, has delivered

important improvements,

similar in significance to that

achieved mechanically with

the "Gamma structure" used

in our cabinets to provide an

inert, neutral foundation for

the drive units.

The need for a simulation tool

Many attempts have been
made in the last 70 years
to avoid the effect of Eddy
currents. Use of a Faraday
ring of copper or aluminium
is common but unsatisfactory
since it embodies no overall
vision of the phenomenon. It
was not until the late ’90s with
the development of the Klippel
Analyzer that we finally had a
tool to reveal their dynamic
behaviour. Moreover, these
devices can have a detrimental
impact on the field strength and
therefore the force factor (BL),
which for Focal is not an option.

Three configurations of Faraday
ring are commonly encountered
(fig G); only the one disposed
within the motor near the
magnet does not affect the
force factor.

Type 1

Type 3

Type 2

Fig G: Three types of Faraday ring that can be fitted to limit the effect of Eddy currents. As
types 1 and 2 have the undesirable side-effect of reducing the force factor (BL), we used
type 3.

The difficulty resides in
the sizing of this ring. Our
tests have shown that low
frequency performance is very
sensitive to the thickness of
the ring, whereas turn-to-turn
distance in the voice coil is very
important at high frequencies.
With our simulation tool, we
were able to work effectively to
optimize both our woofers and
our midrange drivers.

The first step in our research

was to focus on evaluating

the change in inductance

of the voice coil as it moves

within the magnet gap, as

a function of frequency.

Ideally the inductance should

remain constant but does not.

Mathematical tools developed

in-house with reference to

measurements made on

numerous physical prototypes

have led to the development of

simulation software that finally

allows us to work effectively.

(fig H)

Fig G : La puissance de notre outil de simulation nous apporte une vision nouvelle et surtout nous autorise des optimisation des circuits
magnétiques inenvisageable auparavant. En exemple, cette modélisation montrant la variation de l’inductance de la bobine en fonction
de sa position dans l’entrefer pour 5 fréquences différentes.

Fig H: The power of our simulation tool brings new insights and above allows us optimize magnetic circuits that were previously unthinkable.
In this example we model the variation in voice coil inductance with position in the magnet gap at five different frequencies without Faraday
ring (left) and with ring (right)

New woofer magnetic circuit

Fig I: Ideal positioning of the Faraday ring for a woofer motor obtained

through our simulation tool.

The best results were obtained
with a concentric Faraday ring
placed at the base of the pole
piece, 3mm to 20mm high,
having no direct contact with
either the pole piece or the
magnet. Measurements made
using the Klippel Analyzer
reveal the precision of our
mathematical model and its
power to optimize a magnetic
circuit. What might seem
irreconcilable is now possible:

to have inductance that is
independent of the position
of the voice coil, the current
in the voice coil, and signal
frequency. Distortion (harmonic
and intermodulation) is reduced
by 70 per cent! (fig J)

Because this magnetic circuit

lacks any ferromagnetic

element, it is notable for being

insensitive to Eddy currents

induced by the current flowing

through the voice coil, which

changes in relation to the

music signal. It has behaviour

New midrange driver magnetic
circuit

close to that of air, resulting in

a magnetic field that remains

stable up to more than 5kHz

Our historical obsession with
the midrange encouraged

– ideal for a midrange driver.

(fig K)

us to go further, as this is an
area of the spectrum where
the ear is extremely sensitive.
As mentioned previously, in
2009 we developed a concept
loudspeaker that uses no
pole piece (US Patent 2013/

0064413) where the voice coil is
immersed in the direct field of
an annular neodymium magnet.

Fig K: Magnetic circuit without a pole piece. Focal

patent US 2013/0064413

Fig J: Measurements made using a Klippel Analyzer on an

8-inch woofer without a Faraday ring (red trace) and with a
Faraday ring optimized through our simulation tool (blue trace).

Left: Inductance variation as a function of the current through
the voice coil, which is signal-dependent. Total stability.

Right: Inductance variation as a function of the position of
the voice coil in the magnet gap. The improvement here is
spectacular, especially as the coil moves into the driver.

The main reason that this
concept has not yet found
application (in addition to the
manufacturing issues) lies in
the amount of Neodymium
needed to achieve high
sensitivity in a 6-inch midrange
driver. Nevertheless it was a
useful reference we have
named NIC (Neutral Inductance
Concept) when envisaging

the magnetic circuit of an
economically viable midrange
unit.

