National Instruments Car Stereo System 320571 01 User Manual

™

NI-DSP

Software Reference Manual

for LabVIEW for Windows

Digital Signal Processing Software for the PC

December 1993 Edition

Part Number 320571-01

Limited Warranty

The media on which you receive National Instruments software are warranted not to fail to execute programming

instructions, due to defects in materials and workmanship, for a period of 90 days from date of shipment, as

evidenced by receipts or other documentation. National Instruments will, at its option, repair or replace software

media that do not execute programming instructions if National Instruments receives notice of such defects during

the warranty period. National Instruments does not warrant that the operation of the software shall be uninterrupted

or error free.

A Return Material Authorization (RMA) number must be obtained from the factory and clearly marked on the

outside of the package before any equipment will be accepted for warranty work. National Instruments will pay the

shipping costs of returning to the owner parts which are covered by warranty.

National Instruments believes that the information in this manual is accurate. The document has been carefully

reviewed for technical accuracy. In the event that technical or typographical errors exist, National Instruments

reserves the right to make changes to subsequent editions of this document without prior notice to holders of this

edition. The reader should consult National Instruments if errors are suspected. In no event shall National

Instruments be liable for any damages arising out of or related to this document or the information contained in it.

EXCEPT AS SPECIFIED HEREIN, NATIONAL INSTRUMENTS MAKES NO WARRANTIES, EXPRESS OR IMPLIED,

AND SPECIFICALLY DISCLAIMS ANY WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR

PURPOSE. CUSTOMER'S RIGHT TO RECOVER DAMAGES CAUSED BY FAULT OR NEGLIGENCE ON THE PART

OF NATIONAL INSTRUMENTS SHALL BE LIMITED TO THE AMOUNT THERETOFORE PAID BY THE CUSTOMER.

NATIONAL INSTRUMENTS WILL NOT BE LIABLE FOR DAMAGES RESULTING FROM LOSS OF DATA, PROFITS,

USE OF PRODUCTS, OR INCIDENTAL OR CONSEQUENTIAL DAMAGES, EVEN IF ADVISED OF THE POSSIBILITY

THEREOF. This limitation of the liability of National Instruments will apply regardless of the form of action,

whether in contract or tort, including negligence. Any action against National Instruments must be brought within

one year after the cause of action accrues. National Instruments shall not be liable for any delay in performance due

to causes beyond its reasonable control. The warranty provided herein does not cover damages, defects,

malfunctions, or service failures caused by owner's failure to follow the National Instruments installation, operation,

or maintenance instructions; owner's modification of the product; owner's abuse, misuse, or negligent acts; and

power failure or surges, fire, flood, accident, actions of third parties, or other events outside reasonable control.

Under the copyright laws, this publication may not be reproduced or transmitted in any form, electronic or

mechanical, including photocopying, recording, storing in an information retrieval system, or translating, in whole or

in part, without the prior written consent of National Instruments Corporation.

Trademarks

™

LabVIEW , NI-DAQ , RTSI , and NI-DSP are trademarks of National Instruments Corporation.

Product and company names listed are trademarks or trade names of their respective companies.

Warning Regarding Medical and Clinical Use

of National Instruments Products

National Instruments products are not designed with components and testing intended to ensure a level of reliability

suitable for use in treatment and diagnosis of humans. Applications of National Instruments products involving

medical or clinical treatment can create a potential for accidental injury caused by product failure, or by errors on the

part of the user or application designer. Any use or application of National Instruments products for or involving

medical or clinical treatment must be performed by properly trained and qualified medical personnel, and all

traditional medical safeguards, equipment, and procedures that are appropriate in the particular situation to prevent

serious injury or death should always continue to be used when National Instruments products are being used.

National Instruments products are NOT intended to be a substitute for any form of established process, procedure, or

equipment used to monitor or safeguard human health and safety in medical or clinical treatment.

Contents

About This Manual

....................................................................................................................xi

Assumption of Previous Knowledge ........................................................................................................xi

Organization of This Manual ....................................................................................................................xi

Conventions Used in This Manual............................................................................................................xii

Related Documentation ............................................................................................................................xiv

Additional Software ..................................................................................................................................xiv

NI-DAQ for DOS/Windows/LabWindows ................................................................................xiv

Developer Toolkit ......................................................................................................................xv

Compatible Hardware............................................................................................................................... xv

Customer Communication ........................................................................................................................xv

Part 1

Getting Started with NI-DSP

........................................................................................1-1

Product Overview ..................................................................................................................................... 1-1

The NI-DSP Software............................................................................................................................... 1-1

What Your Distribution Diskettes Should Contain ....................................................................1-2

Installing NI-DSP for LabVIEW for Windows ........................................................................................1-2

Board Configuration ................................................................................................................................. 1-3

Installation on an ISA (or AT) Bus Computer ........................................................................... 1-3

Installation on an EISA Bus Computer ......................................................................................1-3

Part 2

Introduction to the NI-DSP Analysis VIs

....................................................1-1

Using the NI-DSP VIs in LabVIEW ........................................................................................................1-1

AT-DSP2200 Software Overview ............................................................................................................1-1

Memory Management and Data Transfer................................................................................................. 1-2

Special Features of the NI-DSP Analysis VIs ..........................................................................................1-5

Hints for Improving the Execution Speed on the DSP Board ....................................................1-7

An Example of Using NI-DSP Analysis VIs............................................................................................1-8

Part 3

NI-DSP Function Reference

Chapter 1

NI-DSP Analysis VI Reference Overview^{............................................................................... 1-1}

The NI-DSP Analysis VI Overview ......................................................................................................... 1-1

Analysis VI Organization ......................................................................................................................... 1-3

Accessing the NI-DSP Analysis VIs ........................................................................................................1-3

About the Fast Fourier Transform (FFT)..................................................................................................1-4

About Filtering..........................................................................................................................................1-5

About Windowing ....................................................................................................................................1-6

Chapter 2

NI-DSP Analysis VI Reference^{........................................................................................................2-1}

Copy Mem(DSP to DSP) ..........................................................................................................................2-1

Copy Mem(DSP to LV) ............................................................................................................................2-2

Copy Mem(LV to DSP) ............................................................................................................................2-3

NI-DSP SRM for LabVIEW for Windows

Contents

DSP Absolute............................................................................................................................................2-4

DSP Add ................................................................................................................................................... 2-5

DSP Allocate Memory..............................................................................................................................2-6

DSP Blackman Window ........................................................................................................................... 2-7

DSP Blackman Harris Window ................................................................................................................2-8

DSP Butterworth Coefficients ..................................................................................................................2-9

DSP Chebyshev Coefficients....................................................................................................................2-10

DSP Clip ................................................................................................................................................... 2-11

DSP Complex FFT....................................................................................................................................2-12

DSP Convolution ......................................................................................................................................2-13

DSP Correlation ........................................................................................................................................2-14

DSP Cross Power......................................................................................................................................2-15

DSP Custom..............................................................................................................................................2-16

DSP Decimate........................................................................................................................................... 2-17

DSP Deconvolution ..................................................................................................................................2-18

DSP Derivative ......................................................................................................................................... 2-19

DSP Divide ............................................................................................................................................... 2-20

DSP Elliptic Coefficients..........................................................................................................................2-21

DSP Equi-Ripple Bandpass ......................................................................................................................2-23

DSP Equi-Ripple Bandstop ......................................................................................................................2-25

DSP Equi-Ripple HighPass ......................................................................................................................2-27

DSP Equi-Ripple LowPass ....................................................................................................................... 2-29

DSP Exact Blackman Window................................................................................................................. 2-30

DSP Exponential Window ........................................................................................................................2-31

DSP FHT ..................................................................................................................................................2-32

DSP Flat Top Window..............................................................................................................................2-33

DSP Force Window ..................................................................................................................................2-34

DSP Free Memory ....................................................................................................................................2-34

DSP Gaussian White Noise ......................................................................................................................2-35

DSP General Cosine Window ..................................................................................................................2-36

DSP Hamming Window ........................................................................................................................... 2-37

DSP Handle to Address ............................................................................................................................2-38

DSP Hanning Window..............................................................................................................................2-39

DSP IIR Filter ........................................................................................................................................... 2-40

DSP Impulse Pattern................................................................................................................................. 2-42

DSP Impulse Train Pattern ....................................................................................................................... 2-43

DSP Index Memory ..................................................................................................................................2-44

DSP Init Memory......................................................................................................................................2-45

DSP Integral..............................................................................................................................................2-46

DSP Inv Chebyshev Coeff ........................................................................................................................2-47

DSP Inverse FFT ......................................................................................................................................2-48

DSP Inverse FHT......................................................................................................................................2-49

DSP Kaiser-Bessel Window ..................................................................................................................... 2-50

DSP Linear Evaluation ............................................................................................................................. 2-51

DSP Load ..................................................................................................................................................2-51

DSP Log....................................................................................................................................................2-52

DSP Max & Min....................................................................................................................................... 2-53

DSP Median Filter ....................................................................................................................................2-54

DSP Multiply ............................................................................................................................................2-55

DSP Parks-McClellan............................................................................................................................... 2-56

DSP Polar to Rectangular ......................................................................................................................... 2-59

DSP Polynomial Evaluation ..................................................................................................................... 2-60

DSP Power Spectrum................................................................................................................................2-61

DSP Product..............................................................................................................................................2-61

DSP Pulse Pattern ..................................................................................................................................... 2-62

DSP Ramp Pattern ....................................................................................................................................2-63

DSP Random Pattern ................................................................................................................................2-64

DSP Rectangular to Polar ......................................................................................................................... 2-65

NI-DSP SRM for LabVIEW for Windows

Contents

DSP ReFFT............................................................................................................................................... 2-66

DSP Reset ................................................................................................................................................. 2-66

DSP Reverse ............................................................................................................................................. 2-67

DSP Sawtooth Pattern............................................................................................................................... 2-68

DSP Set..................................................................................................................................................... 2-69

DSP Shift ..................................................................................................................................................2-69

DSP Sinc Pattern....................................................................................................................................... 2-70

DSP Sine Pattern....................................................................................................................................... 2-71

DSP Square Pattern................................................................................................................................... 2-72

DSP Square Root ......................................................................................................................................2-73

DSP Sort ................................................................................................................................................... 2-74

DSP Start ..................................................................................................................................................2-75

DSP Subset ............................................................................................................................................... 2-75

DSP Subtract............................................................................................................................................. 2-76

DSP Sum................................................................................................................................................... 2-76

DSP TimeOut............................................................................................................................................2-77

DSP Triangle Pattern ................................................................................................................................2-78

DSP Triangular Train................................................................................................................................2-80

DSP Triangular Window ..........................................................................................................................2-81

DSP Uniform White Noise ....................................................................................................................... 2-82

DSP Unwrap Phase................................................................................................................................... 2-83

DSP Zero Padder ......................................................................................................................................2-84

Part 4

NI-DSP Interface Utilities

Chapter 1

Introduction to the NI-DSP Interface Utilities ^{....................................................................1-1}

Overview of the NI-DSP Interface Utilities..............................................................................................1-1

Installing the NI-DSP Interface Utilities ..................................................................................................1-2

Using the NI-DSP Interface Utilities ........................................................................................................1-2

Chapter 2

Getting Started with the NI-DSP Interface Utilities ^{....................................................... 2-1}

Creating Your Custom NI-DSP Library ................................................................................................... 2-1

1. Create Your Source Code of C Functions ............................................................................. 2-1

GMaxMin.c Example..................................................................................................2-1

Guidelines for the Custom Functions ..........................................................................2-2

DSP Board Memory Management............................................................................... 2-3

2. Compile and/or Assemble Source Code................................................................................2-4

3. Add Your Object Filenames to a Linker File (ifile) ..............................................................2-4

4. Add Your New Function Names to a Library Function List File..........................................2-4

Customizing the DSP Library by Deleting Functions ................................................. 2-5

5. Run the Build Dispatch Application to Generate an Assembly Dispatch File......................2-6

6. Compile, Assemble, and Link Your Custom Library............................................................2-7

Creating Your LabVIEW Interface ..........................................................................................................2-8

1. Bundle All of the Input Parameters to Arrays....................................................................... 2-8

2. Call the Custom VI................................................................................................................2-10

3. Index the Output Arrays to Obtain the Results ..................................................................... 2-10

Executing the Custom Function from LabVIEW ..................................................................................... 2-12

Chapter 3

DSP Board Function Overview ^{......................................................................................................3-1}

Data Acquisition Functions ......................................................................................................................3-3

vii

NI-DSP SRM for LabVIEW for Windows

Contents

Chapter 4

Using the DMA VIs^{..................................................................................................................................4-1}

DSP DMA Copy(DSP to LV)................................................................................................................... 4-3

DSP DMA Copy(LV to DSP)................................................................................................................... 4-4

Appendix A

Error Codes

..........................................................................................................................................A-1

Error Conditions........................................................................................................................................A-1

Appendix B

Customer Communication

............................................................................................... B-1

Glossary........................................................................................................................................... Glossary-1

..........................................................................................................................................................Index-1

Index

NI-DSP SRM for LabVIEW for Windows

viii

Contents

Figures

Part 1

Figure 1-1.

Part 2

Development Paths with the NI-DSP Software ............................................................................. 1-1

Figure 1-1.

Figure 1-2.

Figure 1-3.

Figure 1-4.

Figure 1-5.

Figure 1-6.

Figure 1-7.

Figure 1-8.

Figure 1-9.

Communication between the PC and the DSP Board ....................................................................1-1

DSP Handle Cluster ....................................................................................................................... 1-3

The Hexadecimal Encoding of a Typical DSP Handle..................................................................1-3

Front Panel–An Example of How to Allocate a DSP Handle Cluster........................................... 1-4

Block Diagram–An Example of How to Allocate a DSP Handle Cluster..................................... 1-4

DSP Add VI ................................................................................................................................... 1-5

The error in/error out Cluster......................................................................................................1-5

An Example That Does Not Use error in/error out for Sequencing VIs ....................................1-6

An Example of Using the error in/error out Cluster for Sequential VI Execution..................... 1-7

Figure 1-10. Front Panel–An Example of Using NI-DSP Analysis VIs ............................................................1-8

Figure 1-11. Block Diagram–An Example of Using NI-DSP Analysis VIs ......................................................1-8

Part 3

Figure 1-1.

Figure 1-2.

Choosing DSP2200 from the Functions Menu ..............................................................................1-4

Spectral Leakage Demonstrated Using Convolution..................................................................... 1-7

Part 4

Figure 1-1.

Figure 1-2.

NI-DSP for DOS Directory Structure ............................................................................................1-1

Interface Layers to Onboard Functions..........................................................................................1-2

Figure 2-1.

Figure 2-2.

Figure 2-3.

Figure 2-4.

Figure 2-5.

Figure 2-6.

Figure 2-7.

Figure 2-8.

Figure 2-9.

Linker File NIDSPLNK................................................................................................................. 2-4

Library Function List File NIDSP.fnc ........................................................................................... 2-4

Typical Section of NIDSP.fnc ....................................................................................................... 2-5

Signals Group Section in dspfncs.h ............................................................................................... 2-6

Signals Group Section in dispatch.s............................................................................................... 2-6

How to Bundle Parameters in LabVIEW to Call gmaxmin.c ........................................................2-9

How to Connect to Custom VI to Call gmaxmin.c ........................................................................2-10

Block Diagram–How to Index the Output Arrays of the Custom VI ............................................2-11

Block Diagram–Using the Custom VI to Call gmaxmin.c on theDSP Board from LabVIEW..... 2-11

Figure 2-10. Front Panel–Using the Custom VI to Call gmaxmin.c on theDSP Board from LabVIEW........... 2-12

NI-DSP SRM for LabVIEW for Windows

Contents

Tables

Part 1

Table 1-1.

Part 3

Table 1-1.

Part 4

Table 2-1.

Subdirectories Created by SETUP................................................................................................. 1-2

The NI-DSP Analysis VI Groups ..................................................................................................1-1

Files Required to Build the Custom DSP Library Example ..........................................................2-7

Appendix A

Table A-1.

NI-DSP Analysis Library Error Codes ..........................................................................................A-1

NI-DSP SRM for LabVIEW for Windows

About This Manual

The NI-DSP Software Reference Manual for LabVIEW for Windows explains how to use the NI-DSP software

package for the LabVIEW for Windows environment. The NI-DSP software package contains the NI-DSP Analysis

VIs, which are high-level digital signal processing (DSP) VIs that call the functions that execute on the

AT-DSP2200 plug-in board for IBM AT bus and compatible computers. This manual describes how to use the

NI-DSP Analysis VIs to develop applications in LabVIEW using the AT-DSP2200 board.

The NI-DSP software package also contains the NI-DSP Interface Utilities. The NI-DSP Interface Utilities are a set

of tools and examples that help you customize your NI-DSP Analysis VIs and the DSP Library, which is resident on

the board. This manual contains step-by-step instructions and useful examples to help the LabVIEW developer add

custom algorithms to the NI-DSP Analysis VIs using the NI-DSP Interface Utilities.

Assumption of Previous Knowledge

The material in this manual is for users who are familiar with LabVIEW and the IBM family of computers and

compatible computers.

Organization of This Manual

This manual is divided into four parts.

•

Part 1, Getting Started with NI-DSP, contains a brief product overview, information about the NI-DSP for

LabVIEW for Windows package, and the procedure for installing the software.

Part 2, Introduction to the NI-DSP Analysis VIs, describes how to use the NI-DSP Analysis VIs in your

LabVIEW applications. This part also describes how to manage memory on the DSP board from your

LabVIEW application, and how to transfer data between your LabVIEW application and the NI-DSP functions

on the board. This part contains general guidelines for developing NI-DSP applications within LabVIEW.

•

Part 3, NI-DSP Function Reference, is intended as a reference for users familiar with Part 2. Part 3 is organized

as follows:

Chapter 1, NI-DSP Analysis VI Reference Overview, contains an overview of the NI-DSP Analysis VIs and

includes a list of the VIs. This chapter describes how the NI-DSP Analysis VIs are organized, and how to

access them.

Chapter 2, NI-DSP Analysis VI Reference, contains a brief explanation of each NI-DSP Analysis VI. The

VIs are arranged alphabetically.

•

Part 4, NI-DSP Interface Utilities, explains how to customize the NI-DSP Analysis Library on the board and to

create and run interfaces in LabVIEW to your custom library functions. Part 4 is organized as follows:

Chapter 1, Introduction to the NI-DSP Interface Utilities, contains an overview of the NI-DSP Interface

Utilities, installation instructions, and explains how to use the NI-DSP Interface Utilities.

Chapter 2, Getting Started with the NI-DSP Interface Utilities, contains a step-by-step example for building

a custom DSP Library, creating a LabVIEW interface to a custom function, and executing the custom

function from the LabVIEW environment. The chapter demonstrates this concept with an example of how

to add a custom function.

NI-DSP SRM for LabVIEW for Windows

About This Manual

Chapter 3, DSP Board Function Overview, contains an overview of the prototypes of the C-callable

NI-DSP Analysis functions on the DSP board that you can use in your custom programs.

Chapter 4, Using the DMA VIs, describes two special VIs that transfer data between the host computer and

the DSP board without interfering with the DSP board.

•

Appendix A, Error Codes, contains a list of the error codes returned by the NI-DSP Analysis VIs and the

corresponding error messages.

Appendix B, Customer Communication, contains forms you can use to request help from National Instruments

or to comment on our products and manuals.

The Glossary contains an alphabetical list and description of terms used in this manual, including abbreviations,

acronyms, metric prefixes, mnemonics, and symbols.

The Index alphabetically lists topics covered in this manual, including the page where the topic can be found.

Conventions Used in This Manual

The following conventions are used in this manual:

Angle brackets enclose the name of a key on the keyboard–for example, <enter>.

bold

Bold text denotes menus, command names, parameters, function panel items, error

messages, and warnings.

DSP board

DSP board refers to the AT-DSP2200.

DSP Handle

DSP Handle refers to a 32-bit long integer code that constitutes an indirect reference to a

buffer of memory in the memory space of an AT-DSP2200 board. The code contains

information about the slot number of the board on which that buffer is allocated.

DSP Handle Cluster

DSP Library

DSP Handle Cluster refers to a cluster that constitutes two fields–a DSP Handle and a

32-bit size that indicates the number of elements the DSP Handle holds.

DSP Library refers to a common object file format (COFF) that constitutes NI-DSP

software that resides and runs on the AT-DSP2200 board. The DSP Library consists of

the Kernel, memory management routines, execution control routines, data

communication routines, interrupt handling routines, data acquisition routines and a set of

analysis functions.

<enter>

Key names are in lowercase letters.

Interface Library

Interface Library refers to the part of the NI-DSP software that resides on the PC and is

linked with your application in order to communicate data to and from the DSP board.

The Interface Library communicates with the DSP board using the driver.

italic

Italic text denotes emphasis, a cross reference, or an introduction to a key concept.

italic monospace Italic text in this font denotes that you must supply the appropriate words or values in the

place of these items.

NI-DSP SRM for LabVIEW for Windows

xii

About This Manual

LabVIEW Data Types

Each VI description includes a data type picture for each control and indicator, as

illustrated in the following table:

Control

Indicator

Data Type

Boolean

String

Signed 16-bit integer

Array of signed 16-bit integers

Signed 32-bit integer

Array of signed 32-bit integers

32-bit floating-point number; by default,

floating-point numbers are double precision

Array of 32-bit floating-point numbers

Path

DSP Handle Cluster

Array of DSP Handle Clusters

Error Cluster

monospace

Text in this font denotes text or characters that are to be literally input from the keyboard,

sections of code, programming examples, and syntax examples. This font is also used for

the proper names of disk drives, paths, directories, programs, subprograms, subroutines,

device names, array names, structures, variables, filenames, and extensions, and for

statements and comments taken from program code.

NI-DAQ

NI-DSP

NI-DAQ is used throughout this manual to refer to the NI-DAQ software for

DOS/Windows/LabWindows unless otherwise noted.

NI-DSP is used throughout this manual to refer to the NI-DSP software for LabVIEW for

Windows unless otherwise noted.

PC refers to the IBM PC AT and compatible computers, and to EISA personal computers.

WE DSP32C tools

WE DSP32C tools is used throughout this manual to refer to the AT&T WE DSP32C

Developer Toolkit.

Abbreviations, acronyms, metric prefixes, mnemonics, symbols, and terms are listed in the Glossary.

xiii

NI-DSP SRM for LabVIEW for Windows

About This Manual

Related Documentation

The following documentation available from National Instruments contains information that you may find helpful as

you read this manual.

•

AT-DSP2200 User Manual, part number 320435-01

LabVIEW Data Acquisition VI Reference Manual, part number 320536-01

LabVIEW Getting Started Manual for Windows, part number 320533-01

LabVIEW User Manual, part number 320534-01

LabVIEW Utility VI Reference Manual, part number 320543-01

NI-DAQ Function Reference Manual for DOS/Windows/LabWindows, part number 320499-01

NI-DAQ Software Reference Manual for DOS/Windows/LabWindows, part number 320498-01

NI-DSP Software Reference Manual for DOS/LabWindows, part number 320436-01

The following documentation also contains information that you may find helpful as you read this manual:

•

"A Computer Program for Designing Optimum FIR Linear Phase Digital Filters," McClellan, Parks, and

Rabiner, IEEE Transactions on Audio and Electroacoustics, Vol. AU-21, No. 6, pp. 506-525, December 1973

•

Digital Filter Design, Parks and Burrus, Wiley-Interscience

Discrete-Time Signal Processing, Oppenheim and Schafer, Prentice Hall

Numerical Recipes, Cambridge University Press

Additional Software

Additional DSP board software includes NI-DAQ for DOS/Windows/LabWindows and the Developer Toolkit.

NI-DAQ for DOS/Windows/LabWindows

Your AT-DSP2200 is shipped with the NI-DAQ for DOS/Windows/LabWindows software. NI-DAQ has a library

of functions that can be called from your application programming environment. These functions include routines

for analog input (A/D conversion), buffered data acquisition (high-speed A/D conversion), analog output (D/A

conversion), waveform generation, digital I/O, counter/timer, SCXI, RTSI, and self-calibration. NI-DAQ maintains

a consistent software interface among its different versions so you can switch between platforms with minimal

modifications to your code. NI-DAQ comes with language interfaces for Professional BASIC, Turbo Pascal,

Turbo C, Turbo C++, Borland C++, Microsoft C for DOS; and Visual Basic, Pascal, Microsoft C with SDK, and

Borland C++ for Windows. NI-DAQ for DOS/Windows/LabWindows software is on high-density 5.25 in. and

3.5 in. diskettes. You can use the DSP board in conjunction with the National Instruments AT Series data

acquisition boards and software to create a complete solution for integrated data acquisition and data analysis

applications.

NI-DSP SRM for LabVIEW for Windows

xiv

About This Manual

Developer Toolkit

The Developer Toolkit, an optional software package that you can purchase separately from National Instruments, is

required for building custom libraries with the NI-DSP Interface Utilities. The Developer Toolkit contains an

AT&T C compiler, assembler, linker, and documentation. With these tools, you can program AT Series DSP boards

directly, using theboard flexibility to custom tailor the DSP Library. The C compiler optimizes WE DSP32C code

and generates assembly language code that can be assembled and linked into a run-time module. When a run-time

module is completed, use the download tools and the debugger to load, debug, and execute the code, set parameters,

and report results. The Developer Toolkit also includes the WE DSP32C Support Software Library User Manual

and the WE DSP32 and WE DSP32C Language Compiler Library Reference Manual.

Compatible Hardware

You can use DSP boards in conjunction with the National Instruments AT Series data acquisition boards. In

particular, the National Instruments AT-DSP2200 is a high-performance, DSP board with high-accuracy audio

frequency (DC to 51.2 kHz) analog input/output for the PC. The AT-DSP2200 gives the PC a dedicated

high-speed numerical computation engine that can perform scientific calculations faster than the general-purpose

80x86 microprocessor on the PC.

Customer Communication

National Instruments wants to receive your comments on our products and manuals. We are interested in the

applications you develop with our products, and we want to help if you have problems with them. To make it easy

for you to contact us, this manual contains comment and configuration forms for you to complete. These forms are

in Appendix B, Customer Communication, at the end of this manual.

NI-DSP SRM for LabVIEW for Windows

Part 1

Getting Started with NI-DSP

This part contains a brief product overview, information about the NI-DSP for LabVIEW for Windows package, and

the procedure for installing the software.

