Forums

Fourier Transform and Sampling rate jitter

Started by tonydeng July 5, 2007
"Fourier Transform" transform the time domain data into frequency domain
information on the premise that the sampling rate is constant or in some
predictable manner. If the sampling clock jitter, it introduces noises
into the frequency domain information.

I would like to know, is there is formula whereby I can calculate the
amount of noise introduced, based on how much the clock jitter ?

On Jul 5, 8:56 pm, "tonydeng" <deng301...@hotmail.com> wrote:
> "Fourier Transform" transform the time domain data into frequency domain > information on the premise that the sampling rate is constant or in some > predictable manner. If the sampling clock jitter, it introduces noises > into the frequency domain information. > > I would like to know, is there is formula whereby I can calculate the > amount of noise introduced, based on how much the clock jitter ?
Sure, there are many *approximation* methods for this problem. Look up "Leeson's formula" for one of them. Julius
On Jul 5, 6:56 pm, "tonydeng" <deng301...@hotmail.com> wrote:
> "Fourier Transform" transform the time domain data into frequency domain > information on the premise that the sampling rate is constant or in some > predictable manner. If the sampling clock jitter, it introduces noises > into the frequency domain information. > > I would like to know, is there is formula whereby I can calculate the > amount of noise introduced, based on how much the clock jitter ?
Try Analog Devices application notes AN-501 and AN-756 Dale B. Dalrymple http://dbdimages.com
On Thu, 05 Jul 2007 20:56:20 -0500, "tonydeng"
<deng301056@hotmail.com> wrote:

>"Fourier Transform" transform the time domain data into frequency domain >information on the premise that the sampling rate is constant or in some >predictable manner. If the sampling clock jitter, it introduces noises >into the frequency domain information. > >I would like to know, is there is formula whereby I can calculate the >amount of noise introduced, based on how much the clock jitter ?
Hey tonydeng, It would help you to say "Hello" at the beginning of your post. That is just good manners when asking experts to help you. Having respect for your DSP-superiors improves your chances of receiving help. It would also help if your sentences were grammitically correct (in English). I don't say that out of some sort of racism, but merely because if you're going to ask a question in some language (English in this case) it is easier for your readers to understand your question if it is worded correctly. Further, it would be helpful to you to tell us what you have done to solve your problem. I say that because, why should we help you if you have done nothing to help yourself? tonydeng, what have YOU done, so far, to solve YOUR problem? Perhaps someone here is willing to give YOU advice regarding the method YOU used to solve YOUR problem. And finally, you might receive more help if you end your post with some sort of a "Thank You" phrase letting the people here know that you appreciate their help. [-Rick-] Cranky ol' fart.
Rick Lyons wrote:
> On Thu, 05 Jul 2007 20:56:20 -0500, "tonydeng" > <deng301056@hotmail.com> wrote: > > >"Fourier Transform" transform the time domain data into frequency domain > >information on the premise that the sampling rate is constant or in some > >predictable manner. If the sampling clock jitter, it introduces noises > >into the frequency domain information. > > > >I would like to know, is there is formula whereby I can calculate the > >amount of noise introduced, based on how much the clock jitter ? > > Hey tonydeng, > > It would help you to say "Hello" at the > beginning of your post. That is just good > manners when asking experts to help you. > Having respect for your DSP-superiors > improves your chances of receiving help. > > It would also help if your sentences were > grammitically correct (in English). I don't > say that out of some sort of racism, but merely > because if you're going to ask a question in > some language (English in this case) it is > easier for your readers to understand your > question if it is worded correctly. > > Further, it would be helpful to you to tell us > what you have done to solve your problem. > I say that because, why should we help you > if you have done nothing to help yourself? > > tonydeng, what have YOU done, so far, to > solve YOUR problem? Perhaps someone here > is willing to give YOU advice regarding > the method YOU used to solve YOUR problem. > > And finally, you might receive more help if > you end your post with some sort of a > "Thank You" phrase letting the people here > know that you appreciate their help. > > [-Rick-] > Cranky ol' fart.
Hi Rick, I agree that showing some manners is good when posting questions and answers. However, using one of my postings as an example http://groups.google.com/group/comp.dsp/browse_thread/thread/4ff7c4cd15fc1130?hl=en it does not always get one all the help needed to clarify one's understanding of the topic. BR, Phil Cranky old fart in the making... at least if you ask my wife.
On 11 Jul., 16:07, Phil <philguilleme...@alumni.uwaterloo.ca> wrote:
> Rick Lyons wrote: > > On Thu, 05 Jul 2007 20:56:20 -0500, "tonydeng" > > <deng301...@hotmail.com> wrote: > > > >"Fourier Transform" transform the time domain data into frequency domain > > >information on the premise that the sampling rate is constant or in some > > >predictable manner. If the sampling clock jitter, it introduces noises > > >into the frequency domain information. > > > >I would like to know, is there is formula whereby I can calculate the > > >amount of noise introduced, based on how much the clock jitter ? > > > Hey tonydeng, > > > It would help you to say "Hello" at the > > beginning of your post. That is just good > > manners when asking experts to help you. > > Having respect for your DSP-superiors > > improves your chances of receiving help. > > > It would also help if your sentences were > > grammitically correct (in English). I don't > > say that out of some sort of racism, but merely > > because if you're going to ask a question in > > some language (English in this case) it is > > easier for your readers to understand your > > question if it is worded correctly. > > > Further, it would be helpful to you to tell us > > what you have done to solve your problem. > > I say that because, why should we help you > > if you have done nothing to help yourself? > > > tonydeng, what have YOU done, so far, to > > solve YOUR problem? Perhaps someone here > > is willing to give YOU advice regarding > > the method YOU used to solve YOUR problem. > > > And finally, you might receive more help if > > you end your post with some sort of a > > "Thank You" phrase letting the people here > > know that you appreciate their help. > > > [-Rick-] > > Cranky ol' fart. > > Hi Rick, > > I agree that showing some manners is good when posting questions and > answers. However, using one of my postings as an example > > http://groups.google.com/group/comp.dsp/browse_thread/thread/4ff7c4cd... > > it does not always get one all the help needed to clarify one's > understanding of the topic.
The dynamics of usenet is sometimes unpredictable. Posts obeying full netiquette get ignored, and vice versa. In the end, the responses you get depend largely on whether you strike a nerve in the group of people who are willing to spend time replying on usenet. Courtesy and netiquette are only half the rent (as we say in Switzerland :-). A question has to be interesting enough to be debated but not too hard to exceed the working knowledge of the group members and the possibilites of the usenet medium.
> > BR, > Phil > Cranky old fart in the making... at least if you ask my wife.
There seem to be a couple of those around here :-). Regards, Andor
On Wed, 11 Jul 2007 07:07:15 -0700, Phil
<philguillemette@alumni.uwaterloo.ca> wrote:

> >Rick Lyons wrote: >> On Thu, 05 Jul 2007 20:56:20 -0500, "tonydeng" >> <deng301056@hotmail.com> wrote: >> >> >"Fourier Transform" transform the time domain data into frequency domain >> >information on the premise that the sampling rate is constant or in some >> >predictable manner. If the sampling clock jitter, it introduces noises >> >into the frequency domain information. >> > >> >I would like to know, is there is formula whereby I can calculate the >> >amount of noise introduced, based on how much the clock jitter ? >> >> Hey tonydeng, >> >> It would help you to say "Hello" at the >> beginning of your post. That is just good >> manners when asking experts to help you. >> Having respect for your DSP-superiors >> improves your chances of receiving help. >> >> It would also help if your sentences were >> grammitically correct (in English). I don't >> say that out of some sort of racism, but merely >> because if you're going to ask a question in >> some language (English in this case) it is >> easier for your readers to understand your >> question if it is worded correctly. >> >> Further, it would be helpful to you to tell us >> what you have done to solve your problem. >> I say that because, why should we help you >> if you have done nothing to help yourself? >> >> tonydeng, what have YOU done, so far, to >> solve YOUR problem? Perhaps someone here >> is willing to give YOU advice regarding >> the method YOU used to solve YOUR problem. >> >> And finally, you might receive more help if >> you end your post with some sort of a >> "Thank You" phrase letting the people here >> know that you appreciate their help. >> >> [-Rick-] >> Cranky ol' fart. > >Hi Rick, > >I agree that showing some manners is good when posting questions and >answers. However, using one of my postings as an example > >http://groups.google.com/group/comp.dsp/browse_thread/thread/4ff7c4cd15fc1130?hl=en > >it does not always get one all the help needed to clarify one's >understanding of the topic. > >BR, >Phil >Cranky old fart in the making... at least if you ask my wife.
Hi Phil, Welcome To The Club. Yes, it is sometimes mysterious why some posts receive replies and other posts do not. I certainly am not sure why this is so. My only point was that "clarity and good manners" improve a poster's chances of success. See Ya', [-Rick-]