Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

  • State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 81 subscribers
  • Views 264 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Please check the FFT result using "load memory" function on 2837x_RFFT(example code)

Hyun Duk Suh
Prodigy 140 points

Hello, everyone.

I am testing RFFT library on my project.

I have finished that check the RFFT example from controlSUIT.

I also have tried changing input data such as square signal and sine signal, and That result was fine as same like matlab result.

But.. I couldn't check the FFT result, when I input the adc signal into RFFTin1Buffer from on my project.

So, I want to request you one things.

I attached one input data file(test_in.dat) on this web page.

when I tried to build on my project, I failed.. all output data were zero.

so please build example project("2837x_RFFT", C:\ti\workspace_6_1\2837x_RFFT) using load memory and attached input file.

This input file length is 256 samples.

Could you check the FFToutBuffer using attached input data on your example project?

Becuase I also tried but failed to check the FFToutBuff after loading "test_in.dat" file.

I really really want to verify the result of FFT!!!!

and I have one more question.

According to library manual(RFFT, FPU-SW-LIB-UG), the output data were saved in FFToutBuff  such as (real[0]~real[N/2], imag[N-1]~imag[1])

Is there an another type of library which saved the output data included all real & imag data??? (not half of real & imag data)

Or is it better to use CFFT function?

If you know the good methods to save the all real and imag value of RFFT data, please let me know.

Thanks.

test_in.dat

  • 0
    TI__Mastermind 20850 points

    Hi hyun,

    All your input data is of the order 10^-42. This is not even representable in single precision floating point - its zero as far as any C2000(+FPU) device is concerned. MATLAB uses double precision math so these numbers do fall in the normal range, and it is able to run an FFT on it. Your input data is way too small to do any single precision math on it.