Product Overview

The NI-DSP software package comes with a set of LabVIEW VIs that invoke the digital signal processing (DSP)

board-resident high-performance functions that efficiently process large blocks of numerical data and perform

numerically intensive computations. The NI-DSP Analysis VIs include numerical analysis, signal generation, DSP,

windowing, digital filtering, and memory management that are suitable for simulation, modeling, and sophisticated

data processing.

You can use NI-DSP to develop programs in the LabVIEW for Windows environment. This software comes with

the NI-DSP Interface Utilities so you can customize the DSP Library by adding functions to or deleting functions

from the Analysis Library on the DSP board and/or add interfaces to these custom functions in LabVIEW.

Figure 1-1 shows the development path for NI-DSP in the LabVIEW environment.

NI-DSP for

LabVIEW

for Windows

The LabVIEW

for Windows

Development

Environment

Interface

Utilities

User

Application

AT&T

Development

Environment

and Tools

(Optional)

Figure 1-1. Development Paths with the NI-DSP Software

The NI-DSP Software

The NI-DSP software consists of the NI-DSP for LabVIEW for Windows diskettes.

The NI-DSP software contains a Warranty Registration Form. Please fill out this form and return it to National

Instruments. The Warranty Registration Form entitles you to receive product upgrades and technical support.

NI-DSP SRM for LabVIEW for Windows

1-1

Part 1: Getting Started with NI-DSP

Getting Started with NI-DSP

Part 1

What Your Distribution Diskettes Should Contain

The NI-DSP software package contains the NI-DSP for LabVIEW for Windows Disks (for licensed LabVIEW for

Windows users). If your kit is missing any of these components, contact National Instruments.

Installing NI-DSP for LabVIEW for Windows

Note: NI-DSP for LabVIEW for Windows is intended for use with the standard LabVIEW for Windows software.

You must install LabVIEW for Windows before installing NI-DSP. You must install the Data Acquisition

Library of LabVIEW to run the NI-DSP software.

Before beginning the software installation, make backup copies of the NI-DSP for LabVIEW for Windows

distribution disks. Copy each disk onto a correctly labeled backup disk and store the original distribution disks in a

safe place.

You can install NI-DSP for LabVIEW for Windows from the DOS prompt, the Windows File Manager, or with the

Run... command from the File menu of the Program Manager.

1. Insert Disk 1 into the disk drive and run the SETUP.EXEprogram on Disk 1 using one of the following three

methods.

•

From the DOS prompt, type X: \SETUP(where X is the proper drive designation).

From Windows, select Run... from the File menu of the Program Manager. A dialog box appears. Type

X: \SETUP(where X is the proper drive designation).

•

From Windows, launch the File Manager. Click on the drive icon that contains Disk 1. Find SETUP.EXE

in the list of files on that disk and double-click on it.

2. The installer gives you the option of performing a full installation or a custom installation. Unless you do not

have sufficient disk space (approximately 4 megabytes), National Instruments recommends that you perform a

full installation. After you choose an installation, follow the instructions that appear on the screen.

After calling SETUP, the appropriate directories are created and the needed files are copied to your hard drive.

SETUPcan also install the NI-DSP Interface Utilities, discussed in Part 4, NI-DSP Interface Utilities, of this manual.

If you choose Full Installation, SETUPdoes the following things:

1. SETUPcreates a subdirectory called DSP2200of the vi.libsubdirectory of the LabVIEWdirectory.

SETUPdecompresses the NI-DSP Analysis VIs (as LabVIEW .LLBfiles) in the DSP2200directory.

2. SETUPcopies DSP.DLLto your windowsdirectory.

3. SETUPcreates the subdirectories shown in Table 1-1 under the directory you specified during setup. These

subdirectories make up the NI-DSP Interface Utilities.

4. SETUPcreates a subdirectory called DSP2200of the EXAMPLESsubdirectory of the LabVIEWdirectory.

SETUPcopies all of the NI-DSP VI examples there.

Table 1-1. Subdirectories Created by SETUP

Subdirectory Name

C:\NIDSP\LIB

C:\NIDSP\DISPATCH

C:\NIDSP\EXAMPLES

Description

Library files for linking with stand-alone programs

The Dispatch utility and related files

Contains source code for the examples

Part 1: Getting Started with NI-DSP

1-2

NI-DSP SRM for LabVIEW for Windows

Part 1

Getting Started with NI-DSP

NIDSPis the name you specify during setup.

The SETUPprogram prompts you for information including the drive letter and directory in which you have

installed the standard LabVIEW package. The program also verifies that your hard disk has enough space to hold

the NI-DSP for LabVIEW for Windows files.

If you choose Custom Installation, SETUPinstalls only the files you specify.

Board Configuration

There are several board configuration parameters that must be established before an NI-DSP application can execute

properly. These parameters are the board ID number, the board subtype, the base address, the interrupt level, the

DMA channel, and the pathname of the DSP Library files. These parameters are established differently depending

on whether you are installing the AT-DSP2200 in an ISA (or AT) bus computer or an EISA bus computer.

Installation on an ISA (or AT) Bus Computer

A configuration utility is supplied with the LabVIEW data acquisition software for establishing all the configuration

parameters on ISA bus computers. This utility, called WDAQCONF.EXE, saves the configuration parameters in a file

named WDAQCONF.CFG. Use the WDAQCONFutility to assign a board ID number to your AT-DSP2200, to choose

the memory subtype (either 64 Kwords or ≥128 Kwords), to set the base address, interrupt level, and DMA channel,

and to specify the pathname of the DSP Library file.

Two DSP Library files are supplied with your NI-DSP software–LV2200S.OUTand LV2200.OUT.

LV2200S.OUTis intended for use with the 64 Kword version of the AT-DSP2200. LV2200.OUTis intended for

use with any other version of the board. With the WDAQCONFutility, you can enter a complete path that will include

the DSP Library file name. If you enter complete pathnames, you can configure the NI-DSP software to

automatically load custom versions of the DSP Library files.

If you installed the Data Acquisition Library of LabVIEW, you can find WDAQCONF.EXEof your LabVIEW

directory. From Windows, you run WDAQCONF.EXEby double-clicking on its icon.

Installation on an EISA Bus Computer

Installing the AT-DSP2200 board on an EISA bus computer involves two different configuration utilities.

•

A system configuration utility is supplied with your computer by your computer vendor. This utility, along with

the !NIC1100.CFGEISA configuration file installed by the LabVIEW SETUPprogram, is used to assign a

slot number to the AT-DSP2200, to choose the memory subtype (either 64 Kwords or ≥128 Kwords), and to set

the base address, interrupt level, and DMA channel.

There are typically two methods for running your EISA system configuration utility.

1. Boot from the diskette containing the utility and place a copy of the !NIC1100.CFGfile on this diskette.

2. Run the utility from a directory on your hard disk and place a copy of the !NIC1100.CFGfile in this

directory. The EISA system configuration utility is typically named CF.EXE.

•

You must use the WDAQCONFutility to enter the DSP Library file pathname. You can only do this after

completing the EISA system configuration. Refer to the information concerning the establishment of this

pathname in the previous section titled, Installation on an ISA (or AT) Bus Computer.

Before using NI-DSP, you must run WDAQCONF.EXEto configure your DSP board.

For more information about board configuration, refer to Chapter 1, Introduction and Configuration, of the

LabVIEW Data Acquisition VI Reference Manual.

NI-DSP SRM for LabVIEW for Windows

1-3

Part 1: Getting Started with NI-DSP

Part 2

Introduction to the NI-DSP Analysis VIs

This part describes how to use the NI-DSP Analysis VIs in your LabVIEW applications. This part also describes

how to manage memory on the DSP board from your LabVIEW application, and how to transfer data between your

LabVIEW application and the NI-DSP functions on the board. This part contains general guidelines for developing

NI-DSP applications within LabVIEW.

Using the NI-DSP VIs in LabVIEW

LabVIEW users use the NI-DSP Analysis VIs as if they were any other standard VIs, as described in the LabVIEW

User Manual. Notice, however, that the NI-DSP Analysis VIs run analysis code on the DSP board rather than on the

host CPU. One of the major features includes the memory management and data transfer VIs, which are discussed

in detail in the section titled Memory Management and Data Transfer later in this chapter.

AT-DSP2200 Software Overview

The AT-DSP2200 board, working in conjunction with your personal computer, is a powerful numeric processor for

high-speed analysis of data arrays. The NI-DSP for LabVIEW for Windows software includes a number of utilities

and low-level memory management and data transfer VIs that facilitate communication between the DSP board and

the host computer. Figure 1-1 is a block diagram of the software utilities and libraries that control the

AT-DSP2200. The DSP Library and the low-level memory management and data transfer functions reside and

execute on the board. You can customize the DSP Library to optimize performance, as described in Part 4,

Chapter 2, Getting Started with the NI-DSP Interface Utilities, of this manual. Your application programs and the

Interface Library, however, reside on the host PC.

AT-DSP2200

AT or EISABus

DSP Hardware

DSP Software

User

Application

in LabVIEW

Onboard Kernel,

DSP Board

Memory

CPU

Management,

Data Transfer

Functions

Interface

(CIN)

Onboard Memory

DSP Library

Interface

Library

(DSP.DLL)

Figure 1-1. Communication between the PC and the DSP Board

NI-DSP SRM for LabVIEW for Windows

1-1

Part 2: Introduction to the NI-DSP Analysis VIs

Introduction to the NI-DSP Analysis VIs

Part 2

The AT-DSP2200 can process large amounts of data, separately and distinctly from the host PC processor. The

board consists not only of a signal processing chip, but also memory where data that the board processes must

reside. The AT-DSP2200 does not have access to memory locations on the host PC. Therefore, you must download

all data from your application programs to DSP board memory before processing it.

The Interface Library, DSP.DLL, and the Code Interface Node (CIN) interface, which reside on the PC, serve as a

bridge between your application programs in LabVIEW and the DSP software running on the board. When you call

an NI-DSP Analysis VI, the VI passes the parameters to the CIN first, which then passes the parameters to the

Interface Library, DSP.DLL. The Interface Library determines what type of parameters are being passed, decides

how to set up the data in DSP board memory, and then calls the actual functions that will run on the board.

When a function on the DSP board processes data, it assumes the data is resident in DSP board memory. Because

transferring data between the PC and the DSP board slows down processing, none of the NI-DSP Analysis VIs

transfer data back and forth internally except the data transferring VIs. The NI-DSP Analysis VIs process the data

buffers that are already on the board and leave the results on the board.

If the data buffer you want to process using the DSP board is in PC memory, you must copy the data to the DSP

board before you call a function on the DSP board to process the data. To see the results, you must then copy the

data back to the PC. Several special NI-DSP Analysis VIs perform these transfers. For scalars, the NI-DSP

Analysis VIs automatically perform the transfer for you.

The representation of data buffers in the NI-DSP Analysis VIs is not the normal LabVIEW data array representation

because the data buffers indicate the data location on the DSP board instead of the PC address. A special structure,

called a DSP Handle Cluster, represents the data buffer on the DSP board from LabVIEW. The DSP Handle Cluster

is a coded DSP board memory address that indicates where the data buffer is on the DSP board. You must call the

DSP Allocate Memory VI to obtain a valid DSP Handle Cluster. Several VIs can manage the memory on the DSP

board. You can allocate and deallocate memory on the DSP board using these VIs. The next section, Memory

Management and Data Transfer, discusses the VIs used to allocate memory and transfer data buffers to and from the

DSP board.

Memory Management and Data Transfer

This section describes how to manage memory on the DSP board from your LabVIEW application and how to

transfer data between your LabVIEW application and the DSP board.

The NI-DSP for LabVIEW package contains a set of VIs that manage memory space on the DSP board and help

improve data transfers between the DSP board and your application. There are VIs for allocating memory buffers on

the DSP board, for indexing into previously allocated buffers, for deallocating buffers and for copying data between

DSP and LabVIEW. The following VIs, described in greater detail in Part 3 of Chapter 2, NI-DSP Analysis VI

Reference, handle memory management on the DSP board and data transfers between the DSP board and your

LabVIEW application:

•

Copy Mem[DSP to DSP]

Copy Mem[DSP to LV]

Copy Mem[LV to DSP]

DSP Allocate Memory

DSP Free Memory

DSP Index Memory

DSP Init Memory

Part 2: Introduction to the NI-DSP Analysis VIs

1-2

NI-DSP SRM for LabVIEW for Windows

Part 2

Introduction to the NI-DSP Analysis VIs

The DSP Allocate Memory VI allocates memory buffers on the DSP board and returns a DSP Handle Cluster, which

has two fields that uniquely describe this buffer–a DSP Handle and a size.

Figure 1-2. DSP Handle Cluster

DSP Handle is a 32-bit integer containing information that indicates the board on which the allocated buffer resides,

and an index into an onboard Memory Look Up Table (MLUT) that holds the actual DSP address of the buffer that

this handle represents. Figure 1-3 shows how a DSP Handle is encoded. The size field holds the number of

elements in this buffer. An element can be 4 bytes (for 32-bit floating-point data or long integer data) or 2 bytes (for

16-bit integer data) depending on the bytes per element selector used in the DSP Allocate Memory VI.

Index into the

MLUT of the

Owner DSP Board

Board Number of

Owner DSP Board

Special Code

Figure 1-3. The Hexadecimal Encoding of a Typical DSP Handle

The first four hexadecimal numbers (upper 16 bits) of the DSP Handle, shown in Figure 1-3, are a special value.

The interface code for a particular function that your application calls decodes these four hexadecimal numbers to

determine if the argument is a valid DSP Handle.

Notes: Do not change the value of a DSP Handle Cluster. Keep in mind that a DSP Handle Cluster is just an entry

of a table that indicates where the data buffer is on the DSP board. If you want to operate on part of the

data in that buffer, use the DSP Index Memory VI or the DSP Subset VI to obtain a new DSP Handle

Cluster to hold the part of the data. Then operate on the new DSP Handle Cluster.

The Memory Look-Up Table (MLUT) has only 128 entries. You can allocate only a total of 128 different

DSP Handle Clusters. Although you might have physical memory on the DSP board, you will get an error

message for not having enough memory if you already have 128 DSP Handle Clusters in use. Free the DSP

Handle Clusters that are not in use. The DSP Init Memory VI will free all DSP Handle Clusters on the

specified DSP board.

NI-DSP SRM for LabVIEW for Windows

1-3

Part 2: Introduction to the NI-DSP Analysis VIs

Introduction to the NI-DSP Analysis VIs

Part 2

Figures 1-4 and 1-5 show how to allocate a DSP Handle Cluster of 2,048 4-byte-long elements on board 3. The

board number on which the buffer is allocated is important for determining the ownership of the buffer. When

making a VI call, the same DSP board on which the function is to execute must own all of the DSP Handle Clusters

or an error code is returned. Only the DSP Allocate Memory VI and few other VIs that do not have DSP Handle

Clusters as input parameters have a board slot parameter. VIs that have DSP Handle Clusters as input parameters

obtain the board slot information from their own DSP Handle Clusters. All of the DSP Handle Clusters should have

the same slot information, because the DSP VIs assume that all are executing on the same DSP board.

Figure 1-4. Front Panel–An Example of How to Allocate a DSP Handle Cluster

Figure 1-5. Block Diagram–An Example of How to Allocate a DSP Handle Cluster

For all of the NI-DSP Analysis VIs, the array data type is DSP Handle Cluster. Before you call any of these VIs,

call the DSP Allocate Memory VI to obtain a valid DSP Handle Cluster, which you then use as a reference to your

data buffer. The Analysis VIs assume that the data is already on the board and stores the results on the board. If you

want to copy data between the PC and the DSP board, use either the Copy Mem(LV to DSP) VI or the Copy

Mem(DSP to LV) VI to copy data back and forth.

If you use the DSP Allocate Memory VI in your program, use the DSP Free Memory VI to free buffers allocated

when you do not need them any more. The board holds these allocations in memory even after your application has

completed or you exit LabVIEW unless you execute the DSP Init Memory VI or reload the DSP Library. Thus, it is

important to free all buffers your application allocated before you exit the application or you may run out of memory

on the board.

Part 2: Introduction to the NI-DSP Analysis VIs

1-4

NI-DSP SRM for LabVIEW for Windows

Part 2

Introduction to the NI-DSP Analysis VIs

Special Features of the NI-DSP Analysis VIs

This section describes the special features of the NI-DSP Analysis VIs that make them different from other

LabVIEW VIs.

•

DSP Handle Cluster in/out. The way you specify the output data buffers for NI-DSP Analysis VIs is different

from the way you would specify output data buffers for other LabVIEW VIs. DSP Handle Clusters also

represent all the output data buffers in the NI-DSP Analysis VIs. To use valid DSP Handle Clusters for the

NI-DSP VI output data buffers, you must use the DSP Allocate Memory VI to obtain the DSP Handle Clusters

before you use them. Supply all of the output DSP Handle Clusters as inputs to tell the DSP board where the

output buffers are. Every output DSP Handle Cluster is identical to the corresponding input DSP Handle

Cluster. The two DSP Handle Clusters are internally connected. The output is an indicator. The input is a

control. For example, in the DSP Add VI, shown in Figure 1-6, Z is the DSP Handle Cluster that indicates

where to store the results. You use Z in to connect to a valid DSP Handle Cluster that you previously allocated.

Z in tells the DSP board where the results will be stored. Z out is the location where the results have already

been stored. Because Z in and Z out are the same DSP Handle Cluster, you need to free only one of them when

you want to deallocate their DSP board buffer.

Figure 1-6. DSP Add VI

All of the controls and indicators in the NI-DSP Analysis VIs follow the Z in, Z out naming convention and

work in the same way as previously described in the example, except for the error in/error out cluster.

•

Error Handling. All of the NI-DSP Analysis VIs have an error input and an error output for managing and

reporting errors. The error in/error out cluster used by the NI-DSP VIs and many other high-level I/O

operations is a cluster containing a Boolean indicating whether the data should be treated as an error, a 32-bit

error code, and a descriptive string that usually contains the name of the source of the error. The error in/error

out cluster is shown in Figure 1-7.

Figure 1-7. The error in/error out Cluster

NI-DSP SRM for LabVIEW for Windows

1-5

Part 2: Introduction to the NI-DSP Analysis VIs

Introduction to the NI-DSP Analysis VIs

Part 2

The error in/error out cluster contains the following elements:

The boolean value is true if an error occurred, false if no error occurred.

code is the error code.

source is the source of the error. If an error occurs during execution, the VI sets source to the name

of the VI that produced the error.

Every VI checks error in first. If there is an error, the VI does not execute any DSP code but simply passes the

contents of error in to the error out cluster. If there is no error, the VI executes. One advantage of this error

in/error out design is that you can connect several I/O operations together so that, if an error occurs, subsequent VIs

do not perform undesired actions. DSP Free Memory will execute even if an error occurs. This ensures that

allocated buffers are freed even if an error occurred.

Another advantage of this error in/error out design is that you can establish the order of a set of operations, even if

there is no other data flow between the operations. Connecting the error out of the first VI to the error in of the

second VI establishes data flow and therefore execution order. You could do the same thing with a Sequence

structure, but with the error in/error out design, you can establish the order with all of the operations at the top level

of the block diagram. For example, in Figure 1-8, you allocate DSP Handle Clusters X and Y as inputs, and you want

to free X and Y after the DSP Add VI has been executed. If you simply connect X to the DSP Free Memory VI as

shown in Figure 1-8, there is no sequential order between the DSP Add VI and the DSP Free Memory VI. If the DSP

Free Memory VI executes first, the DSP Add VI will receive an invalid handle because that DSP Handle Cluster was

deallocated.

Figure 1-8. An Example That Does Not Use error in/error out

for Sequencing VIs

If you connect the VIs as shown Figure 1-9 instead, you ensure that the DSP Add VI executes before the DSP Free

Memory VI.

Part 2: Introduction to the NI-DSP Analysis VIs

1-6

NI-DSP SRM for LabVIEW for Windows

Part 2

Introduction to the NI-DSP Analysis VIs

error out of the Copy Mem(LV to DSP) VI is

connected to the error in of the DSP Free

Memory VI

error out of the DSP Add VI is

connected to the error in of the

Copy Mem(LV to DSP) VI

Figure 1-9. An Example of Using the error in/error out Cluster for Sequential VI Execution

For more information about the error in/error out cluster, refer to Chapter 2, Error Handler VIs, in the LabVIEW

Utility VI Reference Manual.

Hints for Improving the Execution Speed on the DSP Board

Check each of the following things to maximize your DSP board performance:

•

Allocate as many of the DSP Handle Clusters as you can before you operate on the data. Keep all data on the

DSP board until you finish all of the processing. Reduce the number of data transfers between the DSP board

and the PC as much as possible. The functions that run on the DSP board are very fast, but transferring data

between the DSP board and the PC and memory allocation slows the total processing performance.

•

Use the error in/error out cluster for sequencing VI execution. Be sure all of your VIs run in the correct

sequence. Use the error in/error out cluster to propagate the errors. If an error occurs, you can tell where the

error happens. You can use the LabVIEW error handler VIs in the Utility option of the Functions menu to

obtain pop-up error messages. Refer to the LabVIEW Utility VI Reference Manual for more information about

these VIs.

•

Many analysis routines on the DSP board can be performed in place; that is, the input and output array can be

the same array. This is very important to remember when you are processing large amounts of data. Large

32-bit floating-point arrays consume a lot of memory. If the results you want do not require that you keep the

original array or an intermediate array of data, perform analysis operations in place whenever possible. For

example, use the same DSP Handle Cluster for the input and output data buffers in your diagram in LabVIEW.

This will save your DSP board memory.

Several intermediate-level data acquisition VIs work with DSP Handle Clusters. These VIs can acquire data

and leave it on the board. You can use the NI-DSP Analysis VIs to operate on this data and then copy the

processed results back to the PC. In this way, you dramatically reduce the data transfer overhead between the

PC and the DSP board, and improve the overall performance. For more information about these data

acquisition VIs, refer to the LabVIEW Data Acquisition VI Reference Manual. An example that shows you how

to use a DSP Handle Cluster to acquire data and process this data on the DSP board can be found in the

DSP2200subdirectory of the EXAMPLESsubdirectory of your LabVIEWdirectory.

NI-DSP SRM for LabVIEW for Windows

1-7

Part 2: Introduction to the NI-DSP Analysis VIs

Introduction to the NI-DSP Analysis VIs

Part 2

An Example of Using NI-DSP Analysis VIs

Figures 1-10 and 1-11 show the front panel and block diagram, respectively, of an example using NI-DSP Analysis

VIs.

Figure 1-10. Front Panel–An Example of Using NI-DSP Analysis VIs

Figure 1-11. Block Diagram–An Example of Using NI-DSP Analysis VIs

Part 2: Introduction to the NI-DSP Analysis VIs

1-8

NI-DSP SRM for LabVIEW for Windows

Part 2

Introduction to the NI-DSP Analysis VIs

This example shows you how to obtain the power spectrum of a sine wave signal. First, generate a sine wave that

you want to analyze using the LabVIEW Analysis VIs, then use the Copy Mem(LV to DSP) VI to copy the data of

this sine signal to the DSP board. Before you copy the data, you must call the DSP Allocate Memory VI to allocate

a DSP Handle Cluster that reserves a data buffer on the DSP board. Connect this DSP Handle Cluster to the

destination in terminal of the Copy Mem(LV to DSP) VI to indicate where the data will be stored on the DSP

board. After the data is copied to the DSP board, call the DSP Power Spectrum VI to perform a power spectrum on

the data. After you finish the analysis, the results are stored in the data buffer indicated by the DSP Handle Cluster

you previously allocated. If you want to see the results, call the Copy Mem(DSP to LV) VI to copy data back to

LabVIEW. Figure 1-10 shows the power spectrum of a sine wave. The last step is to call the DSP Free Memory VI

to free the DSP Handle Cluster that you allocated. This example connects all of the error out clusters of the

previous VIs to the error in clusters of the subsequent VIs to establish the proper sequence and to pass through an

error, should it occur, without executing the rest of the VIs.

NI-DSP SRM for LabVIEW for Windows

1-9

Part 2: Introduction to the NI-DSP Analysis VIs

Chapter 1

NI-DSP Analysis VI Reference Overview

This chapter contains an overview of the NI-DSP Analysis VIs and includes a list of the VIs. This chapter describes

how the NI-DSP Analysis VIs are organized and how to access them.

The NI-DSP Analysis VI Overview

The NI-DSP Analysis VIs are a set of high-performance VIs that efficiently process large blocks of numerical data

and perform numerically intensive computations. The NI-DSP Analysis VIs include numerical analysis, signal

generation, digital signal processing, digital filtering, and windowing operations that are suitable for simulation,

modeling, and sophisticated data processing.

The NI-DSP Analysis VIs are presented in alphabetical order in Chapter 2, NI-DSP Analysis VI Reference.

Table 1-1 lists these VIs by group.

Table 1-1. The NI-DSP Analysis VI Groups

Signal Generation

DSP Sine Pattern

DSP Pulse Pattern

DSP Impulse Pattern

DSP Impulse Train Pattern

DSP Ramp Pattern

DSP Sinc Pattern

DSP Square Pattern

DSP Triangle Pattern

DSP Triangular Train

DSP Sawtooth Pattern

DSP Uniform White Noise

DSP Random Pattern

DSP Gaussian White Noise

Frequency Domain

DSP ReFFT

DSP Complex FFT

DSP Inverse FFT

DSP Power Spectrum

DSP Cross Power

DSP FHT

DSP Inverse FHT

DSP Zero Padder

Time Domain

DSP Convolution

DSP Deconvolution

DSP Correlation

DSP Decimate

DSP Derivative

DSP Integral

(continues)

NI-DSP SRM for LabVIEW for Windows

1-1

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference Overview

Chapter 1

Table 1-1. The NI-DSP Analysis VI Groups (Continued)

Filters

DSP Butterworth Coefficients

DSP Chebyshev Coefficients

DSP Inverse Chebyshev Coeff

DSP Elliptic Coefficients

DSP IIR Filter

DSP Equi-Ripple LowPass

DSP Equi-Ripple HighPass

DSP Equi-Ripple BandPass

DSP Equi-Ripple BandStop

DSP Parks-McClellan

DSP Median Filter

Windows

DSP Blackman Window

DSP Exact Blackman Window

DSP Blackman Harris Window

DSP Hanning Window

DSP Hamming Window

DSP Flat Top Window

DSP General Cosine Window

DSP Exponential Window

DSP Force Window

DSP Kaiser-Bessel Window

DSP Triangular Window

Array Functions

DSP Add

DSP Subtract

DSP Multiply

DSP Divide

DSP Absolute

DSP Square Root

DSP Product

DSP Sum

DSP Log

DSP Clip

DSP Reverse

DSP Shift

DSP Sort

DSP Linear Evaluation

DSP Max & Min

DSP Polynomial Evaluation

DSP Subset

DSP Set

DSP Unwrap

DSP Polar to Rectangular

DSP Rectangular to Polar

Memory Management

DSP Allocate Memory

Copy Mem(LV to DSP)

Copy Mem(DSP to LV)

Copy Mem(DSP to DSP)

DSP Free Memory

DSP Index Memory

DSP Init Memory

Part 3: NI-DSP Function Reference

1-2

NI-DSP SRM for LabVIEW for Windows

Chapter 1

NI-DSP Analysis VI Reference Overview

Table 1-1. The NI-DSP Analysis VI Groups (Continued)

Utility Functions

DSP Reset

DSP Load

DSP Start

DSP Timeout

DSP Custom

DSP DMA Copy(DSP to LV)

DSP DMA Copy(LV to DSP)

DSP Handle to Address

Analysis VI Organization

After installation, the NI-DSP Analysis VIs reside in the following VI library files (LabVIEW .LLBfiles) within

the DSP2200 option:

•

Signal Generation contains VIs that generate digital patterns.

