XPERTS CORNER
Understanding Timing Parameters
in Xilinx System Generator
When it comes to model-based design of feedback control systems,
getting a grip on timing parameters is crucial.
by Juergen Wassner
Lecturer
Lucerne University of Applied Sciences and Arts,
School of Engineering and Architecture
juergen.wassner@hslu.ch
Christoph Eck
Lecturer
Lucerne University of Applied Sciences and Arts,
School of Engineering and Architecture
christoph.eck@hslu.ch
Model-based design (MBD) has recently
attracted much attention for its promise of
closing the gap between abstract mathematical
modeling and physical realization
of real-time systems. By using the same
source for algorithm analysis, architecture
exploration, behavioral simulation and
hardware/software design, MBD promises
to shorten the system design cycle.
Several tools for model-based design
have emerged over the past years, many of
which center on the MATLAB ® and
Simulink ® environments from The
MathWorks. An elegant way to leverage the
precious stimuli-generation and dataanalysis
features of MATLAB/Simulink for
MBD is by employing the Xilinx System
Generator for DSP tool. Despite its name,
this tool can be useful for disciplines other
than classical digital signal processing.
Control applications are of particular interest,
since MATLAB/Simulink is the standard
design and simulation environment
for the control-engineering community.
Xilinx System Generator for DSP
enables control engineers to design their
system in the familiar Simulink environment
and then implement it in an FPGA,
without knowledge of a hardware description
language (HDL). To do so, you must
relate the parameter values in the mathematical
model of the controlled system
(often called the plant)—for example, a
continuous- or discrete-time transfer
function or state-space description—to
the sample rate of the digital controller
and the FPGA system clock frequency.
Digital Controllers in FPGA
As with any signal-processing algorithm,
digital controllers can be implemented in
either software or hardware. A combination
of the two—that is, a hardware-accelerated
software implementation—is rarely necessary,
due to the moderate sample-rate
requirements and algorithm complexity of
most control applications. When choosing
FPGAs as implementation technology for
digital controllers, designers can utilize features
not available with MCU-based software
solutions without compromising on
36 Xcell Journal Third Quarter 2009
To view a Ceros magazine you need to have JavaScript enabled in your browser and have the flash player 8 or higher installed. To install the latest flash player, click the icon above, install the flash plugin and then click here to refresh this page.