Our simulation tool was very
helpful in developing an optimal
structure with performance
closer to our NIC reference.
A central Neodymium magnet
is topped with a ferromagnetic
pole piece brought to saturation

(> 1.5T) by a second Neodymium

pellet above it (fig L). The field

is looped by a ferromagnetic

circuit dimensioned to avoid

saturation. Finally a Faraday

ring is carefully positioned

further to reduce distortion

below 1kHz. (fig M)

Fig L: Our simulation tool has led us to this optimized «NIC» magnetic circuit for the

midrange driver. It reconciles efficiency with linearity of the voice coil inductance as
a function of both the position of the coil and the current flowing through it, across the
frequency range.

Inductance over current L (X=0, I)

0,50

0,45

0,40

0,35

0,30

0,25

L [mH]

0,20

0,15

0,10

0,05

0,00

New design Electra Midrange

-3 -2 -1 0 1 2 3
I [A]

KLIPPEL

New design Electra Midrange

1,0

0,9

0,8

0,7

0,6

0,5

L [mH]

0,4

0,3

0,2

0,1

0,0

-4 -2 0 2 4

<< Coil in X [mm] Coil out >>

KLIPPEL

Fig M: Measurements made using a Klippel Analyzer on a

6-inch midrange driver with the new "NIC" magnetic circuit
(blue trace) and a 6-inch midrange unit with a conventional
ferrite motor (red trace).

Left: Inductance variation as a function of the current through
the voice coil, which is signal-dependent. Total stability with
"NIC".

Right: Inductance variation as a function of the position of
the voice coil in the magnet gap. The “NIC” improvement is
spectacular.

Two large plusses ...

We have now applied two major
developments to our upmarket
midrange driver: our harmonic
damper suspension "TMD"
and our new magnetic circuit
"NIC". It is interesting to see
the combined effect of these
two innovations on overall
performance.

The results show: extended
frequency response for better
transient performance and
thus better definition; high
linearity in the critical region
1-3kHz for improved timbral
accuracy; drastic reduction of
resonance in the surround and
nonlinearities in the magnetic
circuit which are responsible
for blurring the stereo image
(fig N).

As for the woofers, distortion is
reduced by about 70 per cent!
This spectacular value shows
that, above all, the focus of the
Focal brand remains on its core
business: the transducers. They
are designed to enhance Focal’s
reputation for expressiveness
of musical rendering!

Let us add that these perfor­mance figures are meaningless
if taken out of their global
context of providing the richest
possible listening experience,
emotional and sensory, with
minimal coloration and the
highest definition. This remains
our crusade...

New Electra

0

-5

-10

-15

-20

-25

-30

-35

[dB]

-40

-45

-50

-55

-60

-65

-70

Fig N: Left, frequency response of our latest-generation midrange driver (blue trace) compared to the previous-generation W midrange (red trace). The extension of the frequency

range, resulting from all the improvements but particularly the exponential cone profile, is important as it promises improved transient response. NB: The dip at 3kHz in the blue
trace is due to the tested driver not being fitted with a dust cap.
Right, distortion analysis using the Klippel multi-tone test signal which gives an overview of nonlinear distortion performance (harmonic and intermodulation). The lowering of
distortion by about 10dB is a reduction of almost 70 per cent.

Relative multitone distortion

Electra 1038

KLIPPEL

500 1k 5k 10k 20k

Frequency [Hz]

SOPRA, A SHOWCASE FOR

THESE TECHNOLOGICAL

ADVANCES

From 2008 to 2010 Utopia III
defined our latest thoughts on
magnetic design (EM/IAL2) and
the side effects caused by the
suspensions of our W sandwich
cones (woofer midrange Diablo
Utopia). Four years later we
have developed solutions
applicable to models that
will reach a wider audience.
Innovation continues, enriched
by the arrival of new talent,
which is what we like to do at
Focal.

Like the research leading
to the Flax sandwich cones
introduced in the Aria range
in 2013, innovation downwards
takes its meaning!
We further confirm that we have
also always had an undeniable
technological leadership with

our latest developments
Sandwich W. Along with the
Beryllium for tweeter cone.
That is reassuring.

It is part of our brand
identity, as mentioned in the
introduction, that exploration
and pragmatism are not
incompatible. On the contrary,
they complement each other
surprisingly well.

In Utopia we have a flagship
range, globally recognized,
selective in nature and
which therefore excludes
many consumers. This is
characteristic of exceptional
products. Its move down­market makes no sense – it’s
not possible because it would
no longer be Utopia.

By contrast, improving the
premium segment occupied by
Electra using new technologies
inspired by what we have
learned from Utopia III makes
perfect sense. The jump in
performance is spectacular...