Frequency Domain contains VIs that perform frequency domain transformations, frequency domain analysis,

and other transforms such as the Hartley transform.

•

Time Domain contains VIs that perform direct time series analysis of signals.

Filters contains VIs that perform IIR and FIR filtering functions.

Windows contains VIs that perform smoothing windowing.

Array contains VIs that perform arithmetic operations on arrays.

Memory Management contains VIs to perform allocating, indexing, copying, and freeing memory on the

AT-DSP2200 board.

•

Utility contains VIs for controlling the operation of the AT-DSP2200 board.

After installation, the eight analysis VI libraries appear in the Functions menu in the order shown in the preceding

list under the DSP2200 option. You can reorganize the folders and the VIs to suit your needs and applications.

Accessing the NI-DSP Analysis VIs

To access the analysis VIs from the block diagram window, choose DSP2200 from the Functions menu as shown in

Figure 1-1, proceed through the hierarchical menus, and select the VI you want. The icon corresponding to that VI

appears in the block diagram and is ready to be wired.

NI-DSP SRM for LabVIEW for Windows

1-3

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference Overview

Chapter 1

Figure 1-1. Choosing DSP2200 from the Functions Menu

About the Fast Fourier Transform (FFT)

The VIs in the Frequency Domain group are based upon the discrete implementation and optimization of the Fourier

Transform integral. The Discrete Fourier Transform (DFT) of a complex sequence X containing n elements is

obtained using the following formula:

n-1

Y[i] = ∑ X[k] * exp (-j2π ik / n), for i = 0, 1, …, n-1

k=0

where Y[i] is the ith element of the DFT of X and j = √-1.

The DFT of X also results in a complex sequence Y of n elements. Similarly, the Inverse Discrete Fourier

Transform (IDFT) of a complex sequence Y containing n elements is obtained using the following formula:

n-1

X[i] = (1/n) ∑ Y[k] * exp (j2π ik / n), for i = 0, 1, …, n-1

k=0

where X[i] is the ith element of the IDFT of Y and j = √-1.

Part 3: NI-DSP Function Reference

1-4

NI-DSP SRM for LabVIEW for Windows

Chapter 1

NI-DSP Analysis VI Reference Overview

The discrete implementation of the DFT is a numerically intense process. However, it is possible to implement a

fast algorithm when the size of the sequence is a power of two. These algorithms are known as FFTs, and can be

found in many introductory digital signal processing (DSP) texts.

The resulting complex FFT sequence has the conventional DSP format as described in this section.

If there are n number of elements in the complex sequence and k = n/2, then the output of the FFT is organized as

follows:

Y[0]

Y[1]

Y[2]

Y[k-1]

Y[k]

Y[k+1]

DC component

Positive first harmonic

Positive second harmonic

Positive k-1 harmonic

Nyquist frequency

Negative k-1 harmonic

Y[n-2]

Y[n-1]

Negative second harmonic

Negative first harmonic

The following conventions and restrictions apply to the VIs in the Frequency Domain folder:

•

All arrays must be a power of two: n = 2 , m = 1, 2, 3, …, 24 (limited by onboard memory).

Complex sequences are manipulated using two arrays. One array represents the real elements. The other array

represents the imaginary elements.

The following notation is used to describe the FFT operations performed in the Frequency Domain class:

•

Y = FFT {X}, the sequence Y is the FFT of the sequence X.

-1

Y = FFT {X}, the sequence Y is the inverse FFT of the sequence X.

X is usually a complex array but can be treated as a real array.

About Filtering

All of the VIs in the Filters group are digital filters that can be represented by the computational algorithm that best

describes the relationship between the input and output discrete time sequences. This computational algorithm is

referred to as the Linear Constant Coefficient Difference Equation. This equation relates the input and output by the

basic operations of addition, delay and multiplication. The following equation relates the input and output

sequences x and y, respectively at the discrete time instant n:

i = N-1

i = M-1

a(0)*y(n) = ∑ (x[n-i]*b(i)) - ∑ (y[n-i]*a(i))

(a)

i = 0

i = 1

where :

is the discrete time input signal to the system represented by the filter

is the discrete time output signal of the system represented by the filter

is the set of coefficients applied to the input in the linear difference equation and represent the

multiplication factors for delays. The equation suggests that there are N such coefficients.

is the set of coefficients applied to the output in the linear difference equation and represent the

multiplication factors for delays. The equation suggests that there are M such coefficients.

NI-DSP SRM for LabVIEW for Windows

1-5

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference Overview

Chapter 1

The set of coefficients a and b are often referred to as the numerator and denominator coefficients, respectively.

Another common way to refer to them is as the feedforward and feedback coefficients. This is due to the

mathematical derivation that led to equation a. Refer to Discrete-Time Signal Processing by Oppenheim and

Schafer for more information. Another frequent assumption is that a(0)=1.0. For example, let us assume M = 4,

N = 4 and n = 2. Using the filtering equation produces:

a(0)*y(2) = b(0)*x(2) + b(1)*x(1) + b(2)* x(0) + b(3)* x(-1) - a(1)*y(1) - a(2)*y(0) - a(3)*y(-1)

(b)

The filtering equation suggests that in order to compute the value of the output at n = 2, you not only need the

coefficients that represent the filter that is producing this output sequence, but also the value of the current input and

output, the values of the input and output one time unit ago, the values of the input and output two time units ago,

and the values of the input and output three time units ago. You need some “history” about the previous outputs of

the filter as well as the previous inputs to the filter. The amount of history (number of previous output and input

samples) depends on the lengths of the arrays a and b (filter coefficients arrays).

At time n = 0, it is important to note that the filtering equation becomes:

a(0)*y(0) = b(0)*x(0) + b(1)*x(-1) + b(2)* x(-2) + b(3)* x(-3) - a(1)*y(-1) - a(2)*y(-2) - a(3)*y(-3)

(c)

Thus, for the function filter to properly operate as of time n = 0, you need to supply some history about previous

behavior. The filter function then updates the history as time goes on, keeping track of previous input values and

corresponding outputs. This history, at time n = 0, is referred to as the initial conditions on the input and output of

the filter.

Digital filters fall into two classes–Infinite Impulse Response filters (IIR filters) and Finite Impulse Response filters

(FIR filters). Notice that IIR filters are represented by equation (a) while the FIR filters can be represented by the

same equation provided all a's are zero except for a(0) as shown in the following equation:

i = N

a(0)*y(n) = ∑ (x[n-i]*b(i))

(d)

i = 0

The NI-DSP Analysis VIs have a set of VIs that implement IIR and FIR filters. Because all digital filters are

approximations of their analog design counterparts, there are several techniques for designing a digital filter.

For the IIR filter design, the NI-DSP Analysis VIs have four approaches representing four different techniques of

obtaining digital filter specifications (coefficients to equation (a))–Butterworth, Chebyshev, inverse Chebyshev, and

elliptic techniques. With each design technique, you can obtain the coefficients for lowpass, highpass, bandpass,

and bandstop filters from the respective NI-DSP Analysis VIs.

For the FIR filters, the NI-DSP Analysis VIs allow the design of a multiband FIR linear phase filter using the Parks-

McClellan algorithm. The frequency response in each band has equal ripples that can be adjusted by a weighting

factor. For more information, please refer to Digital Filter Design by Parks and Burrus, or "A Computer Program

for Designing Optimum FIR Linear Phase Digital Filters," by McClellan, Parks, and Rabiner.

The IIR VIs generate filter coefficients that represent the a's and b's in the equations a and b. No filtering is actually

performed. The IIR filter design coefficients all have a(0) = 1.0. You can use the general-purpose VI that accepts

filter coefficients, initial conditions on the input and output sequences, and an input sequence to filter any of the

filter specifications. This VI filters the input sequence and provides the final conditions on the output and input of

the filter.

About Windowing

Almost every application requires you to use finite length signals. This requires that continuous signals be

truncated, using a process called windowing.

Part 3: NI-DSP Function Reference

1-6

NI-DSP SRM for LabVIEW for Windows

Chapter 1

NI-DSP Analysis VI Reference Overview

The simplest window is a rectangular window. Because this window requires no special effort it is commonly

referred to as the no window option. Remember, however, that a discrete signal and its spectrum is always affected

by a window. Let x[n] be a digitized time-domain waveform that has a finite length of n. w[n] is a window

sequence of n points. The windowed output is calculated as follows:

y[i] = x[i] * w[i]

(1)

If X, Y, and W are the spectra of x, y, and w, respectively, the time-domain multiplication in formula 1 is equivalent

to the convolution shown as follows:

Y[k] = X[k] Θ W[k].

(2)

Convolving with the window spectrum always distorts the original signal spectrum in some way. A window

spectrum consists of a big mainlobe and several sidelobes.

The mainlobe is the primary cause of lost frequency resolution. When two signal spectrum lines are too close to

each other, they may fall in the width of the mainlobe, causing the output of the windowed signal spectrum to have

only one spectrum line. Use a window with a narrower mainlobe to reduce the loss of frequency resolution. It has

been shown that a rectangular window has the narrowest mainlobe, so that it provides the best frequency resolution.

The sidelobes of a window function affect frequency leakage. A signal spectrum line will leak into the adjacent

spectrum if the sidelobes are large. Once again, the leakage results from the convolution process. Select a window

with relatively smaller sidelobes to reduce spectral leakage. Unfortunately, a narrower mainlobe and smaller

sidelobes are mutually exclusive. For this reason, selecting a window function is application dependent. An

example of a windowed spectrum in the continuous case is shown in Figure 1-2.

Signal Spectrum

Window Spectrum

Windowed Signal Spectrum

Figure 1-2. Spectral Leakage Demonstrated Using Convolution

The original signal spectrum is convolved with the window spectrum and the output is a smeared version of the

original signal spectrum. In this example, you can still see four distinctive peaks from the original signal, but each

peak is smeared and the frequency leakage effect is clear.

NI-DSP SRM for LabVIEW for Windows

1-7

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference Overview

Chapter 1

Window definitions used in National Instruments analysis libraries are designed in such a way that the window

operations in the time domain are exactly equivalent to the operations of the same window in the frequency domain.

To meet this requirement, the windows are not symmetrical in the time domain, that is:

w[0] ≠ w[N-1]

(3)

where N is the window length. They are usually symmetrical in the frequency domain, however. For example, the

Hamming window definition uses the formula:

w[i] = 0.54 - 0.46 cos(2πi/N)

(4)

Other manufacturers may use a slightly different definition, such as:

w[i] = 0.54 - 0.46 cos(2πi/N-1)

(5)

The difference is small if N is large.

Formula 4 is not symmetrical in the time domain, but it ensures that the time domain windowing is equivalent to the

frequency domain windowing. If you want to have a perfectly symmetrical sequence in the time domain, you must

write your own windowing function using formula 5.

The choice of a window depends on the application. For most applications, the Hamming or Hanning windows

deliver good performance.

Part 3: NI-DSP Function Reference

1-8

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

This chapter contains a brief explanation of each NI-DSP Analysis VI . The VIs are arranged alphabetically.

Copy Mem(DSP to DSP)

Copies a buffer of data from the Source

buffer on the DSP board that is referred to

by a DSP Handle Cluster to the destination

buffer on the DSP board, which is referred

to by another DSP Handle Cluster. Source

and destination buffers should be on the

same DSP board.

To copy data correctly from one DSP

buffer to another DSP buffer, you must set the data type to the appropriate type to indicate what kind of data is on

the DSP board. The VI has three data types–32-bit floating-point data, 16-bit short integer data, and 32-bit long

integer data.

Source is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the

data that you want to copy to the destination.

data type indicates the type of data that Source contains. The VI has three data types:

0: 32-bit floating-point data.

1: 16-bit short integer data.

2: 32-bit long integer data.

data type defaults to 32-bit floating-point data.

Destination in is a DSP Handle Cluster that indicates the destination memory buffer on the DSP board

that will contain the data copied from source buffer.

Destination out is a DSP Handle Cluster that is identical to the Destination in, but with the source

buffer data already copied to the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-1

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

Copy Mem(DSP to LV)

Copies an entire or partial buffer

of data according to the

entire/partial copy selector from

the Source buffer on the DSP

board that is referred to by a DSP

Handle Cluster to the destination

buffer in LabVIEW.

To copy data correctly from the

DSP board to LabVIEW, you

must indicate what type of data is

stored in the Source buffer on the DSP board. You must set the destination type to the appropriate type and wire to

the corresponding destination terminal. The VI has three destination types–float (32-bit), short (16-bit), and long

(32-bit). To copy different types of data, you must wire to the appropriate destination terminal that corresponds to

the destination type you choose. Remember, you must wire only one kind of destination terminal.

If you set the entire/partial copy selector to entire copy, all the data in the Source buffer will be copied back to the

destination without considering the values of offset and size. Otherwise, only the number of size the data in the

Source buffer, starting from the offset, will be copied back to the destination.

entire/partial copy selector has two types: entire copy and partial copy. It defaults to entire copy.

0: entire copy.

1: partial copy.

Source is a DSP Handle Cluster that indicates the memory buffer on the DSP board which contains the

data that you want copy to the destination in LabVIEW.

destination type indicates the type of the destination data on the DSP board. It has three options:

0: 32-bit floating-point.

1: 16-bit short integer.

2: 32-bit long integer.

destination type defaults to 32-bit floating point.

offset indicates where to start to copy data from the Source buffer on the DSP board to the destination

buffer in the LabVIEW. offset defaults to 0. This parameter is ignored when entire/partial copy

selector is set to entire copy.

size indicates how much data to copy from the Source buffer on the DSP board to the destination buffer

in LabVIEW. size defaults to 0. If offset plus size is greater than the number of elements in the Source

buffer, the VI returns an error. This parameter is ignored when entire/partial copy selector is set to

entire copy.

destination(float) is the terminal to which you should wire the destination buffer if the data you want to

copy is 32-bit floating-point data array.

destination(long) is the terminal to which you should wire the destination buffer if the data you want to

copy is 32-bit long data array.

destination(short) is the terminal to which you should wire the destination buffer if the data you want to

copy is 16-bit short data array.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-2

NI-DSP SRM for LabVIEW for Windows

NI-DSP Analysis VI Reference

Chapter 2

Copy Mem(LV to DSP)

Copies a buffer of data from the source

buffer in LabVIEW to the destination

buffer on the DSP board, which is referred

to by a DSP Handle Cluster.

The source buffer can contain one of three

kinds of data–float (32-bit), short

(16-bit), and long (32-bit). To copy

different types of data, you must wire to the

appropriate source terminal. You must

wire only one kind of source terminal. The

destination buffer must be large enough to contain all of the data from the source buffer.

source(float) is the terminal to which you should wire the source buffer if the data you want to copy is

32-bit floating point data array.

source(long) is the terminal to which you should wire the source buffer if the data you want to copy is

32-bit long data array.

source(short) is the terminal to which you should wire the source buffer if the data you want to copy is

16-bit short data array.

Destination in is a DSP Handle Cluster that indicates the destination memory buffer on the DSP board

that will contain the data copied from the LabVIEW source buffer.

Destination out is a DSP Handle Cluster that is identical to Destination in, but with the source buffer

data already copied to the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-3

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Absolute

Find the absolute value of input array X. The i

element of the output array Y is obtained using

the following formula:

Y(i) = |X(i) |.

for i = 0, 1, 2, … , n-1

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

absolute results of X(i).

Y out is a DSP Handle Cluster that is identical to Y in, but with the absolute results already stored in the

memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input and output arrays can be the same DSP Handle Cluster.

NI-DSP SRM for LabVIEW for Windows

2-4

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Add

Add array X to array Y. The i element of the

output array Z is obtained using the following

formula:

Z(i) = X(i) + Y(i).

for i = 0, 1, 2, … , n-1

where n is the smaller number of elements in X and Y.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array Y.

Z in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of X(i) + Y(i).

Z out is a DSP Handle Cluster that is identical to Z in, but with the added results already stored in the

memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input and output arrays can be the same DSP Handle Cluster.

Part 3: NI-DSP Function Reference

2-5

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Allocate Memory

Allocates a block of memory buffer on

the DSP board specified by slot and

returns a DSP Handle Cluster that

contains the coded DSP board memory

and the number of elements in this

buffer.

The number of bytes allocated for this

buffer depends on size and

bytes/element selector. If bytes/element selector selects 4 bytes, then the number of bytes = size*4. If

bytes/element selector selects 2 bytes, then the number of bytes = size*2.

The allocation routines on the board assure alignment to the nearest 4-byte boundary ≥ number of bytes to guarantee

memory alignment on 4-byte addresses. For example, if you request to allocate a buffer of size 1,022 or 1,024 bytes,

the allocation routines allocate 1,024 bytes in both cases.

slot is the board ID number. slot defaults to 3.

size is the number of elements to allocate for this buffer. size defaults to 0.

bytes/element selector specifies the number of bytes per element. It has two options:

0: 4 bytes.

1: 2 bytes.

bytes/element selector defaults to 4 bytes.

DSP Handle Cluster indicates the allocated memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-6

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Blackman Window

Applies a Blackman window to the input

sequence X. If Y represents the output

sequence Blackman{X}, the elements of Y

are obtained from the following formula:

y_i= x_i[0.42 - 0.50 cos(w)

+ 0.08 cos(2w)]

for i = 0, 1, 2, … , n-1 ,

2πi

w =

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output

Blackman {X}.

Blackman {X} is a DSP Handle Cluster that is identical to X, but with the results of Blackman {X}

already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-7

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Blackman Harris Window

Applies a Blackman Harris window to

the input sequence X. If Y represents

the output sequence Blackman

Harris{X}, the elements of Y are

obtained using the following formula:

y_i= x_i[0.42323 - 0.49755 cos(w)

+0.07922 cos(2w)]

for i = 0, 1, 2, … , n-1

2πi

w =

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output

Blackman Harris {X}.

Blackman Harris {X} is a DSP Handle Cluster that is identical to X, but with the results of Blackman

Harris {X} already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-8

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Butterworth Coefficients

Generates the set of filter

coefficients to implement an

IIR filter as specified by the

Butterworth filter model.

You can then pass these

coefficients to the DSP IIR

Filter VI to filter a sequence

of data.

filter type specifies the passband of the filter. It has four options:

0: lowpass.

1: highpass.

2: bandpass.

3: bandstop.

filter type defaults to lowpass.

sampling freq : fs is the sampling frequency and must be greater than 0. If it is less than or equal to

zero, the VI returns an error. sampling freq : fs defaults to 1.0.

high cutoff freq: fh is the high cutoff frequency. The VI ignores this parameter when filter type is

lowpass. high cutoff freq: fh defaults to 0.45.

low cutoff freq: fl is the low cutoff frequency. The VI ignores this parameter when filter type is

highpass. low cutoff freq: fl defaults to 0.125.

Note: fh and fl must observe the Nyquist criterion: 0 ≤ fl ≤ fh ≤ fs /2.

Forward Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that will contain the forward coefficients.

Feedback Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that will contain the feedback coefficients.

Forward Coefficients out is a DSP Handle Cluster that is identical to the Forward Coefficients in, but

with the forward coefficients already stored in the memory buffer on the DSP board.

Feedback coefficients out is aDSP Handle Cluster that is identical to the Feedback Coefficients in, but

with the feedback coefficients already stored in the memory buffer on the DSP board.

order must be greater than zero. If order is less than or equal to zero, the VI returns an error. order

defaults to 2.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-9

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

Parameter Discussion

The arrays Forward Coefficients in/out and Feedback Coefficients in/out must have a size of at least (order + 1)

for lowpass and highpass filters. The arrays Forward Coefficients in/out and Feedback Coefficients in/out must

have a size of at least (2*order + 1) for bandpass and bandstop filters.

DSP Chebyshev Coefficients

Generates the set of filter

coefficients to implement an

IIR filter as specified by the

Chebyshev filter model.

You can then pass

coefficients to the DSP IIR

Filter VI to filter a sequence

of data.

filter type specifies the passband of the filter. It has four options:

0: lowpass.

1: highpass.

2: bandpass.

3: bandstop.

filter type defaults to lowpass.

sampling freq : fs is the sampling frequency and must be greater than 0. If it is less than or equal to

zero, the VI returns an error. sampling Freq : fs defaults to 1.0.

high cutoff freq: fh is the high cutoff frequency. The VI ignores this parameter when filter type is

lowpass. high cutoff freq: fh defaults to 0.45.

low cutoff freq: fl is the low cutoff frequency. The VI ignores this parameter when filter type is

highpass. low cutoff freq: fl defaults to 0.125.

Note: fh and fl must observe the Nyquist criterion: 0 ≤ fl ≤ fh ≤ fs /2.

ripple must be greater than 0. If ripple is less than or equal to zero, the VI returns an error. ripple

defaults to 0.1.

order must be greater than zero. If order is less than or equal to zero, the VI returns an error. order

defaults to 2.

Forward Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that will contain the forward coefficients.

Feedback Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that will contain the feedback coefficients.

Forward Coefficients out is a DSP Handle Cluster that is identical to the Forward Coefficients in, but

with the forward coefficients already stored in the memory buffer on the DSP board.

NI-DSP SRM for LabVIEW for Windows

2-10

Part 3: NI-DSP Function Reference

Chapter 2

NI-DSP Analysis VI Reference

Feedback Coefficients out is a DSP Handle Cluster that is identical to the Feedback Coefficients in,

but with the feedback coefficients already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Parameter Discussion

The arrays Forward Coefficients in/out and Feedback Coefficients in/out have to have a size of at least (order +

1) for lowpass and highpass filters.The arrays Forward Coefficients in/out and Feedback Coefficients in/out have

to have a size of at least (2*order + 1) for bandpass and bandstop filters.

DSP Clip

Clips the input array values. The range of

the resulting output array is [lower:upper].

Let Y represent the output array. The i

element of the resulting array is obtained by

using the following formulas:

upper if X(i) > upper

lower if X(i) < lower

Y(i) =

X(i) otherwise

for i = 0, 1, 2, …, size-1 ,

where n is the number of elements in X.

The operation can be performed in place; that is, the input and output arrays can be the same DSP Handle Cluster.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

upper limit. upper limit ≥ lower limit. upper limit defaults to 1.0.

lower limit. lower limit defaults to 0.0.

Clipped{X} in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output array of Clipped{X}.

Clipped{X} out is a DSP Handle Cluster that is identical to Clipped {X} in, but with the Clipped{X}

already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-11

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Complex FFT

Computes the Fast Fourier transform of

the complex input sequence X. If Y

represents the complex output sequence,

then:

Y = F{X}.

Re{X} is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the real

part of the input signal array X.

Im{X} is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the

imaginary part of input signal array X.

Notes: The number of elements for the input array must be a power of 2.

If the size of Re{X} is different from the size of Im{X}, the VI uses the smaller number as the

size of the Complex FFT.

The operation is performed in place and the input array Re{X} and Im{X} is overwritten by the

Re FFT{X} and Im FFT{X}.

The largest complex FFT that can be computed depends upon the amount of memory on your

board.

Re FFT{X} is a DSP Handle Cluster that is identical to Re{X}, but with the real part of FFT{X}already

stored in the memory buffer on the DSP board.

Im FFT{X} is a DSP Handle Cluster that is identical to Im{X}, but with the imaginary part of

FFT{X}already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-12

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Convolution

Computes the convolution of the input

sequences X and Y. The convolution Cxy(t), of

the signals x(t) and y(t), is defined as follows:

∞

Cxy(t) = x(t) y(t) = ∫ x(τ) y(t-τ) dτ,

-∞

where the symbol denotes convolution.

For the discrete implementation of the convolution, let Cxy represent the output sequence X * Y, n be the number of

elements in the input sequence X, and m be the number of elements in the input sequence Y. Assuming that indexed

elements of X and Y that lie outside their range are zero,

x_i= 0,

y_j= 0,

i < 0 or i ≥ n

and

j < 0 or j ≥ m,

then you obtain the elements of Cxy using:

n-1

Cxy =

x y

for i = 0, 1, 2, …, size-1 ,

∑

k i-k

k = 0

size = n + m - 1,

where size is the total number of elements in the output sequence X * Y.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array Y.

Cxy in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

convolution of X with Y.

Note: The size of Cxy in must be (n+m-1) elements long. n is the size of X, m is the size of Y. You

cannot perform the operation in place.

Cxy out is a DSP Handle Cluster that is identical to the Cxy in but with the convolution of X and Y

already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-13

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Correlation

Computes the cross correlation of the input

sequences X and Y. The cross correlation R_xy(t)

of the signals x(t) and y(t) is defined as follows:

∞

R_xy(t) = x(t) ⊗ y(t) = x(t) y(t+t) dt ,

∫

-∞

where the symbol ⊗ denotes correlation.

For the discrete implementation of the correlation, let R_xyrepresent the output sequence X ⊗ Y, n be the number of

elements in the input sequence X, and m be the number of elements in the input sequence Y. You then obtain the

elements of R_xyusing the following formula:

n-1

x[k] y[k+m-1]

for i = 0, 1, 2, …, m+n-1 ,

∑

xy i

k=0

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array Y.

