Home

National Instruments Car Stereo System 320571-01 User's Manual

1. If you choose Custom Installation SETUP installs 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 WOAQCONF EXE saves the configuration parameters in a file named WDAQCONF CFG Use the WDAQCONF utility to assign a board ID number to your AT DSP2200 to choose the memory subtype either 64 Kwords or 2128 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 OUT and LV2200 OUT LV2200S OUT is intended for use with the 64 Kword version of the AT DSP2200 LV2200 OUT is intended for use with any other version of the board With the WOAQCONF utility you can enter a complete path that will include the DSP Library file name If you enter complete pathnam
2. hexadecimal Inverse Discrete Fourier Transform Institute of Electrical and Electronic Engineers Inverse Fast Fourier Transform Inverse Fast Hartley Transform infinite impulse response filter input output 1 024 words of memory least mean square least significant bit of a word LabVIEW megabytes of memory 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 32 bits of memory or four bytes unless otherwise stated National Instruments Corporation Glossary 1 NI DSP SRM for LabVIEW for Windows Index A Alloc_Mem function Part 4 2 3 array 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 amp 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 DSP Shift Part 3 2 66 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 assembling custom libraries Part 4 2 7 to 2 8 assembling source code Part 4 2 4 m
3. 271 WwW 2 n where nis the number of elements in X and m is the number of elements in Cosine Coefficients mma 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 ama Cosine Coefficients is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the cosine coefficients Tma 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 H are obtained from the formula error in no error error out Hamming 4 yi xi 0 54 0 46 cos w DSF Hamming Window for 1 0 1 2 n 1 271 n gt I where n is the number of elements in the input sequence X o X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input data array Note The operat
4. 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 ara Band Parameters is a cluster Each cluster element contains the necessary information associated with DPHE i ane each band for the FIR design Each cluster contains four elements as shown in the following figure Band Parameters 2128 Weighted Ripple u The Band Parameters cluster must contain at least one element that is one band ora amplitude is a DSP Handle Cluster that indicates the memory buffer on the DSP board that DPHE contains the desired amplitude for each band ae lower freq is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the lower frequency bands ene 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 ora 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 gt f for i 0 1 2 m l yh where f represents the lower freq in the it band and f
5. Figure 2 4 Signals Group Section in dspfnes h The file dispatch s is 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 Dispatch application uses the files NIDSP fnc and NIESSEN fnc to create dispatch s Figure 2 5 shows a section of dispatch s corresponding to the section given from NIDSP fnc Notice that in this example Signals was 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 a from the LIB directory 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
6. IIR Infinite Impulse Response filters Part 3 1 6 See also filter VIs installation board configuration EISA bus computer Part 1 1 3 ISA or AT bus computer Part 1 1 3 NI DSP for LabVIEW for Windows Part 1 1 2 to 1 3 NI DSP Interface Utilities Part 4 1 2 Interface Utilities See NI DSP Interface Utilities Inverse Discrete Fourier Transform IDFT Part 3 1 4 ISA or AT bus computers Part 1 1 3 K Kaiser Bessel window See DSP Kaiser Bessel Window VI NI DSP SRM for LabVIEW for Windows Index L LabVIEW software See NI DSP for LabVIEW for Windows Linear Constant Coefficient Difference Equation Part 3 1 5 linker file ifile Part 4 2 4 linking custom libraries Part 4 2 7 to 2 8 lowpass filters See filter VIs M makelib bat file Part 4 2 8 manual See documentation Memory Look Up Table MLUT definition Part 2 1 3 number of entries allowed Part 2 1 3 memory management and data transfer Part 2 1 2 to 1 4 DSP Handle Clusters Part 2 1 2 to 1 4 examples of DSP Handle Cluster allocation Part 2 1 4 freeing buffers Part 2 1 4 hexadecimal encoding of DSP Handle Part 2 1 3 lower level routines for customized DSP routines Part 4 2 3 VIs for memory management and data transfer Part 2 1 2 memory management VIs Copy Mem DSP to DSP Part 3 2 1 Copy Mem DSP to LV Part 3 2 2 Copy Mem LV to DSP Part 3 2 3 DSP Allocate Memory Part 3 2 6 DSP Free Memory Part 3 2 34 DSP I
7. but delete functions rather than add functions Follow these guidelines when deleting functions from the DSP Library e 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 DLL is 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 e 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
8. obtained using the formula Exact Blackman i DSP Exact Blackman Window y x 0 42659071 0 49656062 cos w 0 07684867 cos 2w for 1 0 1 2 n 1 271 w n gt where n is the number of elements in X cain 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 ono 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 final value Y are obtained using the formula error in no error error out Exponential 4 y x exp a i DSP Exponential Window for i 0 1 2 n 1 a PP 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 th
9. 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 Time Domain short DSP_Convolution short DSP_Correlation short DSP_Decimate f short DSP_Deconvoluti short DSP_Difference short DSP_Integrate Filters short DSP_Bw_Coef short DSP_Ch_Coef short DSP_Elp_Coef short DSP_EqRip_BPF short DSP_EQRip_BSF short DSP_EQRip_HPF short DSP_EQRip_LPF short DSP_IIR_ Filter short DSP_InvCh_Coef short DSP_Median_Filt short DSP_Parks_McCle Windows short DSP_CosWin floa short DSP_ExpWin flo short DSP_ForceWin f short DSP_GenCosineWi short DSP_KsrWin flo short DSP_TriWin flo Array Functions short DSP_Abs float short DSP_Add float short DSP_Clip float short DSP_Div float short DSP_LinEv floa short DSP_Log float short DSP_MaxMin short DSP_Mul float short DSP_PolyEv flo short DSP_Prod float short DSP_Reverse fl short DSP_Set float short DSP_Shift floa short DSP_Sort float Part 4 NI DSP Interface Utilities Chapter 3 float x long n float y long m float cxy float x long n float y long m float rxy loat x long n long decFact long ave float y on float cxy long n float y long m float x float x long n float dt float xInit float xFinal float y float x long n float dt float xInit f
10. 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 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 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 DSP Log VI Part 3 2 51 DSP
11. 32 bit short sourcellong 16 bit and long 32 bit To copy different types of data you must wire to the Copy Mem L to DSP 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 mo i B p B 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 mo 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 ith element of the outp
12. 34 DSP Gaussian White Noise Part 3 2 35 DSP General Cosine Window Part 3 2 36 DSP Hamming Window Part 3 2 37 DSP Handle Cluster input output Part 2 1 5 DSP Hanning Window Part 3 2 38 DSP IIR Filter Part 3 2 39 to 2 40 DSP Impulse Pattern Part 3 2 41 DSP Impulse Train Pattern Part 3 2 42 DSP Index Memory Part 3 2 43 DSP Init Memory Part 3 2 44 DSP Integral Part 3 2 45 DSP Inv Chebyshev Coeff Part 3 2 46 to 2 47 DSP Inverse FFT Part 3 2 47 DSP Inverse FHT Part 3 2 48 DSP Kaiser Bessel Window Part 3 2 49 DSP Linear Evaluation Part 3 2 50 DSP Load Part 3 2 50 DSP Log Part 3 2 51 DSP Max amp Min Part 3 2 52 DSP Median Filter Part 3 2 53 DSP Multiply Part 3 2 54 DSP Parks McClellan Part 3 2 55 to 2 57 DSP Polynomial Evaluation Part 3 2 58 DSP Power Spectrum Part 3 2 59 DSP Product Part 3 2 59 DSP Pulse Pattern Part 3 2 60 DSP Ramp Pattern Part 3 2 61 DSP Random Pattern Part 3 2 62 DSP ReFFT Part 3 2 63 National Instruments Corporation Index DSP Reset Part 3 2 63 overview Part 1 1 1 DSP Reverse Part 3 2 64 VI library files Part 3 1 3 DSP Sawtooth Pattern Part 3 2 65 NI DSP Interface Utilities DSP Set Part 3 2 66 creating custom DSP library Part 4 2 1 to 2 8 DSP Shift Part 3 2 66 adding function names to library function DSP Sinc Pattern Part 3 2 67 list file Part 4 2 4 to 2 5 DSP Sine Pattern Part 3 2 68 adding object filenames to linker file ifile DS
13. Cluster that indicates the memory buffer on the DSP board that will contain the forward coefficients 2 mo a Ei o Feedback Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the feedback coefficients oro 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 oro 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 Clipped 1 in Let Y represent the output array The ith 7 ele
14. Eland Wares speer eges cpus par ee ip eO sees n ee sees epdcevh esbep oa E S on TOES OEN IESPA EEEE XV Customer Communication osise eine eee ieee aR nea Ea iee Ee Erene a ENE eoeta oaae reres eenias XV Part 1 Getting Started with NI DSP ooo scsi 1 1 PLOCUCE OVERVIEW esinen ere E N E E E E S E Ee REE EE E NEE EEEE 1 1 The NI DSP S Of Ware ereiaro ae en code Dele anda tee eau ac E Ea genio dts TETE caneoas Paden A E E n aia 1 1 What Your Distribution Diskettes Should Contain 0 cecceeescecessecesececeseecesenseceseeceessneeeneaes 1 2 Installing NI DSP for LabVIEW for Windows 0 cc ecceeesecsceeseceseeseceeceseceeceseeseceseeseceseseaseseseaecaaeeaesaes 1 2 Board Configuration s str skouneaas esa cb obbabebauesesuststusaseas lesepebeas eupsbeasasnd ats soeegsaseesbeasces vesy 1 3 Installation on an ISA or AT Bus Computer cece eeceeeceesecesseceeeeceaeeeaeeceneecaeceeeeeeeeenaeeeee 1 3 Installation on an EISA Bus Computer cece ceesecsecssecneceneceecseeeeceseseceeeeeeseaeeeescaaseaeeaee 1 3 Part 2 Introduction to the NI DSP Analysis VIs 000000 1 1 Using the NI DSP VIs in LabVIEW iesiiiies teenin ekse dee EE Ee o RE REESE E ea 1 1 AT DSP2200 Software OVerVieW siscsssscusssiiss vous casssevssscsssshuachenesbsuans obsvegssesypeasieu vecbeabssdepcbvas EEEE EE En EEE 1 1 Memory Management and Data Transfer eeeescecececseceseceeeceseecescecaeeeaeecereeeneceneessecesaeceeeecaeeeneess 1 2 Special Features of the NI DSP Analysis
15. 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 i points Magnitude Phase The it Tos m Phage element of the polar coordinate set is error in no error wo error out obtained by using the following fomulas Magnitude i 4 X i Y Phase i arctan Y i X i DSP Rectangular To Polar Note The operation is perfomed in place and the input arrays X and Y are overwritten by Magnitude and Phase o 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 pSPHE input signal array Y Ema Magnitude is the DSP Handle Cluster that is identical to X but with the caculated polar coordinate BSPHE amplitude values already stored in the memory buffer on the DSP board alae 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 Chapte
16. 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 makelib bat Batch file to build the custom library gmaxmin c C source file dspfncs h C include file generated by Dispatch dispatch s Assembly file generated by Dispatch NIDSPLNK Modified linker command file LIBLVDSP a Archived library of object modules for the DSP Library functions for LabVIEW libc32c a AT amp T C library libap32c a AT amp T Application library libm32c a AT amp T Math library found in the LIB directory in the path set by the environment variable All these archives must be in the LIB directory in your WE DSP32C tools directory Copy the file makelib bat from the LIB directory 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 bat file 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 map and dsp map are located in the LIB directory 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 out and LV2200 out respectively You can rename these files and move them to other directories If you want the drive
17. VIs fax technical support B 1 FFT See Fast Fourier Transform FFT filter VIs DSP Butterworth Coefficients Part 3 2 9 to 2 10 DSP Chebyshev Coefficients Part 3 2 10 to 2 11 DSP Elliptic Coefficients Part 3 2 21 to 2 22 DSP Equi Ripple BandPass Part 3 2 23 to 2 24 DSP Equi Ripple BandStop Part 3 2 25 to 2 26 DSP Equi Ripple HighPass Part 3 2 27 to 2 28 DSP Equi Ripple LowPass Part 3 2 29 DSP IIR Filter Part 3 2 39 to 2 40 DSP Inv Chebyshev Coeff Part 3 2 46 to 2 47 DSP Median Filter Part 3 2 53 National Instruments Corporation Index 3 Index DSP Parks McClellan Part 3 2 55 to 2 57 list of functions Part 3 1 2 prototypes for customizable functions Part 4 3 2 filtering algorithms Part 3 1 5 to 1 6 FIR Finite Impulse Response filters Part 3 1 6 Free_Mem function Part 4 2 3 frequency domain VIs DSP Complex FFT Part 3 2 12 DSP Cross Power Part 3 2 15 DSP FHT Part 3 2 32 DSP Inverse FFT Part 3 2 47 DSP Inverse FHT Part 3 2 48 DSP Power Spectrum Part 3 2 59 DSP ReFFT Part 3 2 63 DSP Zero Padder Part 3 2 81 Fast Fourier Transform FFT Part 3 1 4 to 1 5 list of functions Part 3 1 1 prototypes for customizable functions Part 4 3 1 Functions Menu Part 3 1 4 H Handles See DSP Handle Clusters hardware See boards header files ATDSP h Part 4 2 2 dspfncs h Part 4 2 6 highpass filters See filter VIs I IDFT Inverse Discrete Fourier Transform Part 3 1 4
18. VIs 00 ce eeeeceeeesseseeeeeeseecaecaaecaeceacsaecaeeseceeeseseeeeeesenesaseaes 1 5 Hints for Improving the Execution Speed on the DSP Board 0 0 eee eeeeceeceeeceeeeeeeeeeeeeeenes 1 7 An Example of Using NI DSP Analysis VIs oo eee eecseesecscesseceseaeceecsecseceseesceeseeeeeeaeeeaeeseeaecaaeeaeeaee 1 8 Part 3 NI DSP Function Reference Chapter 1 NI DSP Analysis VI Reference Overview sssssseesessssssssssssssssnnsnseessccccesnnnnnneeeeees 1 1 The NI DSP Analysis VI Overview sscsssissssecsesdssses sss sessnussbagseesassssosctasssas dbahicsuscsscsssees ces seessassnausseseasss Soenase 1 1 Analysis VI Organization oss cess Asis iise ois orire sue checdvaces EEEa dues sca ve buses ESTEE EE E REE EEE tutus SEEE SE fase 1 3 Accessing the NI DSP Analysis VIS ssss scsjesssescssnssepescasdsbuschste step cb pobevsaabeesaeusisdysubsanssdusebens skouaensaeabenaei outs 1 3 About the Fast Fourier Transform FFT ccccccccccccccsssscecceesesscececesesssecccecesseaeececeensaececcseseaeseseeesenseeeees 1 4 About Filtering sisses iie prnaos eieae kersa a EEEa aaor TE Oa EESTE EE O EEEa as sdousesvassvassdpsasdvapsesebes 1 5 About Windowing 2 5 secs co sacres arrate A sek tases E E EE E EEE Mavs cosas eck E ESKES 1 6 Chapter 2 NI DSP Analysis VI Referee c cee ccessesssssssseessesesseseectsnnsnnnnesesececcesnsnnnnnnnseseeeeeeeste 2 1 Copy Mem DSP to DSP aei E aE E E R E E E E E ES EERE 2 1 Copy Mem DSP tOo L
