• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 78 subscribers
  • Views 601 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.

MSP430FR5994: FFT with LEA enabled giving different result from LEA disabled

Kiwan Maeng
Prodigy 110 points
Part Number: MSP430FR5994

I am currently testing the FFT code from the DSPLib example, transform_ex1_fft_fixed_q15.c

My problem is that LEA enabled result (calculated by LEA) is notably different from the LEA disabled one (where the software library kicks in)

I calculated the magnitude of the FFT result by Re(x)^2 + Im(x)^2 and plotted the result

This is a result from LEA enabled. I think it makes sense because It should have one peak at 200 and one at 2100.

Since FS=8192, I assume the result shows from 0 ~ omega/2 (4096), and the middle peak is 2100 (though not sure why the peak is fluctuating..)

Anyway, my concern is this plot. When I disable LEA, the calculation uses the SW library. and the result is notably different, showing high peak at the right. (around 4096Hz?)

Is this some fundamental limitation the software implementation has? Or am I doing something wrong?

Since I don't know what should be the result of the example, I am not even sure if the first plot is correct,

and even more lost about the second result.

Anyway all I did was flip the flag to not use lea...

Thank you!

  • 0
    TI__Expert 5490 points
    Hello Kiwan,

    The plots were not uploaded correctly and get crashed.
    Could you upload them again?

    Also can you check the sampling frequency is set the same when using the SW implementation?

    Thanks,
    Yiding
  • 0 in reply to Yiding Luo
    Prodigy 110 points

    Will this work??

    For your question, everything is the same. Only one flag that enables or disables lea (MSP_USE_LEA) is being changed.

    Thank you.

  • 0 in reply to Kiwan Maeng
    TI__Expert 5490 points
    Hello Kiwan,

    The result of enabling LEA comparing to disabling LEA should be the same.
    Do you mind share with us your code and the the data so we can see how you use the example code?

    Thanks,
    Yiding
  • 0 in reply to Yiding Luo
    Prodigy 110 points

    Hi, thanks for the reply. From your comment I got a hint that the code itself might not be a problem, and was able to find a solution.

    I was not including <math.h> in my code, and I am not sure exactly what happened (my thought is that my mspgcc compiler added its math lib instead of TI math lib),

    but it anyway was the problem.

    Including <math.h> explicitly solved the problem.

    Thank you!

**Attention** This is a public forum