Rxy in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

correlation results.

Note: The size of Rxy in must be at least (n+m-1) elements long. n is the size of X, m is the size

of Y.

Rxy out is a DSP Handle Cluster that is identical to the Rxy in, but with the correlation of X and Y

already stored in the memory buffer on the DSP board.

Note: If X and Y are the same array, an auto correlation is performed, otherwise, a cross correlation is

performed. You cannot perform the operation in place.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-14

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Cross Power

Computes the cross power spectrum of the input

sequences X and Y. The cross power, S_xy(f), of

the signals x(t) and y(t) is defined as follows:

S_xy(f) = X^*(f)Y(f)

where X (f) is the complex conjugate of X(f),

X(f) = F{x(t)}, and

Y(f) = F{y(t)}.

This VI uses the FFT routine to compute the cross power spectrum, that is given by the following formula:

F {X} F{Y} ,

n = 2

for k = 1, 2, 3, …, 23,

where S represents the complex output sequence Sxy, and

n is the number of samples that can accommodate both input sequences X and Y.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array Y.

Notes: If the size of X is different than Y, the VI uses the smaller number as the size of input arrays X

and Y. The number of elements for the input arrays must be a power of two.

The operation is performed in place and the input arrays X and Y are overwritten by the

outputs Re{Sxy} and Im{Sxy}.

The largest Cross Power Spectrum that can be computed depends upon the amount of memory

in your DSP board.

This VI allocates a temporary workspace on the DSP board equal to the size of the input signal

array.

Re{Sxy} is a DSP Handle Cluster that is identical to X, but with the real part of Sxy already stored in the

memory buffer on the DSP board.

Im{Sxy} is a DSP Handle Cluster that is identical to Y, but with the imaginary part of Sxy already

stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-15

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Custom

Use this VI as the interface to

call your own custom functions

written on the DSP board from

LabVIEW. For more details

about how to use this VI, refer

to Part 4, NI-DSP Interface

Utilities, of this manual.

slot is the board ID number. slot defaults to 3.

function code indicates which function on the DSP board LabVIEW is calling. function code defaults

to 0.

Array of Handles in is an array of DSP Handle Clusters that hold all of the references to the input and

output data buffers used in your custom functions on the DSP board.

Array of SGL Scalars in is an array of 32-bit floating-point scalars that hold all of the input 32-bit

floating-point scalars used in your custom functions on the DSP board.

Array of I32 Scalars in is an array of 32-bit long integer scalars that hold all of the input 32-bit long

integer scalars used in your custom functions on the DSP board.

Array of Handles out is an array of DSP Handle Clusters that is identical to Array of Handles in, but

with the output results already stored in the memory buffers on the DSP board.

Array of SGL Scalars out is an array of 32-bit floating-point scalars that is identical to Array of SGL

Scalars in, but with the output results already stored in the array.

Array of I32 Scalars out is an array of 32-bit long integer scalars that is identical to Array of I32

Scalars in, but with the output results already stored in the array.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-16

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Decimate

Decimates the input sequence X by the

decimating factor and the averaging control.

If Y represents the output sequence Decimated

Array, the elements of the sequence Y are

obtained using:

X_im

if ave = no averaging

m−1

Y_i=

if ave = averaging

∑

i(m+k)

m _k=0

for i = 0, 1, 2, …, size -1

size = trunc

(_m)

where n is the number of elements in X,

m is the decimating factor,

ave is the averaging option, and

size is the number of elements in the output sequence Y.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

decimated output array.

Y out is a DSP Handle Cluster that is identical to Y in, but with the results of the decimated array

already stored in the memory buffer on the DSP board.

decimating factor must be greater than zero:

0 < decimating factor ≤ n.

If decimating factor is greater than the number of samples in X or less than or equal to zero, the VI

returns an error. decimating factor defaults to 1.

averaging has two options:

0: averaging.

1: no averaging.

averaging defaults to no averaging.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input X and the output Y can be the same DSP Handle Cluster.

NI-DSP SRM for LabVIEW for Windows

2-17

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Deconvolution

Computes the deconvolution of the input

sequences X and Y. The convolution operation

can be realized using Fourier identities because

x(t) y(t) ⇔ X(f) Y(f)

is a Fourier transform pair, where the symbol

denotes convolution, and the deconvolution is

the inverse of the convolution operation. If h(t) is the signal resulting from the deconvolution of the signals x(t) and

y(t), the VI obtains h(t) using the equation

X(f)

{_Y(f)}

-1

h(t) = F

where X(f) is the Fourier transform of x(t), and

Y(f) is the Fourier transform of y(t).

The VI performs the discrete implementation of the deconvolution using the following steps.

1. Compute the Fourier transform of the input sequence X * Y.

2. Compute the Fourier transform of the input sequence Y.

3. Divide the Fourier transform of X * Y by the Fourier transform of Y. Call the new sequence H.

4. Compute the inverse Fourier transform of H to obtain the deconvoluted sequence X.

Cxy is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array Cxy. The number of elements in Cxy must be greater than or equal to the number of

elements in Y.

Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array Y.

X in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

deconvolution results.

Note: The size of X in must be n elements long, although only (n-m+1) elements are valid. n is the

size of Cxy, m is the size of Y.

X out is a DSP Handle Cluster that is identical to X in, but with the deconvolution results already stored

in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Note: The deconvolution operation is a numerically unstable operation, and it is not always possible to solve the

system numerically. Computing the deconvolution via FFTs is perhaps the most stable generic algorithm

that does not require sophisticated DSP techniques. However, it is not free of errors (for example, when

there are zeros in the Fourier transform of the input sequence Y).

Part 3: NI-DSP Function Reference

2-18

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Derivative

Performs a discrete differentiation of the

sampled signal X. The differentiation f(t) of a

function F(t) is defined as follows:

f(t) =

F(t).

Let Y represent the sampled output sequence

d/dt X. The discrete implementation is given

by the following formula:

2dt

y_i=

(x_i+1- x_i-1)

for i = 0, 1, 2, …, n-1.

where n is the number of samples in x(t)

x_-1is specified by initial condition when i = 0, and

x_nis specified by final condition when i = n-1.

initial condition and final condition minimize the error at the boundaries.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

initial condition defaults to 0.0.

final condition defaults to 0.0.

dt is the sampling interval and must be greater than zero. If dt is less than or equal to zero, the VI

returns an error. dt defaults to 1.0.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of the differentiation of X.

Y out is a DSP Handle Cluster that is identical to Y in, but with the results of differentiation already

stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input X and the output Y can be the same DSP Handle Cluster.

NI-DSP SRM for LabVIEW for Windows

2-19

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Divide

Divide array X by array Y. The i element of

the output array Z is obtained using the

following formula:

Z(i) = X(i) / Y(i).

for i = 0, 1, 2, …, n-1.

where n is the smaller number of elements in

X and Y.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array Y.

Z in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of X(i) / Y(i).

Z out is a DSP Handle Cluster that is identical to Z in, but with the divided results already stored in the

memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input and the output arrays can be the same DSP Handle

Cluster.

Part 3: NI-DSP Function Reference

2-20

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Elliptic Coefficients

Generates the set of filter

coefficients to implement a

digital elliptic IIR filter. You

can then pass these

coefficients to the DSP IIR

Filter VI.

filter type specifies the passband of the filter. filter type has four options:

0: lowpass.

1: highpass.

2: bandpass.

3: bandstop.

filter type defaults to lowpass.

sampling freq : fs is the sampling frequency and must be greater than 0. If it is less than or equal to

zero, the VI returns an error. sampling freq : fs defaults to 1.0.

high cutoff freq: fh is the high cutoff frequency. The VI ignores this parameter when filter type is

lowpass. high cutoff freq: fh defaults to 0.45.

low cutoff freq: fl is the low cutoff frequency. The VI ignores this parameter when filter type is

highpass. high cutoff freq: fh defaults to 0.125.

Note: fh and fl must observe the Nyquist criterion: 0 ≤ fl ≤ fh ≤ fs /2.

passband ripple must be greater than 0, and you must express it in decibels. If passband ripple is less

than or equal to zero, the VI returns an error. passband ripple defaults to 1.0.

order must be greater than zero. If order is less than or equal to zero, the VI returns an error. order

defaults to 2.

stopband attenuation must be greater than 0, and you must express it in decibels. If stopband

attenuation is less than or equal to zero, the VI returns an error. stopband attenuation defaults to 60.0.

Forward Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that will contain the forward coefficients.

Feedback Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that will contain the feedback coefficients.

Forward Coefficients out is a DSP Handle Cluster that is identical to Forward Coefficients in, but

with the forward coefficients already stored in the memory buffer on the DSP board.

Feedback Coefficients out is a DSP Handle Cluster that is identical to Feedback Coefficients in, but

with the feedback coefficients already stored in the memory buffer on the DSP board.

NI-DSP SRM for LabVIEW for Windows

2-21

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Parameter Discussion

The arrays Forward Coefficients in/out and Feedback Coefficients in/out must have a size of at least (order + 1)

for lowpass and highpass filters. The arrays Forward Coefficients in/out and Feedback Coefficients in/out must

have a size of at least (2*order + 1) for bandpass and bandstop filters.

Part 3: NI-DSP Function Reference

2-22

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Equi-Ripple BandPass

Generates a bandpass FIR filter

with equi-ripple characteristics

using the Parks-McClellan

algorithm and the number of taps,

lower stop frequency, higher stop

frequency, lower pass frequency,

higher pass frequency, and

sampling frequency. The VI then

filters the input sequence X to

obtain the bandpass filtered linear-

phase sequence Filtered Data.

The first stopband of the filter

region goes from zero (DC) to the

lower stop frequency. The passband region goes from the lower pass frequency to the higher pass frequency, and

the second stopband region goes from the higher stop frequency to the Nyquist frequency.

higher pass freq must be greater than lower pass freq frequency. If higher pass freq is less than or

equal to lower pass freq, the VI returns an error. higher pass freq defaults to 0.3.

lower pass freq must be greater than the lower stop freq. If lower pass freq is less than or equal to

lower stop freq, the VI returns an error. lower pass freq defaults to 0.2.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

data array.

# of taps must be greater than 0. If the number of taps is less than or equal to zero, the VI returns an

error. # of taps defaults to 32.

lower stop freq must be greater than zero. If lower stop freq is less than or equal to zero, the VI

returns an error. lower stop freq defaults to 0.1.

higher stop freq must be greater than higher pass freq and must observe the Nyquist criterion:

0 < f₀< f₁< f₂< f₃≤ 0.5f_s,

where f₀is lower stop freq, f₁is lower pass freq, f₂is higher pass freq, f₃is the higher stop freq, and

f_sis the sampling frequency. If any of these conditions is violated, the VI returns an error. higher stop

freq defaults to 0.4.

sampling freq: fs defaults to 1.0.

Filtered Data in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the filtered data output.

Filtered Data out is a DSP Handle Cluster that is identical to Filtered Data in, but with the filtered

data already stored in the memory buffer on the DSP board. Because the VI filters via convolution, the

number of elements, k, in Filtered Data is as follows:

k = n + m - 1,

where n is the number of elements in X, and m is the number of taps.

NI-DSP SRM for LabVIEW for Windows

2-23

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

A delay is also associated with the output sequence

m-1

delay =

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation cannot be performed in place; that is, the input X and the output Filtered Data cannot be the same

DSP Handle Cluster.

Part 3: NI-DSP Function Reference

2-24

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Equi-Ripple BandStop

Generates a bandstop FIR digital filter

with equi-ripple characteristics using the

Parks-McClellan algorithm and number

of taps, lower pass frequency, higher

pass frequency, lower stop frequency,

higher stop frequency, and sampling

frequency. The VI then filters the input

sequence X to obtain the bandstop

filtered linear-phase sequence Filtered

Data.

The first passband region of the filter

goes from zero (DC) to the lower pass

frequency. The stopband region goes

from the lower stop frequency to the higher stop frequency, and the second passband region goes from the higher

pass frequency to the Nyquist frequency.

higher pass freq must be greater than higher stop freq and observe the Nyquist criterion:

0 < f₀< f₁< f₂< f₃≤ 0.5f_s,

where f₀is the lower pass freq, f₁is the lower stop freq, f₂is the higher stop freq, f₃is the higher pass

freq, and f_sis the sampling freq. If any of these conditions is violated, the VI returns an error. higher

pass freq defaults to 0.4.

lower pass freq must be greater than zero. If lower pass freq is less than or equal to 0, the VI returns

an error. lower pass freq defaults to 0.1.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

data array.

# of taps must be odd and must be greater than 0. If the number of taps is an even number or is less

than or equal to zero, the VI returns an error. # of taps defaults to 31.

lower stop freq must be greater than lower pass freq. If lower stop freq is less than or equal to lower

pass freq, the VI returns an error. lower stop freq defaults to 0.2.

higher stop freq must be greater than lower stop freq. If higher stop freq is less than or equal to

lower stop freq, the VI returns an error. higher stop freq defaults to 0.3.

sampling freq: fs defaults to 1.0.

Filtered Data in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the filtered data output.

Filtered Data out is a DSP Handle Cluster that is identical to Filtered Data in, but with the filtered

data already stored in the memory buffer on the DSP board. Because the VI filters via convolution, the

number of elements, k, in Filtered Data is as follows:

k = n + m - 1,

where n is the number of elements in X, and m is the number of taps.

NI-DSP SRM for LabVIEW for Windows

2-25

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

A delay is also associated with the output sequence:

m-1

delay =

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation cannot be performed in place; that is, the input X and the output Filtered Data cannot be the same

DSP Handle Cluster.

Part 3: NI-DSP Function Reference

2-26

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Equi-Ripple HighPass

Generates a highpass FIR filter with

equi-ripple characteristics using the

Parks-McClellan algorithm and the

number of taps, high frequency, stop

frequency, and sampling frequency. The

VI then filters the input sequence X to

obtain the highpass filtered linear-phase

sequence Filtered Data.

The stopband of the filter goes from zero

(DC) to the stop frequency. The

transition band goes from the stop

frequency to the high frequency, and

the passband goes from the high frequency to the Nyquist frequency.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

data array.

# of taps must be greater than 0 and must be an odd number. If the number of taps is an even number or

is less than or equal to zero, the VI returns an error. # of taps defaults to 31.

stop freq must be greater than zero. If stop freq is less than or equal to 0, the VI returns an error. stop

freq defaults to 0.2.

high freq must be greater than stop freq and observe the Nyquist criterion:

0 < f₀< f₁≤ 0.5f_s,

where f₀is the stop freq, f₁is the high freq, and f_sis the sampling frequency. If any of these conditions

is violated, the VI returns an error. high freq defaults to 0.3.

sampling freq: fs defaults to 1.0.

Filtered Data in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the filtered data output.

Filtered Data out is a DSP Handle Cluster that is identical to Filtered Data in, but with the filtered

data already stored in the memory buffer on the DSP board. Because the VI filters via convolution, the

number of elements, k, in Filtered Data is:

k = n + m - 1,

where n is the number of elements in X, and m is the number of taps.

A delay is also associated with the output sequence:

m-1

delay =

NI-DSP SRM for LabVIEW for Windows

2-27

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation cannot be performed in place; that is, the input X and the output Filtered Data out cannot be the

same DSP Handle Cluster.

Part 3: NI-DSP Function Reference

2-28

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Equi-Ripple LowPass

Generates a lowpass FIR filter with equi-

ripple characteristics using the Parks-

McClellan algorithm and the number of

taps, pass frequency, stop frequency, and

sampling frequency. The VI then filters

the input sequence X to obtain the

lowpass filtered linear-phase sequence

Filtered Data.

The passband of the filter goes from zero

(DC) to pass freq. The transition band

goes from pass freq to stop freq, and

the stopband goes from stop freq to the

Nyquist frequency.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

data array.

# of Taps must be greater than 0. If the number of taps is less than or equal to zero, the VI returns an

error. # of Taps defaults to 31.

pass freq must be greater than 0. If pass freq is less than or equal to zero, the VI returns an error. pass

freq defaults to 0.2.

stop freq must be greater than the pass freq and observe the Nyquist criterion:

0 < f₀< f₁≤ 0.5f_s,

where f₀is the pass freq, f₁is the stop freq, and f_sis the sampling frequency. If any of these conditions

is violated, the VI returns an error. stop freq defaults to 0.3.

sampling freq: fs defaults to 1.0.

Filtered Data in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the filtered data output.

Filtered Data out is a DSP Handle Cluster that is identical to Filtered Data in, but with the filtered

data already stored in the memory buffer on the DSP board. Because the VI filters via convolution, the

number of elements, k, in Filtered Data out is:

k = n + m - 1,

where n is the number of elements in X, and

m is the number of taps.

m-1

A delay equal to

is also associated with the output sequence.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation cannot be performed in place; that is, the input X and the output Filtered Data cannot be the same

DSP Handle Cluster.

NI-DSP SRM for LabVIEW for Windows

2-29

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Exact Blackman Window

Applies an Exact Blackman window to

the input sequence X. If Y represents

the output sequence Exact

Blackman{X}, the elements of Y are

obtained using the formula:

y_i= x_i[0.42659071

- 0.49656062 cos(w)

+ 0.07684867 cos(2w)]

for i = 0, 1, 2, … , n-1

2πi

w =

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output Exact

Blackman {X}.

Exact Blackman {X} is a DSP Handle Cluster that is identical to X, but with the results of Exact

Blackman {X} already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-30

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Exponential Window

Applies an exponential window to the input

sequence X. If Y represents the output

sequence Exponential{X}, the elements of

Y are obtained using the formula:

y_i= x_iexp(a * i)

for i = 0, 1, 2, … , n-1,

ln |f|

a =

n - 1

where f is the final value, and

n is the number of elements in X.

You can use the Exponential Window VI to analyze transients.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output

Exponential {X}.

final value defaults to 0.10.

Exponential {X} is a DSP Handle Cluster that is identical to X, but with the results of Exponential {X}

already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-31

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP FHT

Computes the fast Hartley transform (FHT) of

the input sequence X. The Hartley transform

of a function x(t) is defined as follows:

∞

X(f) = ∫ x(t) cas(2πft) dt

-∞

where cas(x) = cos(x) + sin(x).

If Y represents the output sequence FHT {X} obtained via the FHT, then Y is obtained through the discrete

implementation of the Hartley integral:

n-1

2π ik

( _n)

Y_k=

X_icas

for k = 0, 1, 2, … , n-1 .

∑

i=0

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Notes: The number of elements for the input array must be a power of two.

The operation is performed in place and the input array X is overwritten by the output

FHT{X}.

The largest FHT that can be computed depends upon the amount of memory on your DSP

board.

FHT{X} is a DSP Handle Cluster that is identical to X, but with the results of FHT{X} already stored in

the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The Hartley transform maps real-valued sequences into real-valued frequency domain sequences. You can use it

instead of the Fourier transform to convolve signals, deconvolve signals, correlate signals, and find the power

spectrum. Furthermore, you can derive the Fourier transform from the Hartley transform.

When the sequences to be processed are real-valued sequences, the Fourier transform produces complex-valued

sequences in which half of the information is redundant. The advantage of using the Hartley transform instead of

the Fourier transform is that the Hartley transform uses half the memory to produce the same information the FFT

produces. Further, the FHT is calculated in place and is as efficient as the Fourier transform.

Part 3: NI-DSP Function Reference

2-32

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Flat Top Window

Applies a flat top window to the input sequence

X. If Y represents the output sequence

Flattop{X}, the elements of Y are obtained

using the formula:

y_i= x_i[0.2810639 - 0.5208972 cos(w)

+ 0.1980399 cos(2w)]

for i = 0, 1, 2, … , n-1

2πi

w =

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output

Flattop {X}.

Flattop {X} is a DSP Handle Cluster that is identical to X, but with the results of Flattop {X} already

stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-33

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Force Window

Applies a force window to the input sequence X

If Y represents the output sequence Force{X},

the elements of Y are obtained using the

formula:

x _iif 0 ≤ i ≤ d

y_i=

elsewhere

for i = 0, 1, 2, …, n-1 ,

d = (0.01)(n)(duty)

where n is the number of elements in X.

You can use the Force Window VI to analyze transients.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output

Force {X}.

duty. 0.0 ≤ duty ≤ 100.0. duty defaults to 50.0.

Force {X} is a DSP Handle Cluster that is identical to X, but with the results of Force {X} already

stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

DSP Free Memory

Frees the memory space referred by DSP

Handle Cluster on the specified DSP board.

DSP Handle Cluster is a DSP Handle Cluster that indicates the memory buffer to free.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and the VI also frees the DSP Handle Cluster.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-34

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Gaussian White Noise

Generates a Gaussian distributed

pseudorandom pattern whose

statistical profile is as follows:

(µ, σ) = (0, s) ,

where s is the absolute value of the

specified standard deviation.

standard deviation defaults to 1.0.

seed < 65536.0. If seed > 0.0, the generated noise will be the same in repeated invocations if the seed

value does not change. If seed ≤ 0.0, the VI generates a random value to use as the seed, and the noise

will differ in repeated invocations although the value in seed does not change. seed defaults to 0.0.

Gaussian Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

will contain the output Gaussian pattern.

Gaussian Pattern out is a DSP Handle Cluster that is identical to Gaussian Pattern in, but with the

generated pattern already stored in the memory buffer on the DSP board. The largest Gaussian pattern

that can be generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-35

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP General Cosine Window

Applies a general cosine window to the input

sequence X. If A represents the Cosine

Coefficients input sequence and Y represents

the output sequence GenCos{X}, the elements

of Y are obtained using the formula:

y_i= x_i^m-1(-1)^ka cos (kw)

∑

k=0

for i = 0, 1, 2, … , n-1

2πi

w =

where n is the number of elements in X, and

m is the number of elements in Cosine Coefficients.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output

GenCos {X}.

Cosine Coefficients is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

contains the cosine coefficients.

GenCos {X} is a DSP Handle Cluster that is identical to X, but with the results of GenCos {X} already

stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-36

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Hamming Window

Applies a Hamming window to the input

sequence X. If Y represents the output

sequence Hamming {X}, the elements of Y

are obtained from the formula:

y = x [0.54 - 0.46 cos(w)]

for i = 0, 1, 2, …, n-1 ,

2πi

w =

where n is the number of elements in the input sequence X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

data array.

Note: The operation is performed in place and the input array X is overwritten by the output

Hamming {X}.

Hamming {X} is a DSP Handle Cluster that is identical to X, but with the results of Hamming {X}

already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-37

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Handle To Address

Finds the actual DSP address value of

DSP Handle Cluster that indicates a

memory buffer on the DSP board.

You can use the output DSP Address as

the input of the DSP Address terminal in

the DSP DMA Copy(DSP to LV) VI or

the DSP DMA Copy(LV to DSP) VI.

DSP Handle Cluster indicates a memory buffer on the DSP board.

DSP Address is the actual DSP address of the memory buffer on the DSP board referred to by DSP

Handle Cluster.

error in(no error) contains the error information from a previous VI. If an error occurs, it is passed

out error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-38

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Hanning Window

Applies a Hanning window to the input

sequence X. If Y represents the output

sequence Hanning {X}, the elements of Y are

obtained using the formula:

y_i= 0.5 x_i[1 - cos(w)]

for i = 0, 1, 2, …, n-1 ,

2πi

w =

where n is the number of elements in the input sequence X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

data array.

Note: The operation is performed in place and the input array X is overwritten by the output

Hanning {X}.

Hanning {X} is a DSP Handle Cluster that is identical to X, but with the results of Hanning {X} already

stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-39

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP IIR Filter

Performs IIR filtering on the

X input array and reports the

result in Y. It uses the arrays

a and b of sizes sza and szb

respectively in implementing

the linear difference equation

that describes IIR filtering:

k = sza - 1

k = szb-1

y[i] = ∑ a(k) * x(i-k) - ∑ b(k) * y(i-k)

i = 0,1 ..... ,n-1

k = 0

k = 1

where:

a is the array of forward coefficients describing an IIR filter (obtained from DSP Butterworth Coefficients

VI for example)

b is the array of feedback coefficients describing an IIR filter (obtained from the same filter design VI)

Notes: b(0) = 1.0in the above equation.

This VI may be called in a loop to perform filtering on data frames that are part of the same data stream.

At time i = 0, the initial conditions on X (for example, X(1-sza)through X(-1)) are obtained from the array

Initial Conditions on input. At exit (the end of the VI execution), this array holds the final conditions on

X that could be used as the initial conditions for the next DSP IIR Filter VI call if you were to call it in a

loop.

At time i = 0, the initial conditions on Y (for example, Y(1-szb) through Y(-1)) are obtained from the array

Initial Conditions on output. At exit (the end of the VI execution), this array holds the final conditions on

Y that could be used as the initial conditions for the next DSP IIR Filter VI call if you were to call it in a

loop.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

output array of filtered data.

Initial Conditions on input is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that contains the initial conditions on input at time zero.

Initial Conditions on output is a DSP Handle Cluster that indicates the memory buffer on the DSP

board that contains the initial conditions on output at time zero.

Forward Coefficients is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

contains the forward coefficients of the IIR filter.

Feedback Coefficients is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

contains the feedback coefficients of the IIR filter.

Part 3: NI-DSP Function Reference

2-40

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

Y out is a DSP Handle Cluster that is identical to Y in, but with the filtered data already stored in the

memory buffer on the DSP board.

Final Conditions on input is a DSP Handle Cluster that is identical to Initial Conditions on input, but

contains the final conditions on input.

Final Conditions on output is a DSP Handle Cluster that is identical to Initial Conditions on output,

but contains the final conditions on output.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input X and the output Y can be the same DSP Handle Cluster.

Parameter Discussion

The arrays X and Y should be the same size. The arrays Forward Coefficients and Initial/Final Conditions on

input should be the same size. The arrays Feedback Coefficients and Initial/Final Conditions on output should

be the same size. In successive calls to this VI, while performing filtering on large data streams, one buffer at a

time, the final conditions of the filter are stored in Final Condition on input and in Final Condition on output, and

are used as the initial conditions of the filter in the subsequent call. For example, if you designed a filter with all of

the initial conditions set to zero, and you then wanted to filter a stream of 8,192 data points, 1,024 points at a time,

after the first call, the final conditions of the filter are stored in Final Condition on input and in Final Condition

on output. These can be used as the initial conditions to the filter in the filtering of the next 1,024 points, and so on.