19. 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 O x O 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 ee d i 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 repr
20. allocations of memory on the specified DSP parenn slot board ISPNEN error out error in no error DSF Init Memory 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 H HE 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 vin function f t is defined as follows X i initial condition 3 ae F t fee dt Tinal condition me error out dt error in no error DSP Integral Let Y represent the sampled output sequence Integral X The VI obtains the elements of Y using the following formula y x 4x x dt for i 0 1 2 n 1 where nis the number of elements in X x is specified by initial condition when i 0 and x is 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 f
21. 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 0 0 Delay s lt lt gt lt Width s q a delta t s The following figure illustrates how the pattern can vary with
22. different values for the parameters Delay 0 Delay delta t lt gt gt gt Width s lt lt lt gt 4 gt delta t s delta t s lt gt 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 7 4 1 1 1 7 1 Delay s lt gt lt amp Width s Delay lt gt lt gt Width s delta t s lt gt delta t s lt gt 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 4 cles train is represented by the sequence amplitude Triangular Train out Y the pattern is generated according delay to the following formula Triangular Train in m error out error in no error DSP Triangular Train l mala o ag ie gt H IA a IA WwW BS yi fori 0 1 2 n 1 Za k T if lt k lt T where k i d modulo T __ n cycles n is the number of eleme
23. 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 Filtered Data in McClellan algorithm and the number of y iF Filtered Data out taps pass frequency stop frequency and of taps Aaa mut sampling frequency The VI then filters error in no error the input sequence X to obtain the lowpass filtered linear phase sequence pass freq Filtered Data stop freq sampling freq fs The passband of the filter goes from zero DC to pass freq The transition band DSP Equi Ripple LowPass goes from pass freq to stop freq and the stopband goes from stop freq to the Nyquist frequency ama 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 freq defaults to 0 2 pass freq must be greater than 0 If pass freq is less than or equal to zero the VI returns an error pass stop freq must be greater than the pass freq and observe the Nyquist
24. 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 J 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 4 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 27i N 4 Other manufacturers may use a slightly different definition s
25. hard drive SETUP can also install the NI DSP Interface Utilities discussed in Part 4 NI DSP Interface Utilities of this manual If you choose Full Installation SETUP does the following things 1 SETUP creates a subdirectory called DSP2200 of the vi lib subdirectory of the LabVIEW directory SETUP decompresses the NI DSP Analysis VIs as LabVIEW LLB files in the DSP2200 directory 2 SETUP copies DSP DLL to your windows directory 3 SETUP creates the subdirectories shown in Table 1 1 under the directory you specified during setup These subdirectories make up the NI DSP Interface Utilities TUP creates a subdirectory called DSP2200 of the EXAMPLES subdirectory of the LabVIEW directory TUP copies all of the NI DSP VI examples there 53 m Table 1 1 Subdirectories Created by SETUP C NIDSP LIB Library files for linking with stand alone programs C NIDSP DISPATCH The Dispatch utility and related files C NIDSP EXAMP LES 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 NIDSP is the name you specify during setup The SETUP program 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
26. 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 ina 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 ith element of the output array Z is obtained using the PA gt oust following formula oo in error out Zi X Y for i 0 1 2 n 1 error in no error where n is the smaller number of elements in X DSP Subtract o X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array X To Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array Y ae 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 ro Z out is a DSP Handle Cluster that is identical to Z in but with the subtracted results already stored in the memory bu
27. information for this call 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 Sinc Pattern in by the sequence Y then the pattern is amplitude generated according to the following ae ay A Sinc Pattern out formula delta t 7 A error out error in no error y asinc iAt d DSP Sinc Pattern for 1 0 1 2 n 1 sin TX where sinc x sin x TX a is the amplitude At 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 lt x lt 1 When Ixl 1 the sinc x 0 0 and the peak value of the sinc function occurs when x 0 You can show using l H pital 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
28. is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the initial conditions on input at time zero aa 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 we au 7 aaa 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 we au 7 Ta 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 mo a 7 Part 3 NI DSP Function Reference 2 40 NI DSP SRM for LabVIEW for Windows Chapter 2 NI DSP Analysis VI Reference mo 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 o 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 plac
29. long imax float min long imin x float y long n float z at x long n float coeff long k float y x long n float prod oat x long n x long n float a t x long n long shift float y x long n long direction float y 3 2 NI DSP SRM for LabVIEW for Windows Chapter 3 DSP Board Function Overview short DSP_Sgqrt 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 h file in your program dspd
30. nkErr FilterDesignE tPowerOfTwo EM HRDWARE_ERR ERO_ALLOCATION TooBigAnOffset HandleNotFound TooManyAllocs PtrNotFou ParamSizel nd Err Error Name Description WinDutyCyclesE The window duty cycle value must be between 0 0 and 100 0 dt must be greater than zero The duty cycle must meet 0 lt duty cycle lt 100 The number of cycles must be greater than zero and less than or equal to the number of samples 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 The cutoff frequency fc must meet O lt fcs fs 2 The following conditions must be met 0 lt f_low lt f_high lt fs 2 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 u must meet 0 lt u lt 0 1 The leakage coefficient Leak must meet 0 lt Leak lt u The filter cannot be designed with the specified input parameters The number of coeffi
31. operations on arrays e Memory Management contains VIs to perform allocating indexing copying and freeing memory on the AT DSP2200 board e 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 Structs amp Constants Arithmetic Trig amp Log Comparison Conversion string Array amp Cluster File 170 Dialog amp Date Time Miscellaneous VI ANALYSIS DAQ DSP2200 GPIB Getting Started serial UTILITY UTILS rr SF FF SF FS eS FF F signal Generation Frequency Domain Time Domain Filters Windows Array Memory Management Utili FF FF Fa Y rr FS FF FF Fh 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
32. 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 Forward Coefficients in IIR filter as specified by the filter type Butterworth filter model sampling freq fs T Forward Coefficients out You can then pass these high cutoff freq fh Feedback Coefficients out coefficients to the DSP IIR low cutoff freq fl i A error out Filter VI to filter a sequence error in no error an of data order Feedback Coefficients in DSP Butterworth Coefficients 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 lowpass high cutoff freq fh defaults to 0 45 high cutoff freq fh is the high cutoff frequency The VI ignores this parameter when filter type is 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 lt fl lt fh lt fs 2 aa Forward Coefficients in i
33. 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 number of elements FFFFTITIT frin ince ol Figure 2 6 How to Bundle Parameters in LabVIEW to Call gnaxmin 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 max value min value function ID max index min index 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 c
34. represents the higher freq in the ith band gt fh for i 0 1 2 m 2 The higher freq in the last band must observe the Nyquist criterion fis lt OS m 1 where f is 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 inp
35. second harmonic Y n 1 Negative first harmonic The following conventions and restrictions apply to the VIs in the Frequency Domain folder e All arrays must be a power of two n 2 m 1 2 3 24 limited by onboard memory e 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 e Y FFT X the sequence Y is the FFT of the sequence X e 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 inatoa any naad a 1 I5 where x is the discrete time input signal to the system represented by the filter y is the discrete time output signal of the system represented by the filter a is the set of coefficients ap
36. 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 iJ X k exp j2mik n for i 0 1 n 1 where Y i is the ith element of the DFT of X and j V 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 2mik n for i 0 1 n 1 k 0 where X i is the ith element of the IDFT of Y andj V 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 DC component Y 1 Positive first harmonic Y 2 Positive second harmonic Y k 1 Positive k 1 harmonic Y k Nyquist frequency Y k 1 Negative k 1 harmonic Y n 2 Negative
37. the memory buffer on the DSP board that contains the input signal array Y To 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 ooo 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 biy tes felerment selector returns a DSP Handle Cluster that contains the coded DSP board memory ai ISPHEH DSF Handle Cluster and the number of elements in this error in no error Allocate error out buffer DSP Allocate Memory 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 gt number of byt
38. 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
39. 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 a 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
40. use any pointer Remember to allocate the DSP Handle Cluster before you call your custom function through the DSP Custom Function VI e 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 e 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 Examples subdirectory 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 e 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
41. where nis the number of elements in the input sequence X and r is the filter rank eae 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 gt r20 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 oro Filtered Data out is a DSP Handle Cluster that is identical to Filtered Data in but with the filtered data L 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 54 NI DSP SRM for LabVIEW for Windows Chapter 2 NI DSP Analysis VI Reference DSP Multiply Multiply array X by array Y The ith element of the output array Z is obtained using the following formula Yo Z out cin a error out Zi X YG for i 0 1 2 n 1 error in no error where n is the smaller number of elements in X DSP Multiply io X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that c
42. 100 gt 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 ae Square Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the output square pattern oe 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 ss Yin error in no error error out of each element of the input array X The ith Y out element of the output array Y is obtained using the following formula DSP Square Root Y i VIX 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 mma Y in is a DSP Handle Cluste
43. 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 DSPrET EUSPrIEN Free Fp DSPHEM Allacaite error out of the Copy Mem LV to DSP VI is TFTFIFFFTA i connected to the error in of the DSP Free Memory VI Allocate 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 e 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 e 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 o
44. Applies a Hanning window to the input sequence X If Y represents the output sequence Hanning X the elements of Y are Hanning H obtained using the formula error in no error error out y 0 5 x 1 cos w DSP Hanning Window for 1 0 1 2 n 1 where n is the number of elements in the input sequence X ama 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 mmn 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 UR filtering on the X input array and reports the Initial Conditions on input result in Y It uses the arrays Forward Coefficients A a and b of sizes sza and szb ms ui Final Conditions on input respectively in implementing Yin Aeee Final Conditions on output the linear difference equation error in no error i that describes IIR filtering Se Feedback Coefficients Initial Conditions on output DSP
45. Call the Custom VP nerean a era E E r A S 2 10 3 Index the Output Arrays to Obtain the Results essssesssseesssesessessressseesrsrerrseerrsrrererensrerese 2 10 Executing the Custom Function from LabVIEW sssesssseesssseeseesreessreeresrsresresesresrerrsrersseeresrsresresrerrereereee 2 12 Chapter 3 DSP Board Function Overview scsssssssssssssssssssssssssssseeseeseeceeeeeececeesnesnsnnnnnnnnnnnnnnnenmnse 3 1 Data Acquisition FUNCHONS 3452 sess EE SEEE E EE EEEE E E E E EE 3 3 National Instruments Corporation vii NI DSP SRM for LabVIEW for Windows Contents Chapter 4 Using the DN VA IW TS vic sea tosiie caine ca aa es a aa ea 4 1 DSP DMA Copy DSP to LV Jerene ce eaue tutes ent ecvuctee Meesescany dacessuiecs Athan peel peestgent dates ea i 4 3 DSP DMA Copy ey to DSP vevecscs vscesscetacctasscbieuesvsoce leas EEE E sup davsesaviedalepataee laasnotecsncevassensbenve 4 4 Appendix A Error kK Ood es n A 1 IKE KOI CONIC ONS RA EE EEEE cea E E AE A EEA E EAEE EEEE AE A 1 Appendix B Customer CommuliCation nr nnn B 1 GIOS AVY aa E eb careanees Glossaryed Index A Reena SURE CER En COTA AB EERE GNA SEE AMEE SIRS Cnt a ROT E EEA E Ie CLE AL Tn Sen EAE EE ea ke Index 1 NI DSP SRM for LabVIEW for Windows viii National Instruments Corporation Contents Figures Part 1 Figure 1 1 Development Paths with the NI DSP Software o oo eeseceeceseeeeceseeeeceeeeeeceeeeeeeeneseaeeaeesaeenes 1 1 Part 2 Figure 1 1
