I've been working with the Anatoliy Sergienko, Volodya Lepeha unicore 256 pt. FFT core with some limited success. The simulation runs OK with a 700 Hz sinusoid input taken from ADC data; there is a nice output produced.

When I implement it in a Spartan 3E, the output is skewed up to about 900 Hz, and the output has almost an "inverted" profile. It almost seems as though the sign bit getting lost in the implementation. There are a host of warnings from ISE when implementing.

Has anyone else successfully implemented this core, or tried?

Please respond here and/or to tim@techsystemsembedded.com.

Thanks