User manual

- 113 -
PS-8810
DIGITAL SIGNAL PROCESSOR
Technical Information
audio as a 'brick' of data. After the input DSPs have processed the
data, they transfer the audio bricks to shared memory. The two
output DSPs then retrieve the audio from memory and do the
required output processing. The audio is then sent serially from
the DSPs serial ports either to the Output board or the CobraNet
outputs.
The DSP processors have a 30 MHz clock, while serial digital
audio is locked to serial clock (3 MHz) and frame sync clock (48
KHz) supplied from the Input board.
An interface to the System Controller board allows programming
of the DSPs, control updates, and metering information to be
passed to the outside world. The DSPs depend upon the System
Controller for programming upon reset or initial power, as there is
no nonvolatile memory on board.
Output Section
Output serial digital audio from the DSP board is sent to the
output board for conversion and analogue processing. The serial
digital audio is comprised of two channels of 24-bit signals. Each
DAC converts the audio data stream into two differential output
channels. A differential amplifier filters the DAC’s audio output
and provides single-ended audio to a gain stage that sets the full-
scale output of the DAC. A single-ended-to-differential stage
provides a 50-ohm output. Optional transformers are available
for the Main Outputs.
System Controller
The System Controller board provides interface to the outside
world as well as providing all of the program storage and
initialization.
At reset, the control processor uses boot code from nonvolatile
flash memory. Program code is then transferred to fast RAM and
the processor begins to run. Each DSP processor is then booted
via the system interface and upon successful booting of all the
DSPs, audio processing begins.
Control and
Interface