46. Communication between the PC and the DSP Board 00 0 0 eee eee eeeeseeeeeeseeceecseecaecnseeaeenaeenees 1 1 Figure 1 2 DSP Handle Clustetcs sais cases fags celtics Seats a AS eb a ba Sach ee chat ias os es ab deta eh EG isk lls cai aes 1 3 Figure 1 3 The Hexadecimal Encoding of a Typical DSP Handle o00 eee eee eeeeeeeeeeeecseecaecnseeseenseesees 1 3 Figure 1 4 Front Panel An Example of How to Allocate a DSP Handle Cluster ee eeeeeeeseeeeeeeee 1 4 Figure 1 5 Block Diagram An Example of How to Allocate a DSP Handle Cluster cece eeeeeeeeeeeeeeeeee 1 4 Figure 1 6 DSP Add Wi cnc veccatidi niin aa na nthe lewis aa E A EE A Ass E E stele EE Ee Eana BiB atei 1 5 Figure 1 7 The error in error out Clusters sscsssesie dosent a chaise ists ois aiiweeeasign ately saat en 1 5 Figure 1 8 An Example That Does Not Use error in error out for Sequencing VIS eeeeeeeeteeeeeeeees 1 6 Figure 1 9 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 00 0 0 cece ceseeceseeeeceeeeeeceseeeeeeneeeneeaee 1 8 Figure 1 11 Block Diagram An Example of Using NI DSP Analysis VIs 000 cece eeeeceeseeeeeseeeeeeneeeneeaee 1 8 Part 3 Figure 1 1 Choosing DSP2200 from the Functions Menu 0 cece ce ceececeseeesceeeeeeceseeeeeeseecsecseecaecsaeeaeesaeenees 1 4 Figure 1 2 Spectral Leakage Demonstrated Using Convolution 00 0 c
47. DSP Derivative meoo aee a E EEE E AE E EEEE EEEE E A TE EE E EE ESSE 2 19 IDES A DIAS Te CAERE AS EE EE copay AE EA EEEE EEE E E N E ERE 2 20 DSP Elliptic Coefficients senn n aig od Seed A ae A RAR es 2 21 DSP Equt Rippl Bandpass ss sings ocets de ued n E eas veeu Us seas tas boi bvees S 2 23 DSP Equi Ripple Bandstopi necesn o enas ae E E N a N E E a E AE 2 25 DSP Equi Rippl m EKA 01 at RR AREE E AEE S EEN E E E A E E EE A 2 27 DSP Equi Ripple LOWwPa S unei eececceeseeeeeeeecaceseecacesaecsecsaecsecsseeeeeseeesesesseeesesenecaaecaecaaesaecsaeeaeeaeenees 2 29 DSP Exact Blackman WINdOW issviesssitiswesseaceasopostussvetnve sei cengedssnnsayisouesdenceasupobesd sdvigaseassncbaviennessbinneudessospense 2 30 DSP Exponential Window ccissie ccsiitesaciedsersecpdaseesdenstnces ca tadcodoespesi cb cues EEE EEES E E E KEE aE e EEE EES 2 31 DSP BH i E E EE TE REE 2 32 DSP Flat Top Window scrii nsan aaan a EARL See Seatac EE a REN E REEE 2 33 DSP Force Window meiri E KE E EAE SAAE TE EAE E AEE 2 34 DSP Bree Memory sern ouier rroa aia E EE EE EEE r oe E EEES ES EEEE E S E E OEKE EASES ESTES 2 34 DSP Gatissiam White Noise cvisiscccccesteassivscocceusvcecedes lasvace see doves a a e a a a a a a a 2 35 DSP General Cosme WiNdOW seios a oria e a EE E RAA ERE 2 36 DSP Hamming Windowsen eee ree a E a Ae EEE aa AE a EEES E EEE PUSE SE 2 37 DSP Handle to Addressa enren rene eaea e RE E OR EETA ORNE AE REEE OE Eae 2 38 DSP Hating Window moere n e eE E REEE E T c
48. EEE E EE EEE EEE EE EEA ERES 2 55 DSP Parks Me Clea a A a E E E a E a A E EEE Tee 2 56 DSP Polar t Rectangular scree eni Risvesssesbe sess sstesdsckssbensscanssbssdstadaehasveads soapssadedsousdvenseapvecnscds sans 2 59 DSP Polynomial Evaltation esee ree nr aoia aeee Eeee i E VEe TE caus da sk EDENE EEEE lens EEEk 2 60 DSE Power Spectrum n stata init ais hays ath nena elaine 2 61 DSP ProdUCte ny era cu cebances od cous ofuadecnseevacdcetde se ecisy uses oe Bie Macnentueds foes a a e a AEE rte cages 2 61 DSP Pulse Pattern anren niia e N Segue tesans as A Seek E i eR ele A hash Seed 2 62 DSP Ramp Patteri aisaen e E wig E E E E A A 2 63 DSP Random Patteri icreto eene eena a E E E A E A E O RR 2 64 DSP Rectans larto PolaT iorns se chess eree sees ocssacsetped coh codhess egesed veda esatins amp suet ected Ea SERE OE Ee POESIEN verses 2 65 NI DSP SRM for LabVIEW for Windows vi National Instruments Corporation Contents DSP REFER P erna ane ea ioanes S ane E EE E EEE E e aa AREE EA E E EE EEE 2 66 DSP LT A E E E T E A E A 2 66 DSP Reverse e a aa ERRA E EE E e EE AAE EA Ar A ESS SEEE TO CE Sa AEEA 2 67 DSP Sawtooth Patterns se a a fe Ses tase a E Ge aa Eai 2 68 DSP SA S stents E OG RL II 5 A HE ey 2 69 DSP STIL PE NE SNE A ENEE EE E EON EEEE EAE A Raseravedsaeadtves E EEE Boros 2 69 DSP Sine Pattern aene aioi e EK septa agus N OE A E E E EEE Ea R 2 70 DSP Sije Patteri eensgesenn enr wend omens r ban dotebds caters nedugeadencaauecsnneette v
49. Fee Flattop H using the formula error in no error ili error out y x 0 2810639 0 5208972 cos w GSP Flat Top Window 0 1980399 cos 2w for 1 0 1 2 n 1 271 w _ n 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 aro 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 y F seh the elements of Y are obtained using the dut is formula error in no ack error out yi 0 elsewhere if O lt i lt d DSP Force Window for 1 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 T
50. HR Filter k sza 1 k szb 1 yli a k xG k b k y i k i 0 1 m l k 0 k 1 where a is the array of forward coefficients describing an IR 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 0 in 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 HI i X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array 2 mo a Ei 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
51. I 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 H BHE 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 Triangle Pattern in Pattern is represented by the amplitude A sequence Y the pattern is generated dela Triangle Pattern out according to the following formula delta t i error out error in no error yi atri Xi width for i 0 1 2 n 1 DSP Triangle Pattern where l Ix if Ixl lt 1 tri x 0 elsewhere ais 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
52. Max amp Min VI Part 3 2 52 DSP Median Filter VI Part 3 2 53 DSP Multiply VI Part 3 2 54 DSP Parks McClellan VI Part 3 2 55 to 2 57 DSP Polar to Rectangular VI Part 3 59 DSP Polynomial Evaluation VI Part 3 2 58 DSP Power Spectrum VI Part 3 2 59 DSP Product VI Part 3 2 59 DSP Pulse Pattern VI Part 3 2 60 National Instruments Corporation DSP Ramp Pattern VI Part 3 2 61 DSP Random Pattern VI Part 3 2 62 DSP Rectangular to Polar VI Part 3 65 DSP ReFFT VI Part 3 2 63 DSP Reset VI Part 3 2 63 DSP Reverse VI Part 3 2 64 DSP Sawtooth Pattern VI Part 3 2 65 DSP Set VI Part 3 2 66 DSP Shift VI Part 3 2 66 DSP Sinc Pattern VI Part 3 2 67 DSP Sine Pattern VI Part 3 2 68 DSP Sort VI Part 3 2 71 DSP Square Pattern VI Part 3 2 69 DSP Square Root VI Part 3 2 70 DSP Start VI Part 3 2 72 DSP Subset VI Part 3 2 72 DSP Subtract VI Part 3 2 73 DSP Sum VI Part 3 2 73 DSP Timeout VI Part 3 2 74 DSP Triangle Pattern VI Part 3 2 75 to 2 76 DSP Triangular Train VI Part 3 2 77 DSP Triangular Window VI Part 3 2 78 DSP Uniform White Noise VI Part 3 2 79 DSP Unwrap Phase VI Part 3 2 80 DSP Zero Padder VI Part 3 2 81 dspfnes h file Part 4 2 6 E EISA bus computers Part 1 1 3 error codes ATDSP h file Part 4 2 2 error conditions A 1 to A 4 error in error out cluster Part 2 1 5 to 1 7 F Fast Fourier Transform FFT Part 3 1 4 to 1 5 See also frequency domain
53. NI DSP Software Reference Manual for LabVIEW for Windows Digital Signal Processing Software for the PC December 1993 Edition Part Number 320571 01 Copyright 1993 National Instruments Corporation All Rights Reserved National Instruments Corporate Headquarters 6504 Bridge Point Parkway Austin TX 78730 5039 512 794 0100 Technical support fax 800 328 2203 512 794 5678 Branch Offices Australia 03 879 9422 Austria 0662 435986 Belgium 02 757 00 20 Canada Ontario 519 622 9310 Canada Qu bec 514 694 8521 Denmark 45 76 26 00 Finland 90 527 2321 France 1 48 14 24 24 Germany 089 741 31 30 Italy 02 48301892 Japan 03 3788 1921 Netherlands 03480 33466 Norway 32 848400 Spain 91 640 0085 Sweden 08 730 49 70 Switzerland 056 20 51 51 U K 0635 523545 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 obtain
54. P 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 EXE program on Disk 1 using one of the following three methods e From the DOS prompt type X SETUP where X is the proper drive designation e 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 e 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
55. P Allocate Memory e DSP Free Memory e DSP Index Memory e 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 DSP Handle Cluster 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 Special Code Owner DSP Board 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
56. P Sort Part 3 2 71 Part 4 2 4 DSP Square Pattern Part 3 2 69 assembling source code Part 4 2 4 DSP Square Root Part 3 2 70 compiling assembling and linking DSP Start Part 3 2 72 Part 4 2 7 to 2 8 DSP Subset Part 3 2 72 compiling source code Part 4 2 4 DSP Subtract Part 3 2 73 creating source code Part 4 2 1 to 2 3 DSP Sum Part 3 2 73 deleting functions from NIDSP fnc DSP Timeout Part 3 2 74 Part 4 2 5 DSP Triangle Pattern Part 3 2 75 to 2 76 files required to build example library DSP Triangular Train Part 3 2 77 Part 4 2 7 DSP Triangular Window Part 3 2 78 generating assembly dispatch file DSP Uniform White Noise Part 3 2 79 Part 4 2 6 DSP Unwrap Phase Part 3 2 80 GMaxMin c example Part 4 2 1 to 2 2 DSP Zero Padder Part 3 2 81 guidelines for custom functions error handling Part 2 1 5 to 1 7 Part 4 2 2 to 2 3 examples Part 2 1 8 to 1 9 including error codes Part 4 2 2 Fast Fourier Transform FFT Part 3 1 4 to 1 5 makelib bat file Part 4 2 8 filtering Part 3 1 5 to 1 6 memory management Part 4 2 3 filters Part 3 1 2 parameter guidelines Part 4 2 2 frequency domain VIs Part 3 1 1 prototypes for customizable functions function groups Part 3 1 1 to 1 3 Part 4 3 1 to 3 3 hints for improving execution speed Part 2 1 7 steps for creating Part 4 2 1 LabVIEW LLB files Part 3 1 3 using WE DSP32C assembler memory management VIs Part 3 1 2 Part 4 2 3 to 2 4 overview Part 1 1 1
57. Part 3 1 1 definition xi prototypes for customizable functions installation Part 4 1 2 Part 4 3 1 to 3 3 interface layers to onboard functions Part 4 1 2 signal VIs Part 3 1 1 LabVIEW interface creating Part 4 2 8 to 2 12 special features Part 2 1 5 to 1 7 bundling input parameters to arrays time domain VIs Part 3 1 1 Part 4 2 8 to 2 9 using in LabVIEW Part 2 1 1 calling Custom VI Part 4 2 10 utility VIs Part 3 1 3 executing custom function from LabVIEW window VIs Part 3 1 2 Part 4 2 12 windowing Part 3 1 6 to 1 8 indexing output arrays to obtain results Z in Z out naming convention Part 2 1 5 Part 4 2 10 to 2 12 NI DSP for LabVIEW for Windows See also overview Part 4 1 1 to 1 2 software using Part 4 1 2 board configuration NI DSP software See NI DSP for LabVIEW for EISA bus computer Part 1 1 3 Windows ISA or AT bus computer Part 1 1 3 NIC1100 CFG configuration file Part 1 1 3 contents of distribution diskettes Part 1 1 2 NIDSP fnce file development paths Part 1 1 1 adding new function names Part 4 2 4 to 2 5 installation Part 1 1 2 to 1 3 deleting replacing function names Part 4 2 5 LabVIEW interface creating Part 4 2 8 to 2 12 NIDSPLNK file bundling input parameters to arrays adding object filenames Part 4 2 4 Part 4 2 8 to 2 9 modifying Part 4 2 4 calling Custom VI Part 4 2 10 executing custom function from LabVIEW Part 4 2 12 indexing output arrays to obtain results
58. Part 4 2 10 to 2 12 National Instruments Corporation Index 5 NI DSP SRM for LabVIEW for Windows Index O object filenames adding to linker file Part 4 2 4 output arrays indexing Part 4 2 10 to 2 12 output data buffers Part 2 1 5 P parameters creating LabVIEW interface Part 4 2 8 to 2 9 guidelines for custom functions Part 4 2 2 prototypes for customizable functions Part 4 3 1 to 3 3 R random number generation See DSP Gaussian White Noise VI DSP Uniform White Noise VI S SETUP program Part 1 1 2 to 1 3 Part 4 1 1 signal truncation See windowing signal VIs DSP Gaussian White Noise Part 3 2 35 DSP Impulse Pattern Part 3 2 41 DSP Impulse Train Pattern Part 3 2 42 DSP Pulse Pattern Part 3 2 60 DSP Ramp Pattern Part 3 2 61 DSP Random Pattern Part 3 2 62 DSP Sawtooth Pattern Part 3 2 65 DSP Sinc Pattern Part 3 2 67 DSP Sine Pattern Part 3 2 68 DSP Square Pattern Part 3 2 69 DSP Triangle Pattern Part 3 2 75 to 2 76 DSP Triangular Train Part 3 2 77 DSP Uniform White Noise Part 3 2 79 list of functions Part 3 1 1 prototypes for customizable functions Part 4 3 1 software See also NI DSP for LabVIEW for Windows Developer Toolkit xv NI DAQ for DOS Windows LabWindows xv T technical support B 1 time domain VIs DSP Convolution Part 3 2 13 DSP Correlation Part 3 2 14 DSP Decimate Part 3 2 17 NI DSP SRM for LabVIEW for Windows U DSP Deconvolution Pa
59. SF DSF DSF DSF DSF DSF DS DS Time Domain DS DS DS DS DS DS NI DSP SRM for LabVIEW for Windows DSF P Sine Patter P Pulse Patte P Impulse Pat P Impulse Tra P Ramp Patter P Sinc Patter P Square Patt P Triangle Pa P Triangular P Sawtooth Pa P Uniform Whi Random Patt p D n rn tern in Pattern n n ern ttern Train ttern te Noise ern ite Noise Gaussian Wh ReFFT Complex FFT Inverse FFT Power Spect Cross Power FHT P Inverse FHT P Zero Padder P Convolution P Deconvoluti P Correlation P Decimate P Derivative P Integral rum on continues 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 DS DS DS DS DS DS DS DS DS DS DS Windows DS DS DS DS DS DS DS DS DS Butterworth Coefficients Chebyshev Coefficients Inverse Chebyshev Coeff Elliptic Coefficients IIR Filter Equi Ripple LowPass Equi Ripple HighPass Equi Ripple BandPass Equi Ripple BandStop Parks McClellan edian Filter TT OS Os OO Te Blackman Window Exact Blackman Window Blackman Harris Window Hanning Window Hamming Window Flat Top Window General Cosine Window Exponential Window Force Window DSP Kaiser Bessel Window DSP Triangular Window Array Functions DSP Add DSP Subtract DS ultiply DSP Divide DSP Absolute DSP Square Root DSP Produc
60. SP 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 lo
61. T 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 f
62. The following documentation also contains information that you may find helpful as you read this manual e 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 e Digital Filter Design Parks and Burrus Wiley Interscience e Discrete Time Signal Processing Oppenheim and Schafer Prentice Hall e 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 i
63. V j eneen e a E EEA O A E EE ESEA 2 2 Copy Mem EV t DSP svi ii ccctvesicessetcescia sss seus a aa e aa aE i a Ea EA a a E ERN 2 3 National Instruments Corporation v NI DSP SRM for LabVIEW for Windows Contents DSP ADSOlULG rre eaer tig oo ea ach os pan ERE a E bens vecnontiv stash EARE A cbbea Wine oeapeatnss E REER AE 2 4 DI oaa Ca Ko AREER A E state ase Se E E E EE E 2 5 DSP Allocate MOM Ory ssiisisssssessscvssscosestecssonssanzasescdodssbaasvanssavspsnedvevosssbsescdsves e Ei Eas oE Eeoae EEPE nEs aai K EESE 2 6 DSP Bl ckman Wind Wi siones menaa E ei cee Se ce bea E A REE e E doa E En 2 7 DSP Blackman Harris Window siesena iecit ie i i E E E E E E a 2 8 DSP Butterworth Coefficients sene eree tire epore eneore e TOE See p ERR e San Epa Ee peiie TETEE ee a 2 9 DSP Chebyshev Coefficients oi cies aces estes scsech cee chtes ni e E e E Ee EEE EEEE EE RKE Raa 2 10 DSP CUP ea e E du ees E N A a E EEE eshte E Ae E SA REESEN 2 11 DSP Complex FE Dt a thin a a ed en oR ee ee E E A TES 2 12 DSP Com volt eea E E A E o eset ea EN a eiea e sdgsecmoade ete eves 2 13 DSP Correlation etones cose en eva Sockets cua de sos o A O E S EEE aE OA AE a TE ESES N 2 14 DSP Cross POWEr aeeti eE EE E E EEE EREE RE AEE TERE EAE E AEE E 2 15 DSP CUstOMi n e e a E A shen te EE E O S NA cons R E e E E 2 16 DSP Decimate senn hinia iee E A RA E EEE E EE aa AA OEE EATE EAA O OERA EAEN 2 17 DSP D c nVvoluti n oriec raan n a a a a a Coane doped tgs a a a 2 18
