I want to use MATLAB to simulate the FFT function designed by Altera. I run the Matlab testbench generated by Quartus, but there is always the error
Undefined function or method 'Sfftmodel' for input arguments of type 'double'. Error in ==> fft256_model at 41 [roc,ioc,eoc] = Sfftmodel(rin,iin,N,THROUGHPUT,ARCH,DATA_PREC,TWIDDLE_PREC,INVERSE); Dose anyone could help me solve this problem? ThanksLink Copied
I have solved the problem. The Quartus only creates the 32bit function, so we must use the 32bit matlab.
you might file an enhancement request with Altera on that. DSP Builder finally supports 64-bit MATLAB, it would be nice if the MegaCore simulation models did too
--- Quote Start --- you might file an enhancement request with Altera on that. DSP Builder finally supports 64-bit MATLAB, it would be nice if the MegaCore simulation models did too --- Quote End --- March 5th 2013 and the problem is still unsolved?!?!?!
--- Quote Start --- I want to use MATLAB to simulate the FFT function designed by Altera. I run the Matlab testbench generated by Quartus, but there is always the error Undefined function or method 'Sfftmodel' for input arguments of type 'double'. Error in ==> fft256_model at 41 [roc,ioc,eoc] = Sfftmodel(rin,iin,N,THROUGHPUT,ARCH,DATA_PREC,TWIDDLE_PREC,INVERSE); Dose anyone could help me solve this problem? Thanks --- Quote End --- I got the sfftmodel.mexw32 file from altera's website, and simply changed the extention to .dll. My computer is of 32 bits, but I keep getting the error message: tsfftmodel.dll can't be loaded correctly. Could somebody tell me how to fix the problem? Besides, I simply put this sfftmodel.dll file in the same folder along with all the other files such as fft_32k_tb.m and fft_small_tb.m. Also, is there a way to change this sfftmodel.mexw32 (or sfftmodel.dll) into sfftmodel.m file? Thanks in advance.
For more complete information about compiler optimizations, see our Optimization Notice.