Download Electronics Projects Volume 15.Bak PDF

TitleElectronics Projects Volume 15.Bak
TagsCompact Cassette Equalization (Audio) Electronic Circuits Electronic Oscillator
File Size8.0 MB
Total Pages178
Document Text Contents
Page 89

77ELECTRONICS PROJECTS-15

Fig. 3: PCB layout for the active amplifier system.

Fig. 4: Component layout for the active amplifier system.

Page 90

78 ELECTRONICS PROJECTS-15

cover the desired range of 50 Hz to 20
kHz, because of the following facts:

The useful frequency range of a
speaker drive unit over which a linear
output can be expected is primarily gov-
erned by two things. The low frequency
limit is defined by the resonance fre-
quency of the cone in the suspension
system. The cone has a mass which is
coupled with a spring, combining the
compliance of the suspension system
with the compliance or ‘springiness’ of
the air in the enclosure.

Below the resonant frequency, the
unit is very inefficient and therefore
requires a lot of power from the ampli-
fier to gain any appreciable output.
Above the resonant frequency, the cone
operates in the ‘compliance’ region,
giving a reasonably linear and efficient
transfer of energy. It is in this range
that the drive unit can be used.

As the frequency increases, a point
is reached when the wavelength of the
sound reaches half the circumference
of the speaker cone. Above this fre-
quency, different areas of the cone start
moving in different ways, resulting in
phase anomalies and unwanted reso-
nances. This then forms the upper fre-
quency limit of the useful range of the
drive units.

Considering the above limitations,
the reproduction of the lower frequen-
cies, especially the base, requires a mas-
sive cone. For example, a 20cm bass
unit may have a resonant frequency of
around 40 Hz, but the upper frequency
limit would be 3 kHz. The reproduc-
tion of the middle order and higher
frequencies requires a lighter cone.

So, in the simplest system, we would
have two drive units: one for reproduc-
tion of the bass frequencies, which we
call the ‘woofer’ and the other one suit-
able for high frequencies, which we call
the ‘tweeter’. In the most popularly
used approach, we use a ‘crossover net-
work’, which splits the audio spectrum
and feeds the relevant drive units.

The crossover network consists of
a passive filter network of inductors,
capacitors and resistors, dividing the
frequency range into a low frequency
band which is fed to the woofer, a mid
frequency band which is fed to the
squawker, and a high frequency band
which is fed to the tweeter.

While this system works perfectly
and adequately, and witnesses a num-
ber of excellent passive speakers in the

Fig. 5: Power supply for the circuit shown in Fig. 2.

market, it has several inherent disad-
vantages. The problem arises in the de-
sign of such a filter.

To design such a filter, two factors
need to be known—the impedance of
the source and the impedance of the
load. The source impedance is the out-
put impedance of the amplifier and the
connecting loads. With good quality
leads, this should be less than a tenth
of an ohm, and no problems present.

However, the load impedance is the
impedance of the drive unit itself. Fig.6
shows the impedance of a typical loud-
speaker. As you can see, the imped-
ance varies considerably with fre-
quency, and is, in fact, a filter designer’s
nightmare. As you can imagine, cross-
over design is an art in itself.

The second problem arises from the
high currents involved in driving a loud-
speaker, which can reach tens of amps.
At low currents, capacitors, inductors
and resistors are linear components.
However, at high currents these can be
far from linear, introducing their own
distortions.

This is particularly true when high
values of inductance are required. To
construct a coil with an inductance of
over, say, l0 mH, a ferrite core is re-
quired, if the size of the inductor is not
prohibitive. A ferrite core will impose
its nonlinear hysteresis curve on
the circuits, which could introduce
serious distortion at high currents.

As we have seen, the imped-
ance of a drive unit by itself is far
from linear, having an impedance
value and phase angle that is vari-
able with frequency. However, ca-
pacitors and inductors also intro-
duce considerable phase (voltage)
lags and leads. A crossover net-
work, while having a linear fre-
quency response, could present a
very complex load to the amplifier.

Another major problem con-

cerns damping factor. A drive unit
works just as effectively as a
generator’s motor. When a transient
peak occurs, the amplifier drives the
cone outwards, and then applies a brak-
ing force. However, the cone will in-
evitably overshoot, and as it settles
back, it generates a current that is fed
back to the amplifier. If the impedance
presented by the amplifier is very low,
or damping factor high, then the cur-
rent will disappear quickly. If, how-
ever, the impedance of the amplifier
output is fairly high, the current will
affect the performance of the ampli-
fier. Effectively, then, a high damping
factor increases the control of the am-
plifier over the movement of the
speaker cone.

In a passive speaker the crossover
network forms a part of the impedance
that the drive unit ‘sees’ as it looks
back at the amplifier output. In the pass
band of the crossover filter, the imped-
ance of the crossover is fairly small,
and the damping factor high. However,
in the cut-off regions, the impedance
of the crossover rises, as it cuts out the
frequencies outside the pass band. This
means that the damping factor de-
creases, and the amplifier progressively
loses control over the speaker cone.

On the other hand, in an active sys-

Fig. 6: Characteristics of impedance
of a typical speaker.

Similer Documents