Two problems about FFT V9.1 IP core.

[Altera_Forum]

Hello , all. 

I am using the FFT V9 IP core in my desing . After some try , It seems to work. But there are two problems. The first one is about the scale of block floating point number . I use the IP core with 2048 points and streaming I/O data flow. The mannual shows that the maximum exponent output is -16 and the minimum exponent output is 0. If I use a test signal like 0.99*sin(2*pi*100/2048 *n)( I appoint the input data to the FFT core as Q15 format ), the exponet output is 11, Since the singnal's amplitude is the maximum the single-tone can get . I am considering what signal I can feed into the IP core in order to I can get the maximum exponent output. After some thinking . I think constant signal whose amplitude close to 1 will get the maximum exponent output . I tried this , But I find now the exponent output is 12. I am confusing. Is the maximum exponent output too reserved or there are some signals which can make exponent output get the maximum? By the way , I just process real signal . I think if I feed a complex signal into the IP core , the exponent output will increament one. Can you give me some advice how to get the maximum exponent output? Thanks.  The second problem is about the dynamic range . The datasheet shows the block floating point can improve dynamic range. But I am confusing about how to define the dynamic range of the FFT core and how to estimate the dynamic range of the core. Can you give me some advice or reference? Thanks. Best Regards. Agump

[Altera_Forum]

Don't have time to think about it for now, but I did have a post that talks about how the block floating point works... Perhaps you can check that out and see if it helps?  

 http://www.alteraforum.com/forum/showthread.php?t=20&highlight=block+floating+point