64. W 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 Lab VIEW 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
65. Y i a X i b for i 0 1 2 D 1 b o error out error in no error DSP Linear Evaluation zal 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 mma Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the results of output array cE 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 slot the full path name of a valid COFF file out path name error out file error in no error DSP Load slot is the board ID number slot defaults to 3 path name is the path name of the out file 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 1 EOE error out contains the error information for this call NI DSP SRM for L
66. _ BandPass_ BandStop ofBands m 2 d2 da _ gt oftaps odd even odd odd even fl 0 amp n 7 0 Odor even taps gt 0 0 0 0 0 for even taps gt 0 0 20 0 for odd taps 20 0 for odd taps fl O 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 Diff odd Diff even Hilbert odd Hilbert even bands n PEREN a m fu fo fom fo odd or even in the table refers to the value of the of taps fI 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 Ais 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 Fun
67. abVIEW 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 Y out obtained by using the following formula mult Yin ome f error out y i log10 X i mult error inno error for 1 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 o 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 aon Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the results of logarithm X mo 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 amp Min Finds the maximum and minimum values in the input array as well as the respective indices of max value the occurrence of the maximum and minimum s ma
68. akelib 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 board configuration EISA bus computer Part 1 1 3 ISA or AT bus computer Part 1 1 3 capabilities xv 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 software overview Part 2 1 1 to 1 2 ATDSP h file Part 4 2 2 AT amp T WE DSP32C digital signal processor Part 4 1 1 National Instruments Corporation Index 1 B bandpass filters See filter VIs 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 C 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 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 t
69. alls 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 gnaxmin 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 DSPHEM TEPME B E E m t H Figure 2 8 Block Diagram How to Index the Output Arrays of the Cust
70. alysis VI Reference DSP Decimate Decimates the input sequence X by the decimating factor and the averaging control averaging If Y represents the output sequence Decimated 4 Array the elements of the sequence Y are decimating factor rout obtained using error in no error ae error out Yin X im if ave no averaging DSP Decimate Y 1 m 1 i co y xX Toget i m k if ave averaging mM k 0 fori 0 1 2 size 1 n size trunc m where nis 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 lt decimating factor lt 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 pas
71. aq h contains all the prototypes of functions and the data structures that the data acquisition functions use You can find this header file in the LIB directory 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 s
72. as two options 4 bytes for 32 bit long and floating point data and 2 bytes for 16 bit short data ae 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 defaults to 0 offset is the index where to start to index into the array referenced by DSP Handle Cluster offset Handle Cluster starting from offset Indexed DSP Handle Cluster is a DSP Handle Cluster that contains the number of size data in DSP 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 out and no other calls are made error in no error contains the error information from a previous VI If an error occurs it is passed out 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
73. ay 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 conventions used in manual xii xiii organization of manual xi xii related documentation xiv DSP Absolute function Part 3 2 4 DSP Add function DSP Handle Cluster input output example Part 2 1 5 purpose and use Part 3 2 5 DSP Allocate Memory function Part 3 2 6 DSP Blackman Harris Window VI Part 3 2 8 DSP Blackman Window VI Part 3 2 7 DSP Butterworth Coefficients VI Part 3 2 9 to 2 10 DSP Chebyshev Coefficients VI Part 3 2 10 to 2 11 DSP Clip VI Part 3 2 11 DSP Complex FFT VI Part 3 2 12 DSP Convolution VI Part 3 2 13 DSP Correlation VI Part 3 2 14 DSP Cross Power VI Part 3 2 15 DSP Custom VI calling from LabVIEW interface Part 4 2 10 definition Part 4 2 9 executing from LabVIEW interface Part 4 2 12 purpose and use Part 3 2 16 DSP Decimate VI Part 3 2 17 DSP Deconvolution VI Part 3 2 18 DSP Derivative VI Part 3 2 19 NI DSP SRM for LabVIEW for Windows 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
74. cation 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 NJ DSP Analysis VI Reference handle memory management on the DSP board and data transfers between the DSP board and your LabVIEW application e Copy Mem DSP to DSP e Copy Mem DSP to LV e Copy Mem LV to DSP e DS
75. ccurs 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 e 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 e 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 DSP2200 subdirector
76. cients must be greater than or equal to one The rank of the filter must meet 1 lt 2 rank 1 lt size The size of the input array must be a valid power of two size 2 0 lt m lt 14 The onboard hardware diagnostic software found a memory hardware error Running functions on the board is unreliable There are no allocated buffers on the DSP board The indexing requested falls outside of the buffer into which indexing is requested 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 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 A 2 National Instruments Corporation Appendix A Error Codes Table A 1 NI DSP Analysis Library Error Codes Continued Error Error Name Description Number TransferSizeE The size of requested block transfer does not have enough space allocated for it on the board tDSPHandle The DSP Handle specified is not valid for this board it has either been freed or never allocated ndleSlotErr The DSP Handle used does not belong to the same board on which the function request is made ndleAllocErr Failed to allocate a handle in the interface code imeOutErr The Kernel did not respond or finish functi
77. criterion 0 lt fo lt f lt 0 5f where f is the pass freq f is the stop freq and f is 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 mma Filtered Data in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the filtered data output oro 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 l where n is the number of elements in X and m is the number of taps m 1 A delay equal to J 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 ie Blackman X the elements of Y are error in no error error out
78. ction 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 Magnitude coordinate points X Y The ith elements Phase mr of the rectangular set is obtained using the error in no error error out following formulas DSP Polar To Rectangular 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 ao 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 pSPHE the input signal array Phase ro 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 ro 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 Refe
79. ctures variables filenames and extensions and for statements and comments taken from program code NI DAQ NI DAQ is used throughout this manual to refer to the NI DAQ software for DOS Windows LabWindows unless otherwise noted NI DSP NI DSP is used throughout this manual to refer to the NI DSP software for LabVIEW for Windows unless otherwise noted PC 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 amp T WE DSP32C Developer Toolkit Abbreviations acronyms metric prefixes mnemonics symbols and terms are listed in the Glossary National Instruments Corporation 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 e AT DSP2200 User Manual part number 320435 01 e LabVIEW Data Acquisition VI Reference Manual part number 320536 01 e LabVIEW Getting Started Manual for Windows part number 320533 01 e LabVIEW User Manual part number 320534 01 e LabVIEW Utility VI Reference Manual part number 320543 01 e NI DAQ Function Reference Manual for DOS Windows LabWindows part number 320499 01 e NI DAQ Software Reference Manual for DOS Windows LabWindows part number 320498 01 e NI DSP Software Reference Manual for DOS LabWindows part number 320436 01
80. d 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 Enables the DSP board to run Use DSP Start with the DSP Load and DSP Reset VIs after ae slot downloading a custom application error out error in no error DSP Start 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 HE DSP Subset Extracts a subset of array X of length beginning at index and stores it in array Y Yin a Y out index i length ee error out error in no error DSP Subset 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 mo 4 mo length is the length of the output subset length defaults to 1 aa Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the subset output array are Y out is a DSP Handle Cluster that is
81. d 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 Examples subdirectory 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 size destination short the DSP board to one of the slat destination float destination arrays in LabVIEW sourcelDSP Address using the onboard DMA transfer error in no error error out method destination ty pe destination long DSP DMA Copy DSP to LY 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 t
82. dle 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 of Pee Shifted 8 out number of shifts selected can be in the Shifted 8 in ome ii errr out positive right or negative left direction error in no error _ DSP Shift 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
83. e 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 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 a ifi d Xj 0 elsewhere for 1 0 1 2 n 1 where ais the amplitude d is the delay and amplitude Chapter 2 delay Impu
84. e input array with zero from starting index to the end of the input array a Zero Padded starting index This VI is useful when the size of the error in no error error out acquired data buffers is not a power of two and you want to take advantage of fast DSP Zero Padder 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 To 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 oo 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 Utilit
85. e 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 Fandom Pattern in Random Pattern out are in the range 0 1 The sequence is seed generated using the Very Long Cycle error in no error os 1 error out random number generator algorithm DSP Random Pattern seed lt 65536 0 If seed gt 0 0 the generated noise will be the same in repeated invocations if the seed value does not change If seed lt 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 ao Random Pattern out is a DSP Handle Cluster that is identical to Random Pattern in but with the 2 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
86. e 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 ma 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 error in no error error out X J x t cas 2mft dt DSP FHT 00 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 v 0x cas 75 for k 0 1 2 1 i 0 where n is the number of elements in X an 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 ou
87. e parteris Sawtooth Pattern in m error out generated according to the following formula error in no error cycles DSP Sawtooth Pattern 2a T bona lt 2 Tk if0 lt k lt 5 yi for i 0 1 2 n 1 2a aD za lt TK T if sk lt T where k i d moduloT and T cycles d is the delay and nis the number of elements in the sawtooth pattern amplitude defaults to 1 0 cycle defaults to 1 0 delay defaults to 0 Sawtooth in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the output sawtooth pattern aoe Sawtooth out is a DSP Handle Cluster that is identical to Sawtooth in but with the generated pattern 2 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 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 E Set Hh error in no error error out is represented by the sequence Y then y set value for 1 0 1 2 n 1 DSP Set where n is the number of elements in X X is a DSP Han
88. eDSP Board from LabVIEW 2 12 National Instruments Corporation ix NI DSP SRM for LabVIEW for Windows Contents Tables Part 1 Table 1 1 subdirectories Created Dy SETUP iiiter iene EEEE RE R OE EE E sbeeeneenes 1 2 Part 3 Table 1 1 The NI DSP Analysis VI Groups iire ererat oro eao ie e oa ans EE E E EEE EEE suede CENE SINEES Eoee 1 1 Part 4 Table 2 1 Files Required to Build the Custom DSP Library Example sseessssesseseeeesesesessrrrrsreesrrerereeserrssene 2 7 Appendix A Table A 1 NI DSP Analysis Library Error Codes 0 ccc cecceceseeseceseeseceeceseceeceeeeeeeeseneeseecsessaecaecaeeaessaeeaees A 1 NI DSP SRM for LabVIEW for Windows x National Instruments Corporation 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 T
89. ece ceeeseeseceseeeeceeceeeceeceeeseneeeaeeaeesaeenee 1 7 Part 4 Figure 1 1 NI DSP for DOS Directory Structure siyete eceecseeseeeececseeecesecseeececsaeeesseceecnevseesesaecaeeseaesaeeees 1 1 Figure 1 2 Interface Layers to Onboard Functions cc eeeesseceececeseeeseceecececesacesseeecsecesceceseeesaeceaeecsaeeeneeesees 1 2 Figure 2 1 Tinker File NIDSPENK is secissctiii advance kee scitiind het crc E aE EE EEE a steiives decisis as ae 2 4 Figure 2 2 Library Function List File NIDSP Ane c cssucssciegsssescaevetheessse ereere caste sre i ea oea Seia 2 4 Figure 2 3 Typical Section of NIDSP Ane 5 isis esere ieee EEEE EEE EEKE E OEE EES EEEE EEEE 2 5 Figure 2 4 Signals Group Section in Aspnes oo eee eeeecceseeeseeseecseceeecaecsaeesecseeseceseeseseeeeeesaeseaeeneseaeenee 2 6 Figure 2 5 Signals Group Section in dispatch s eee eee eeeeeceseecseeseecaeceecaecesessecsseeseceseseseeeeeeseseseaseaeeeneenes 2 6 Figure 2 6 How to Bundle Parameters in LabVIEW to Call gmaxmin C 00 eee eceeeeeeeseeeeeceeenecsseeseenseesees 2 9 Figure 2 7 How to Connect to Custom VI to Call gmaxmin C 0 eee ce eeeeeceeeceeceseeeneeseecsecseecaecsaeeaeesaeeaees 2 10 Figure 2 8 Block Diagram How to Index the Output Arrays of the Custom VI oo eee eee ceeeseeeeeeees 2 11 Figure 2 9 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 th
90. ed 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 incl