NI-DSP SRM for LabVIEW for Windows

2-41

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Impulse Pattern

Generates an array containing an

impulse pattern. If the Impulse

Pattern is represented by the

sequence X, the VI generates the

pattern according to the following

formula:

if i = d

x_i=

elsewhere

for i = 0, 1, 2, … , n-1 ,

where a is the amplitude,

d is the delay, and

n is the number of elements in Impulse Pattern.

amplitude defaults to 1.0.

delay must be greater than or equal to 0. If delay is less than zero or greater than or equal to the size of

Impulse Pattern, the VI returns an error. delay defaults to 0.

Impulse Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

will contain the output impulse pattern.

Impulse Pattern out is a DSP Handle Cluster that is identical to Impulse Pattern in, but with the

generated pattern already stored in the memory buffer on the DSP board. The largest impulse pattern

that can be generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-42

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Impulse Train Pattern

Generates a train of impulses of value

amplitude at sample delay.

If the impulse train pattern is represented

by the sequence X, the VI generates the

pattern according to the following

formula:

amplitude if i modulo P = d

x_i=

0.0 elsewhere

for i = 0, 1, 2, … , n-1 ,

where P is the period,

d is the delay, and

n is the number of elements in the impulse train.

amplitude defaults to 1.0.

delay defaults to 0.

period defaults to 1.

Impulse Train in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output impulse train.

Impulse Train out is a DSP Handle Cluster that is identical to Impulse Train in, but with the generated

pattern already stored in the memory buffer on the DSP board. The largest impulse train that can be

generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-43

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Index Memory

Indexes into a DSP buffer

allocated in the memory

space of the specified DSP

board. The return value is

another DSP Handle Cluster.

This VI does not allocate

memory.

In order to index into a buffer

correctly, you need to indicate

what type of data is in the DSP

buffer. You must set data type to the appropriate type. data type has two options–4 bytes (for 32-bit long and

floating-point data) and 2 bytes (for 16-bit short data).

DSP Handle Cluster is a DSP Handle Cluster into which you want to index.

data type indicates the type of data in the buffer on the DSP board that is referred to by DSP Handle

Cluster. data type has two options:

0: 4 bytes (for 32-bit long and floating-point data).

1: 2 bytes (for 16-bit short data).

data type defaults to 4 bytes.

offset is the index where to start to index into the array referenced by DSP Handle Cluster. offset

defaults to 0.

Indexed DSP Handle Cluster is a DSP Handle Cluster that contains the number of size data in DSP

Handle Cluster starting from offset.

Notes: If you free a DSP Handle Cluster using the DSP Free Memory VI, then all other DSP Handle

Clusters obtained by using DSP Index Memory VI will no longer be valid.

This VI returns a new DSP Handle Cluster. If you no longer need this DSP Handle Cluster,

remember to free it.

size indicates how many elements you want to index starting from the offset. size defaults to 0.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-44

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Init Memory

Initializes the memory heaps and frees all

allocations of memory on the specified DSP

board.

slot is the board ID number. slot defaults to 3.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-45

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Integral

Performs the discrete integration of the

sampled signal X. The integral F(t) of a

function f(t) is defined as follows:

F(t) = ∫ f(t) dt .

Let Y represent the sampled output sequence Integral X. The VI obtains the elements of Y using the following

formula:

y_i=

(x_j-1+ 4x_j+ x_j+1) dt

for i = 0, 1, 2, …, n-1 ,

∑

j=0

where n is the number of elements in X,

x_-1is specified by initial condition when i = 0, and

x_nis specified by final condition when i = n-1.

The initial condition and final condition minimize the overall error by increasing the accuracy at the boundaries,

especially when the number of samples is small. Determining boundary conditions before the fact enhances

accuracy.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

initial condition defaults to 0.0.

final condition defaults to 0.0.

dt is the sampling interval and must be greater than zero. If dt is less than or equal to zero, the VI

returns an error. dt defaults to 1.0.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of the integration of X.

Y out is a DSP Handle Cluster that is identical to Y in, but with the results of integration already stored

in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input X and the output Y can be the same DSP Handle Cluster.

Note: You can also use the DSP Integral x(t) VI to numerically evaluate the finite integral:

∫ f(t) dt = F(b) - F(a) ≈ y_n-1

by extracting the last element of the output sequence Y.

Part 3: NI-DSP Function Reference

2-46

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Inv Chebyshev Coeff

Generates the set of filter

coefficients to implement an

inverse IIR filter as specified

by the Chebyshev II Filter

mode. You can then pass

these coefficients to the DSP

IIR Filter VI to filter a

sequence of data.

filter type specifies the passband of the filter. filter type has four options:

0: lowpass.

1: highpass.

2: bandpass.

3: bandstop.

filter type defaults to lowpass.

sampling Freq : fs is the sampling frequency and must be greater than 0. If it is less than or equal to

zero, the VI returns an error. sampling Freq : fs defaults to 1.0.

high cutoff freq: fh is the high cutoff frequency. The VI ignores this parameter when filter type is

lowpass. high cutoff freq: fh defaults to 0.45.

low cutoff freq: fl is the low cutoff frequency. The VI ignores this parameter when filter type is

highpass. low cutoff freq: fl defaults to 0.125.

Note: fh and fl must observe the Nyquist criterion: 0 ≤ fl ≤ fh ≤ fs /2.

attenuation must be greater than 0.0. If attenuation is less than or equal to zero, the VI returns an

error. attenuation defaults to 60.0.

order must be greater than zero. If the filter order is less than or equal to zero, the VI returns an error.

order defaults to 2.

Forward Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that will contain the forward coefficients.

Feedback Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that will contain the feedback coefficients.

Forward Coefficients out is a DSP Handle Cluster that is identical to Forward Coefficients in, but

with the forward coefficients already stored in the memory buffer on the DSP board.

Feedback Coefficients out is a DSP Handle Cluster that is identical to Feedback Coefficients in, but

with the feedback coefficients already stored in the memory buffer on the DSP board.

NI-DSP SRM for LabVIEW for Windows

2-47

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Parameter Discussion

The arrays Forward Coefficients in/out and Feedback Coefficients in/out must have a size of at least (order + 1)

for lowpass and highpass filters. The arrays Forward Coefficients in/out and Feedback Coefficients in/out must

have a size of at least (2*order + 1) for bandpass and bandstop filters.

DSP Inverse FFT

Computes the inverse Fourier transform of the

complex input sequence FFT {X}. If Y

represents the output sequence, then:

-1

Y = F {X}.

Re FFT{X} is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains

the real part of input signal array FFT {X}.

Im FFT{X} is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains

the imaginary part of input signal array FFT {X}.

Notes: The number of elements for the input array must be the power of two. If the size of Re

FFT{X} is different from the size of Im FFT{X}, the VI uses the smaller number as the size of

the inverse FFT.

The operation is performed in place and the input array Re FFT{X } and Im FFT{X} is

overwritten by Re {X} and Im {X}.

The largest inverse FFT that can be computed depends upon the amount of memory on your

DSP board.

Re {X} is a DSP Handle Cluster that is identical to Re FFT{X }, but with the real part of X already

stored in the memory buffer on the DSP board.

Im {X} is a DSP Handle Cluster that is identical to the Im FFT{X} but with the imaginary part of X

already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-48

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Inverse FHT

Computes the inverse fast Hartley transform of

the input sequence FHT {X}. The inverse

Hartley transform of a function X(f) is defined

∞

x(t) = ∫ X(f) cas(2πft) df

-∞

where cas(x) = cos(x) + sin(x).

If Y represents the output sequence X, the VI calculates Y through the discrete implementation of the inverse

Hartley integral:

n-1

2π ik

( _n)

Y_k=

X_icas

for k = 0, 1, 2, … , n-1

∑

i=0

where n is the number of elements in X.

The inverse Hartley transform maps real-valued frequency sequences into real-valued sequences. You can use it

instead of the inverse Fourier transform to convolve, deconvolve, and correlate signals. Furthermore, you can derive

the Fourier transform from the Hartley transform.

See the description of the DSP FHT VI for a comparison of the Fourier and Hartley transforms.

FHT {X} is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the

input signal array.

Notes: The number of elements for the input array must be a power of 2.

The operation is performed in place and the input array FHT {X} is overwritten by

the output array X.

The largest inverse FHT that can be computed depends upon the amount of memory in your

DSP board.

X is a DSP Handle Cluster that is identical to FHT {X}, but with the results of inverse FHT {X } already

stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-49

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Kaiser-Bessel Window

Applies a Kaiser-Bessel window to the

input sequence X. If Y represents the

output sequence Kaiser-Bessel{X}, the

elements of Y are obtained using the

formula:

I_oβ 1. 0 - a ²

I_o

( )

y_i= x_i

for i = 0, 1, 2, … n - 1

i - k

a =

n-1

k =

where n is the number of elements in X, and

I ( ) is the zeroth-order modified Bessel function.

•

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output

Kaiser-Bessel {X}.

beta is proportional to the sidelobe attenuation–that is, the larger beta is, the greater the sidelobe

attenuation is. beta defaults to 0.0.

Kaiser-Bessel {X} is a DSP Handle Cluster that is identical to X, but with the results of

Kaiser-Bessel {X} already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-50

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Linear Evaluation

Performs a linear evaluation of the input array

X. The i element of the output array Y is

obtained using the following formula:

Y(i) = a* X(i) + b for i = 0, 1, 2, …, n-1,

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

a is the multiplicative constant. a defaults to 1.0.

b is the additive constant. b defaults to 0.0.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of output array.

Y out is a DSP Handle Cluster that is identical to Y in, but with the results already stored in the memory

buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input and output arrays can be the same DSP Handle Cluster.

DSP Load

Downloads a library specified by path name to

the DSP board in slot. path name must contain

the full path name of a valid COFF file (.out

file).

slot is the board ID number. slot defaults to 3.

path name is the path name of the .outfile.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-51

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Log

Computes the logarithm base 10 of the X input

array. The ith element of resulting array is

obtained by using the following formula:

y(i) = log10 (X(i)) * mult

for i = 0, 1, 2, …, n-1,

where n is the number of elements in X.

This VI is useful for converting values that represent power to decibels. This VI returns the most negative number

for any input less than or equal to zero.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

mult defaults to 1.0.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of logarithm X.

Y out is a DSP Handle Cluster that is identical to Y in, but with the results of logarithm X already stored

in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-52

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Max & Min

Finds the maximum and minimum values in the

input array, as well as the respective indices of

the occurrence of the maximum and minimum

values.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

max value is the maximum value of the input array.

max index is the index into the input array where maximum occurred.

min value is the minimum value of the input array.

min index is the index into the input array where minimum occurred.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-53

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Median Filter

Applies a median filter of rank to the

input sequence X. The median filter is a

nonlinear filter that combines lowpass

filters characteristics (to remove high-

frequency noise) and high-frequency

characteristics (to detect edges).

If Y represents the output sequence

Filtered Data, if J_irepresents a subset

of the input sequence X centered about the i element of x:

J_i= {x_i-r, x_i-r+1, …, x_i-1, x_i, x_i+1, …, x_i+r-1, x_i+r},

and if the indexed elements outside the range of X equal zero, the VI obtains the elements of Y using:

y_i= Median(J_i)

for i = 0, 1, 2, …, n-1,

where n is the number of elements in the input sequence X, and

r is the filter rank.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

data array. The number of elements in X must be greater than the rank:

n>r≥0.

If the number of elements in X is less than or equal to rank, the VI returns an error.

rank must be greater than or equal to zero. If rank is less than zero, the VI returns an error. rank

defaults to 2.

Filtered Data in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the filtered data output.

Filtered Data out is a DSP Handle Cluster that is identical to Filtered Data in, but with the filtered data

already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation cannot be performed in place; that is, the input X and the output Filtered Data cannot be the same

DSP Handle Cluster.

Part 3: NI-DSP Function Reference

2-54

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Multiply

Multiply array X by array Y. The i element

of the output array Z is obtained using the

following formula:

Z(i) = X(i) * Y(i) for i = 0, 1, 2, …, n-1,

where n is the smaller number of elements in X

and Y.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array Y.

Z in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of X(i) * Y(i).

Z out is a DSP Handle Cluster that is identical to Z in, but with the multiplied results already stored in

the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input and output arrays can be the same DSP Handle Cluster.

NI-DSP SRM for LabVIEW for Windows

2-55

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Parks-McClellan

Generates a set of linear-phase finite impulse

response multiband digital filter coefficients

using the number of taps, sampling

frequency, filter type, and Band Parameters.

# of taps is the total number of coefficients in H. A tap corresponds to a multiplication and an addition.

If there are n taps, every filtered sample requires n multiplications and n additions. # of taps must be

greater than 0. If # of taps is less than or equal to zero, the VI returns an error. # of taps defaults to 32.

sampling freq : fs defaults to 1.0.

Band Parameters is a cluster. Each cluster element contains the necessary information associated with

each band for the FIR design. Each cluster contains four elements, as shown in the following figure:

The Band Parameters cluster must contain at least one element, that is, one band.

amplitude is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

contains the desired amplitude for each band.

lower freq is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

contains the lower frequency bands.

higher freq is a DSP Handle Cluster that indicates the memory buffer on the DSP board

that contains the upper frequency bands.

Part 3: NI-DSP Function Reference

2-56

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

weighted ripple is a DSP Handle Cluster that indicates the memory buffer on the DSP

board that contains the weighting factor for each band.

For each band, higher freq must be greater than lower freq, and for adjacent bands, lower

freq in the higher band must be greater than higher freq in the lower band,

f_hi> f_li,

for i = 0, 1, 2, …, m-1,

f_li+1> f_hi, for i = 0, 1, 2, …, m-2,

where f_lirepresents the lower freq in the i band, and f_hirepresents the higher freq in the i

band.

The higher freq in the last band must observe the Nyquist criterion:

f_hm-1≤ 0.5f ,

where f_sis the sampling frequency.

If Band Parameters does not contain any elements, or if any of the preceding frequency

conditions is violated, the VI returns an error.

filter type has three options:

0: multiband.

1: differentiator.

2: Hilbert.

filter type defaults to multiband.

ripple is the optimal ripple the VI computes and is a measure of deviation from the ideal filter

specifications.

H in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

FIR coefficients.

H out is a DSP Handle Cluster that is identical to H in, but with the coefficients already stored in the

memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Note: The DSP Parks-McClellan VI finds the coefficients using iterative techniques based upon an error criterion.

Although you specify valid filter parameters, the algorithm may fail to converge.

The DSP Parks-McClellan VI generates only the filter coefficients. It does not perform the filtering

function. To filter a sequence X using the set of FIR filter coefficients H, use the DSP Convolution VI with

X and H as the input sequences.

NI-DSP SRM for LabVIEW for Windows

2-57

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

The equi-ripple filters use a similar technique to filter the data.

Part 3: NI-DSP Function Reference

2-58

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

Parameter Discussion

The weights are usually the same for every band and inversely proportional to frequency f for a differentiate. The

amplitudes of the bands are usually the same for every band and form a slope for a differentiation. The designed

filter is unconstrained, that is the frequency response in transitional period specifications is not guaranteed. Adjust

the parameters if there are discrepancies between specifications and results.

Generally, when type = multiband and bands ≥ 2, the DSP Parks-McClellan VI designs a multiband filter. When

filter type = differentiator and bands = 1, the DSP Parks-McClellan VI designs a differentiation. When

filter type = Hilbert and bands = 1, the DSP Parks-McClellan VI designs a Hilbert transform. The following tables

contain the brief requirements you must meet when you design different types of filters. For more information,

please refer to Digital Filter Design by Parks and Burrus, or "A Computer Program for Designing Optimum FIR

Linear Phase Digital Filters," by McClellan, Parks, and Rabiner.

When filter type = Multiband:

Filter Band Type

# of Bands (n)

# of taps

LowPass

HighPass

BandPass

BandStop

≥3

odd/even

≥3

odd

odd/even

0.0

odd

0.0

fl[0]

0.0

0.5fs

≥0.0

0.0

0.5fs

>0.0

fh[n-1]

0.5fs

amp[0]

>0.0

≥0.0

amp[n-1]

0.0(for even taps) >0.0

≥0.0(for odd taps)

0.0(for even taps) >0.0

≥0.0(for odd taps)

fl[0] represents the first value in the array of Lower Freq.

fh[n-1] represents the last value in the array of Higher Freq.

amp[0] represents the first value in the array of Amplitude.

amp[n-1] represents the last value in the array of Amplitude.

fs is the sample frequency.

When filter type = Differentiator or Hilbert:

Filter Type

bands(n)

fl[0]

Diff(odd)

Diff(even)

Hilbert(odd)

Hilbert(even)

≥1

0.0

0.5fs-∆

≥1

0.0

0.5fs

≥1

0.0+∆

0.5fs-∆

≥1

0.0+∆

0.5fs

fh[n-1]

odd or even in the table refers to the value of the # of taps.

fl[0] represents the first value in the array of Lower Freq.

fh[n-1] represents the last value in the array of Higher Freq.

fs is the sample frequency.

∆ is a small number that specifies the transitional band.

Although the DSP Parks-McClellan VI is the most flexible way to design a FIR linear phase filter, it has more

complex parameters and requires some DSP knowledge. You may find it more convenient to use DSP Equi-Ripple

LowPass, DSP Equi-Ripple HighPass, DSP Equi-Ripple BandPass, and DSP Equi-Ripple BandStop. These

functions, which provide lowpass, highpass, bandpass and bandstop FIR filters with equal weighting factors in all

bands, are special cases of the DSP Parks-McClellan VI with simplified parameters.

NI-DSP SRM for LabVIEW for Windows

2-59

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Polar to Rectangular

Converts a set of polar coordinate points

(Magnitude, Phase) to a set of rectangular

coordinate points (X, Y). The i elements

of the rectangular set is obtained using the

following formulas:

X(i) = Magnitude(i) * cos(Phase(i))

Y(i) = Magnitude(i) * sin(Phase(i))

Note: The operation is performed in place and the input arrays Magnitude and Phase are overwritten by X and

Magnitude is the DSP Handle Cluster that indicates the memory buffer on the DSP board that

contains the input signal array Magnitude.

Phase is the DSP Handle Cluster that indicates the memory buffer on the DSP board that contains

the input signal array Phase.

X is the DSP Handle Cluster that is identical to Magnitude but with the caculated rectangular

coordinate values already stored in the memory buffer on the DSP board.

Y is the DSP Handle Cluster that is identical to Phase but with the caculated rectangular coordinate

values already stored in the memory buffer on the DSP board.

error in(no error) contains the error information from a previous VI. If an error occurs, it is passed

out error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-60

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Polynomial Evaluation

Performs a polynomial evaluation on the input

array X. The i element of the output array Y

is obtained using the following formula:

k-1

Y(i) = ∑ (Coefficients(j) X(i) )

j=0

for i = 0, 1, 2, …, n-1,

where n is the number of elements in X and k is the number of elements in Coefficients.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Coefficients is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains

the coefficients array.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of output array.

Y out is a DSP Handle Cluster that is identical to Y in, but with the results already stored in the memory

buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input X and the output Y can be the same DSP Handle Cluster.

NI-DSP SRM for LabVIEW for Windows

2-61

Part 3: NI-DSP Function Reference

Chapter 2

NI-DSP Analysis VI Reference

DSP Power Spectrum

Computes the Power Spectrum of the

input sequence X. The Power Spectrum

S_xx(f) of a function x(t) is defined as

S_xx(f) = X^*(f)X(f) = | X(f) | ²,

where X(f) = F{x(t)}, and X (f) is the

complex conjugate of X(f).

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Notes: The number of elements for the input array must be a power of two.

The operation is performed in place and the input array X is overwritten by the output Power

Spectrum.

The largest power spectrum that can be computed depends upon the amount of memory on your

DSP board.

This VI allocates a temporary workspace on the DSP board equal to the size of the input signal

array.

Power Spectrum is a DSP Handle Cluster that is identical to X, but with the results of Power

Spectrum already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

DSP Product

Finds the product of the elements of the input

array X. The product of the elements is

obtained using the following formula:

product = ⁿ⁻¹X(i)

∏

i=0

where n is the smaller number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

product is the product of the elements in X.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-62

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Pulse Pattern

Generates an array containing a pulse

pattern. If Pulse Pattern is represented

by the sequence X, then the pattern is

generated according to the following

formula:

a if d ≤ i < last

x_i=

0 elsewhere

for i = 0, 1, 2, … , n-1 ,

last = d + w ,

where a is the amplitude,

d is the delay,

w is the width, and

n is the number of elements in Pulse Pattern.

amplitude defaults to 1.0.

delay must be greater than or equal to 0. delay defaults to 0.

width must be greater than or equal to 0. width defaults to 1.

Pulse Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output pulse pattern.

Pulse Pattern out is a DSP Handle Cluster that is identical to Pulse Pattern in, but with the generated

pattern already stored in the memory buffer on the DSP board. The largest pulse pattern that can be

generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-63

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Ramp Pattern

Generates an array containing a ramp

pattern. If the Ramp Pattern is

represented by the sequence X, then the

pattern is generated according to the

formula:

x_i= x₀+ i ∆x

for i = 0, 1, 2, … , n-1 ,

x_n-1- x₀

where ∆x =

n - 1

x_n-1is the final,

x₀is the init, and

n is the number of elements in Ramp Pattern.

The Ramp Pattern VI does not impose conditions on the relationship between init and final. The VI can therefore

generate ramp-up and ramp-down patterns.

final defaults to 0.0.

init defaults to 1.0.

Ramp Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output ramp pattern.

Ramp Pattern out is a DSP Handle Cluster that is identical to Ramp Pattern in, but with the generated

pattern already stored in the memory buffer on the DSP board. The largest ramp pattern that can be

generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-64

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Random Pattern

Generates a uniformly distributed

pseudorandom pattern whose values

are in the range [0:1]. The sequence is

generated using the Very-Long-Cycle

random number generator algorithm.

seed < 65536.0. If seed > 0.0, the generated noise will be the same in repeated invocations if the seed

value does not change. If seed ≤ 0.0, the VI generates a random value to use as the seed, and the noise

differs in repeated invocations although the value in seed does not change. seed defaults to 0.0.

Random Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

will contain the output random pattern.

Random Pattern out is a DSP Handle Cluster that is identical to Random Pattern in, but with the

generated pattern already stored in the memory buffer on the DSP board. The largest random pattern

that can be generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-65

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Rectangular To Polar

Converts a set of rectangular coordinate

points (X, Y) to a set of polar coordinate

points (Magnitude, Phase). The i

element of the polar coordinate set is

obtained by using the following fomulas:

Magnitude(i) = X(i) +Y(i)

Phase(i) = arctan(Y(i)/X(i))

Note: The operation is perfomed in place and the input arrays X and Y are overwritten by Magnitude and Phase.

X is the DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the

input signal array X.

Y is the DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the

input signal array Y.

Magnitude is the DSP Handle Cluster that is identical to X but with the caculated polar coordinate

amplitude values already stored in the memory buffer on the DSP board.

Phase is the DSP Handle Cluster that is identical to Y but with the caculated polar coordinate phase

values already stored in the memory buffer on the DSP board.

error in(no error) contains the error information from a previous VI. If an error occurs, it is passed

out error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-66

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP ReFFT

Computes the Fast Fourier transform of a

real input sequence X. If Y represents

the complex output sequence, then:

Y = F{X}.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Notes: The number of elements for input X must be the power of two.

The operation is performed in place and the input array X is overwritten by Re FFT{X}.

The largest FFT that can be computed depends upon the amount of memory in your DSP board.

Im FFT{X} in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the imaginary part of FFT{X}.

Re FFT{X} is a DSP Handle Cluster that is identical to X, but with the real part of FFT{X}already

stored in the memory buffer on the DSP board.

Im FFT{X} out is a DSP Handle Cluster that is identical to Im FFT{X} in, but with the imaginary part

of FFT{X} already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

DSP Reset

DSP Reset aborts any currently running

function on the board, stops the library, and

reloads the DSP Library from the directory

specified in the WDAQCONF.EXEfile.

slot is the board ID number. slot defaults to 3.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-67

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Reverse

Reverse the order of the elements of the input

array X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

The operation can only be performed in place; that is, the input X is overwritten by the output Reverse {X}.

Reverse {X} is a DSP Handle Cluster that is identical to the X, but with the reverse {X} already stored

in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-68

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Sawtooth Pattern

Generate a sawtooth pattern with

positive-slope, zero-crossing at

sample delay. If the sawtooth

pattern is represented by the

sequence Y, then the pattern is

generated according to the

following formula:

if 0 ≤ k <

y =

for i = 0, 1, 2, ..., n - 1

(k -T) if ≤ k < T

cycles

where k = (i - d) modulo T and

d is the delay, and

T =

n is the number of elements in the sawtooth pattern.

amplitude defaults to 1.0.

delay defaults to 0.

cycle defaults to 1.0.

Sawtooth in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output sawtooth pattern.

Sawtooth out is a DSP Handle Cluster that is identical to Sawtooth in, but with the generated pattern

already stored in the memory buffer on the DSP board. The largest sawtooth pattern that can be

generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-69

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Set

Set the elements of the input array X to the

constant value set value. If the output Set {X}

is represented by the sequence Y, then:

y_i= set value

for i = 0, 1, 2, …, n-1 ,

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X. The operation is performed in place; that is, the input X is overwritten by the output

Set {X}.

set value defaults to 0.0.

Set {X} is a DSP Handle Cluster that is identical to X, but with the set {X} already stored in the memory

buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

DSP Shift

Shifts the elements in the input array X,

replacing the new values with zeros. The

number of shifts selected can be in the

positive (right) or negative (left) direction.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

# of shifts defaults to 0.

Shifted {X} in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output array of shifted {X}.

Shifted {X} out is a DSP Handle Cluster that is identical to Shifted {X} in, but with the

shifted {X} already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input and output arrays can be the same DSP Handle Cluster.

Parameter Discussion

Shifting can be performed right or left by setting # of shifts to a positive or negative value, respectively. If # of

shifts is greater than n, ( n is the number of elements of input array X ), the outputs are all zero.

NI-DSP SRM for LabVIEW for Windows

2-70

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Sinc Pattern

Generates an array containing a sinc

pattern. If the Sinc Pattern is represented

by the sequence Y, then the pattern is

generated according to the following

formula:

y_i= a sinc(i∆ t - d) ,

for i = 0, 1, 2, …, n-1 ,

sin(πx)

where sinc(x) =

πx

a is the amplitude,

∆t is the sampling interval delta t,

d is the delay, and

n is the number of elements in Sinc Pattern.