THE DESIGN FLOWS NATURALLY FROM
THE TWEETER REQUIREMENTS...

Compactness, strength,
innovation ; these are some of
the many terms which, from
the beginning, have fuelled
our design brief for Sopra.
The first sketches soon took
shape by integrating the key
"focus time" element of the
brand image. The tweeter
remains central, the "eye"
of the product that captures
the eye of the onlooker. Its
positioning is highly critical as
it predetermines the speaker’s
spatiality.

With the Utopia IAL 2 tweeter
we worked on improving the
flow of the rear-directed wave
from the dome. In Utopia
III the tweeter has its own
enclosure with a large volume
to absorb the back wave very
gently, as would an infinite
space (fig R). With Sopra,
the need for compactness
meant we could not devote
the necessary volume for this.
So our thinking has led us to
imagine a rear acoustic load
on the tweeter that approaches
the ideal by gradually
absorbing the rear radiation
(fig O, P). Located on the back
of the tweeter a progressively
damped horn-shaped duct
evacuates smoothly the rear

radiation of the dome and
has the advantage of being
the most effective solution
in terms of compactness,
thereby preserving internal
volume for the bass section.
Our horn leading to the back of
the speaker became an obvious
visual signature for this new
line.

This device (baptised Infinite
Horn Loader (IHL)) provides
a distortion reduction of 30%
(fig Q) in the critical area 1.5­4kHz and thus complements the
work done to reduce distortion
in the midrange driver. A patent
is pending.

Around this central part of the
cabinet design care was taken
to remove all obstacles to the
spread of the sound wave. Any
sharp or rough edges were
banished as they create low­level reflections inconsistent
with a high level spatiality.

We called the new range
Sopra, from the Latin Supra,
symbolizing the "plus" brought
by this new line in our crusade
to always offer more! Moreover,
it is the root soprano, probably
the most delicate voice to
reproduce, which occupies

the part of the frequency range
on which we have focused our
efforts and in which Sopra
brings major advances.
Further enriching the sound
signature Focal characterized
her expressive medium.

Fig O: The tweeter intermediate section,

shown in pink. Of monobloc construction it
incorporates a horn section leading to the rear,
in which damping material is placed to absorb
the rear radiation from the tweeter dome. The
structure is open in order to maximise the
internal volume available to the bass driver
while maintaining a compact cabinet (see below
illustration).

Fig P: Cross-section showing the structure of
the tweeter section with the horn absorbent
highlighted in orange. The internal volume
of the bass enclosure extends to the back of
the midrange section to preserve optimal
compactness.

Fig Q: Comparison of Klippel multi-tone distortion performance between our
IAL1 (red trace) and the new IHL (blue trace). A reduction of 30 per cent is
achieved between 1.5 and 4kHz.

Fig R: The IAL tweeter is changed for a first cavity designed for different density foam. At the
centre, a circular hole leads to the horn of the central mono bloc part.

LISTENING AND TUNING

The first listening assessment,
made in mid-July 2014 and
conducted with prototypes
of the midrange driver
incorporating the magnetic
and mechanical innovations,
were outstanding. The quasi
"holographic" quality of vocals
could not have been more
encouraging, as were focus,
articulation, expression ; in
short, a surprising realism.

Beginning in November 2014,
the first tuning sessions began
with transducers and cabinet
in their completed forms to
finalize crossover filtering.
Sopra 2 was chosen as the
reference point because it
is the most complicated
product for fine tuning. With
a high-midrange-class all the
difficulty lies in the connection
of the serious track / lower
midrange. The compactness
sought on Sopra 2 imposes
limits in the low range in one

hand, but especially in the other
transitional arrangements to
have the "fitting" the most
successful in harmonic
structure. In other words,
to have the most accurate
height and amplitude of each
vis-à-vis the original musical
message harmonics according
reproduced by the serious way
or through a medium. In the
past the complexity of tuning
mainly resided at connection
upper midrange. Having set
the source limitations (end
band of the medium through
TMD and tweeter BOT with IHL)
has significantly simplified the
upper midrange filtering and
naturally moved the issue to the
serious connection medium.

By early December we had
versions of Sopra 2 that we
could evaluate with different
electronics and in various
acoustics. Its incredibly neutral,
natural, high-definition sound

and high-midrange spatiality
revealed that the speakers
have extreme sensitivity to
other components in the
system. This point certainly
very positive, but requires
a particular discipline when
tuning not to compensate for
faults upstream of the speaker.
We had to redouble our efforts
and try endless combinations
to achieve near-final definition
of Sopra 2 by mid-February.

