[SI-LIST] Re: Pk-Pk jitter
- From: olaney@xxxxxxxx
- To: hmurray@xxxxxxxxxxxxxxx
- Date: Sat, 22 Dec 2007 08:30:49 -0800
The gaussian assumption is often not a good one. For instance, take the
case of a control signal with crosstalk from a nearby sine wave. If high
enough, the jitter will begin to take on the bimodal distribution of the
sine wave. Statistical tools are quite useful for troubleshooting cases
like this. Also note that gaussian tails do not go on forever in circuit
applications. They can't, because once the signal excursion hits the
rails it is clamped by the nonlinearities of the protection networks.
Even for infinite rails, a reasonable observation period must be used,
lest the observer grow old and retire before the product is shipped. 100
seconds is probably equivalent to forever, depending on the bandwidth of
course. Various standards have specifications for this, whether explicit
or implicit (e.g. a number of bits at a given bit rate).
Since some data sheets quote only the lower and therefore more impressive
looking RMS value, a useful rule of thumb is that, absent contamination
such as mentioned above, a working number for 'random' P-P jitter is ~7X
RMS. Where both figures are shown on data sheets, the ratio is rarely
less than 6 or greater than 7. The energy in the 'lost' tails is
meaningless.
Orin
On Sat, 22 Dec 2007 00:43:11 -0800 Hal Murray <hmurray@xxxxxxxxxxxxxxx>
writes:
>
> > The vendor often provides PK to PK jitter parameter of PLL,
> but as
> > we know, a reference clock on board will be feed to PLL via ASIC
> > package and a input IO buffer, so we how to do simulation to get
> > reference clock jitter requirement. could you help to give
> simulation
> > methodology? thanks!
>
> What does Peak-to-peak mean in this context?
>
> If the distribution is Gausian (which is the usual assumption), then
> the
> tails go on forever. It's just a matter of how long you wait.
>
> Is there an implied cutoff at 6 sigma or something like that? I
> think that
> would make Pk-Pk a useful description.
>
> Is the typical distribution actually Gaussian? All sorts of things
> get more
> interesting if it isn't quite Gaussian.
>
>
> --
> These are my opinions, not necessarily my employer's. I hate spam.
>
>
>
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject
> field
>
> or to administer your membership from a web page, go to:
> //www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
>
> List technical documents are available at:
> http://www.si-list.net
>
> List archives are viewable at:
> //www.freelists.org/archives/si-list
> or at our remote archives:
> http://groups.yahoo.com/group/si-list/messages
> Old (prior to June 6, 2001) list archives are viewable at:
> http://www.qsl.net/wb6tpu
>
>
>
>
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
