�ݺ�ߣ

M. S. Ramaiah School of Advanced Studies
1
M. Sc. (Engg.) in Electronics System Design Engineering
GREESHMA S
CWB0913004 , FT-20136thModule Presentation
Module code : ESE2511
Module name : Microcontrollers and Interfacing
Module leader: Mr. Nagananda S.N.
Presentation on : 07/05/2014
ARM Boards for DSP Applications

2
•INTRODUCTION
•ARM9E-S
•DM3730
•FUNCTIONALBLOCKDIAGRAM
•BLOCKDIAGRAM
•SOFTWAREARCHITECTURE
•CHARACTERISTICSOFDSPPROCESSORS
•FEATURESOFDM3730
•REPRESENTINGADIGITALSIGNAL
•ADDITIONANDSUBTRACTIONOFFIXED-POINTSIGNAL
Overview

3
•MULTIPLICATIONANDDIVISIONOFFIXED-POINTSIGNAL
•SQUAREROOTOFFIXEDPOINTSIGNAL
•DSPONARM9E
•DSPONARM10E
•FIRFILTER
•IIRFILTER
•THEDISCRETEANDFASTFOURIERTRANSFORM
•APPLICATIONS
•CONCLUSION
•REFERENCES
Overview

4Introduction
Emergingstandardsforalgorithmsinmanyapplicationareashaveputfurtherdemandsontheabilityofprocessingplatformstodeliverefficientcontrolcapability
ARM’sapproachhasbeentodesignRISCcorearchitectureswithinstructionsetsthatprovideefficientsupportforparticularapplications,withoptimalbalancebetweenhardwareandsoftwareimplementation
Toacceleratesignal-processingalgorithmsARMaddsnewDSPinstructionstotheARMinstructionset
ARMDSPextensionsbroadenthesuitabilityoftheARMCPUfamilytoapplicationsthatrequireintensivesignalprocessingandatthesametimeretainingthepowerandefficiencyofahighperformanceRISCmicrocontroller
TheARMDSPextensionshavealreadybeenimplementedintheARM926EJ-S, ARM946E-S,ARM966E-S,ARM9E-S

5Introduction
Processing digitalized signals requires high memory bandwidths and fast multiplyaccumulate operations
A microcontroller handles the user interface, and a separate DSP processor manipulate digitalized signals such as audio
A single-core design can reduce cost and power consumption over a two-core solution
The ARMv5TE extensions available in the ARM9E and later cores provideefficient multiply accumulate operations
DSP applications are typically multiply and load-store intensive
Filtering is most commonly used signal processing operation
Another very common algorithm is the Discrete Fourier Transform

6
Introduction

7ARM9E-S
The ARM9E-S core has the ARM architecture v5TE
This includes an enhanced multiplier design for improved DSP performance
It is a 32-bit microcontroller
It offers high performance for very low power consumption and gate count
The ARM architecture is based on Reduced Instruction Set Computer (RISC) principles
The reduced instruction set and related decode mechanism are much simpler than those of Complex Instruction Set Computer (CISC) designs
This simplicity gives
•a high instruction throughput
•an excellent real-time interrupt response
•a small, cost effective, processor macrocell

8DM3730
Based on enhanced device architecture
Integrated on TI’s advanced 45-nm technology
Device supports HLOS and RTOS
Fully backward compatible

9Functional Block DiagramFigure 1 : DM3730 Functional Block Diagram

10
Block Diagram
Benefits
•2000DMIPS for Oss like linux, Win CE, RTOS
•3-D graphics up to 20M polygons per second for robust GUIs
•Backward compatible with OMAP3530
Figure 2 : DM3730 BlockDiagram
Application
•Smart connected devices
•Patient monitoring
•Media Player

11Software ArchitectureFigure 3 : Software Architecture of DM3730
Industry Standard OS component
TI provider component
Open Source

12
Characteristics of DSP processor
Harvard Architecture
High performance MAC
Saturating math
SIMD instruction for parallel computation
Barrel shifters
Floating point hardware

