Question about the FFT

Hi all!

A have a little probleme with FFT.

I use the GRC without USRP. I would like to messeaure a triangle
signal, but i never becam the right effect. I attached the picture
about the measuring.

The problem are the followed:

I don’t why but, i have allways a peak at 0Hz. - Independent from the
type of signal
On the picture i messaure a triangle, but the next peak in the FFT
failed at 12kHz.

Attached picture:
http://img8.imageshack.us/img8/5620/messung5.png

Have you any Idea?

Thank: Gergely

Hi,

in your picture is a peak at -12KHz.

Is it possible, that you have to calculate the absolut value from the
FFT?
On a normal calculator it’s the function abs().

I don’t why but, i have allways a peak at 0Hz. - Independent from the


Discuss-gnuradio mailing list
[email protected]
Discuss-gnuradio Info Page


Neu: GMX DSL bis 50.000 kBit/s und 200,- Euro Startguthaben!

Kóczián Gergely wrote:

I don’t why but, i have allways a peak at 0Hz. - Independent from the
type of signal

The imaginary part of your signal has a dc offset.

On the picture i messaure a triangle, but the next peak in the FFT
failed at 12kHz.

Try making it a real signal (float outputs, not complex outputs).

The second peak have to be:

                                       Frequency: 3kHz
                                       Amplitudo: -45dB

The second peak is 20*math.log10(9) dB down from the first one. Thats
exactly what you should expect:

The amplitudo of the first peak also not correct.

You should take a look at logpwrfft.py

-Josh

Hello!

Thanks for the help. Now the frequency is correct. But the amplitudo
is wrong. I messauring in the real life a triangle signal and it’s not
similar.

Attached picture:

http://img7.imageshack.us/img7/8049/messung9.png

The second peak have to be:

                                       Frequency: 3kHz
                                       Amplitudo: -45dB

The amplitudo of the first peak also not correct.

Best regreads

2009/10/22 Josh B. [email protected]:

about the measuring.


Discuss-gnuradio mailing list
[email protected]
Discuss-gnuradio Info Page