Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
It seems to me that there is not other way (than oversampling) in order to simulate jitering in time domain. As soon as you use uniform sampling time grid to represent the jiter you should have more dense (oversampling) time grid.
In frequency domain the jiter is represents as smearing of the energy around main tones (but most probably you know this very well)
So it seems to me that oversampling is the only way. Note that to get precise jiter simulation you need big oversampling ratio and long simulation.
>MAy I use fractional delay line filters with ramdom delay on a sample by sample basis?
Can you describe your idea more deeply? If you think that it will be not interesting for the rest you can send me private message. Sorry for my slow responce... recent days I am checking my e-mails only 2 times a week.
I found an interesting topic which is talking sampling time uncertainty...
Basically, from the figure 3-2, you will see what its relationship inbetween the sampling time uncertainty and ENOB.
This ebook also available on edaboard, if you are interesting, have your time to look at...
Hi Satellite,
Have you got an answer?
What is the signal that you want to sample. You are interested only in simulation ?
You have to build an elaborate model if you want to use some non-white process.
Say you ahve to sample an AR(1) process. This is simple but will illustrate the point. Generate time as I had written earlier model. Plug it in the AR equation. But remember that the driving noise becomes more uncorrelated as the time interval increases.
my problem consists in modelling a generic ADC subject to aperture time uncertainty.
This model should be independent on the knowledge of the input signal that at the moment is supposed to be a FDM stream.
Currently I am thinking to use a variable FIR filter which at each instant represents a different group delay characteristic (parametric FIR or time dependent FIR).
The jitter statistics should define instant by instant the required group delay the FIR should exhibit.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.