Motorola DSP Developer's Kit

ifft-r.asm

Motorola DSP Developer's Kit

ifft-r.asm

Y = mot###_ifft( X )

This function returns the discrete Fourier transform (DFT) of the input for real vector X

Input/Output

Input: Vector Xe (even index input data, located in X memory), and vector Xo (odd index input data, located in Y memory)

Output: Vector Yr (real data of output), and vector Yi (imaginary data of output)

Algorithm

First, use the algorithm in FFT-C.ASM to calculate the length/2 complex data FFT.

Then, use the split algorithm to calculate the final results.

No scaling is required for the input data; the output data should be scaled up by 2^(r2+1).

For example, after the FFT is done and r2=7, to get true FFT values, every output item has to be shifted left 8 bits.

The scaling algorithm is described below:

Note    The variable "size" used below is half the size of the input of the real vector X

for (k =0; k < size/2 -1; k++)
{
   H1r = (Xr[k] + Xr[size-k]) / 2;
   H1i = (Xi[k] - Xi[size-k])  / 2;
   H2r = (Xi[k] + Xi[size-k]) / 2;
   H2i = (Xr[size-k] - Xr[k]) / 2;
   Yr[k] = H1r + (C2r[k]*H2r - C2i[k]*H2i);
   Yr[size-k] = H1r - (C2r[k]*H2r - C2r[k]*H2i);
   Yi[k] = H1i + (C2i[k]*H2r- C2r[k]*H2i);
   Yi[size-k] = -(H1i) + (C2i[k]*H2r - C2r[k]*H2i);
}

Memory & Register

Memory allocations:

• X memory: IN_REAL stores the location of the real part of the input data
• Y memory: IN_IMAG stores the location of the imaginary part of the input

The above two symbols are also used as output symbols.

p:$F000 = half the length of the input real vector

p:$F001 = the address offset of the coefficient lookup table, C1 (used by Fft-c)

p:$F002 = the address offset of the coefficient lookup table, C2 (used by split)

p:$F003 = the address offset of the output vector

Register usage:

• R7 is used for passnum
• R2 is used for the scaling exponent

Status Register

The assembly function ifft-r.asm does not explicitly set any status registers/bits during the function execution.

Data Size Limit

The length of vector X can't be larger than the continuous available data memory size.

Data Range Limit

The value of input vector X must be between -1.0 and +1.0.

filter-c.asm

ifft-c.asm