Gérard Chrétien

February 12

th

2015

ANNEXES

W Sandwich Cone

The three physical properties
of most importance in a cone
material are stiffness, density
and internal damping (loss
factor):

• High stiffness ensures

that the drive unit operates
pistonically over a wide
frequency range, i.e. that the
cone moves as a rigid whole
without breakup resonances.
High stiffness is particularly
important at low frequencies
where the cone has also to
resist internal cabinet pressure.

• Low density facilitates low

moving mass which ensures
that the force from the voice
coil results in high acceleration.
This maximises sensitivity and
ensures the best transient
performance and resolution
of fine details.

• High internal damping means

that where cone resonances
do occur they are effectively
suppressed, preventing them
causing audible coloration.

Combining these properties in
a single material is impossible,
so conventional cone materials
are all compromises. Paper

has low density and reasonable
internal damping but it is not
very stiff and its sound often
suffers from a "cardboard"
coloration. Polypropylene and
other plastics benefit from good
internal damping properties but
have relatively high density and
are not especially stiff; their
sound tends to lack detail
and precision. A composite of
woven aramid fibres in a resin
matrix achieves the required
rigidity but tends to suffer a
dull, "plastic" coloration.

Unlike a homogeneous cone, a
cone of sandwich construction
with skins and a core of
different materials) allows
high stiffness, low density and
high damping to be combined
in an ideal way. So as long
ago as the 1980s Focal began
developing sandwich cones,
our first proprietary "Poly-K
sandwich" material combining
woven aramid skins with a core
of hollow glass microspheres
embedded in resin. This
structure exhibited extremely
high rigidity and low mass, with
controllable internal damping.
It was a breakthrough, but still
we searched for better.

First, we replaced the glass
microball core with a structural
foam from the aerospace
industry; no other foam offers
the same high stiffness-to­mass ratio. Then, to create our
W sandwich (W is derived from
Verre-Verre ("verre" means
"glass" in French)) we replaced
the aramid skins with very fine
tissues of woven glass fibres
that are lighter and thinner

than aramid fibres and adhere
better to the foam core. The
result is a cone structure that
is mechanically more stable
and has superior stiffness,
which allows us to optimize
further the transmission
speed of the sound wave in
the cone. The relationship
between the thickness of the
glass skins and the foam core
allows us to adapt the cone
structure to the particular
application and the frequency
range being reproduced.
Internal damping can be very
accurately controlled by varying
the thickness of the foam: the
thicker it is, the higher the
damping.
As a result the W cone
generates an extremely
transparent, neutral sound,
free from the colorations and
distortions normally associated
with loudspeakers. Its only
downside is its price: more
than ten times that of a quality
paper cone.

Gamma Structure

Because the surface area of a
loudspeaker cabinet is many
times greater than that of the
drive unit diaphragms, it is all
too easy for the enclosure to
radiate sound at a level similar

to that of the drivers, sound
that is coloured by resonances
within the cabinet structure.
This muddies the speaker’s
sound and blurs the stereo
image.

To prevent this, it is vital that
the loudspeaker cabinet be
as inert as possible. At Focal,
we use MDF (medium-density
fibreboard) to achieve this. It
may seem a "low-tech" solution
compared to some cabinet
materials employed today but
MDF has inherent advantages
that we believe make it the
optimum material from which
to construct a loudspeaker.

First, it is dense enough and
stiff enough – when used in a
thick, heavy front baffle – to
resist the magnet reaction
force from the drive units.
As the driver diaphragm is
forced forwards by the voice
coil, an equal force acts in the
opposite direction on the drive
unit chassis. This is one of the
major inputs of vibrational
energy to the cabinet and it
must be resisted. This requires
not just a thick baffle but also
meticulously placed internal
bracing. Too stiff a cabinet,
though, can be as bad as
one which is not stiff enough
because it pushes structural
resonances up in frequency to
a part of the spectrum where
the ear is more sensitive.

Second, MDF has something
resembling a sandwich
structure, in which the faces
on each side of the board are
denser than its core. As well as

contributing to stiffness, this
endows MDF with good internal
damping to help suppress
vibrations when they occur.
Sopra and Utopia front baffles
use MDF plates thick laminated
to multiply the sandwich effect
(69mm thick for Sopra).

Third, MDF can easily
formed into curved cabinet
forms. These are good both
acoustically, because they
allow the radiated sound to
diffract smoothly around the
cabinet without secondary
radiation from sharp cabinet
edges, and structurally because
curved panels are stiffer than
conventional flat ones.
This combination of a thick
front baffle, extensive internal
bracing and curved panel forms
(all constructed using MDF) we
call Focal’s Gamma Structure.
It is as important to the sound
of our top loudspeakers as the
drive units themselves.