91. edback 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 Re FFT represents the output sequence then Im FFT 8 error in no error Y F X DSP Inverse FFT ara 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 arma Im FFT X is a DSP Handle Cluster that indicates the memory buffer o
92. el Computer Bus XT AT ISA or EISA Microprocessor Clock Frequency Type of Video Board Installed DOS Version Programming Language e 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 Lab VIEW for Windows Edition Date December 1993 Part Number 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 National Instruments Corporation 6504 Bridge Point Parkway MS 53 02 MS 53 02 Austin TX 78730 5039 512 794 5678 Glossary ee e MSE Mwords pt RTSI SCXI s word one dimensional two dimensional complex decibels Discrete Fourier Transform direct memory access digital signal processing Fast Fourier Transform Fast Hartley Transform finite impulse response filter
93. erformed 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 ith element of the output array Z is obtained using the PA Z out following formula oo in a error out Zi X i YG error in no error fori 0 1 2 n 1 DSP Divide where n is the smaller number of elements in X and Y mmm X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array X mmm Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array Y ora Z in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the DSPHC i results of X i Y i Pal 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 El
94. es 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 WOAQCONF EXE of your LabVIEW directory From Windows you run WOAQCONF EXE by 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 e 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 SETUP program is used to assign a slot number to the AT DSP2200 to choose the memory subtype either 64 Kwords or 2128 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 CFG file on this diskette 2 Run the utility from a directory on your hard disk and place a copy of the NIC1100 CFG file in this directory The EISA system configuration utility is typically named CF EXE e You must use the WOAQCONF utility 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 Installati
95. es 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 Branch Offices Australia Austria Belgium Denmark Finland France Germany Italy Japan Netherlands Norway Spain Sweden Switzerland U K 800 328 2203 512 794 5678 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 91 640 0085 08 730 49 70 056 20 51 51 0635 523545 National Instruments Corporation B 1 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 91 640 0533 08 730 43 70 056 27 00 25 0635 523154 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 ha
96. es 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 returns an error lower stop freq defaults to 0 1 lower stop freq must be greater than zero If lower stop freq is less than or equal to zero the VI higher stop freq must be greater than higher pass freq and must observe the Nyquist criterion O0 lt f lt f lt f lt f lt 0 5f where f is lower stop freq f is lower pass freq f is higher pass freq f is the higher stop freq and f is 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 o Filtered Data in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the filtered data output oro 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 l 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 2 delay error in
97. es 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 d B E bytes element selector defaults to 4 bytes ma DSP Handle Cluster indicates the allocated memory buffer on the DSP board 2 mo a 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 i Blackman is are obtained from the following formula error in no error error out y x 0 42 0 50 cos w DSP Blackman Window 0 08 cos 2w for 1 0 1 2 n 1 271 gt n I where n is the number of elements in X ane 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 t
98. esenting 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 fo
99. evious 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 init represented by the sequence X then the final Ramp Patternmout pattern is generated according to the Ramp Pattern in cmt ery error out formula error in no error Xi Xo i Ax DSP Ramp Pattern for 1 0 1 2 n 1 Xn 1 7 Xo where Ax 5 n 1 X 1 Is the final Xo is 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 aan Ramp Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the output ramp pattern ooo Ramp Pattern out is a DSP Handle Cluster that is identical to Ramp Pattern in but with the generated r 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 ar
100. fX 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 f of the signals x t and y t is defined as follows Y Im Sey error in no error A error out Sy XAYA DSP Cross Power where X f is the complex conjugate of X f X f F x t and Y F y o This VI uses the FFT routine to compute the cross power spectrum that is given by the following formula Sxy 3 F X FY n 2k for k 1 2 3 23 where Sxy represents the complex output sequence Sxy and nis the number of samples that can accommodate both input sequences X and Y ooo 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 Ifthe 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
101. fer 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 amp 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 AEDSP2200 CI CV a a a co CJC CJC E Memor Lib CS gowo o o doooe 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 execu
102. ferentiation f t of a Yin function F t is defined as follows y initial condition f t i F t final suse i t error out error inno error Let Y represent the sampled output sequence d dt X The discrete implementation is given DSP Deriratire by the following formula uz vat Gye fori 0 1 2 1 0 1 where nis the number of samples in x t X is specified by initial condition when i 0 and X is 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 mo 4 mo initial condition defaults to 0 0 E scr final condition defaults to 0 0 F 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 El 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 mo 19 o ooo 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 a a Ca 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 p
103. ffer 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 f A zum the following formula gt error in no error error out DSP Sum where n is the number of elements in X ara 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 out and no other calls are made error in no error contains the error information from a previous VI If an error occurs it is passed out 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 slot execution timeout error out error in no error The default timeout setting at startup and after a DSP Reset call is 10 s CSP TimeOut 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 V
104. ffer 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 freq defaults to 0 2 stop freq must be greater than zero If stop freq is less than or equal to 0 the VI returns an error stop high freq must be greater than stop freq and observe the Nyquist criterion 0 lt f lt f lt 0 5f where f is the stop freq f is the high freq and f is 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 aan Filtered Data in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the filtered data output oro 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 l 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 2 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
105. fl ma Feedback Coefficients out sequence of data BAH E E E il i 2 Hea a mt mo 4 mo error in no error wm error out attenuation order DSP Inv Chebysher Coeff 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 lt fl lt fh lt 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 fe
106. 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 o 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 o 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 Slot written on the DSP board from Array of Handles in Array of Handles aut LabVIEW For more details Array of SGL Scalars in Array of SGL Scalars out about how to use this VI refer Array of IZ2 Scalars in oe ane Eor Array of IZ2 Scalars out to Part 4 NI DSP Interface error in no error error out Utilities of this manual function code DSP Custom slot is the board ID number slo
107. g 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 nnan a Assumption of Previous Knowledge escesccesseceneeceseeesseesececeeeesnecseeesaeceaeecsaeeesaeeeneecseeseaeceeeesaeeeaeees xi Organization of This Manual iss orsnsrer niestier eo irou teete vsi Ee E Ea E EEEa EEEE A ES E SE EEES xi Conventions Used in This Man dlise smere en a a E E E E ES xii Related Documentation r E E a a a EAE re Ce aus TEKES EE SE LESENE xiv Additonal Softwares esinen eer eeir E E ie Ee E eria ea Seaia oiee aari E eE Eis xiv NI DAQ for DOS Windows LabWindows cccessccessscecensecesnececesaeceseceeceseeeesesaecesesseessneeeneaes xiv Developer Toolkit reais ninnerae iei eai i E E E eE KEE S a XV Compatible
108. gnal array Cxy The number of elements in Cxy must be greater than or equal to the number of elements in Y ain Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array Y we au 7 oon X in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the deconvolution results we au T 7 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 ooo X out is a DSP Handle Cluster that is identical to X in but with the deconvolution results already stored L 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 dif
109. he 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
110. he input array X is overwritten by the output Blackman X oo Blackman X is a DSP Handle Cluster that is identical to X but with the results of Blackman X L 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 a Harris X the elements of Y are error in no error error out obtained using the following formula Blackman Harris 4 DSP Blackman Harris Window y x 0 42323 0 49755 cos w 0 07922 cos 2w for 1 0 1 2 n 1 2ni n gt where n is the number of elements in X are 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 Tmn 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
111. he 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 Wind
112. he operation is performed in place and the input array X is overwritten by the output Force X duty 0 0 lt duty lt 100 0 duty defaults to 50 0 oo 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 DSF Handle Cluster error in no error DSP Free Memory aoe 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 Gaussian Pattern in statistical profile is as follows standard deviation Bese crane seed EE error out m 0 s error in no error where s is the absolute value of the specified standard deviation DSP Gaussian White Hoise standard deviation default
113. her 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 e 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 gnaxmin 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 e 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 e 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 h should be included at the begin
114. his 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 e 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 e Part2 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 e Part3 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 a
115. holds the input array sorted tif include atdsp h short DSP_GMaxMin z y max min n imax imin register float z max min 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 long n imin imax long i j float x small float localmin localmax if n lt 0 return 0 X Z localmin x localmax localmin ximax 0 ximin 0 for i O i lt n i x if localmax lt x localmax x ximax i if localmin gt x localmin x ximin i min localmin max localmax for i 0 i lt n i yli zl il for i 0 i lt n itt for j itle JS Sny J if y i gt y j small y j ylj3 ylil y i small return noError Chapter 2 Note The file ATDSP h is the header file containing all of the error codes used by the DSP Library You can find it in the LIB directory 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 Include directory of your WE DSP32C tools Otherwise you should specify the correct path of ATDSP h in your program Guidelines for the Custom Functions When adding functions to build your custom DSP Library follow these guidelines e Pass all parameters by address All input output and input output parameters whet
116. hort 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 e 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
117. ies 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 amp 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 amp 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 subdirectorie
118. 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 sourcelshort buffer on the DSP board which is referred to by a DSP Handle Cluster sourcel floaty Destination out l Destination in A The source buffer can contain one of three error in no error AAEREN kinds of data float
119. inal 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 as 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 ooo Y out is a DSP Handle Cluster that is identical to Y in but with the results of integration already stored L 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 J f t dt F b F y 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 Feedback Coefficients in Forward Coefficients in filter type aves sampling freq f F these coefficients to the DSP high cae pee th T Forward Coefficients out IIR Filter VI to filter a low cutoff freq
120. ion is performed in place and the input array X is overwritten by the output Hamming X aro Hamming X is a DSP Handle Cluster that is identical to X but with the results of Hamming X L 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 DSP Handle Cluster DSP Address error in no error error out You can use the output DSP Address as the input of the DSP Address terminal in DSP Handle To Address 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 out error out and no other calls are made error out contains the error information for this call oH error in no error contains the error information from a previous VI If an error occurs it is passed 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
121. 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 nokError unknownDeviceE badDeviceErr deviceSupportE memFullErr configFileErr cmosConfigErr DLLInterfaceErr SamplesGTZero SamplesGEZero SamplesGETwo SizesGTZero xSizeGl EySize OutSizeGEInSize OutOfMe DecFactErr WidthLTSamples IndexLTSamples DelayGI WidthGl EZero EZero DelayWidthErr National Instruments Corporation No error the call was successful The board specified is not a National Instruments DSP board The board number used in function call should be 1 lt board lt 8 Function cannot be executed by specified board Ins
