Prev: contour area
Next: dspsigops
From: juho salminen on 1 Nov 2009 08:33
Hi,

is there any way to take angle information of noisy sine wave? data vectors are hundreds/thousands periods long. I have tried FFT, with bad and unaccurate results,

Thanks already for answers,

-Juho
From: Greg Heath on 1 Nov 2009 19:48
On Nov 1, 9:33 am, "juho salminen" <jssal...(a)cc.hut.fi> wrote:
> Hi,
>
> is there any way to take angle information of noisy sine wave? data vectors are hundreds/thousands periods long. I have tried FFT, with bad and unaccurate results,
>
> Thanks already for answers,

If, in spite of very long periods, the FFT is unreliable,
perhaps a noise filter (lowpass or bandpass) could
be used before the FFT.

If the the filter cutoff frequencies are not known apriori,
they could be estimated using a pre-filter FFT to estimate
the sine wave frequency.

Hope this helps.

Greg
From: dpb on 1 Nov 2009 20:54
juho salminen wrote:
> Hi,
>
> is there any way to take angle information of noisy sine wave? data
vectors are hundreds/thousands periods long. I have tried FFT, with bad
and unaccurate results,
....

averaging multiple samples should help cancel random noise (assuming it
is random, of course)...

Bendat & Piersoll, Random Data: Analysis and...Procedures, Wiley

is one classic engineering reference
--
 | 
Pages: 1
Prev: contour area
Next: dspsigops