The main lobe of the sinc function, sinc(x), is the part of the sinc curve bounded by the region -1 ≤ x ≤ 1.

When |x| = 1, the sinc(x) = 0.0, and the peak value of the sinc function occurs when x = 0. You can show using

l'Hôpital's Rule that sinc(0) = 1 and that it is also its peak value. Thus, the main lobe is the region of the sinc curve

encompassed by the first set of zeros to the left and the right of its peak value.

amplitude defaults to 1.0.

delay shifts the peak value within the Sinc Pattern as the VI generates the pattern. This condition is

determined from the preceding formula and occurs when i∆t = d . delay defaults to 0.0.

delta t is the sampling interval. It is a floating-point number inversely proportional to the width of the

main sinc lobe. That is, the smaller the sampling interval, the wider the main lobe, and the larger the

sampling interval, the smaller the main lobe. Notice that when delta t is 1 and d is an integer value, the

VI sets Sinc Pattern to zero except at the point where i = d, at which point the value is equal to

amplitude. The recommended range of values for the sampling interval is 0 < delta t ≤ 1 . delta t must

be greater than 0.0. If delta t is less than or equal to zero, the VI returns an error. delta t defaults to 1.0.

Sinc Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output sinc pattern.

Sinc Pattern out is a DSP Handle Cluster that is identical to the Sinc Pattern in, but with the generated

pattern already stored in the memory buffer on the DSP board. The largest sinc pattern that can be

generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-71

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Sine Pattern

Generates an array containing a sinusoidal

pattern. If the Sine Pattern is represented

by the sequence Y, the pattern is

generated according to the following

formula:

y_i= a sin(x_i) ,

for i = 0, 1, 2, …, n-1 ,

2πi k

π ø_o

180

where x_i=

a is the amplitude,

k is the number of cycles in the pattern,

ø_ois the initial phase in degrees, and

n is the number of elements in Sine Pattern.

amplitude defaults to 1.0.

phase defaults to 0.0.

Note: phase must be in degrees rather than radians. If phase is in radians, make sure you convert it to

degrees, as shown in the following figure, before using the Sine Pattern VI.

cycles defaults to 1.0.

Note: Because cycles is a floating-point number, fractional cycles are possible for the Sine Pattern.

Furthermore, setting cycles to a negative number does not generate an error condition because it

is mathematically correct and useful to consider negative frequencies in Fourier and spectral

analysis.

Sine Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output sine pattern.

Sine Pattern out is a DSP Handle Cluster that is identical to Sine Pattern in, but with the generated

pattern already stored in the memory buffer on the DSP board. The largest sine pattern that can be

generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-72

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Square Pattern

Generates an array containing a square

pattern. If the Square Pattern is

represented by the sequence X, then

the pattern is generated according to

the following formula:

if 0 ≤ remainder

< 0.01 dT

(_T)

x_i=

for i = 0, 1, …, n-1 ,

0.0 elsewhere

where T =

is the time period of one cycle of the Square Pattern,

cycles

a is the amplitude,

d is the duty cycle in percent, and

n is the number of elements in Square Pattern.

amplitude defaults to 1.0.

duty cycle must be greater than or equal to 0 and less than or equal to 100. If duty cycle is less than

zero or greater than 100, the VI returns an error. duty cycle defaults to 50.0.

Note: duty cycle must be a percentage. Make sure you convert the fractions of a cycle into

percentages, as shown below, before using the DSP Square Pattern VI in a block diagram.

Special Cases: Two special cases can occur when the duty cycle assumes its extreme values of 0 and

100. The VI sets Square Pattern to zero:

duty cycle = 0 or 100

⇒

X = 0.

cycles must be greater than 0. If cycles is less than or equal to zero, the VI returns an error. cycles

defaults to 1.0.

Note: Because cycles is a floating-point number, fractional cycles of the Square Pattern are

permitted.

Square Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output square pattern.

Square Pattern out is a DSP Handle Cluster that is identical to Square Pattern in, but with the

generated pattern already stored in the memory buffer on the DSP board. The largest square pattern that

can be generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-73

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Square Root

Find a square root estimate of the absolute value

of each element of the input array X . The i

element of the output array Y is obtained using

the following formula:

Y(i) = |X(i)|

for i = 0, 1, 2, …, n-1 ,

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

square root results of X(i).

Y out is a DSP Handle Cluster that is identical to Y in, but with the square root results already stored in

the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input and output arrays can be the same DSP Handle Cluster.

If the input value is negative, it calculates the square root of the absolute value of that number.

NI-DSP SRM for LabVIEW for Windows

2-74

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Sort

Sort the input array X in ascending or

descending order.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

direction is the direction to sort:

0: ascending.

1: descending.

direction defaults to ascending.

Sorted {X} in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output array of sorted {X}.

Sorted {X} out is a DSP Handle Cluster that is identical to Sorted {X} in, but with the sorted {X}

already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input and output arrays can be the same DSP Handle Cluster.

Part 3: NI-DSP Function Reference

2-75

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Start

Enables the DSP board to run. Use DSP Start

with the DSP Load and DSP Reset VIs after

downloading a custom application.

slot is the board ID number. slot defaults to 3.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

DSP Subset

Extracts a subset of array X of length beginning

at index and stores it in array Y.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

index defaults to 0.

length is the length of the output subset. length defaults to 1.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

subset output array.

Y out is a DSP Handle Cluster that is identical to Y in, but with the subset array already stored in the

memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

This VI is particularly useful when performing onboard processing with large arrays, with index different from zero,

or Y in a memory space different from X. If index is less than zero, then the first element of the subset is the first

element of the input array.

NI-DSP SRM for LabVIEW for Windows

2-76

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Subtract

Subtract array Y from array X. The i element

of the output array Z is obtained using the

following formula:

Z(i) = X(i) -Y(i). for i = 0, 1, 2, …, n-1 ,

where n is the smaller number of elements in X

and Y.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array Y.

Z in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of X(i) - Y(i).

Z out is a DSP Handle Cluster that is identical to Z in, but with the subtracted results already stored in

the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

The operation can be performed in place; that is, the input and output arrays can be the same DSP Handle Cluster.

DSP Sum

Find the sum of the elements of the input array

X. The sum of the elements is obtained using

the following formula:

n-1

sum =

X(i)

∑

i=0

where n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array X.

sum is the sum of the elements in X.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-77

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP TimeOut

Selects the timeout limit in seconds to wait

for a function on DSP board to complete

execution.

The default timeout setting at startup and

after a DSP Reset call is 10 s.

All subsequent calls from the VI to the

onboard functions on the specified board return timeout errors if function execution on the board calling the VI is

not completed in the timeout limit.

slot is the board ID number. slot defaults to 3.

timeout is the timeout in seconds. timeout defaults to 10.0 s.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-78

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Triangle Pattern

Generates an array containing a

triangle pattern. If the Triangle

Pattern is represented by the

sequence Y, the pattern is generated

according to the following formula:

y_i= a tri (x_i) ,

for i = 0, 1, 2, …, n-1 ,

where

i∆t - d

x_i=

, and

1 - | x |

if | x | ≤ 1

tri(x) =

elsewhere

a is the amplitude,

d is the delay,

w is the width, and

n is the number of elements in Triangle Pattern.

amplitude is the value of the waveform at the peak. amplitude defaults to 1.0.

delay is the distance in seconds between the beginning of the pattern and the peak. delay shifts the peak

value within Triangle Pattern. delay defaults to 0.0.

delta t is the duration of the pattern in seconds, or the sampling interval, and must be greater than 0.0 to

avoid undefined arguments. If delta t is less than or equal to zero, the VI returns an error. delta t

defaults to 1.0.

width is the distance in seconds between the peak and end of the pattern. In other words, width sets the

width from the peak value to the first zero value in the pattern. Thus the actual duration of the pattern is

2width (twice the value of width). width must be greater than 0.0 to avoid undefined arguments. If

width is less than or equal to zero, the VI returns an error. width defaults to 1.0.

Triangle Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

will contain the output triangular pattern.

Triangle Pattern out is a DSP Handle Cluster that is identical to Triangle Pattern in, but with the

generated pattern already stored in the memory buffer on the DSP board. The largest triangle pattern

that can be generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

You use this VI to create patterns based on an isosceles triangle. The following figure shows how the VI parameters

relate to the generated pattern.

Part 3: NI-DSP Function Reference

2-79

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

Amplitude

Delay (s)

Width (s)

delta t (s)

The following figure illustrates how the pattern can vary with different values for the parameters.

Delay = 0 Delay = delta t

Width (s)

delta t (s)

Notice that when the Delay is 0, the signal is a ramp with a slope of -Amplitude/Width, and when Delay equals

delta t, the signal is a ramp with a slope of Amplitude/Width. Remember that the Width parameter is actually the

width of the ramp rather than the width of the triangle. delta t is the duration of the pattern in seconds. In the

previous and following illustrations, the box shows the pattern limits.

Delay = Width = 0.5 delta t

Delay + Width = delta t

Delay(s)

delta t (s)

Width (s)

Delay

delta t (s)

Width (s)

NI-DSP SRM for LabVIEW for Windows

2-80

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Triangular Train

Generates a train of triangular

pattern crossing value zero at delay

with positive slope. If the triangular

train is represented by the sequence

Y, the pattern is generated according

to the following formula:

if 0 ≤ k ≤

- 4a

if ≤ k ≤

y =

for i = 0, 1, 2, ..., n-1

( )

(k- T)

≤ k < T

where k = (i - d) modulo T,

T =

cycles

n is the number of elements in the triangular train, and

d is the delay.

amplitude defaults to 1.0.

delay defaults to 0.

cycle defaults to 1.0.

Triangular Train in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that

will contain the output triangular train.

Triangular Train out is a DSP Handle Cluster that is identical to Triangular Train in, but with the

generated pattern already stored in the memory buffer on the DSP board. The largest triangular train

that can be generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-81

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Triangular Window

Applies a triangular window to the input

sequence X. If Y represents the output

sequence Triangle{X}, the elements of Y are

obtained from the formula:

y_i= x_itri(w)

for i = 0, 1, 2, …, n-1 ,

2i-n

w =

where tri(w) = 1 - |w|, and

n is the number of elements in X.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output Triangle

{X}.

Triangle {X} is a DSP Handle Cluster that is identical to X, but with the results of Triangle {X} already

stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

2-82

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Uniform White Noise

Generates a uniformly distributed

pseudorandom pattern whose values are in

the range [-a:a], where a is the absolute

value of amplitude. The pseudorandom

sequence is generated using a modified

version of the Very-Long-Cycle random

number generator algorithm.

amplitude defaults to 1.0.

seed < 65536.0. If seed > 0.0, the generated noise will be the same in repeated invocations if the seed

value does not change. If seed ≤ 0.0, the VI generates a random value to use as the seed, and the noise

differs in repeated invocations although the value in seed does not change. seed defaults to 0.0.

White Noise in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will

contain the output white noise.

White Noise out is a DSP Handle Cluster that is identical to White Noise in, but with the generated

pattern already stored in the memory buffer on the DSP board. The largest white noise signal that can be

generated depends upon the amount of memory on your DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-83

NI-DSP SRM for LabVIEW for Windows

Chapter 2

NI-DSP Analysis VI Reference

DSP Unwrap Phase

Unwraps the Phase array by eliminating

discontinuities whose absolute values exceed π.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the

results of the integration of X.

Y out is a DSP Handle Cluster that is identical to Y in, but with the unwrapped output array already

stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Both the input array and the output array are in radians. The operation can be performed in place; that is, the input

X and the output Y can be the same DSP Handle Cluster.

NI-DSP SRM for LabVIEW for Windows

2-84

Part 3: NI-DSP Function Reference

NI-DSP Analysis VI Reference

Chapter 2

DSP Zero Padder

Pads the input array with zero from

starting index to the end of the input array.

This VI is useful when the size of the

acquired data buffers is not a power of two

and you want to take advantage of fast

processing algorithms in the analysis VIs.

These algorithms include Fourier

transforms, power spectrum, and fast Hartley transforms, which are extremely efficient for buffer sizes that are a

power of two.

X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input

signal array.

Note: The operation is performed in place and the input array X is overwritten by the output Zero

Padded {X}.

starting index is the index from which the input array is padded with zero. starting index defaults to 0.

Zero Padded {X} is a DSP Handle Cluster that is identical to {X}, but with the zero-padded array

already stored in the memory buffer on the DSP board.

error in (no error) contains the error information from a previous VI. If an error occurs, it is passed out

error out and no other calls are made.

error out contains the error information for this call.

Part 3: NI-DSP Function Reference

2-85

NI-DSP SRM for LabVIEW for Windows

Chapter 1

Introduction to the NI-DSP Interface Utilities

This chapter contains an overview of the NI-DSP Interface Utilities, installation instructions, and explains how to

use the NI-DSP Interface Utilities. You should familiarize yourself with the material in this chapter before

beginning subsequent chapters in this part.

The National Instruments AT-DSP2200, a high-performance digital signal processing (DSP) plug-in board for the

PC, is based on the AT&T WE DSP32C 32-bit, floating-point, digital signal processor. With the NI-DSP

Interface Utilities, you can customize the DSP Library that resides on the board by adding or deleting functions to

and from the common object file format (COFF) file that constitutes the onboard-resident software. This COFF

file is made of the Kernel, memory management routines, execution control routines, data communication

routines, interrupt handling routines for data acquisition, and a set of more than 60 analysis functions. Throughout

the chapters in this part, the COFF file is referred to as the DSP Library. By customizing the DSP Library, you

can change the number of analysis functions that are included. The other parts of the DSP Library always remain

unchanged.

Note: To customize the DSP Library that resides on the board, you need to have the Developer ToolKit,

available through National Instruments, which contains an AT&T C compiler, assembler, linker, and

documentation.

Overview of the NI-DSP Interface Utilities

The NI-DSP Interface Utilities are on the NI-DSP for LabVIEW for Windows distribution disks.

Unless otherwise stated, all references to directories in this part of the manual are to subdirectories under the path

you specified during installation of the NI-DSP Interface Utilities. If you chose the path D:\DEVEL, then the

installer will create the directory DEVEL and copy all files and directories to it. Figure 1-1 shows directory

structure created by SETUP.

D:\DEVEL

Dispatch Examples

Lib

Figure 1-1. NI-DSP for DOS Directory Structure

The Dispatch directory contains the utilities necessary for creating a custom DSP Library dispatcher. The

Dispatch directory contains the Dispatch application and the files NIDSP.fnc, dispatch.s., and

NIESSEN.fnc.

Dispatch is a PC application that uses the files NIDSP.fnc and NIESSEN.fnc and generates the

WE DSP32C assembly code needed to properly dispatch to a custom DSP Library based on the grouped function

names.

NI-DSP SRM for LabVIEW for Windows

1-1

Part 4: NI-DSP Interface Utilities

Introduction to the NI-DSP Interface Utilities

Chapter 1

The Examples directory contains the files used in Part 4, Chapter 2, Getting Started with the NI-DSP Interface

Utilities, to build a custom DSP Library.

The LIB directory contains LIBLVDSP.a, the archived library of object modules containing all of the analysis

functions and Kernel routines used in building the downloadable DSP Library files, as well as the two DSP

Library files LV2200S.out and LV2200.out.

Installing the NI-DSP Interface Utilities

For instructions on installing the NI-DSP Interface Utilities, refer to Part 1, Getting Started with NI-DSP, under

the section titled Installing NI-DSP for LabVIEW for Windows.

For instructions on installing the AT&T WE DSP32C software, refer to the WE DSP32C Support Software Library

User Manual and other related documentation.

Using the NI-DSP Interface Utilities

Use the NI-DSP Interface Utilities to customize the DSP Library and execute your own functions on the DSP board.

Figure 1-2 shows a diagram of the interface layers used to access functions in your custom library.

LabVIEW VI

Call to a CIN

Driver Software

(DSP.DLL)

PC Bus

AT-DSP2200

KERNEL

Dispatcher

Memory

Manager

Onboard

Library

Figure 1-2. Interface Layers to Onboard Functions

When your library is downloaded and a LabVIEW interface to your function has been created, you can execute a

function in your onboard custom library by making a call to the interface function in your program from LabVIEW.

This call passes control to the interface software which calls the NI-DSP driver. The NI-DSP driver communicates

with the DSP Library to pass parameters, indicates which onboard function to execute using the dispatcher, and

returns results to LabVIEW.

Part 4: NI-DSP Interface Utilities

1-2

NI-DSP SRM for LabVIEW for Windows

Chapter 2

Getting Started with the NI-DSP Interface

Utilities

This chapter contains a step-by-step example for building a custom DSP Library, creating a LabVIEW interface to a

custom function, and executing the custom function from the LabVIEW environment. The chapter demonstrates this

concept with an example of how to add a custom function. The custom function finds the maximum and minimum

values in the input array, as well as the respective indices of the occurrence of the maximum and minimum value.

This example custom function also returns an output array with the input array sorted.

Creating Your Custom NI-DSP Library

To build your own custom library, follow these steps:

1. Create your source code of C functions (or assembly functions).

2. Compile and/or assemble source code.

3. Add your object filenames to a linker file (ifile).

4. Add your new function names to a library function list file.

5. Run the Dispatchapplication to generate an assembly dispatch file.

6. Compile, assemble, and link your custom library.

These steps are detailed in the following pages.

Note: You can use the source files included in the Examplesdirectory instead of creating the files in this

example.

1. Create Your Source Code of C Functions

The first step in building a custom DSP Library is to create your WE DSP32C C-callable custom function(s). The

gmaxmin.cexample file in your examplesdirectory contains the source code.

GMaxMin.c Example:

/* DSP_GMaxMin(z,y,max,min,n,imax,imin)

* This function returns the maximum and minimum of array z, as well as the

* respective indices of the occurrence of the maximum and minimum. The output

* array y holds the input array sorted.

#include "atdsp.h"

short DSP_GMaxMin(z,y,max,min,n,imax,imin)

float *y;

NI-DSP SRM for LabVIEW for Windows

2-1

Part 4: NI-DSP Interface Utilities

Getting Started with the NI-DSP Interface Utilities

Chapter 2

long *n,*imin,*imax;

{

long i,j;

float *x, small;

float localmin ,localmax;

if ((*n)<= 0) return(0);

x = z;

localmin = *x;

localmax = localmin;

*imax = 0;

*imin = 0;

for (i=0;i< *n;i++,*x++) {

if (localmax < *x) {

localmax = *x;

*imax = i;

}

if (localmin > *x) {

localmin = *x;

*imin = i;

}

*min = localmin;

*max = localmax;

for(i=0; i< *n; i++) y[i] = z[i];

for(i=0; i< *n; i++){

for(j=i+1; j< *n; j++)

if(y[i] > y[j]) {

small = y[j];

y[j] = y[i];

y[i] = small;

}

return(noError);

}

Note: The file ATDSP.his the header file containing all of the error codes used by the DSP Library. You can

find it in the LIBdirectory. To efficiently and correctly compile WE DSP32C code that uses error codes

defined in ATDSP.h, you may want to copy this header file to the Includedirectory of your WE

DSP32C tools. Otherwise, you should specify the correct path of ATDSP.hin your program.

Guidelines for the Custom Functions

When adding functions to build your custom DSP Library, follow these guidelines:

•

Pass all parameters by address–All input, output, and input/output parameters, whether arrays or scalars, must

be passed by address (pointer). For instance, in the example gmaxmin.c, although the length n of an input

array is only an input scalar, it is passed by address.

•

You must pass parameters in a certain order–Pass all pointers to arrays, then pass all of the pointers to 32-bit

floating-point scalars, and then pass all of the pointers to 32-bit long integer scalars. In gmaxmin.c, the

parameters are passed in the following order–z (input array), y (output array), max (output floating-point

scalar), min (output floating-point scalar), n (input integer scalar), imax (output integer scalar), and imin

(output integer scalar).

Part 4: NI-DSP Interface Utilities

2-2

NI-DSP SRM for LabVIEW for Windows

Chapter 2

Getting Started with the NI-DSP Interface Utilities

•

Parameters must be 32-bit floating-point, 32-bit integer scalars, or pointers to arrays of data (any type)–

Supported parameter types are 32-bit floating-point scalars and multidimensional arrays, and 32-bit integer

scalars and multidimensional arrays.

NI-DSP SRM for LabVIEW for Windows

2-3

Part 4: NI-DSP Interface Utilities

Getting Started with the NI-DSP Interface Utilities

Chapter 2

•

Return a 16-bit short integer error code–Every function should return an integer error code. A list of error codes

that the existing DSP Library returns is given in Appendix A, Error Codes. If any of these error codes are used,

ATDSP.hshould be included at the beginning of your source code. A function that completes with no error

should return noErroror zero.

Use the memory management routines–Perform any dynamic onboard memory allocation using the functions

Alloc_Memand Free_Mem, described in the next section.

DSP Board Memory Management

To write your own customized DSP routines that become integrated with the DSP Library, you need some lower

level memory management routines. Two onboard calls are included with the DSP Library to dynamically allocate

and deallocate onboard memory–Alloc_Memand Free_Mem. Use these functions instead of the standard C

routines mallocand free.

Call the function Alloc_Memwhenever your routines require memory space. Alloc_Memattempts to allocate

memory of the requested size (in bytes) in any available memory bank. If the requested buffer size (in bytes) is not a

multiple of four, the allocation routines ensure alignment to the nearest radix 4 boundary >= NumBytes. This is to

guarantee memory alignment on 4-byte addresses. For example, if you request the allocation of a buffer of size

1,021, 1,022, 1,023 or 1,024 bytes, then the allocation routines allocate 1,024 bytes in all four cases. The allocation

routines first attempt to allocate the buffer in on-chip memory (all DSP boards have 4 kilobytes of on-chip memory).

If the routine fails, the allocation routine tries to allocate the buffer in onboard memory (DSP boards have between

256 kilobytes and 1.5 megabytes of onboard memory). The function and input parameters for Alloc_Memare

defined as follows:

void * Alloc_Mem(size)

long

size;

where sizeis the requested block size in bytes.

Alloc_Memreturns a pointer to the allocated buffer or returns NULLif the allocation failed.

Call the function Free_Memwhen you deallocate memory buffers. Use this function instead of the standard C

function free. The function and input parameters for Free_Mem are defined as follows:

short

long

Free_Mem(ptr)

ptr;

where ptris the address of the buffer to be deallocated.

Free_Memdeallocates the buffer pointed to by ptrif that buffer was allocated previously using Alloc_Mem.

Free_Memreturns NULLif deallocation succeeded and returns an error code if deallocation did not succeed.

Each of the NI-DSP Analysis VIs call a function that is part of the DSP Library that resides on the board. Your own

custom functions can call any of the DSP Library functions included in the NIDSP.fncfile described later in this

chapter. The DSP functions that you can call from your own functions are prototyped in Part 4, Chapter 3, DSP

Board Function Overview.

Note for Assembly Language Programmers: If you are using WE DSP32C assembly language, each function

must accept and return C-style parameters and preserve all

registers used in the WE DSP32C environment. For example, the

function parameters are pushed onto the stack from right to left and

the return value is placed in Register r1. Refer to the WE DSP32C

C Language Compiler Library Reference Manual for details.

Part 4: NI-DSP Interface Utilities

2-4

NI-DSP SRM for LabVIEW for Windows

Chapter 2

Getting Started with the NI-DSP Interface Utilities

2. Compile and/or Assemble Source Code

Compile all new C source files and assemble all new assembly source files using the WE DSP32C C compiler and

assembler. Remember to use only floatand longdata types for scalars. If you use any of the error codes from

Appendix A, Error Codes, include ATDSP.hat the beginning of your source code. A function that completes with

no error should return noErroror zero. The ATDSP.hfile is in the Libdirectory. After you have installed the

WE DSP32C tools, you may want to copy ATDSP.hto the Includedirectory in the root directory for those tools.

In this example, use d3cc -c gmaxmin.cto generate WE DSP32C object code.

3. Add Your Object Filenames to a Linker File (ifile)

The next step is to modify the file NIDSPLNKto add the names of your custom object files to the list of modules to

be linked into your custom DSP Library. NIDSPLNKis a source file containing link editor directives for the WE

DSP32C tools. This modification is illustrated in Figure 2-1.

essent.o

stackbld.o

/*** Add all your file names after this comment line ***/

gmaxmin.o

/*** Add all your file names above this comment line ***/

-lLVDSP

Figure 2-1. Linker File NIDSPLNK

Copy the NIDSPLNKfile from the LIBdirectory to your working directory and then modify the copy. This ensures

that you maintain the original ifile. Also, you need to copy the files essent.o, dsp_glob.o, and

stackbld.s, from the LIBdirectory to your working directory.

Note: Add your function(s) name(s) to this file between the comments instructing you to do so. Do not modify

this file in any other way.

4. Add Your New Function Names to a Library Function List File

Copy the files NIDSP.fncand NIESSEN.fncfrom the Dispatchdirectory to your working directory.

In Figure 2-2, the file NIDSP.fncshows the addition of the function DSP_GMaxMinto the library function list.

This is done by editing the file NIDSP.fncin the current directory. Add this function to a new group, My

Functions, at the end of this file.

** My Functions

DSP_GMaxMin

Figure 2-2. Library Function List File NIDSP.fnc

NI-DSP SRM for LabVIEW for Windows

2-5

Part 4: NI-DSP Interface Utilities

Getting Started with the NI-DSP Interface Utilities

Chapter 2

Notice that functions accepted by the Dispatchapplication should have acceptable C syntax, that is, names may

contain letters, numbers, and the underscore character but must start with a letter or underscore. A function name

can contain up to 31 characters. You must divide your functions into groups with a maximum of 64 (40 hex)

functions per group. The Dispatchapplication returns an error message if function names are not acceptable or if

you have more than 64 functions per group. You may have up to 24 groups of 64 functions each. The Dispatch

application assigns a unique ID number to each function depending on how the functions are grouped. Groups start

at zero and are consecutive (Group 0, 1, 2, ...). The ID number of the first function of a group is the group number

multiplied by 64. The function IDs are then consecutive within that group. The function ID, therefore, has group

information as well as index information for functions within each group. This information is necessary for the

dispatcher to execute a particular function efficiently based on the function ID. The first group of functions are the

Essential Functions group. The Essential Functions are always linked into the DSP Library. These functions are

specified in NIESSEN.fnc. Do not modify this file. In the library function list file, NIDSP.fnc, group

beginnings are marked by two asterisks (** ) followed by a group name. Figure 2-3 shows a typical section of

NIDSP.fnc.