122. l 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 float and long data types for scalars If you use any of the error codes from Appendix A Error Codes include ATDSP h at the beginning of your source code A function that completes with no error should return noError or zero The ATDSP h file is in the Lib directory After you have installed the WE DSP32C tools you may want to copy ATDSP h to the Include directory in the root directory for those tools In this example use d3cc c gmaxmin c to generate WE DSP32C object code 3 Add Your Object Filenames to a Linker File ifile The next step is to modify the file NIDSPLNK to add the names of your custom object files to the list of modules to be linked into your custom DSP Library NIDSPLNK is 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 1 LVDSP Figure 2 1 Linker File NIDSPLNK Copy the NIDSPLNK file from
123. lates Y through the discrete implementation of the inverse Hartley integral n 1 y t ox cas R 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 ama 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 ma 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 2 mo a a 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 Ap
124. liptic Coefficients Generates the set of filter coefficients is implement a Feedback Coefficients in digital elliptic IIR filter You Forward Coefficients in can then pass these filter type ii coefficients to the DSP IIR sampling freq fs 7 Forward Coefficients out uler high cutoff freq fh Feedback Coefficients out low cutoff freq fl tinder error out error in no error passband ripple order stopband attenuation DSP Elliptic Coefficients 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 AAG Note fh and fl must observe the Nyquist criterion 0 lt fl lt fh lt 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 retu
125. lities 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 or jii Bus AT DSP2200 User DSP Software DSP Hardware Application Onboard Kernel in LabVIEW DSP Board Memory Management Data Transfer VI Functions Interface CIN DSP Library Onboard Memory Interface Library DSP DLL Figure 1 1 Communication between the PC and the DSP Board NI DSP SRM for LabVIEW for Windows l 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 D
126. loat xFinal float y float fs float f_low float f_hi long order float a float b long type float fs float f_low float f_hi long order float ripple float a float b long type float fs float f_low float f_hi long order float ripple float atten float a float b long type float x long n long taps float fs float fhl float 12 float fh2 float 13 float y float x long n long taps float fs float fhl float 12 float fh2 float 13 float y float x long n long taps float fs float fhl float 12 float y float x long n long taps float fs float fhl float 12 float y float x long n float a float Condx long sza float b float Condy long szb float y float fs float f_low float f_hi long order float ripple float a float b long type er float x long n long rank float y lian long n float fs long bands float Amp float flow float fHi float wRipple long filterType float h float delta t x long n long type at x long n float finalval loat x long n float duty n float x long n float coeff long m at x long n float beta at x long n x rong n float y xy Eloat yp long n float z x long n float upper float lower float y x float y long n float z t x long n float a float b float y x long n float mult float y float x long n float max
127. lse Pattern out Impulse Pattern in error out error in no error n is the number of elements in Impulse Pattern DSP Impulse Pattern 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 amplitude defaults to 1 0 ro Impulse Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the output impulse pattern ooo 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 period amplitude If the impulse train pattern is represented Padela Impulse Train out by the sequence X the VI generates the Impulse Train in mE error out pattern according to the following error in no error formula DSP Impulse Train Pattern amplitude if i
128. lter coefficients to implement an IIR filter as specified by the Feedback Coefficients in Forward Coefficients in Chebyshev filter model filter type You can then pass sampling freq fs L coefficients to the DSP IIR high cutoff freq th Forward Coefficients out Filter VI to filter a sequence low cutoff freq cf s Po Feedback Coefficients out of data error inno error wa error out ripple order DSP Chebysher Coefficients 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 Aa GE Note fh and fl must observe the Nyquist criterion 0 lt fl lt fh lt 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 df HI 0 Forward Coefficients in is a DSP Handle
129. ly one of them when you want to deallocate their DSP board buffer x od E out E in oe es error out error in no error DSP Add 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 e 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 Source Source Cd Ld 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 TF 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
130. ment of the resulting array is obtained by upper limit Clipped A out using the following formulas lower limit error out error in no error upper if X i gt upper Y lower if X i lt lower DSP Clip X i otherwise for 1 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 EAR 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 mmn 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 ara Clipped X out is a DSP Handle Cluster that is identical to Clipped X in but with the Clipped X L 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 Ir th En error in no error Y F X DSF Complex FF
131. modulo P d Xi 0 0 elsewhere for 1 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 period defaults to 1 delay defaults to 0 RER Impulse Train in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the output impulse train arma Impulse Train out is a DSP Handle Cluster that is identical to Impulse Train in but with the generated 2 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 DSP Index Memory Indexes into a DSP buffer allocated in the memory offset space of the specified DSP DSP Handle Cluster pereen board The return value is data type DSPMEN Indexed DSP Handle another DSP Handle Cluster error in no error error out This VI does not allocate memory size In order to index into a buffer DSP Index Memory correctly you need to indicate what type of data is in the DSP Chapter 2 Cluster buffer You must set data type to the appropriate type data type h
132. mory 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 e 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 on
133. n X error out and no other calls are made error out contains the error information for this call error in no error contains the error information from a previous VI If an error occurs it is passed out NI DSP SRM for LabVIEW for Windows 2 62 Part 3 NI DSP Function Reference NI DSP Analysis VI Reference 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 lt i lt last Xi 0 elsewhere for 1 0 1 2 n 1 last d w where ais the amplitude d is the delay w is the width and Chapter 2 Pulse Pattern in amplitude Pulse Pattern out delay width Saw error out error in no error DSP Pulse Pattern n is the number of elements in Pulse Pattern amplitude defaults to 1 0 width must be greater than or equal to 0 width defaults to 1 delay must be greater than or equal to 0 delay defaults to 0 ma Pulse Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the output pulse pattern io 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 pr
134. n array containing a sinusoidal pattern If the Sine Pattern is represented Sine Pattern in by the sequence Y the pattern is amplitude Sine Pattern out ne according to the following phase degrees on error in no error error out yj asin x cycles for i 0 1 2 n 1 DSP Sine Pattern 2ni k H T o n 180 where x a is the amplitude k is the number of cycles in the pattern is the initial phase in degrees and nis 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 E 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 b
135. n 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 ro 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 o 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 as error in no error Heatley error out DSP Inverse FHT x t X f cas 2nft df where cas x cos x sin x If Y represents the output sequence X the VI calcu
136. nboard memory The function and input parameters for Al 1oc_Mem are defined as follows void Alloc_Mem size long size where size is the requested block size in bytes Alloc_Mem returns a pointer to the allocated buffer or returns NULL if the allocation failed Call the function F ree_Mem when 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 Free _Mem ptr long ptr where pt r is the address of the buffer to be deallocated Free_Mem deallocates the buffer pointed to by ptr if that buffer was allocated previously using Al loc_Mem Free Mem returns NULL if 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 fnc file 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 r
137. ndex Memory Part 3 2 43 DSP Init Memory Part 3 2 44 list of functions Part 3 1 2 prototypes for customizable functions Part 4 3 2 to 3 3 memory map files Part 4 2 7 MLUT See Memory Look Up Table MLUT N NI DAQ for DOS Windows LabWindows software xv NI DSP Analysis VIs accessing Part 3 1 3 to 1 4 array VIs Part 3 1 2 Copy Mem DSP to DSP Part 3 2 1 Copy Mem DSP to LV Part 3 2 2 Copy Mem LV to DSP Part 3 2 3 definition xi DSP Absolute Part 3 2 4 NI DSP SRM for LabVIEW for Windows Index 4 DSP Add Part 3 2 5 DSP Allocate Memory Part 3 2 6 DSP Blackman Harris Window Part 3 2 8 DSP Blackman Window Part 3 2 7 DSP Butterworth Coefficients Part 3 2 9 to 2 10 DSP Chebyshev Coefficients Part 3 2 10 to 2 11 DSP Clip Part 3 2 11 DSP Complex FFT Part 3 2 12 DSP Convolution Part 3 2 13 DSP Correlation Part 3 2 14 DSP Cross Power Part 3 2 15 DSP Custom Part 3 2 16 DSP Decimate Part 3 2 17 DSP Deconvolution Part 3 2 18 DSP Derivative Part 3 2 19 DSP Divide Part 3 2 20 DSP Elliptic Coefficients Part 3 2 21 to 2 22 DSP Equi Ripple BandPass Part 3 2 23 to 2 24 DSP Equi Ripple BandStop Part 3 2 25 to 2 26 DSP Equi Ripple HighPass Part 3 2 27 to 2 28 DSP Equi Ripple LowPass Part 3 2 29 DSP Exact Blackman Window Part 3 2 30 DSP Exponential Window Part 3 2 31 DSP FHT Part 3 2 32 DSP Flat Top Window Part 3 2 33 DSP Force Window Part 3 2 34 DSP Free Memory Part 3 2
138. ning of your source code A function that completes with no error should return noError or zero e Use the memory management routines Perform any dynamic onboard memory allocation using the functions Alloc_Mem and 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_Mem and Free_Mem Use these functions instead of the standard C routines malloc and free Call the function Al 1loc_Mem whenever 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 gt 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 o
139. 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 Filtered Data in Parks McClellan algorithm and number higher pass freg i of taps lower pass frequency higher lower pass freq Fz Filtered Data out pass frequency lower stop frequency i a ER error out higher stop frequency and sampling of taps frequency The VI then filters the input error in no error sequence X to obtain the bandstop lower stop freq filtered linear phase sequence Filtered higher stop freq Data sample freq fs The first passband region of the filter DSP Equi Ripple BandStop 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 E 0 lt fi lt fi lt f lt f lt O5f where f is the lower pass fre
140. nt on our products and manuals e The Glossary contains an alphabetical list and description of terms used in this manual including abbreviations acronyms metric prefixes mnemonics and symbols e 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 lt gt Angle brackets enclose the name of a key on the keyboard for example lt enter gt 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 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 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 lt enter g
141. nts 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 ooo 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 f sequence Triangle X the elements of Y are F Triangle tx obtained from the formula error in no error error out y x tri w DSP Triangular Window for 1 0 1 2 n 1 2i n n I where tri w Iwl 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 T
142. o 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 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 files required to build example library Part 4 2 7 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 calling Custom VI Part 4 2 10 executing custom function from LabVIEW Part 4 2 12 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 Custom VI See DSP Custom VI customer communication xv B 1 D 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 arr
143. o 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 SGL uy destination sh
144. oard 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 duty ane nt Square Pattern aut Square Pattern in er error out cycles the pattern is generated according to the following formula error in no error DSP Square Pattern a ifO0 lt remainder y lt 0 01dT X 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 nis 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 0or
145. ollows Cxy in cm error out 00 error in no error Cxy t x t t T y t T dt TOEA Jx IEN DSP Convolution 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 0 i lt O or i n y 0 j lt O or j2m then you obtain the elements of Cxy using n 1 Cxyi xn yix for i 0 1 2 size 1 k 0 size n m l ma X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array X To Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array Y To 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 fo
146. om VI to Obtain Results of gnaxmin 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 m ae czt mm t coed ALL Free im ISPHEM ii Allocate function ID SGL S60 mase value max value min value SG J roin value max index min index i max index index Figure 2 9 Block Diagram Using the Custom VI to Call gmaxmin c on 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 funetion ID number of elements min index Figure 2 10 Front Panel Using the Custom VI to Call gmaxmin c on 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
