Re: [SI-LIST] : HSTL voltage levels

D. C. Sessions (dc.sessions@vlsi.com)
Wed, 04 Aug 1999 14:15:37 -0700

Gerald Johnson wrote:
>
> D.C.
>
> What you say is certainly true if one only worries about DC
> noise margin. However, if the offending noise were to be so
> impolite as to happen during a series of high speed transitions
> such as real data, then it would indeed subtract from the
> AC levels wouldn't it? Therefore the AC noise margin could be
> smaller than the DC noise margin. Seems like both are important.
> Different, but both important.

I don't think so. The whole point of the two sets of thresholds
is to facilitate this kind of distinction. The names are
admittedly somewhat confusing.

In any case, the idea is that IF you swing to or beyond the "AC
threshold" then you are sure to have provided enough input energy
to guarantee a timely transition (remember, the speed of the
response is at least partly a function of the magnitude of the
stimulus.) The other side of that coin is that no matter HOW
long you sit between the AC and DC thresholds the input is
guaranteed not to switch.

Think of the "DC threshold" as being related to a combination of
the offset voltage and gain of a comparator, and the "AC threshold"
as being the so-called "input overdrive" required to achieve spec
timing. Which is exactly what they are....

Regardless of the environment, the noise margin is the difference
between the noise-free state of the signal (Volmax or whatever) and
the nearest value that MIGHT be mistakenly recognized as a different
state. In the case of HSTL, that "nearest state" is the so-called
"DC threshold."

> >
> > Watch out -- noise margin to the AC numbers is irrelevant. The AC
> thresholds
> > are guaranteed to cause a change of state, but for noise margin you only
> care
> > about the DC thresholds; they are the limits of what is guaranteed to NOT
> > cause a change of state.
> >
> > --
> > D. C. Sessions
> > dc.sessions@vlsi.com
>
> **** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****

-- 
D. C. Sessions
dc.sessions@vlsi.com

**** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****