QFlex - Brochure

Amplifiers
Each QFlex module incorporates 8 or 16
channels of high performance amplication.
The choices available to the design team
were traditional class AB, integrated Class D
‘chip amps’ or build a bespoke professional
performance Class D multichannel amplier
from the ground up. Class AB was deemed
impractical primarily for physical and thermal
reasons; it wouldn’t t and if it did it would
roast! Class D chip ampliers obviously
solve the physical issue but do not provide
the performance required. Typically these
packages are specied at 10% THD which
we wouldn’t entertain, and they offer less
efciency, possibly 85%. A good gure until
you compare with QFlex, which achieves
97 – 98%.
QFlex amplier technology is proven in the
most demanding sectors of the industry.
Cable management
Further evidence of the attention to detail our
engineers have lavished on QFlex is revealed
in the cable management. Although this
aspect of the product would typically have
been left to last, the Tannoy approach of
total system integration is in evidence again.
Installation is swift, aesthetically pleasing
and error free by design. Two ying cables
interface power and data between column
elements. The Cat5 data cable carries
control signals, network communications
messaging and audio, and an IEC cable
for the mains power, both carried internally
within the column. The simplistic solution of
passing analog audio from master to slave
elements would almost
certainly result in signicant crosstalk
between data and audio, and would
degrade the signal due to an additional D/A,
A/D process. Our elegant solution satises
all functional requirements while maintaining
optimal signal quality.
DSP
A new member of the 3rd generation of
Sharc Digital Signal Processors is used to
perform the sophisticated processing for
all sixteen channels in the QFlex product.
This device has enough processing power
to run the entire product at a brisk 96
kHz sampling rate without the need to
compromise the efcacy of the processing,
and still has enough spare power to run
the protocol engine, the monitoring and
protection systems.
The audio processing algorithms we run in
the DSP use a mixture of xed-point and
oating-point processing, and various bit-
widths at various stages of the processing,
depending on the particular needs of
the algorithm, but always with a view to
maximizing sonic performance rather than
just to ‘squeeze more processing in’.