147. on execution in the selected timeout FF_FilePathl Could not find the DSP Library file in the directory set by the DAQCONF utility CoffHdrErr The file you are attempting to load to the DSP board is not created using the WE DSP32C tools not an acceptable COFF file CoffSectErr Found an unimplemented section flag in the COFF file AddrSpaceErr 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 274 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 DAQCONF utility FNC_E 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 NotFindFunctionInDLL An error occurs when loading dsp d11 library in CIN LoadDLLLibErr An error occurs when loading a function in dsp d1l1 in CIN ErrInGetFunctionHandle An error occurs when calling the Get IndirectFunct ionHand_1e function in CIN InvalidBytesSelect bytes element selector can only select between 4 bytes 0 and 2 bytes 1 AllocArrayFailed Failed to allocate a LabVIEW array InvalidArrayType The array type can only be 32 bit floating poin
148. on on an ISA or AT Bus Computer Before using NI DSP you must run WOAQCONF EXE to 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 uti
149. onse dips EEE RNE REEE ESSER 2 39 DSP LERA SEA LIE E E E E E E E 2 40 DSP Impulse Patterns eoin Seige ee ees E EN E EEE EE aS E E EEES E EE ERE ETSk 2 42 DSP Impulse Train Pattern ssscscs ssssccsssdscassdststasassssnsevehe ostesdscsestvacecansebssdbusdecha sveadssuapasaseisessdvansecsvesnscdisans 2 43 DSP Index Memory cioten eri a Ara a este had he Se habeas Aovidesst Aig ated desta a 2 44 DSP Init Memory ocorre ein a S E E EE E a EE EEE a EE A E EEES 2 45 DSP Integral E E E E E 2 46 DSP Inv Chebyshev Coeff scscssssseesscsscsssesssesseansecteshasscasevsnassassisvessvasesseseseveeaiouss oa scdnseaeasveadeveassnasveseves 2 47 DSP Universe FE E AE edie tisk Soest ce tothe wea bea Senco se waves botee E A E 2 48 DSP Inverse FAT ae ceeds deveavenialisiyctececaeveveaaageesiscaecesseteancece Sada aE AEE A TE E RTE 2 49 DSP K iser Bessel Window o e eei iee ee e a a e e E e Buadeuedeoceres 2 50 DSP Linear Eval ation es neinn eenei ie E E ise Dessous EE E E Eiin 2 51 DSP Ladies eii e nr roves ongtve cectote r AES deedasbeneehs dates Maeeh luae has teeveee couse ennai EEA 2 51 DSP Log AEE iain hen ahaha E gel i RRA Ae ie 2 52 DSP Max Mani ssc sess sscesee sssssaeicsyssesssstsescaasdsaceavhssuasaivecdovapsessscebstesdeosbassdbaieashssueaseseassnapetsescbesteeds sebevcdesady 2 53 DSP Median Filtet sscis s cc Sicdeet nce inceuats Pus anae ces de cos bac sta subch duase sndods bavi a aA tutus Soshsdh acs i a a 2 54 DSP Multiply sissie rrer aa aa E a R
150. ontains the input signal array X o Y is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array Y o Z in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the results of XG YQ ooo 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 H in using the number of taps sampling frequency filter type and Band Parameters of taps sampling freq fs Band Parameters sod A error out error in no error filter type DSP Parks McClellan of taps is the total number of coefficients in H A tap corresponds to a multiplication and an addition 132 of taps is th ber of coeffici in H A tap pond Itiplicati d dditi If there are n taps every filtered sample requires n multiplications and n additions
151. 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 function function group function function Figure 2 3 Typical Section of NIDSP fne 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
152. ort 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 132 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 op 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 ET 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 sourcelshort Address on the DSP board using the slot onboard DMA method sourcel float ip error out destination DSP Address This VI uses only the DMA controller error in no error on the DSP board to transfer data sourcellong Therefore it does not interfere with DSP DMA Copy L to DSP 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 kind
153. ory 2 makelib bat rebuilds the DSP Library file with name 1ibfile 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 bat lists 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 t
154. ows 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 LabVIE
155. 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 DSP Handle Cluster number of points 045 bytes element selector Figure 1 4 Front Panel An Example of How to Allocate a DSP Handle Cluster DSP Handle Cluster SEMEN number of points ae bytes element selector 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 me
156. plied to the input in the linear difference equation and represent the multiplication factors for delays The equation suggests that there are N such coefficients b isthe 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 4andn 2 Using the filtering equation produces a 0 y 2 b O 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 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
157. plies a Kaiser Bessel window to the input sequence X If Y represents the y TEE output sequence Kaiser Bessel X the bets halser Bessel ua elements of Y are obtained using the error in no error i error out formula DSP Kaiser Bessel Window 1 By1 0 a Yj Xj 1 B for 1 0 1 2 n 1 a E k n 1 k z where nis the number of elements in X and Io 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 ro 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 ith element of the output array Y is obtained using the following formula Y out
158. preceding formula and occurs when iAt 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 lt delta t lt 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 A AG o Sinc Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the output sinc pattern aoe Sinc Pattern out is a DSP Handle Cluster that is identical to the Sinc Pattern in but with the generated g 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 a
159. 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 e 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 processe
160. q f is the lower stop freq f is the higher stop freq f is the higher pass freq and f is 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 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 lower stop freq must be greater than lower pass freq If lower stop freq is less than or equal to lower o Filtered Data in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the filtered data output oro Filtered Data out is a DSP Handle Cluster that is identical to Filtered Data in but with the filtered L 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 l where n is
161. r H HE DSP Reset 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 Reverse error in no error error out DSP Reverse aa 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 amo Reverse X is a DSP Handle Cluster that is identical to the X but with the reverse X already stored rEPHE 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 i pattern is represented by the amplitude Sawtooth Pattern aut delay sequence T thin t
162. r 2 DSP ReFFT Computes the Fast Fourier transform of a real input sequence X If Y represents ad m Fe FFT the complex output sequence then Im FFT in FAH Inn FFT 3 out error in no error P error out Y F X DSP ReFFT are 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 slot reloads the DSP Library from the directory error out specified in the WDAQCONF EXE file error in no erro
163. r 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 yli xli wh 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 WIKI 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
164. r that indicates the memory buffer on the DSP board that will contain the square root results of X i faa 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 Sorted x out direction Sorted W in mmf r error out error in no error o 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 o 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 ro 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 boar
165. r 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 t F of the signals x t and y t is defined as follows Soy Ray in wees Vl error out 00 error in no error Rat USO fo yer ut DSP Correlation where the symbol denotes correlation For the discrete implementation of the correlation let R 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 You then obtain the elements of R using the following formula 1 5 Rxyi 1 x k y k m 1 for i 0 1 2 m n 1 k 0 ama 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 ao al 19 19 o o 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 nt m 1 elements long nis the size of X m is the size of Y HI o 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 2 mo a T Note I
166. r to load this new DSP Library file make sure you run WOAQCONE as 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 map or dsp map as input to makelib bat Ifthe 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 map and out for the fields memmap and 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 bat batch file performs the following l makelib bat assembles the file stackbld s in the current directory using the AT amp T d3as exe assembler You need to assemble this file because it includes the file stackbld s created by the Dispatch program and contains the assembly code that determines which function the DSP board executes The file dispatch s reflects the organization and content of the DSP Library file as is reflected by the library function list file NIDSP fnc in the current direct
167. rdware or software products related to this problem include the configuration forms from their user manuals Include additional pages if necessary Name Company Address Fax Phone Computer brand Model Processor Operating system Speed MHz RAM MB Display adapter Mouse yes no Other adapters installed Hard disk capacity MB Brand Instruments used National Instruments hardware product model Revision Configuration National Instruments software product Version Configuration 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 e 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 Mod
168. rence DSP Polynomial Evaluation Performs a polynomial evaluation on the input array X The ith element of the output array Y Y aut is obtained using the following formula Coefficients a Yin a error out k 1 error in no error i YA E Coefficients X o a Oe XOD DSP Polynomial Evaluation for 1 0 1 2 n 1 o X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array X an Coefficients is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the coefficients array o Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the results of output array io 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 Power Snecirirn S f of a function x t is defined as F error in no error fe er
169. 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 NJ DSP Analysis VI Reference Table 1 1 lists these VIs by group Table 1 1 The NI DSP Analysis VI Groups Signal Generation DS DS DS DS DS DS DS DS DS DS DS DS Frequency Domain D
170. riangle X Seo 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 DSP Uniform White Noise Generates a uniformly distributed pseudorandom pattern whose values are in the range a a where a is the absolute amplitude defaults to 1 0 Chapter 2 White Noise in White Noise out amplitude seed a error out error in no error DSP Uniform White Noise seed lt 65536 0 If seed gt 0 0 the generated noise will be the same in repeated invocations if the seed value does not change If seed lt 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 Ta White Noise in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the output white noise ma 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
171. rns 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 IRER il 0 Forward Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the forward coefficients 2 mo a Ei mma Feedback Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the feedback coefficients ama 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 Tma 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 ou
172. ror out Sx XOXO IXH1 where X f F x t and X f is the complex conjugate of X f DSP Power Spectrum nma 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 cE 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 2 mo a 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 ss ee obtained using the following formula i gt proge error in no error A error out 1 product Tl X i i 0 DSP Product where n is the smaller number of elements in X on 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 i
173. rranged alphabetically e Part4 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 National Instruments Corporation xi 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 e Appendix A Error Codes contains a list of the error codes returned by the NI DSP Analysis VIs and the corresponding error messages e Appendix B Customer Communication contains forms you can use to request help from National Instruments or to comme
174. rt 3 2 18 DSP Derivative Part 3 2 19 DSP Integral Part 3 2 45 list of functions Part 3 1 1 prototypes for customizable functions Part 4 3 2 utility VIs Vv DSP Custom Part 3 2 16 DSP Handle to Address Part 3 2 38 DSP Load Part 3 2 50 DSP Reset Part 3 2 63 DSP Start Part 3 2 72 DSP Timeout Part 3 2 74 DSP DMA Copy DSP to LV Part 4 4 3 DSP DMA Copy LV to DSP Part 4 4 4 list of functions Part 3 1 3 VIs See NI DSP Analysis VIs W WDAQCONF EXE utility Part 1 1 3 Part 4 2 7 window VIs DSP Blackman Harris Window Part 3 2 8 DSP Blackman Window Part 3 2 7 DSP Exact Blackman Window Part 3 2 30 DSP Exponential Window Part 3 2 31 DSP Flat Top Window Part 3 2 33 DSP Force Window Part 3 2 34 DSP General Cosine Window Part 3 2 36 DSP Hamming Window Part 3 2 37 DSP Hanning Window Part 3 2 38 DSP Kaiser Bessel Window Part 3 2 49 DSP Triangular Window Part 3 2 78 list of functions Part 3 1 2 prototypes for customizable functions Part 4 3 2 windowing Part 3 1 6 to 1 8 mainlobes Part 3 1 7 sidelobes Part 3 1 7 spectral leakage Part 3 1 7 National Instruments Corporation
175. s a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the forward coefficients ae Feedback Coefficients in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the feedback coefficients oro Forward Coefficients out is a DSP Handle Cluster that is identical to the Forward Coefficients in but L with the forward coefficients already stored in the memory buffer on the DSP board Tma 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 out and no other calls are made error in no error contains the error information from a previous VI If an error occurs it is passed out 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 fi
176. s 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 ma ma g m Sanr Amant 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 lie 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
177. s 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 National Instruments Corporation 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 amp 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 frequenc