** group name

function name

** group name

function name

Figure 2-3. Typical Section of NIDSP.fnc

Customizing the DSP Library by Deleting Functions

Another way to customize the DSP Library is to delete functions from NIDSP.fnc. This decreases the size of the

DSP Library because the object code for those deleted functions are not included. To customize the DSP Library by

deleting functions, follow the instructions in this chapter, but delete functions rather than add functions. Follow

these guidelines when deleting functions from the DSP Library:

•

When modifying the NIDSP.fnc, replace the name of each function you want to delete with the function name

DSP_NOP. This replacement ensures that the remaining functions in the library are addressable by the

dispatcher using the same function ID. The Interface Library DSP.DLLis built using the organization of the

original library function list file, and thus function IDs are hard coded into the Interface Library. If you

reorganize the original functions and groups without replacing the function name with DSP_NOP, the interface

to the function calls a different DSP Library function, which causes the DSP Library to fail. For example, if

you had a function group with 10 functions and you wanted to delete the first and third functions of that group,

if you replace their names with DSP_NOP, you can ensure that all of the other functions in the group maintain

the same function ID. This ensures that the interface code for those functions remain valid. If you do not

follow these guidelines, functions become disorganized, you could call the wrong function, and an unexpected

error could happen.

•

Never delete any of the function group names that form the original DSP Library that you received with the

NI-DSP software package. The function group names ensure that the remaining functions in the DSP Library

have the same function IDs. If you want to delete all functions of a group, replace each function name of that

group in the NIDSP.fncwith DSP_NOP.

Part 4: NI-DSP Interface Utilities

2-6

NI-DSP SRM for LabVIEW for Windows

Chapter 2

Getting Started with the NI-DSP Interface Utilities

5. Run the Build Dispatch Application to Generate an Assembly Dispatch File

The next step is to generate an assembly dispatch file. In the current directory, run dispatch.exe. You can find

the executable file under the dispatchdirectory. The Dispatchapplication automatically generates the WE

DSP32C assembly code file dispatch.sand a header file dspfncs.hin your current directory.

The dspfncs.hfile has define statements assigning function codes for all the functions in the library. Each

function code is used by the LabVIEW interface when calling the DSP Library. The onboard Kernel uses these

function codes (passed to it by the interface code) to execute the proper functions. Look at the file dspfncs.hto

see the resulting function IDs assigned to your custom functions. The function codes for DSP_GMaxMinis 448.

Remember this code. You will use the code later to build your LabVIEW interface VI. Figure 2-4 shows a section

of dspfncs.hcorresponding to the section given from NIDSP.fnc. Notice that in this example, Signalswas

the seventh group (Group 6) listed in NIDSP.fncand hence the function IDs.

/** Signals

**/

#define DSP_Sine_NUM 384

#define DSP_Square_NUM 385

#define DSP_Sinc_NUM 386

Figure 2-4. Signals Group Section in dspfncs.h

The file dispatch.sis an assembly file that the Kernel uses to execute the proper function. This file has

information about the number of groups and the number of functions per group that are linked into the software that

resides on the board. The Dispatchapplication uses the files NIDSP.fncand NIESSEN.fncto create

dispatch.s. Figure 2-5 shows a section of dispatch.scorresponding to the section given from

NIDSP.fnc. Notice that in this example, Signalswas the seventh group (Group 6) listed in NIDSP.fnc.

Group zero is the group of Essential Functions.

/** Signals

**/

call DSP_Sine(r18)

nop

goto FNC_CALL_END

nop

call DSP_Square(r18)

nop

goto FNC_CALL_END

nop

call DSP_Sinc(r18)

nop

goto FNC_CALL_END

nop:

Figure 2-5. Signals Group Section in dispatch.s

NI-DSP SRM for LabVIEW for Windows

2-7

Part 4: NI-DSP Interface Utilities

Getting Started with the NI-DSP Interface Utilities

Chapter 2

6. Compile, Assemble, and Link Your Custom Library

The last step in building a custom DSP Library is to compile, assemble, and link your custom library using the WE

DSP32C tools. You must have installed those tools as described in the WE DSP32C Support Software Library User

Manual. You must copy the file LIBLVDSP.afrom the LIBdirectory in the stand-alone root directory to the LIB

directory of the WE DSP32C tools as set by the environment variable recommended in the WE DSP32C Support

Software Library User Manual. Table 2-1 lists the files you need to build the custom NI-DSP example.

Table 2-1. Files Required to Build the Custom DSP Library Example

File

Description

makelib.bat

gmaxmin.c

dspfncs.h

dispatch.s

NIDSPLNK

Batch file to build the custom library

C source file

C include file generated by Dispatch

Assembly file generated by Dispatch

Modified linker command file

Archived library of object modules for the DSP Library functions

for LabVIEW

LIBLVDSP.a*

libc32c.a*

libap32c.a*

libm32c.a*

AT&T C library

AT&T Application library

AT&T Math library found in the LIBdirectory in the path set by

the environment variable

All these archives must be in the LIBdirectory in your WE DSP32C tools directory.

Copy the file makelib.batfrom the LIBdirectory. This is a batch file that rebuilds the DSP Library file to

reflect the customization and changes needed. You must enter in the memory map file names to this .batfile to

use in linking the DSP Library file, as well as the name of the output file for the DSP Library file. The two memory

map files used for linking the DSP Library files that come with NI-DSP, dsp64.mapand dsp.map, are located in

the LIBdirectory. Copy the memory map file that you will use to your current directory. The memory map files

are used when linking the DSP Library files LV2200S.outand LV2200.out, respectively. You can rename

these files and move them to other directories. If you want the driver to load this new DSP Library file, make sure

you run WDAQCONFas described in the section titled Board Configuration in Part 1, Getting Started with NI-DSP.

You should use one of the memory map files, dsp64.mapor dsp.map, as input to makelib.bat. If the size of

your library cannot be linked using the section sizes specified in these memory map files, you must edit these files

accordingly. Refer to the WE DSP32C Support Software Library User Manual for more information.

To begin rebuilding the customized DSP Library, type the following command:

makelib memmap libfile

where

memmap

is the name of the memory map file used

libfile is the name of the desired DSP Library file name

For example,

makelib dsp64 LV2200S

Do not use the extension names .mapand .outfor the fields memmapand libfile.

Part 4: NI-DSP Interface Utilities

2-8

NI-DSP SRM for LabVIEW for Windows

Chapter 2

Getting Started with the NI-DSP Interface Utilities

The makelib.batbatch file performs the following:

1. makelib.batassembles the file stackbld.sin the current directory using the AT&T d3as.exe

assembler. You need to assemble this file because it includes the file stackbld.s, created by the

Dispatchprogram, and contains the assembly code that determines which function the DSP board executes.

The file dispatch.sreflects the organization and content of the DSP Library file as is reflected by the library

function list file, NIDSP.fnc, in the current directory.

2. makelib.batrebuilds the DSP Library file with name libfile.out(the name you decided to give to the

new DSP Library file). This new DSP Library file always has the Kernel, memory management routines,

execution control routines, data communication routines, interrupt handling routines for data acquisition, and

the functions listed in the library function list file NIDSP.fnc.

3. makelib.batlists all undefined symbols from the DSP Library file. These symbols could be undefined

because you did not link all the object modules.

Any other error messages that are encountered are WE DSP32C error messages. Refer to the WE DSP32C Support

Software Library User Manual for information.

At this point, the WE DSP32C tools have generated a custom DSP Library file.

Creating Your LabVIEW Interface

Now that you have added a function to the DSP Library, you need to be able to call this function from LabVIEW

using a VI. The NI-DSP Analysis VIs contains a VI called Custom VI in the Utility folder. You can use the Custom

VI to call any function on the DSP board that follows the guidelines listed in the Guidelines for the Custom

Functions section earlier in this chapter.

To create your LabVIEW interface, follow these steps:

1. Bundle all the input parameters to arrays.

2. Call the Custom VI.

3. Index the output arrays to get the results.

1. Bundle All of the Input Parameters to Arrays

Use the Custom VI in the Utility folder of the DSP2200 folder as the interface for calling custom functions on the

DSP board from LabVIEW. You can use three types of data for input/output parameters in the Custom VI. The data

types correspond to the different types of parameters that you can use in your custom functions on the DSP board.

Custom functions have three groups of parameters–pointers to arrays of data (any type; 32-bit floating point, 32-bit

integer, and 16-bit integer), pointers to 32-bit floating-point scalars, or pointers to 32-bit integer scalars. All of the

parameters in a group must be the same type. You must arrange the custom function parameters so that the group of

all array pointers (if the function has any) are first, followed by the group of all pointers to 32-bit floating-point

scalars (if the function has any), followed by the group of all pointers to 32-bit integer scalars. The LabVIEW

interface to such a custom function, Custom VI, should reflect the same parameter order.

The number of parameters varies for different functions. To use the Custom VI as the interface for different custom

functions, you must bundle the same types of parameters to an array. LabVIEW can check the array sizes in the CIN

to determine how many parameters to pass to a custom function on the DSP board from LabVIEW. For this reason,

Custom VI uses arrays as input/output parameter types. The number of parameters of a custom function is the sum

of the sizes of three arrays of the Custom VI.

NI-DSP SRM for LabVIEW for Windows

2-9

Part 4: NI-DSP Interface Utilities

Getting Started with the NI-DSP Interface Utilities

Chapter 2

The array of DSP Handle Clusters holds all the references to arrays of data used by the custom DSP functions. You

should bundle the DSP Handle Clusters together to an array in the order they appear in the first group of parameters

in the custom function on the DSP board. For example, if the first three parameters of the custom function are the

arrays X, Y, and Z, which correspond to the DSP Handle Clusters in LabVIEW, hdl1, hdl2, and hdl3, respectively,

then you must bundle hdl1, hdl2, and hdl3 to an array in that order. The same order requirements apply to the array

of 32-bit floating-point scalars that holds the LabVIEW parameters corresponding to the second group of the

parameters in the custom function. Likewise, the array of 32-bit integer scalars that holds the LabVIEW parameters

corresponding to the third and final group of parameters in the custom function must also have the same order

requirements as the other arrays.

Bundle all of the parameters of the same type to an array before you call Custom VI. Remember that the order in

which you bundle the parameters should correspond to the order of the terminals in the corresponding group of the

custom function. The first element in the array representing a group should correspond to the first parameter of the

same group in the custom function, and so on.

The Custom VI is the interface from LabVIEW to all of the custom functions on the DSP board. Each parameter of

the custom functions on the DSP board require two corresponding terminals in LabVIEW–an input control to

reserve the space on the DSP board for the data, and an output indicator to hold the result. All of the output arrays in

the Custom VI are internally connected to the corresponding input arrays.

The Custom VI treats all of the parameters in the custom functions as input/output parameters. The parameters that

represent outputs of the custom function must also be bundled in LabVIEW in the array that holds all of the other

input parameters of the same type. Although the output parameters, whether scalars or arrays, do not contain any

valid or useful information on input, they will be overwritten by the return values. The parameters that represent

inputs to the custom functions also exist in the corresponding output arrays, although the values of the output arrays

are the same values as on input–the custom function will not change the parameter values.

Figure 2-6 shows how to bundle the parameters for gmaxmin.c. Allocate two DSP Handle Clusters corresponding

to the input array z and the output array y. The bundling order is as follows–the DSP Handle Cluster representing z

is first, followed by the DSP Handle Cluster representing y. y is the output array. The data in this array is not valid

as an input, but you must still allocate it first and bundle it with other DSP Handle Clusters that represent input

arrays. Notice that although the parameters max value, min value, max index, and min index are outputs only in

gmaxmin.c, you must first bundle the additional four input scalars with the actual input parameters to reserve

space in memory for the parameters on output. The initial values are not important. The parameter n (the number of

elements) is only an input parameter in gmaxmin.c, but you will still have an output corresponding to this

parameter, although the output and input values are the same.

Figure 2-6. How to Bundle Parameters in LabVIEW to Call gmaxmin.c

Part 4: NI-DSP Interface Utilities

2-10

NI-DSP SRM for LabVIEW for Windows

Getting Started with the NI-DSP Interface Utilities

Chapter 2

2. Call the Custom VI

After you bundle all of the parameters to arrays, connect each array to the corresponding terminals of the Custom

VI. Figure 2-7 shows how to connect to the Custom VI for the gmaxmin.c.

Figure 2-7. How to Connect to Custom VI to Call gmaxmin.c

If your custom function does not have a certain type of parameter, leave the corresponding terminal unwired. This is

important for passing your parameters correctly from LabVIEW to the DSP board.

The value of the function ID is the custom function ID that you obtain from dspfncs.h. This ID determines

which custom function on the DSP board the Custom VI calls. To call the correct function on the DSP board, you

must supply the appropriate function ID. In this example, the function ID is 448, which you obtained in step 5 of

creating your custom NI-DSP Library. Finally, you must indicate the slot number on which to execute the custom

VI. In Figure 2-6, the slot parameter terminal is left unwired. This causes all VIs to execute on the default DSP

board, which is set to slot 3.

3. Index the Output Arrays to Obtain the Results

All of the output results are in the output arrays of the Custom VI. You can read the output arrays to see the results.

If you want to operate on some output parameters, use the method for indexing an array to extract an element from

an array. The order of the output parameters in an output array is identical to the order in the corresponding input

array.

Figure 2-8 shows how to index the output arrays of the Custom VI to obtain the results of gmaxmin.c. The

Custom VI leaves the output data buffers on the board, and returns only the DSP Handle Clusters that indicate the

locations of these buffers on the board. You must use the Copy Mem(DSP to LV) VI to copy data back to

LabVIEW. The Custom VI always copies scalars back to LabVIEW.

Part 4: NI-DSP Interface Utilities

2-11

NI-DSP SRM for LabVIEW for Windows

Chapter 2

Getting Started with the NI-DSP Interface Utilities

Figure 2-8. Block Diagram–How to Index the Output Arrays of the Custom VI

to Obtain Results of gmaxmin.c

Figure 2-9 shows the whole block diagram that uses the Custom VI to call the custom function gmaxmin.con the

DSP board from LabVIEW. Figure 2-10 shows the front panel of the block diagram shown in Figure 2-9.

Figure 2-9. Block Diagram–Using the Custom VI to Call gmaxmin.con the

DSP Board from LabVIEW

NI-DSP SRM for LabVIEW for Windows

2-12

Part 4: NI-DSP Interface Utilities

Getting Started with the NI-DSP Interface Utilities

Chapter 2

Figure 2-10. Front Panel–Using the Custom VI to Call gmaxmin.con the

DSP Board from LabVIEW

At this point, finish creating the VI interface to call your custom function on the DSP board from LabVIEW. You

can customize this VI as a subVI using the method described in the LabVIEW User Manual if you want to use this

VI from your other applications in LabVIEW.

Executing the Custom Function from LabVIEW

Run this VI from LabVIEW. Change the input array and display the results.

Part 4: NI-DSP Interface Utilities

2-13

NI-DSP SRM for LabVIEW for Windows

Chapter 3

DSP Board Function Overview

This chapter contains an overview of the prototypes of the C-callable NI-DSP Analysis functions on the DSP board

that you can use in your custom programs.

Every NI-DSP Analysis VI calls a function on the DSP board. When you write your own custom functions, you can

call these functions from your program. These C functions include numerical analysis, signal generation, digital

signal processing, digital filtering, windowing, and memory management.

For more information about these functions, refer to the NI-DSP Software Reference Manual for DOS/LabWindows.

Note: The functions listed in this chapter are the same functions described in the NI-DSP Software Reference

Manual for DOS/LabWindows, except for the following differences:

•

For all of the functions, ignore the slot number parameter specified in the tables of parameters. The

slot number is not meaningful when you call these functions from the DSP board.

The integer parameters are all 32 bits long. The corresponding parameter type in the NI-DSP Software

Reference Manual for DOS/LabWindows is 16-bit integer. When you call these functions in your

custom program, use the 32-bit long type defined in this manual. Use the data types defined in this

manual rather than the data types for similar functions listed in any other manual. The parameter order

and meaning and the purpose of the functions, however, are equivalent with the functions described in

the NI-DSP Software Reference Manual for DOS/LabWindows.

•

The functions listed are all of the functions that you can call from your custom program. Not all the

functions described in the NI-DSP Software Reference Manual for DOS/LabWindows are available in

NI-DSP for LabVIEW for Windows.

The following is a list of the prototypes of each function that you can customize for use on the DSP board.

Signals

short DSP_Gaussian (long n, float sDev, float* noise, float seed)

short DSP_ImpTrain (long n, float amp, long delay, long period, float * x)

short DSP_Impulse (long n, float amp, long index, float * x)

short DSP_Pulse (long n, float amp, long delay, long width, float *x)

short DSP_Ramp (long n, float first, float last, float * x)

short DSP_Sawtooth (long n, float amp, long delay, float cycles, float * x)

short DSP_Sinc (long n, float amp, float delay, float dt, float *x)

short DSP_Sine (long n, float amp, float phase, float cycles, float * x)

short DSP_Square (long n, float amp, float duty, float cycles, float * y)

short DSP_Triangle (long n, float amp, float delay, float width,

float dt, float * x)

short DSP_TriTrain (long n, float amp, long delay, float cycles, float * x)

short DSP_Uniform (long n, float * noise, float seed)

short DSP_WhiteNoise (long n, float amp, float * noise, float seed)

Frequency Domain

short DSP_CrossPower (float * x, float * y, long n)

short DSP_CxFFT (float * x, float * y, long n)

short DSP_FHT (float * x, long n)

short DSP_InvFFT (float * x, float * y, long n)

short DSP_InvFHT (float * x, long n)

short DSP_ReFFT (float * x, float * y, long n)

short DSP_Spectrum (float * x, long n)

short DSP_ZeroPad (float * x, long n, long size)

NI-DSP SRM for LabVIEW for Windows

3-1

Part 4: NI-DSP Interface Utilities

DSP Board Function Overview

Chapter 3

Time Domain

short DSP_Convolution (float * x, long n, float * y, long m, float * cxy)

short DSP_Correlation (float * x, long n, float * y, long m, float * rxy)

short DSP_Decimate (float * x, long n, long decFact, long ave, float * y)

short DSP_Deconvolution (float * cxy, long n, float * y, long m, float * x)

short DSP_Difference (float * x, long n, float dt, float xInit,

float xFinal, float * y)

short DSP_Integrate (float * x, long n, float dt, float xInit,

float xFinal, float * y)

Filters

short DSP_Bw_Coef

(float fs, float f_low, float f_hi, long order,

float * a, float * b, long type)

short DSP_Ch_Coef

(float fs, float f_low, float f_hi, long order,

float ripple, float * a, float * b, long type)

short DSP_Elp_Coef (float fs, float f_low, float f_hi, long order, float

ripple, float atten, float * a, float * b, long type)

short DSP_EqRip_BPF (float * x, long n, long taps, float fs, float fh1,

float fl2, float fh2, float fl3, float * y)

short DSP_EqRip_BSF (float * x, long n, long taps, float fs, float fh1,

float fl2, float fh2, float fl3, float * y)

short DSP_EqRip_HPF (float * x, long n, long taps, float fs, float fh1,

float fl2, float * y)

short DSP_EqRip_LPF (float * x, long n, long taps, float fs, float fh1,

float fl2, float * y)

short DSP_IIR_Filter (float * x, long n, float * a, float * Condx, long sza,

float * b, float * Condy, long szb, float * y)

short DSP_InvCh_Coef (float fs, float f_low, float f_hi, long order, float

ripple, float *a, float * b, long type)

short DSP_Median_Filter (float * x, long n, long rank, float * y)

short DSP_Parks_McClellan (long n, float fs, long bands, float * Amp, float

* fLow, float * fHi, float * wRipple, long

filterType, float * h, float * delta)

Windows

short DSP_CosWin(float * x, long n, long type)

short DSP_ExpWin (float * x, long n, float finalval)

short DSP_ForceWin (float * x, long n, float duty)

short DSP_GenCosineWin (float * x, long n, float * coeff, long m)

short DSP_KsrWin (float * x, long n, float beta)

short DSP_TriWin (float * x, long n)

Array Functions

short DSP_Abs (float * x, long n, float * y)

short DSP_Add (float * x, float * y, long n, float * z)

short DSP_Clip (float * x, long n, float upper, float lower, float * y)

short DSP_Div (float * x, float * y, long n, float * z)

short DSP_LinEv (float * x, long n, float a, float b, float * y)

short DSP_Log (float * x, long n, float mult, float * y)

short DSP_MaxMin

(float * x, long n, float * max, long * imax, float *

min, long * imin)

short DSP_Mul (float * x, float * y, long n, float * z)

short DSP_PolyEv (float * x, long n, float * coeff,long k, float * y)

short DSP_Prod (float * x, long n, float * prod)

short DSP_Reverse (float * x, long n)

short DSP_Set (float * x, long n, float a)

short DSP_Shift (float * x, long n, long shift, float * y)

short DSP_Sort (float * x, long n, long direction, float * y)

Part 4: NI-DSP Interface Utilities

3-2

NI-DSP SRM for LabVIEW for Windows

Chapter 3

DSP Board Function Overview

short DSP_Sqrt (float * x, long n, float * y)

short DSP_Sub (float * x, float * y, long n, flaot * z)

short DSP_Subset (float * x, long n, long index, long length, float * y)

short DSP_Sum (float * x, long n, float * sum)

short DSP_Unwrap (float * x, long n, float * y)

short DSP_To Polar (float * x, float * y, long n, float * mag,

float * phase)

short DSP_To Rect (float * mag, float * phase, long n, float * x,

float * y)

Memory Management Data Transfer

void * Alloc_Mem (long numbytes)

short DSP_CopyMem

(void * Source, void * Destination, long type, long

count)

short Free_Mem (long ptr)

Data Acquisition Functions

Some low-level data acquisition NI-DSP functions are also available for you to call in an application on the DSP

board. You can use these low-level data acquisition functions to develop acquisition and waveform generation

applications that occur on the DSP board, such as acquiring and sending data over the input and output channels. If

you use any of the low-level data acquisition functions in your custom functions, please include the dspdaq.hfile

in your program. dspdaq.hcontains all the prototypes of functions and the data structures that the data

acquisition functions use. You can find this header file in the LIBdirectory. For more information about these

functions, refer to Part 4, Chapter 5, Data Acquisition Functions for the AT-DSP2200, in the NI-DSP Software

Reference Manual for DOS/LabWindows. All of the information in that chapter applies to the NI-DSP software for

LabVIEW for Windows.

NI-DSP SRM for LabVIEW for Windows

3-3

Part 4: NI-DSP Interface Utilities

Chapter 4

Using the DMA VIs

____________________________________________________________________________________________

This chapter describes two special VIs that transfer data between the host computer and the DSP board without

interfering with the DSP board.

Most VIs in the NI-DSP library run sequentially on the DSP board. In other words, if you run a VI from LabVIEW,

even though the actual function is running on the board, LabVIEW must wait for the function to finish before it can

do anything else. If you have a custom function that runs indefinitely on the board, LabVIEW will continue to wait.

However, every VI has the ability to check for timeout errors, and you can use the DSP Timeout VI to resolve this

problem by setting a timeout limit. The host will wait for the DSP board to finish an operation for only the amount

of time you designate; if the DSP board does not complete the function in the specified amount of time, the VI will

return a timeout error. If your custom function runs a long time or indefinitely, you can set a very short timeout

period and ignore the timeout error in your LabVIEW program. However, under these conditions you normally lose

communication with the DSP board until you reset it with the DSP Reset VI.

Two VIs–DSP DMA Copy(LV to DSP) and DSP DMA Copy(DSP to LV)–solve this communication problem.

These two VIs can transfer data between the host and the DSP board using onboard DMA without interfering with

the DSP board. You can use the DSP DMA Copy(LV to DSP) VI to send data to a designated memory location on

the DSP board from LabVIEW when the DSP board is still running. The function running on the board can read

data from that location and update operations that depend on that data. LabVIEW can also read data back from the

DSP board using the DSP DMA Copy(DSP to LV) VI to check the status of the board or read results while the board

is still running. In other words, you can still communicate with the board.

When you use the DSP DMA Copy(DSP to LV) and DSP DMA Copy(LV to DSP) VIs, remember the following:

•

These two VIs require a real DSP address to transfer data. You can determine this address in two ways:

Give a valid DSP address, such as FFF000 hex, if you know the exact DSP memory address of a buffer or

variable that your DSP program will use.

Allocate a normal DSP Handle Cluster and use the DSP Handle To Address VI to obtain the valid DSP

address for this DSP Handle Cluster. You can use this DSP address in your program on the DSP board as

you would use any pointer. Remember to allocate the DSP Handle Cluster before you call your custom

function through the DSP Custom Function VI.

•

The data format will not change when you transfer data. If the data is floating point, you need to convert the

data format in your code between DSP format and IEEE format. When you send floating-point data from

LabVIEW to the DSP board, convert the data to DSP format before using it in your program on the DSP board.

When you copy floating-point data back from the DSP board to LabVIEW, convert the data to IEEE format

before you save it to the DSP memory location from which you will copy data.

Do not use data acquisition VIs to acquire data to PC memory or send data from PC to DSP memory while

using these two VIs, because these VIs use the same DSP board register, PDR, to transfer data.

Two examples shipped with this package show you how to make use of these VIs. The examples are installed in the

Examplessubdirectory of your NI-DSP package.

NI-DSP SRM for LabVIEW for Windows

4-1

Part 4: NI-DSP Interface Utilities

Using the DMA VIs

Chapter 4

The first example is a simple spectral analyzer. The main VI is called Analyzer VI. It uses the DSP Custom

Function VI to call a custom function running on the board. The scheme is as follows:

•

Custom function on the board:

1. Set the data acquisition parameter and initialize the board.

2. Start the data acquisition.

while(TRUE) {

when data is ready, copy it to the process buffer

read processing information from DSP memory location 1

depending on the processing information

apply the selected window to the data

apply the selected filter to the data

compute the FFT magnitude or Power Spectrum

Transfer the data format to IEEE format.

Save processed data in DSP memory location 2.