13Features of DM3730
ARM microprocessor subsystem
Enhanced direct memory access controller
Video hardware accelerators
Tile based architecture delivering up to 20MPoly/sec
DSP instructions/data little Endian
NEON multimedia architecture
Load store architecture with Non-aligned support
64 32-Bit General purpose registers
Six ALUs, each supports single 32-bit, dual 16-bit, or quad-8 bit , Arithmetic per clock cycle

14Representing a Digital Signal Figure 4 : Digitalizing an Analogue Signal
xis signal and t is time
In an analogue signal x[t ], the index tand the value x are both continuous real variables
ARM uses fixed point representation

15
Addition and Subtraction of Fixed-Point Signals
The general case is to convert the signal equation
Fixed-point format
or in integer C
n = m = d. Therefore normal integer addition gives a fixed-point
Provided d = m or d = n

16Contd…
There are four common ways you can prevent overflow
•Ensure that the X[t ]and C[t ] representations have one bit of spare headroom each
•Use a larger container type for Y than for X and C
•Use a smaller Q representation for y[t ]
•For example, if d = n − 1 = m − 1, then the operation becomes
•Use saturation

17Multiplication of Fixed-Point Signals
Fixed point format
or in integer CDivision of Fixed-Point Signals
fixed point format
or in integer C

18Square Root of a Fixed-Point Signals
Fixed point format
or in integer C

19DSP on the ARM9E
The ARM9E core has a very fast pipelined multiplier array that performs a 32-bit by 16-bit multiply in a single issue cycleWriting DSP Code for the ARM9E
The ARMv5TE architecture multiply operations are capable of unpacking 16-bit halvesfrom 32-bit words and multiplying them
The multiply operations do not early terminate. Therefore use MUL and MLA for multiplying 32-bit integers. For 16-bit values use SMULxy and SMLAxy
Multiply is the same speed as multiply accumulate. Use the SMLAxy instructionrather than a separate multiply and add

20DSP on the ARM10E
The ARM10E implements a background loading mechanism to accelerate load and storemultiples
It uses a 64-bit-wide data path that can transfer two registers on every background cycleWriting DSP Code for the ARM10E
Load and store multiples run in the background to give a high memory bandwidth
Ensure data arrays are 64-bit aligned so that load and store multiple operations canTransfer two words per cycle
The multiply operations do not early terminate. Therefore use MUL and MLA for multiplying 32-bit integers. For 16-bit values use SMULxy and SMLAxy
The SMLAxy instruction takes one cycle more than SMULxy

21FIR filters
The finite impulse response (FIR) filter is a basicbuilding block of many DSP applications
FIR filter to remove unwanted frequency ranges, boostcertain frequencies, or implement special effects
The FIR filter is the simplest type of digital filter
The filtered sample y(t)depends linearly on afixed, finite number of unfilteredsamples x(t)
Calculating accumulated values A[t ]

22IIR filters
An infinite impulse response (IIR) filter is a digital filter that depends linearly on a finite number of input samplesand a finite number of previous filter outputs
Mathematically
Factorize the filter into a series of bi quads—anIIR filter with M = L = 2
Z-Transform

23The Discrete Fourier TransformThe Fast Fourier Transform
The DiscreteFourier Transform (DFT)converts a time domain signal to a frequency domain signal
A FFT is an algorithm to compute the discrete Fourier transform and its inverse

24Applications
Portable data terminals
Navigation
Auto Infotainment
Gaming
Medical Imaging
Home automation
Single board

25Conclusion
DM3730 cost effective
It is low power and has high performance
DM3730 delivers a nearly 40% increase in ARM performance
Over 50% increase in DSP performance
Has twice the graphics capability, while reducing power consumption
Use a fixed-point representation for DSP applications where speed is critical withmoderate dynamic range

26Reference
1.DM3730, http:// www.ti.com/lit/ds/symlink/dm3730.pdf
2.DM3730, http://www.ti.com/lit/ml/sprt571/sprt571.pdf
3.DM3730, http://media.digikey.com/pdf/ DM3730_AM3703TorpedoSOMBrief.pdf

27

�ݺ�ߣ

ARM Boards for DSP Applications

More Related Content

ARM Boards for DSP Applications