Hello All:
I am designing cic filter and compensation filter. Fs is the signal sample rate
and R is the down sample rate with cic filter.I search most of articles about
cic+compensation filter design. And most of Amplitude to Frequence feature for
cic+compensation filter is shown in the Fs frequency but not Fs/R.
In my understanding of cic+compensation filter, the Amplitude to Frequence
feature should be shown in Fs/R frequency. Because after cic filter, sample rate
has been down to Fs/R and compensation filter work at Fs/R frequency. So I think
Amplitude to Frequence feature for cic+compensation filter should be in Fs/R
frequency. Am I right?
I really hope to listen to your idear and appreciate your reply.
jiangli
cic filter and cic compensation filter
Started by ●April 22, 2010
Reply by ●April 25, 20102010-04-25
>I am designing cic filter and compensation filter. Fs
is the signal sample rate and R is the down sample rate with cic filter.I search
most of articles about cic+compensation filter design. And most of Amplitude to
Frequence feature for cic+compensation filter is shown in the Fs frequency but
not Fs/R.
>In my understanding of cic+compensation filter, the Amplitude to Frequence feature should be shown in Fs/R frequency. Because after cic filter, sample rate has been down to Fs/R and compensation filter work at Fs/R frequency. So I think Amplitude to Frequence feature for cic+compensation filter should be in Fs/R frequency. Am I right?
>I really hope to listen to your idear and appreciate your reply.
>
Another way to look at it is from the normalized sample rate. This way we don't need to be confused about the sample rate. You can take the results from any paper/book and normalize.
It is valid to look at the response from either Fs or Fs/R perspective in absolute terms. You should notice that even though it is from the original Fs perspective there should be nothing of interest above the new Nyquist, Fs/(2*R). But most will feel more comfortable from the Fs/R perspective which you should easily be able to convert the axis units.
.chris
>In my understanding of cic+compensation filter, the Amplitude to Frequence feature should be shown in Fs/R frequency. Because after cic filter, sample rate has been down to Fs/R and compensation filter work at Fs/R frequency. So I think Amplitude to Frequence feature for cic+compensation filter should be in Fs/R frequency. Am I right?
>I really hope to listen to your idear and appreciate your reply.
>
Another way to look at it is from the normalized sample rate. This way we don't need to be confused about the sample rate. You can take the results from any paper/book and normalize.
It is valid to look at the response from either Fs or Fs/R perspective in absolute terms. You should notice that even though it is from the original Fs perspective there should be nothing of interest above the new Nyquist, Fs/(2*R). But most will feel more comfortable from the Fs/R perspective which you should easily be able to convert the axis units.
.chris