}

•

In LabVIEW:

1. Initialize the board and set parameters.

2. Allocate DSP Handle Cluster 1; get the address for this handle (this is DSP memory location 1).

3. Allocate DSP Handle Cluster 2; get the address for this handle (this is DSP memory location 2).

4. Set a timeout limit to a very small number (2 ms).

5. Pass DSP Handle Cluster 1 and 2 with the other necessary parameters to the DSP Custom Function VI and

call the Analyzer VI.

6. Ignore the timeout error returned from the DSP Custom Function VI.

While(STOP button not pressed) {

send new processing information (you can change certain processing

controls from the front panel) to DSP memory location 1 using the

DSP DMA Copy(LV to DSP) VI.

copy the processed data from DSP memory location 2 to LabVIEW and

plot the data.

}

7. Call the DSP Reset VI to reset the DSP board.

The second example is the audio equalizer example. The VI loads the audio equalizer kernel called audio.out

using the DSP Load VI. It uses the DSP DMA Copy(LV to DSP) VI to send the new gains for each channel to the

DSP board and uses the DSP DMA Copy(DSP to LV) VI to copy the processed audio data back to LabVIEW and

then plots the data.

You can find all of the source codes and VIs for these two examples in the Examplessubdirectory of your NI-DSP

package.

These examples show you how to use the DSP DMA Copy(DSP to LV) and DSP DMA Copy(LV to DSP) VIs to

communicate between the DSP board and the host computer. They also show you how to run parallel DSP and

LabVIEW applications, which makes the board more flexible and powerful. With a well-designed system, you can

gain both speed and flexibility.

Part 4: NI-DSP Interface Utilities

4-2

NI-DSP SRM for LabVIEW for Windows

Chapter 4

Using the DMA VIs

DSP DMA Copy(DSP to LV)

Copies a buffer of size elements

from source(DSP Address) on

the DSP board to one of the

destination arrays in LabVIEW

using the onboard DMA transfer

method.

This VI uses only the DMA controller on the DSP board to transfer data. Therefore, it does not interfere with the

other program that is running on the DSP board.

Note: This VI does not convert floating-point data to the IEEE format while copying. When you copy floating

data, convert to IEEE format in your program on the DSP board. However, the Copy Mem(DSP to LV) VI

automatically converts data format during the copying procedure.

To copy data correctly from the DSP board to LabVIEW, you must indicate what type of data is stored in the

source(DSP Address) buffer on the DSP board, set the destination type to the appropriate type, and wire to the

corresponding destination terminal. This VI has three destination types–float (32-bit), short (16-bit), and long

(32-bit). Remember, you must wire only one destination terminal.

slot is the board ID number. slot defaults to 3.

source(DSP Address) is the actual DSP address from which you copy your data. It should be a

24-bit integer (such as FFF000 hex) less than or equal to FFFFFF hex. If you know the exact DSP

address where your data is stored, use this address directly. Alternatively, you can use the DSP

Handle To Address VI to convert a DSP Handle Cluster to an actual DSP address.

size indicates the amount of data you want to copy back from the source(DSP Address) buffer on

the DSP board to the destination buffer in the LabVIEW. size defaults to 0.

destination type indicates the type of the data in the source(DSP Address) buffer on the DSP board.

It has three options:

0: 32-bit floating-point.

1: 16-bit short integer.

3: 32-bit long integer.

destination type defaults to 32-bit floating-point.

destination(float) is the terminal to which you must wire the destination buffer if the data you want

to copy is a 32-bit floating-point data array. Remember, the program on the DSP board must convert

this data to IEEE format before it is copied.

destination(short) is the terminal to which you must wire the destination buffer if the data you want

to copy is a 16-bit short data array.

destination(long) is the terminal to which you must wire the destination buffer if the data you want

to copy is a 32-bit long data array.

error in(no error) contains the error information from a previous VI. If an error occurs, it is passed

out error out and no other calls are made.

error out contains the error information for this call.

NI-DSP SRM for LabVIEW for Windows

4-3

Part 4: NI-DSP Interface Utilities

Using the DMA VIs

Chapter 4

DSP DMA Copy(LV to DSP)

Copies the data in the LabVIEW array

source to the destination(DSP

Address) on the DSP board using the

onboard DMA method.

This VI uses only the DMA controller

on the DSP board to transfer data.

Therefore, it does not interfere with

the other program that is running on

the DSP board unless you copy the

data to the DSP memory location that

is being used by the other program.

Note: This VI does not convert floating-point data to DSP format while copying. When you copy floating data,

convert to DSP format in your program on the DSP board before using this data. However, the Copy

Mem(LV to DSP) VI automatically converts data format during the copying procedure.

The source buffer can contain one of three kinds of data–float (32-bit), short (16-bit), and long (32-bit). To copy

different types of data, you must wire the source data buffer to the appropriate source terminal. For example, if you

want to copy floating data, you must wire the data buffer to the terminal source(float). You must wire only one

source terminal. The destination buffer must be large enough to contain all of the data from the source buffer.

slot is the board ID number. slot defaults to 3.

source(float) is the terminal to which you must wire the source buffer if the data you want to copy is

a 32-bit floating-point data array. Remember, the program on the DSP board must convert this data

to DSP format before using it.

source(long) is the terminal to which you must wire the source buffer if the data you want to copy is

a 32-bit long data array.

source(short) is the terminal to which you must wire the source buffer if the data you want to copy

is a 16-bit long data array.

destination(DSP Address) is the actual DSP address to which you want to copy your data. It should

be a 24-bit integer (such as FFF000 hex) less than or equal to FFFFFF hex. If you know the exact

DSP address where you want to store the data, you can use this address directly. Alternatively, you

can use the DSP Handle To Address VI to convert a DSP Handle Cluster to an actual DSP address.

error in(no error) contains the error information from a previous VI. If an error occurs, it is passed

out error out and no other calls are made.

error out contains the error information for this call.

Part 4: NI-DSP Interface Utilities

4-4

NI-DSP SRM for LabVIEW for Windows

Appendix A

Error Codes

This appendix contains a list of the error codes returned by the NI-DSP Analysis VIs and the corresponding error

messages.

Error Conditions

If an error condition occurs during execution of any of the VIs in the NI-DSP Analysis, the VI returns an error code.

This code is a value which indicates the type of error that occurred. The currently defined error codes and their

associated meanings are given in Table A-1. For error codes other than those listed, refer to the NI-DAQ Software

Reference Manual for DOS/Windows/LabWindows.

Table A-1. NI-DSP Analysis Library Error Codes

Error

Error Name

Description

Number

-10401

-10005

-10403

-10444

-10243

-10247

-10459

-21204

-21205

-21206

-21207

-21208

-21209

noError

No error; the call was successful.

unknownDeviceErr

badDeviceErr

deviceSupportErr

memFullErr

The board specified is not a National Instruments DSP board.

The board number used in function call should be 1<board<8.

Function cannot be executed by specified board.

Insufficient memory or disk space.

configFileErr

cmosConfigErr

DLLInterfaceErr

SamplesGTZero

SamplesGEZero

SamplesGETwo

SizesGTZero

Board configuration file not found.

EISA system configuration invalid.

The DLL could not be called due to an interface error.

The number of samples must be greater than zero.

The number of samples must be greater than or equal to zero.

The number of samples must be greater than or equal to two.

The sizes of the input sequences must be greater than zero.

The size of x must be greater than or equal to the size of y.

xSizeGEySize

OutSizeGEInSize

The output array size must be greater than or equal to the input

array size.

-21210

-21212

OutOfMem

There is not enough space left on the DSP board for onboard

processing.

DecFactErr

The decimating factor must meet: 0 < decimating factor ≤

samples.

-21213

-21214

-21215

-21216

-21217

WidthLTSamples

IndexLTSamples

DelayGEZero

The width must meet: 0 < width < samples.

The index must meet: 0 ≤ index < samples.

The delay must be greater than or equal to zero.

The width must be greater than or equal to zero.

WidthGEZero

DelayWidthErr

The following condition must be met:

0 ≤ (delay + width) < samples.

(continues)

A-1

NI-DSP SRM for LabVIEW for Windows

Error Codes

Appendix A

Table A-1. NI-DSP Analysis Library Error Codes (Continued)

Error

Error Name

Description

Number

-21218

-21219

-21220

-21221

WinDutyCyclesErr

dtGTZero

The window duty cycle value must be between 0.0 and 100.0.

dt must be greater than zero.

DutyCycleErr

CyclesErr

The duty cycle must meet: 0 ≤ duty cycle ≤ 100.

The number of cycles must be greater than zero and less than or

equal to the number of samples.

-21223

-21225

-21226

-21227

-21228

-21229

UpperGELower

IntervalNumErr

MixedSignErr

SizeGTOrder

OrderGTZero

ConvSizeErr

The upper value must be greater than or equal to the lower value.

The number of intervals must be greater than zero.

The sign of y in Ax=y values must be all positive or negative.

The array size must be greater than the order.

The order must be greater than zero.

The number of elements of convolved arrays should be greater

than zero.

-21233

-21234

NyquistErr

Nyquist2Err

The cutoff frequency, fc, must meet: 0 ≤ fc≤ fs/2.

The following conditions must be met:

0 ≤ f_low ≤ f_high ≤ fs/2.

-21235

-21236

-21237

-21242

-21243

-21245

-21246

-21248

-21249

-21250

-21251

-21297

RippleGTZero

AttenGTZero

WidthGTZero

SizeGTZero

NullVectorErr

AttenGTRipple

StepSizeErr

LeakErr

The ripple amplitude must be greater than zero.

The attenuation must be greater than zero.

The width must be greater than zero.

The size of the input sequence must be greater than zero.

The input vector is null. The unit vector does not exist.

The attenuation must be greater than the ripple amplitude.

The step-size parameter µ must meet: 0 ≤ µ ≤ 0.1.

The leakage coefficient, Leak, must meet: 0 ≤ Leak ≤ µ.

The filter cannot be designed with the specified input parameters.

The number of coefficients must be greater than or equal to one.

The rank of the filter must meet: 1 ≤ (2 rank +1) ≤ size.

FilterDesignErr

LenGEOne

RankErr

NotPowerOfTwo

The size of the input array must be a valid power of two:

size = 2^m, 0 < m < 14.

-21301

MEM_HRDWARE_ERR

The onboard hardware diagnostic software found a memory

hardware error. Running functions on the board is unreliable.

-21302

-21303

ZERO_ALLOCATION

TooBigAnOffset

There are no allocated buffers on the DSP board.

The indexing requested falls outside of the buffer into which

indexing is requested.

-21304

-21305

HandleNotFound

TooManyAllocs

The DSP Handle specified is not found on this board: it has

either been freed or never allocated.

The Memory Look Up Table is full: no more allocations allowed

before freeing up some DSP handles (maximum allowed = 128).

-21306

-21307

PtrNotFound

ParamSizeErr

The DSP address pointer could not be found to free the buffer.

One of the parameters does not have enough space allocated for

it on the board.

(continues)

NI-DSP SRM for LabVIEW for Windows

A-2

Appendix A

Error Codes

Table A-1. NI-DSP Analysis Library Error Codes (Continued)

Error Name Description

Error

Number

-21308

-21309

-21310

TransferSizeErr

NotDSPHandle

The size of requested block transfer does not have enough space

allocated for it on the board.

The DSP Handle specified is not valid for this board:

it has either been freed or never allocated.

HandleSlotErr

The DSP Handle used does not belong to the same board on

which the function request is made.

-21311

-21312

HandleAllocErr

TimeOutErr

Failed to allocate a handle in the interface code.

The Kernel did not respond or finish function execution in the

selected timeout.

-21313

-21314

COFF_FilePathErr

CoffHdrErr

Could not find the DSP Library file in the directory set by the

DAQCONFutility.

The file you are attempting to load to the DSP board is not

created using the WE DSP32C tools (not an acceptable

COFF file).

-21315

-21317

CoffSectErr

AddrSpaceErr

Found an unimplemented section flag in the COFF file.

The COFF file you are trying to load is linked with a memory

map that exceeds the maximum memory space allowed by the

WE DSP32C chip (2²⁴).

-21318

NoBrdRespToTransfer

The DSP board on which you are trying to load the COFF file is

not responding to the transfer of the file. Check if there is a

board in the slot desired and verify with the DAQCONFutility.

-21324

-21325

FNC_ERR

The function ID specified does not correspond to a function that

is part of the DSP Library currently running on the board.

GROUP_ERR

The function ID specified does not belong to any group that is

part of the DSP Library currently running on the board.

-21331

-21332

-21333

NotFindFunctionInDLL

LoadDLLLibErr

An error occurs when loading dsp.dlllibrary in CIN.

An error occurs when loading a function in dsp.dll in CIN.

ErrInGetFunctionHandle An error occurs when calling the

GetIndirectFunctionHandlefunction in CIN.

-21334

InvalidBytesSelect

bytes/element selector can only select between 4 bytes(0) and 2

bytes(1).

-21336

-21337

AllocArrayFailed

InvalidArrayType

Failed to allocate a LabVIEW array.

The array type can only be 32-bit floating point data(0), 16-bit

long integer data(1) or 32-bit short integer data(2).

-21338

-21339

PathnameTooLong

InvalidCopyType

The path name should be less than 256 characters.

The entire/partial copy selector can only select between entire

copy (0) and partial copy(1).

-21340

-21341

-21342

CopyNumGTZero

OffsetGEZero

The number of data to copy should be greater than zero.

The offset should be greater than or equal to zero.

IndexSizeGTZero

The size of the New DSP Handle Cluster should be greater than

zero.

(continues)

A-3

NI-DSP SRM for LabVIEW for Windows

Error Codes

Appendix A

Table A-1. NI-DSP Analysis Library Error Codes (Continued)

Error Name Description

Error

Number

-21343

IndexSizeOffsetErr

The size+offset should be less than or equal to the size of the

DSP Handle Cluster that you index into.

-21344

-21345

InvalidDataType

NumBytesGTZero

The data type can only be 4 bytes long(0) or 2 bytes long(1) .

The number of bytes to allocate should be greater than zero.

NI-DSP SRM for LabVIEW for Windows

A-4

Appendix B

Customer Communication

For your convenience, this appendix contains forms to help you gather the information necessary to help us solve

technical problems you might have as well as a form you can use to comment on the product documentation. Filling

out a copy of the Technical Support Form before contacting National Instruments helps us help you better and faster.

National Instruments provides comprehensive technical assistance around the world. In the U.S. and Canada,

applications engineers are available Monday through Friday from 8:00 a.m. to 6:00 p.m. (central time). In other

countries, contact the nearest branch office. You may fax questions to us at any time.

Corporate Headquarters

(512) 795-8248

Technical support fax:

(800) 328-2203

(512) 794-5678

Branch Offices

Australia

Austria

Belgium

Denmark

Finland

France

Germany

Italy

Phone Number

(03) 879 9422

(0662) 435986

02/757.00.20

45 76 26 00

(90) 527 2321

(1) 48 14 24 00

089/741 31 30

02/48301892

(03) 3788-1921

03480-33466

32-848400

Fax Number

(03) 879 9179

(0662) 437010-19

02/757.03.11

45 76 71 11

(90) 502 2930

(1) 48 14 24 14

089/714 60 35

02/48301915

(03) 3788-1923

03480-30673

32-848600

Japan

Netherlands

Norway

Spain

Sweden

Switzerland

U.K.

(91) 640 0085

08-730 49 70

056/20 51 51

0635 523545

(91) 640 0533

08-730 43 70

056/27 00 25

0635 523154

B-1

NI-DSP SRM for LabVIEW for Windows

Technical Support Form

Photocopy this form and update it each time you make changes to your software or hardware, and use the completed

copy of this form as a reference for your current configuration. Completing this form accurately before contacting

National Instruments for technical support helps our applications engineers answer your questions more efficiently.

If you are using any National Instruments hardware or software products related to this problem, include the

configuration forms from their user manuals. Include additional pages if necessary.

Name

Company

Address

Fax (

Computer brand

Operating system

)

Phone (

Model

)

Processor

Speed

MHz

RAM

Display adapter

Mouse

yes

Other adapters installed

Brand

Hard disk capacity

Instruments used

National Instruments hardware product model

Configuration

Revision

National Instruments software product

Configuration

Version

The problem is

List any error messages

The following steps will reproduce the problem

NI-DSP for LabVIEW for Windows

Hardware and Software Configuration Form

Record the settings and revisions of your hardware and software on the line located to the right of each item.

Complete this form each time you revise your software or hardware configuration, and use this form as a reference

for your current configuration. Completing this form accurately before contacting National Instruments for technical

support helps our applications engineers answer your questions more efficiently.

National Instruments Products

•

DSP Hardware

_________________________________________________

Interrupt Level of Hardware

DMA Channels of Hardware

Base I/O Address of Hardware

NI-DSP Version

NI-DAQ Version

LabVIEW Version

Other Products

•

Computer Make and Model

_________________________________________________

Computer Bus (XT/AT/ISA or EISA)

Microprocessor

Clock Frequency

Type of Video Board Installed

DOS Version

Programming Language

Programming Language Version

Other Boards in System

Base I/O Address of Other Boards

DMA Channels of Other Boards

Interrupt Level of Other Boards

Documentation Comment Form

National Instruments encourages you to comment on the documentation supplied with our products. This

information helps us provide quality products to meet your needs.

™

Title: NI-DSP Software Reference Manual for LabVIEW for Windows

Edition Date:

Part Number:

December 1993

320571-01

Please comment on the completeness, clarity, and organization of the manual.

If you find errors in the manual, please record the page numbers and describe the errors.

Thank you for your help.

Name

Title

Company

Address

Phone

(

)

Mail to:

Technical Publications

Fax to:

Technical Publications

National Instruments Corporation

MS 53-02

National Instruments Corporation

6504 Bridge Point Parkway, MS 53-02

Austin, TX 78730-5039

(512) 794-5678

Glossary

___________________________________________________

Prefix

Meaning

Value

-9

-6

-3

nano-

micro-

milli-

kilo-

µ-

M-

mega-

one-dimensional

two-dimensional

complex

decibels

Discrete Fourier Transform

direct memory access

digital signal processing

Fast Fourier Transform

Fast Hartley Transform

DFT

DMA

DSP

FFT

FHT

FIR

hex

finite impulse response (filter)

hexadecimal

IDFT

IEEE

IFFT

IFHT

IIR

Inverse Discrete Fourier Transform

Institute of Electrical and Electronic Engineers

Inverse Fast Fourier Transform

Inverse Fast Hartley Transform

infinite impulse response (filter)

input/output

I/O

Kwords

LMS

LSB

1,024 words of memory

least mean square

least significant bit (of a word)

LabVIEW

megabytes of memory

MLUT

MSE

Mwords

RTSI

SCXI

Memory Look Up Table

mean squared error

1,024 x 1,024 words of memory

point

Real-Time System Integration

Signal Conditioning eXtensions for Instrumentation

seconds

word

32-bits of memory or four bytes, unless otherwise stated

Glossary-1

NI-DSP SRM for LabVIEW for Windows

Index

Alloc_Mem function, Part 4: 2-3

array VIs

bandpass filters. See filter VIs.

DSP Absolute, Part 3: 2-4

DSP Add, Part 3: 2-5

DSP Clip, Part 3: 2-11

DSP Divide, Part 3: 2-20

DSP Linear Evaluation, Part 3: 2-50

DSP Log, Part 3: 2-51

DSP Max & Min, Part 3: 2-52

DSP Multiply, Part 3: 2-54

DSP Polar to Rectangular, Part 3 3-59

DSP Polynomial Evaluation, Part 3: 2-58

DSP Product, Part 3: 2-59

DSP Rectangular to Polar, Part 3: 3-65

DSP Reverse, Part 3: 2-64

DSP Set, Part 3: 2-66

bandstop filters. See filter VIs.

boards. See also AT-DSP2200 board.

compatibility of DSP boards with National

Instruments data acquisition boards, xv

configuration

EISA bus computer, Part 1: 1-3

ISA (or AT) bus computer, Part 1: 1-3

build Dispatch application. See Dispatch application.

Butterworth filter. See DSP Butterworth

Coefficients VI.

DSP Shift, Part 3: 2-66

Chebyshev filter. See DSP Chebyshev Coefficients

VI; DSP Inv Chebyshev Coeff VI.

Code Interface Node (CIN) interface, Part 2: 1-2

COFF (common object file format) file, Part 4: 1-1

compatibility of DSP boards with National

Instruments data acquisition boards, xv

compiling and assembling custom libraries,

Part 4: 2-4, 2-7 to 2-8

DSP Sort, Part 3: 2-71

DSP Square Root, Part 3: 2-70

DSP Subset, Part 3: 2-72

DSP Subtract, Part 3: 2-73

DSP Sum, Part 3: 2-73

DSP Unwrap Phase, Part 3: 2-80

list of functions, Part 3: 1-2

prototypes for customizable functions,

Part 4: 3-2 to 3-3

configuration. See boards.

Copy Mem(DSP to DSP) VI, Part 3: 2-1

Copy Mem(DSP to LV) VI, Part 3: 2-2

Copy Mem(LV to DSP) VI, Part 3: 2-3

custom NI-DSP library, creating, Part 4: 2-1 to 2-8

adding function names to library function list

file, Part 4: 2-4 to 2-5

assembling custom libraries, Part 4: 2-7 to 2-8

assembling source code, Part 4: 2-4

makelib.bat file, Part 4: 2-8

procedure, Part 4: 2-7 to 2-8

required files, Part 4: 2-7

using WE DSP32C assembly language,

Part 4: 2-3 to 2-4

assembly dispatch file, generating, Part 4: 2-6

AT-DSP2200 board

adding object filenames to linker file (ifile),

Part 4: 2-4

assembling source code, Part 4: 2-4

compiling and assembling, Part 4: 2-7 to 2-8

compiling source code, Part 4: 2-4

deleting (replacing) function names in

NIDSP.fnc, Part 4: 2-5

board configuration

EISA bus computer, Part 1: 1-3

ISA (or AT) bus computer, Part 1: 1-3

capabilities, xv

files required to build example library,

Part 4: 2-7

DSP library files, Part 1: 1-3

improving execution speed of DSP VIs,

Part 2: 1-7

memory management and data transfer,

Part 2: 1-2 to 1-4

generating dispatch file, Part 4: 2-6

GMaxMin.c example, Part 4: 2-1 to 2-2

guidelines for adding functions, Part 4: 2-2 to 2-3

including error codes, Part 4: 2-2

LabVIEW interface, creating, Part 4: 2-8 to 2-12

bundling input parameters to arrays,

Part 4: 2-8 to 2-9

software overview, Part 2: 1-1 to 1-2

ATDSP.h file, Part 4: 2-2

AT&T WE DSP32C digital signal processor,

Part 4: 1-1

calling Custom VI, Part 4: 2-10

executing custom function from LabVIEW,

Part 4: 2-12

Index-1

NI-DSP SRM for LabVIEW for Windows

Index

indexing output arrays to obtain results,

Part 4: 2-10 to 2-12

linking, Part 4: 2-7 to 2-8

makelib.bat file, Part 4: 2-8

memory management, Part 4: 2-3

parameter guidelines, Part 4: 2-2 to 2-3

prototypes for customizable functions,

Part 4: 3-1 to 3-3

source code creation, Part 4: 2-1 to 2-3

steps for creating, Part 4: 2-1

using WE DSP32C assembly language,

Part 4: 2-3 to 2-4

DSP Divide VI, Part 3: 2-20

DSP DMA Copy(DSP to LV) VI, Part 4: 4-3

DSP DMA Copy(LV to DSP) VI; Part 4: 4-4

DSP Elliptic Coefficients VI, Part 3: 2-21 to 2-22

DSP Equi-Ripple BandPass VI, Part 3: 2-23 to 2-24

DSP Equi-Ripple BandStop VI, Part 3: 2-25 to 2-26

DSP Equi-Ripple HighPass VI, Part 3: 2-27 to 2-28

DSP Equi-Ripple LowPass VI, Part 3: 2-29

DSP Exact Blackman Window VI, Part 3: 2-30

DSP Exponential Window VI, Part 3: 2-31

DSP FHT VI, Part 3: 2-32

DSP Flat Top Window VI, Part 3: 2-33

DSP Force Window VI, Part 3: 2-34

DSP Free Memory VI, Part 3: 2-34

DSP Gaussian White Noise VI, Part 3: 2-35

DSP General Cosine Window VI, Part 3: 2-36

DSP Hamming Window VI, Part 3: 2-37

DSP Handle Clusters

Custom VI. See DSP Custom VI.

customer communication, xv, B-1

allocation examples, Part 2: 1-4

array data type, Part 2: 1-4

bundling for LabVIEW interface,

Part 4: 2-8 to 2-9

definition, Part 2: 1-2, Part 2: 1-3

hexadecimal encoding, Part 2: 1-3

illustration, Part 2: 1-3

improving execution speed of DSP VIs,

Part 2: 1-7

input/output, Part 2: 1-5

memory management and data transfer,

Part 2: 1-2 to 1-4

obtaining valid DSP Handle Cluster, Part 2: 1-4

output data buffers, Part 2: 1-5

data acquisition functions, Part 4: 3-3

data buffers. See DSP Handle Clusters.

data transfer. See memory management and data

transfer.

data types

DSP Handle Cluster as array data type,

Part 2: 1-4

icons representing (table), xiii

deleting (replacing) functions in NIDSP.fnc file,

Part 4: 2-5

Developer Toolkit, xv

DFT (Discrete Fourier Transform), Part 3: 1-4 to 1-5

Dispatch application, Part 4: 1-1, Part 4: 2-6

documentation

values not to be changed, Part 2: 1-3

Z in, Z out naming convention, Part 2: 1-5

DSP Handle to Address VI; Part 3: 2-38

DSP Hanning Window VI, Part 3: 2-38

DSP IIR Filter VI, Part 3: 2-39 to 2-40

DSP Impulse Pattern VI, Part 3: 2-41

DSP Impulse Train Pattern VI, Part 3: 2-42

DSP Index Memory VI, Part 3: 2-43

DSP Init Memory VI, Part 3: 2-44

DSP Integral VI, Part 3: 2-45

DSP Inv Chebyshev Coeff VI, Part 3: 2-46 to 2-47

DSP Inverse FFT VI, Part 3: 2-47

DSP Inverse FHT VI, Part 3: 2-48

DSP Kaiser-Bessel Window VI,Part 3: 2-49

DSP Linear Evaluation VI, Part 3: 2-50

DSP Load VI, Part 3: 2-50

conventions used in manual, xii-xiii

organization of manual, xi-xii