This document is a technical note on using Analog Devices' DSP components and development tools. It provides information on using the 21xx C-FFT library for fast Fourier transforms (FFT).
This document is an Engineer-to-Engineer Note (EE-90) providing technical guidance on using the Analog Devices 21xx C-FFT library. It explains the theory of programming FFTs, specifically focusing on the correct placement of data buffers to ensure bit-reversal works properly on ADSP-21xx processors. The note details how to define circular buffers using assembler macros or inline assembler since the standard C-compiler does not handle this automatically.
| Supported array sizes | 8 to 1024 points |
|---|---|
| Function prototype | fftN(r_inp,i_inp,r_outp,I_outp) |
| Twiddle table precision | 16bit integer |
Add the directive #include < ffts.h> to the main file header.
The linker utility ld21 must be advised to use the qualifier CIRC.