Quartus FFT IP problem

[patrick.melet]

Hi everybody,

I'm trying to use the FFT IP core from Quartus 6.0...

I would like to know what are the representation of the mantissa (2's
complement or other ...) and where is the sign bit ?

My problem is : when I send a 1 MHz sine with 10 MHz sampling
frequency to the real part input of the FFT, I obtain peaks at 1 MHz
at the output and at 9 MHz, that's good but my peaks (if I represent
the mantissa in 2's complement) have one part positive and one part
negative !!! not a pure negative or positive peak... whats' happen...

Sure If I take FFT_real ^ 2 + FFT_imag ^2 then my peaks are all
positives...

Thanks