Need Help - CCStudio TMS32064xx- Goertzel project

Discussion in 'Homework Help' started by nick_533, Mar 28, 2006.

  1. nick_533

    Thread Starter New Member

    Mar 28, 2006
    1
    0
    Hello Guys any help frm u is really appreciated

    i got struck up with 2 problems .I use Code Compsoser Studio.

    1) I have to design goertzel algorithm on tMS32064xx and build a floatiing point and fixed point algorithm and also working assembly algorithm...finally i hav to build a threshold detector...

    2) I have to design a fir filter 2nd order in matlab and hav to take coefficients frm it and using them I have to write a main program and build a fixed and floatiing point and assembly algortihm for it.


    Please need help urgently.
     
  2. Papabravo

    Expert

    Feb 24, 2006
    9,912
    1,724
    Tell me about Code Composer Studio. Does it use a high level language, or assembly language? Can it support a mix of the two programming styles?

    On to your problems.
    1. You did not say if you know and understand what the goertzel algorithm is. If you do, that's fine. If not you should do a Google search and read up on the algorithm.

    2. Do you understand the difference between fixed point and floating point algorithms? You should write the floating point algorithm first so you can check your understanding of the goertzel algorithm without having to worry about messy details like overflow. Writing a fixed point algorithm involves the same steps but you have to manage both undeflow and overflow. If you don't know what these terms mean it's back to Google. I don't know where "assembly algorithm" fits into this context, but the description of Code Composer should resolve this confusion.

    3. A threshold detector is just a comparison of a signal to a constant. The single bit output is one if the signal is greater than the constant, and zero if the signal is less than or equal to the constant. You could also invert the single bit output. It's a digital version of an analog comparator.

    4. An FIR filter is one without feedback so it produces a weighted sum of samples of the signal. There needs to be some criteria that the filter needs to satisfy in order to have some idea of how to approach the problem.

    5. The same comments from item 2. above apply here, and again I'm not sure if we are talking about writing the algorithm in some high level language first and then writing it in assembly language. If that's what you have to do then that's the way I would approach the problem.

    In short:
    Write floating point algorithms first, then write fixed point algorithms.
    Write High level language first, then write assembly language algorithms.

    Hope this helps
     
Loading...