Re: Eye Patterns

Ray Anderson (raymonda@radium)
Mon, 22 Jan 1996 16:16:18 +0800

>
> If I understand eye patterns correctly, a key component
> is jitter. Wouldn't you need a REALLY good statistical
> model for the IC, including the clock distribution, as
> well as a simulator that efficiently handles Monte Carlo
> analysis? The only IC vendor who handles Monte Carlo
> analysis on a regular basis that I've seen is IBM, and
> they have their own simulation language (ASTAP).
>
> I'd be willing to bet not many people simulate eye patterns;
> they probably try to get as close as they can and then
> build some prototype hardware...
>
> --
> Greg Edlund o---/\/\/\---+-----+---o

I am able to run eye-pattern simulations on Contec Spice. While it
is true that jitter is a key component of an eye pattern as mentioned above
by Greg, other factors such as inter-symbol interference due to group delay
through band limited structures, unequal rise and fall times, frequency
dependent losses (which can induce frequency dependent phase delay) and many
other related factors can cause signal degradation on signals. Cross talk,
impedance mismatch and differential skew also can and will cause measurable
distortion of the eye pattern. These degradations are a function of the
transmission media (i.e. interconnects) and are a separate issue from the eye
pattern degradation induced by source jitter.

Ray Anderson
raymonda@radium.eng.sun.com