Digital Signal Processing

Size: px

Start display at page:

Download "Digital Signal Processing"

Claude Weaver
5 years ago
Views:

1 Real-Time Second Edition Digital Signal Processing from MATLAB to C with the TMS320C6X DSPs Thad B. Welch Boise State University, Boise, Idaho Cameron H.G. Wright University of Wyoming, Laramie, Wyoming Michael G. Morrow University of Wisconsin, Madison, Wisconsin CRC Press Taylor &. Francis Group Boca Raton London New York CRC Press is an imprint of the Taylor & Francis an Croup, Informa business

2 Contents List of Figures xix List of Tables xxv List of Program Listings xxvii Preface xxxi Acknowledgments xxxv Section I: Enduring Fundamentals 1 1 Introduction and Organization Why Do You Need This Book? Other DSP Books Demos and DSP Hardware Philosophy of This Book Real-Time DSP How to Use This Book Supported Boards Host Computer to DSP Board Communication Transition to Real-Time Chapter Coverage Hardware and Software Installation Reading Program Listings Get Started Problems 12 2 Sampling and Reconstruction Theory Choosing a Sampling Frequency Input/Output Issues: Samples or Frames? The Talk-Through Concept windsk Demonstration Starting windsk Talk-Thru Application Talk-Through Using Windows Talk-Through Using MATLAB and Windows Talk-Through Using MATLAB Only 24 xi

3 xii CONTENTS Talk-Through Using MATLAB and the DSK DSK Implementation in C Follow-On Challenges Problems 29 3 FIR Digital Filters Theory Traditional Notation FIR Filters Compared to IIR Filters Calculating the Output of a Filter windsk Demonstration Graphic Equalizer Application Notch Filter Application Audio Effects Application MATLAB Implementation Built-in Approach Creating Your Own Filter Algorithm DSK Implementation in C Brute-Force FIR Filtering in C: Part Brute-Force FIR Filtering in C: Part Circular Buffered FIR Filtering Follow-On Challenges Problems 52 4 IIR Digital Filters Theory windsk Demonstration: Notch Filter Application MATLAB Implementation Filter Design and Analysis IIR Filter Notation Block Diagrams Built-in Approach Creating Your Own Filter Algorithm DSK Implementation in C Brute-Force IIR Filtering More Efficient IIR Filtering Follow-On Challenges Problems 81 5 Periodic Signal Generation Theory Periodic Signals in DSP Signal Generation windsk Demonstration Arbitrary Waveform DTMF MATLAB Implementation Direct Digital Synthesizer Technique Table Lookup Technique DSK Implementation in C Direct Digital Synthesizer Technique 96

4 CONTENTS xiii Table Lookup Technique Table Lookup Technique with Table Creation Digital Resonator Technique Pseudonoise Sequences Theory windsk Demonstration MATLAB Implementation DSK Implementation in C Follow-On Challenges Problems Frame-Based DSP Theory Drawbacks of Sample-Based DSP What Is a Frame? windsk Demonstration MATLAB Implementation DSK Implementation in C Triple Buffering A Frame-Based DSP Example Using Direct Memory Access Summary of Frame-Based Processing Follow-On Challenges Problems Digital Filters Using Frames Theory windsk Demonstration MATLAB Implementation DSK Implementation in C Understanding the FIR Process for Frames How to Avoid the "Edge" Problems Explanation of the C Code Follow-On Challenges Problems The Fast Fourier Transform Theory Denning the FFT The Twiddle Factors The FFT Process Bit-Reversed Addressing Using the FFT for Filtering Avoiding Circular Convolution Real-Time Fast Convolution windsk Demonstration MATLAB Implementation Implementation in C Follow-On Challenges Problems 161

5 xiv CONTENTS 9 Spectral Analysis and Windowing Theory Power Spectrum of a Signal The Need for Windowing Window Characteristics windsk Demonstration MATLAB Implementation DSK Implementation in C Conclusion Follow-On Challenges Problems 174 Section II: Projects Project 1: Guitar Special Effects Introduction to Projects Theory Background How the Effects Work windsk Demonstration MATLAB Implementation FIR Comb Filter IIR Comb Filter Notch Filter Flanger Tremelo DSK Implementation in C Real-Time Comb Filters Other Real-Time Special Effects Follow-On Challenges Project 2: Graphic Equalizer 11.1 Theory 11.2 windsk Demonstration Graphic Equalizer Application Effect of the Graphic Equalizer MATLAB Implementation 11.4 DSK Implementation in C Applying Gain to Filter Bands GEL File Slider Control Follow-On Challenges Project 3: Peak Program Meter Theory windsk Demonstration: commdsk MATLAB Implementation DSK Implementation in C Example PPM Code DSK LED Control Another PPM Code Version Follow-On Challenges

