I am trying to find RMS jitter of the clock coming out of my ADC ( in post layout sim). I can plot eye diagram, but am not sure how to get the RMS value of jitter. Is there a way (may be using calculator or otherwise) to get the RMS value? I think the eye diagram gives the peak-peak value of jitter.
(posting again, as the pervious post was in wrong category)
I believe what you need to do is to compute the time interval error (TIE) as a function of zero-crossing and then determine its rms value. Please note, however, that "jitter" is defined over a specific bandwidth. Hence, the rms value you compute is limited by the sample size of the TIE you compute. For example, the minimum frequency of the time interval error is determined by the time span of the simulation you analyze and the maximum frequency is essentially 1/(2*your clock period). If you are trying to compute the rms jitter over a specific frequency band, you will need to set the simulation time and your analysis period accordingly. There is an calculator function called "abs_jitter" that will compute the time interval error on a periodic waveform. An example of its syntax for a clock waveform named hsclki_clipped with period 1/fin and switching threshold level vth is as follows:
hsclki_tie = abs_jitter( hsclki_clipped "rising" vth ?yUnit "s" ?Tnom 1.0/fin )
I hope this helps!
Unfortunately abs_jitter will only work with a periodic signal, not with a random bit stream. The Eye Diagram assistant (in IC615 onwards) gives the option of computing various metrics such as the standard deviation of the edge, but if you need more general calculations, you can use the SKILL code in solution 11395772.
I beleve34892 is looking to compute the time interval error of a periodic signal as he states: "I am trying to find RMS jitter of the clock coming out of my ADC". This is why I suggested he try the abs_jitter() function. Your added information, however, is quite useful!
So, I used the abs_jitter function to find the RMS jitter. I got a reasonable value, but I am wondering, what if I use period_jitter function in the calculator. I know there is a difference in so called 'period jitter' and 'TIE' jitter measurement, but I am not sure which one to use when and which result is more accurate.
FYI: I did use period_jitter function, and the RMS jitter value was 217fs (vs. that from abs_jitter was 249fs). They are close, but still different.
I haven't used the eyeDiagram method Andrew suggested, but will check that as well.