IAL & IHL

In beryllium we have the very
best dome material for our
tweeter. Because of beryllium’s
low density the dome
weighs just 21mg – literally
featherweight. This low moving
mass is advantageous because

it increases tweeter sensitivity
and ensures excellent transient
response, resulting in the
detailed, nuanced sound that
is Focal’s trademark.

But the low diaphragm mass
also poses a problem. In a
conventional tweeter a small
volume of air is enclosed
behind the dome, and this
acts as a spring, just as the
larger volume of air in a bass
enclosure does. The mass
of the dome and compliance
(springiness) of the air behind
it together form a resonance
system, and we need to keep
the frequency of that resonance
low in order to achieve a
tweeter whose frequency
response extends down to 2kHz
or lower, to allow an optimum
crossover frequency to the
midrange driver. The low mass
of the beryllium dome works
against us here because, for a
given volume of enclosed air, it
results in a higher resonance
frequency.

The solution is to increase
the volume of air behind the
tweeter dome by venting its
rear radiation into a chamber
behind. That solves the
moving mass/air compliance
resonance issue but it is vital
that standing wave resonances
and reflections within the rear
chamber are suppressed so
that they cannot find their
way back to the tweeter dome
where they would muddy its
forward output and compromise
sound quality. Ideally, the rear
radiation from the tweeter
should seem to disappear into

ANNEXES

an infinite acoustic space, from
where it never returns.

To achieve this requires careful
design of the rear chamber and
meticulous application of sound
absorbent materials within it.
This is what we achieved with
IAL (Infinite Acoustic Loading)
first used in the Electra 1000 Be
series. In IAL 2, used in Utopia,
we further refined this concept
by improving the flow of sound
from the rear of the tweeter
dome into the absorbent
chamber. But this solution
takes up a lot of space, more
than was available to us for
Sopra. So we developed Infinite
Horn Loader (IHL) to achieve
equally effective absorption
of the rear radiation but in a
smaller volume.

Beryllium Tweeter

marked ultrasonic resonances,
typically at below 30kHz.
Some of our competitors
have sought to address this
by adding a supertweeter but
Focal does not consider this a
satisfactory solution. Because
the distance between the
tweeter and supertweeter is
typically many wavelengths at
the highest audio frequencies
(the wavelength of sound in air
is just 1.7cm at 20kHz), beaming
and lobing effects occur where
the two units are crossed
over. For a tweeter to provide
the full benefit of extended
frequency response and clean
transient performance, and
be capable of crossing over
to the midrange driver before
the latter’s output becomes
significantly directional, it has
to be a single drive unit with a
working bandwidth of 1kHz to
40kHz, a remarkable range of
more than five octaves.

sound velocity, it has higher
density than beryllium, which
adversely affects sensitivity;
Focal ’s Beryllium tweeter,
because of the low mass of
its dome, achieves the high
sensitivity of 95dB SPL for

2.83V at 1m.

By using an inverted (concave)
dome rather than the more
typical convex dome, the
Focal Beryllium tweeter also
achieves a better mechanical
interface between the voice coil
and the dome, ensuring that
all of the vibrational energy
is transmitted into the dome
and radiated as sound, rather
than partially lost as heat in the
dome surround.

Particularly with today’s high­resolution digital sound sources
there is a need to extend
loudspeaker high frequency
response out to 40kHz and to do
so in such a way as to maintain
a "clean" impulse response,
without ultrasonic resonance.
Conventional metal dome
tweeters, using aluminium,
magnesium or titanium domes,
don’t achieve this: they have

Focal has achieved this
by selecting the very best
material available for making
a pure beryllium tweeter dome
regardless of its high cost, and
by manufacturing the beryllium
domes ourselves in France to
guarantee the highest quality.
Beryllium’s combination of
high tensile modulus and low
density endow it with a speed
of sound 2.5 times higher than
that of aluminium, magnesium
or titanium, and that translates
into a higher dome resonance
frequency. Beryllium’s higher
internal damping then helps
control that resonance when
it occurs. Although synthetic
diamond offers even higher

Focal-JMlab® - BP 374 - 108, rue de l’Avenir - 42353 La Talaudière cedex - France.

Tél . [phone] - Fax [phone] - © 2013 Focal-JMlab® - SCEB -140327/1

Due to constant technological advances, Focal-JMlab® reserves its right to modify specifications without notice. - Images may not conform exactly to specific product - Photos L’Atelier Sylvain Madelon