[SI-LIST] Re: Ring back
- From: "Andrew Ingraham" <a.ingraham@xxxxxxxx>
- To: <si-list@xxxxxxxxxxxxx>
- Date: Tue, 23 May 2006 09:53:51 -0400
> In Intel FSB specs ...
...
> I first hit this several years ago doing slot 2 designs and it's always
> been counter-intuitive to me - I would prefer the terms "positive
> overshoot" and "negative overshoot" instead of "overshoot" and
> "undershoot". But, there was too much momentum behind the existing
> terminology to change it. So, I've adapted my terminology accordingly.
Indeed, Intel was the large semiconductor company I referred to earlier,
that consistently uses 'undershoot' incorrectly.
> To me, the bottom line is that these terms, like many in our discipline,
> are merely conventions, and ambiguous ones at that.
Not necessarily. The IEEE, at the least, has actual definitions for
overshoot and undershoot. Intel chooses to ignore IEEE's definitions and
use undershoot to describe overshoot. Which has helped a whole generation
of engineers to use those words incorrectly and perpetuate confusion and
misunderstanding.
> It's best to very
> carefully spell out exactly what we mean when using these terms (which
> this spec does).
Yes, but it would be better to also use them correctly.
I'm sure we all use the wrong terms sometimes. I don't carry an IEEE
dictionary and I know I am frequently wrong. I just get annoyed by Intel's
misuse of 'undershoot' because the error is so glaring, being so frequently
used and exactly opposite to its correct meaning.
(Climbing off my soapbox now...)
Back to the original question:
> How to solve ringback problem ...
You got a couple of concise suggestions already. I'll add that each case
can be a little different, some situations require different remedies.
Maybe you need to route your traces differently (daisy-chain vs. star), or
maybe altering impedances / terminations / drive strengths does the trick.
> ... and what is the disadvantages of ringback
Maybe nothing (if, as Steve says, you meet your specs), but it depends.
If it's a clock or has clock-like behavior, could be double-clocking.
If the ringback is wide, the input buffer may be sitting near threshold for
a long time, which can sometimes cause oscillations or other unpredictable
behavior.
Ringback, if it crosses thresholds, means the signal takes more time to
settle on a valid level. Which means your wire delays are longer than they
might have been without the ringback, and you need to account for that in
your timing analysis.
Ringback also implies there is overshoot, which is a hint that you need to
make sure the amount of overshoot you have is acceptable.
Regards,
Andy
------------------------------------------------------------------
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 FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
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:
