AnsweredAssumed Answered

FFT accelerator question

Question asked by ClaireCanada on Mar 26, 2012
Latest reply on Dec 8, 2013 by dougcl

Hi

 

I am doing a class project for my supervisor Dr. Smith to show the use of SHARC FFT accelerators. I have down loaded the plug-in to generate the PLL init files and the FFT init files

 

Two questions

Init_PLL_ddr_ami file contains the include file call

#include<def21469.h>

#include<Cdef21469.h>

#include<cs_macros.h>

 

where do I get cs_macros.h -- I am using Visual DSP version  5.0.10.0 and that does not pick up the include file from the system includes

 

When you use the FFT code generation tool, there is a global setting for all channels  -- data format floating

 

However when I set the format to floating -- the code generated is for   extern int buffers and not extern float buffers

 

How can I ensure that the FFT is being performed in floats?

 

Going back to the 21469 example code 256fft

I would have expected that the output file would be complex -- is it stored as an array of real and then array of imaginary

or real and imaginary alternately -- I have used plot to try and recognize the format -- (stride 1 or 2) but the results plotted don't seem to make sense -- I am obviously not getting magnitude as there are positive and negative values.

 

It would really make my day

if you could direct me to an FFT example -- completely working in C rather than assembly code

 

Thanks

Claire

Outcomes