178. s to 1 0 seed lt 65536 0 If seed gt 0 0 the generated noise will be the same in repeated invocations if the seed value does not change If seed lt 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 o Gaussian Pattern in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the output Gaussian pattern oro 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 Gentes 8 Coefficients input sequence and Y represents Cosine Coefficients the output sequence GenCos X the elements error in no error error out of Y are obtained using the formula DSP General Cosine Window y X ye 1 cos kw K 0 for 1 0 1 2 n 1
179. s 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 Dispatch 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 re
180. sdspescvoavaschasnveeoipessossdauaesspassausesassvapserebes 1 2 Chapter 2 Getting Started with the NI DSP Interface Utilities 0000000 2 1 Creating Your Custom NI DSP Library ieis ornen nea e a eeaeee reog ge eE E o E E aE 2 1 1 Create Your Source Code of C Functions eesesseeessseesesresrsrestsresrersreerssreresreeesernrerrerrerssee 2 1 GMaxMin c Example ssscsresie eesse rr aa a E En EE Ek EEE aea odes 2 1 Guidelines for the Custom Functions eseseesesseesesseesrsreererteerrsresrssrerrsreererenerrreresssee 2 2 DSP Board Memory Management eseseessseeessseeesseeeressesresesresteresrenrerenresensesesserees 2 3 2 Compile and or Assemble Source Code essssesessseeesesreesseeessstsrestsstrteserrrnreerssesesrneerreseeees 2 4 3 Add Your Object Filenames to a Linker File ifile oe eee ceececeeeeeseeeeeeeeeeeeeeeceeecaeeeeeees 2 4 4 Add Your New Function Names to a Library Function List File cece eeeseeereeeeeeeeeeeee 2 4 Customizing the DSP Library by Deleting Functions 000 00 eee eeeeeeeeeeeeeeeeeee 2 5 5 Run the Build Dispatch Application to Generate an Assembly Dispatch File 2 6 6 Compile Assemble and Link Your Custom Library eeeeceseeesceeeeeeceseeeeeeneeeneenes 2 7 Creating our LabVIEW Interface sesent ri vesus reste vcess saves oc tus PEE E EE E EEEE T desu suede dl EE ETE E Ee 2 8 1 Bundle All of the Input Parameters to Arrays esesessssesssresrsrrersrrssrersreeresreresresesteererrerreresee 2 8 2
181. sed 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 Yoo u a out in r A error out x t yt S XA YO error in no error is a Fourier transform pair where the symbol DSP Deconrolution 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 ho FL v5 E 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 in Cxy is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input si
182. 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 DSPMEM 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
183. sinda values min value error in no error min index error out DSP Hax amp Min 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 f f Filtered Data Out nonlinear filter that combines lowpass Fitered Data in filters characteristics to remove high rank es error out frequency noise and high frequency characteristics to detect edges error in no error DSP Median Filter If Y represents the output sequence Filtered Data if J represents a subset of the input sequence X centered about the ith element of x Ji Xie Xin y Xii Xi Xiro 709 Kiera Xij and if the indexed elements outside the range of X equal zero the VI obtains the elements of Y using y MediandJ for 1 0 1 2 n 1
184. t DSP Sum DSP Log DSP Clip DSP Reverse DSP Shift DSP Sort DSP Linear Evaluation DSP Max amp 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 PM A TD OT TT STE oT TTS C FU AS TT TT To SB FT Se TE TD 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 DSP DMA Copy LV to DSP Handle to Address Analysis VI Organization After installation the NI DSP Analysis VIs reside in the following VI library files LabVIEW LLB files within the DSP2200 option e Signal Generation contains VIs that generate digital patterns e Frequency Domain contains VIs that perform frequency domain transformations frequency domain analysis and other transforms such as the Hartley transform e Time Domain contains VIs that perform direct time series analysis of signals e Filters contains VIs that perform IIR and FIR filtering functions e Windows contains VIs that perform smoothing windowing e Array contains VIs that perform arithmetic
185. t 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 National Instruments Corporation 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 Indicator Data Type Boolean TF ka 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 jE HEBERBE 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 stru
186. t 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 i lower stop frequency higher stop lower pass freq Filtered Data out frequency lower pass frequency iii error out higher pass frequency and of taps sampling frequency The VI then error anino riar filters the input sequence X to higher Aa free obtain the bandpass filtered linear sample freq fs phase sequence Filtered Data Filtered Data in higher pass treg DSP Equi Ripple BandPass 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 aro X is a DSP Handle Cluster that indicat
187. t data 0 16 bit long integer data 1 or 32 bit short integer data 2 PathnameTooLong The path name should be less than 256 characters InvalidCopyType The entire partial copy selector can only select between entire copy 0 and partial copy 1 CopyNumGTZero The number of data to copy should be greater than zero OffsetGEZero 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 National Instruments Corporation A 3 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 IndexSizeOffsetE The size offset should be less than or equal to the size of the DSP Handle Cluster that you index into InvalidDataType The data type can only be 4 bytes long 0 or 2 bytes long 1 NumBytesGTZero The number of bytes to allocate should be greater than zero NI DSP SRM for LabVIEW for Windows A 4 National Instruments Corporation 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 provid
188. t 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 32 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 32 Scalars out is an array of 32 bit long integer scalars that is identical to Array of 132 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 An
189. the LIB directory 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 LIB directory 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 fnc and NIESSEN fnc from the Dispatch directory to your working directory In Figure 2 2 the file NIDSP fnc shows the addition of the function DSP_GMaxMin to the library function list This is done by editing the file NIDSP fnc in 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 fne 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 Dispatch application 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 Dispatch application returns an error message if function names are not acceptable
190. 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 Unwraps the Phase array by eliminating discontinuities whose absolute values exceed 7 y Y in Y out error inno error Hrena error out DSF Unwrap Phase ae X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array ian 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 Plat 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 th
191. 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 2 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 Filtered Data in Parks McClellan algorithm and the i Filtered Data out number of taps high frequency stop of taps m error out frequency and sampling frequency The error in no error i VI then filters the input sequence X to obtain the highpass filtered linear phase stop freq sequence Filtered Data high freq sampling freq fs The stopband of the filter goes from zero ea 7 DC to the stop frequency The DSP Equi Ripple HighPass 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 bu
192. ting 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 i file 4 Add your new function names to a library function list file 5 Run the Dispatch application 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 Examples directory 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 c example file in your examples directory 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
193. to delete all functions of a group replace each function name of that group in the NIDSP fnc with 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 dispatch directory The Dispatch application automatically generates the WE DSP32C assembly code file dispatch s and a header file dspfnes h in your current directory The dspfncs h file 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_GMaxMin is 448 Remember this code You will use the code later to build your LabVIEW interface VI Figure 2 4 shows a section of dspfncs h corresponding to the section given from NIDSP fnc Notice that in this example Signals was the seventh group Group 6 listed in NIDSP fnc and hence the function IDs Signals ae define DSP_Sine_NUM 384 define DSP_Square_NUM 385 define DSP_Sinc_NUM 386
194. tput FHT X The largest FHT that can be computed depends upon the amount of memory on your DSP board oo FHT X is a DSP Handle Cluster that is identical to X but with the results of FHT X already stored in eerue 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
195. uch as w i 0 54 0 46 cos 27i 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 data type 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 Source i Destination out Destination in L error in no error error out Copy Mem DSP to DSP 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 mmn Source is a DSP Handle Cluster that indicates the memor
196. uding 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 Copyright 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 involvin
197. ufficient memory or disk space 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 The output array size must be greater than or equal to the input array size There is not enough space left on the DSP board for onboard processing The decimating factor must meet 0 lt decimating factor lt samples The width must meet 0 lt width lt samples The index must meet 0 lt index lt samples The delay must be greater than or equal to zero The width must be greater than or equal to zero The following condition must be met 0 lt delay width lt samples continues A 1 NI DSP SRM for LabVIEW for Windows Error Codes Error Number NI DSP SRM for LabVIEW for Windows Appendix A Table A 1 NI DSP Analysis Library Error Codes Continued dtGTZero DutyCycleE CyclesErr UpperGELower IntervalNumErr MixedSignErr SizeGTOrder OrderGTZero ConvSizeErr NyquistErr Nyquist2Err RippleGTZero AttenGTZero WidthGTZero SizeGTZero NullVectorErr AttenGTRipple StepSizeErr LeakErr nGEOne
198. ut array Y is obtained using a Y out the following formula ins error in no error error out Y i IX DSP Absolute for 1 0 1 2 n 1 where n is the number of elements in X ae X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array X ro Y in is a DSP Handle Cluster that indicates the memory buffer on the DSP board that will contain the absolute results of X ro 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 ithelement of the output array Z is obtained using the following E i 2 out formula Y m in error out Z i X i YG error in no error for i 0 1 2 n l DSP Add where n is the smaller number of elements in X and Y To X is a DSP Handle Cluster that indicates the memory buffer on the DSP board that contains the input signal array X To Y is a DSP Handle Cluster that indicates
199. ut sequences NI DSP SRM for LabVIEW for Windows 2 57 Part 3 NI DSP Function Reference NI DSP Analysis VI Reference Chapter 2 Band Parameters 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 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 _ LowPass_ HighPass_
200. utieantene cee taees a aiN 2 71 DSP Square Pae littl eae ia et ata hal eae Meee ee 2 72 DSP Square ROOt aN EAE E EEE AE S EE EE EEE EA E EEE cited pebbepzed eee 2 73 DSP SOD u ERE E EE EEEE E E E 2 74 DSP Start iapa E EE E asus EEE EEEE E EERE 2 75 DSP S bsetisii e a E EE E N E E E E A EEE eE EE E EEEE EE E ES 2 75 DSP Subtr t i snein eea E R A EEA O RA EEEN EAA OA NAE TEEDE EE EEA SSE 2 76 PSP OUM eea i E ine Minsk wees E OESE Mea betes Wee TE ASEE EET ESEESE EEE Abe ERES 2 76 DSP TimeOutrs r ra a nea eea ERE EAE E E E R EE AEE EA RE EES 2 77 DSP Triansle Patteri eneren eere ee oe eE E E saen aE i Bes epo eRe EEEE EEEE PEE Ea AT EEEE Spe eE S EEES pays 2 78 DSP Triangular TAA tive iG Biel A RA AR i ek I a eM Re ahead 2 80 DSP Transilar Windows ii ccces cis Si e a R a E E eet eat once A tua 2 81 DSP Uniform White Noise sc ccssccoNeceesces deii e E E E E a A E OE 2 82 DSP Unwrap sa E I SAAE AE E EEE E E EE NEEE E EE E E EEEE E EES 2 83 DSP Zero Padder nioen NE EE E A E E A R a etaa 2 84 Part 4 NI DSP Interface Utilities Chapter 1 Introduction to the NI DSP Interface Utilities 2000000000 sssseseeseceeeeeeeeneee 1 1 Overview of the NI DSP Interface UUHtES sissiciiresicscterners sser iesiri sercis seesi knee esee ekb ese ei esi 1 1 Installing the NI DSP Interface Utilities o oo ce cseececssensecseaecsecaecsecesesseeseseeeeaeseaseaeseaeesaeeaesaee 1 2 Using the NI DSP Interface Utilities is sccssc ssseess sidsshsscoessseas ssta
201. vironment 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 The LabVIEW LabVIEW for Windows for Windows Development Environment Interface Utilities User Application AT amp 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 DS
202. y 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 National Instruments Corporation xv 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 en
203. y 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 Ta 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 oma Destination out is a DSP Handle Cluster that is identical to the Destination in but with the source L 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 destination short entire partial copy selector from Source om destination float the Source buffer on the DSP size board that is referred to by a DSP offset Handle Cluster to the destination error in no error ra out buffer in LabVIEW destination type destination long To copy data correctly from the Copy Mem DSP to LY DSP board to LabVIEW you must indicate what
204. y of the EXAMPLES subdirectory of your LabVIEW directory 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
205. 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 NJ DSP Software Reference Manual for DOS LabWindows Note The functions listed in this chapter are the same functions described in the NJ DSP Software Reference Manual for DOS LabWindows except for the following differences e 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 e The integer parameters are all 32 bits long The corresponding parameter type in the NJ 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 e The functions listed are all of the functions that you can call from your custom program Not all the functions described in t

National Instruments Car Stereo System 320571-01 User's Manual

Contents

Download Pdf Manuals

Related Search

Related Contents