6 CONTENTS xv 13 Project 4: AM Transmitters Theory windsk Demonstration MATLAB Implementation DSK Implementation in C Follow-On Challenges Project 5: AM Receivers Theory Envelope Detector The Hilbert-Based AM Receiver windsk Demonstration MATLAB Implementation DSK Implementation in C Follow-On Challenges Project 6: Phase-Locked Loop Theory windsk Demonstration MATLAB Implementation PLL Simulation A Few Updates to the MATLAB Implementation DSK Implementation in C Components of the PLL System Testing Follow-On Challenges Project 7: BPSK Digital Transmitters Theory Random Data and Symbol Generation BPSK Using Antipodal Rectangularly Shaped Bits BPSK Using Impulse Modulated Raised-Cosine Shaped Bits windsk Demonstration commdsk: Unfiltered BPSK commdsk: Raised-Cosine Filtered BPSK MATLAB Implementation Rectangular Shaped BPSK Signal Generator Impulse Modulated Raised-Cosine BPSK Signal Generator DSK Implementation in C A Rectangular Pulse Shaped BPSK Transmitter A Raised-Cosine Pulse Shaped BPSK Transmitter Summary of Real-Time Code Follow-On Challenges Project 8: BPSK Digital Receivers Theory The Output of the Matched Filter The Eye-Pattern Maximum Likelihood Timing Recovery windsk Demonstration MATLAB Implementation 286

7 xvi CONTENTS 17.4 DSK Implementation in C Components of the Digital Receiver System Testing Follow-On Challenges Project 9: MPSK and QAM Digital Transmitters Theory I- and Q-Based Transmitters A Few Constellation Diagrams windsk Demonstration commdsk: Root-Raised-Cosine Filtered QPSK MATLAB Implementation Impulse Modulated Root-Raised-Cosine QPSK Signal Generator DSK Implementation in C A Root-Raised-Cosine Pulse Shaped QPSK Transmitter A More Efficient RRC Pulse Shaped QPSK Transmitter Summary of Real-Time Code Higher-Order Modulation Schemes Follow-On Challenges Project 10: QPSK Digital Receivers Theory windsk8 Demonstration MATLAB Implementation Through the AGC A complete QPSK receiver DSK Implementation in C Through the AGC A complete QPSK receiver System Testing Follow-On Challenges 340 Section III: Appendices 343 A Code Composer Studio: An Overview 345 A.l Introduction 345 A.2 Starting Code Composer Studio 345 A. 3 Conclusion 346 B DSP/BIOS 349 B. l Introduction 349 B.l.l DSP/BIOS Major Features 349 B.1.2 DSP/BIOS Threads 349 B. 2 DSP/BIOS Sample Projects 350 C Numeric Representations 351 C. l Endianness 351 C.2 Integer Representations 352 C.3 Integer Division and Rounding 353 C.4 Floating-Point Representations 354 C.5 Fixed-Point Representations 356

8 CONTENTS xvii C.6 Summary of Numeric Representations 357 D TMS320C6x Architecture 359 D. l Computer Architecture Basics 359 D.l.l Instruction Set Architecture 360 D.1.2 Register Architectures 360 D.l.3 Memory Architectures 361 D.1.4 Fetch-Execute Model 362 D.l.5 Pipelining 362 D.l.6 Single- versus Multiple-Issue 365 D.1.7 Scheduling 365 D.2 TMS320C671x Architecture 366 D.2.1 Memory System 368 D.2.2 Pipeline and Scheduling 369 D.2.3 Peripherals 370 D. 2.4 Host Port Interface 370 D. 3 TMS320C674x Architecture 370 E Related Tools for DSKs 373 E. l Introduction 373 E.2 Windows Control Applications 373 E. 2.1 Sample Windows Control Application 374 E.3 MATLAB Exports 374 E.3.1 Exporting Direct-Form II Implementations 374 E.3.2 Exporting Second-Order Section Implementations 375 E.4 MATLAB Real-Time Interface 376 F Programming Perils and Pitfalls 377 F. l Debug versus Release Builds 377 F.2 The Volatile Keyword 377 F.3 Function Prototypes and Return Types 378 F.4 Arithmetic Issues 379 F.5 Controlling the Location of Variables in Memory 380 F.6 Real-Time Schedule Failures 381 F.7 Variable Initialization 382 F.8 Integer Data Sizes 383 G Abbreviations, Acronyms, and Symbols 385 References 391 Index 397

REAL-TIME DIGITAL SIGNAL PROCESSING from MATLAB to C with the TMS320C6x DSK

REAL-TIME DIGITAL SIGNAL PROCESSING from MATLAB to C with the TMS320C6x DSK Thad B. Welch United States Naval Academy, Annapolis, Maryland Cameron KG. Wright University of Wyoming, Laramie, Wyoming Michael