[SI-LIST] : IBIS Accuracy Sub-Committee Minutes

Greg Edlund (Greg.Edlund@digital.com)
Fri, 6 Mar 1998 12:44:05 -0500

IBIS Accuracy Sub-Committee Minutes

Thursday, March 26, 1998, held at Digital Equipment Corp. in Maynard, MA

Attendees:

Greg Edlund, Digital Equipment (chair)
Fawn Engelmann, EMC
Bob Haller, Digital Equipment
Bruce Heilbrunn, Stratus
Peter LaFlamme, Fairchild
Harvey Stiegler, Texas Instruments
Mike Ward, Fairchild

We welcomed three new members to the IBIS Accuracy Sub-Committee at our
second meeting. Harvey Stiegler, who works for TI's memory division,
flew up from Houston to join us. Peter LaFlamme and Mike Ward, both
from Fairchild, drove down from Portland, Maine. We now have a
well-rounded group with representatives from the semiconductor, EDA, and
system industries.

Agenda

The agenda for the meeting was to draft a BIRD (Buffer Issue Resolution
Document) proposing an IBIS Accuracy Specification. There was much
discussion, but after all was said nothing was done. It has been
suggested that we not attempt accomplish any writing during these
meetings but rather assign parts of the specification to each member as
homework and focus on discussion and decision making during the
meetings. Greg Edlund will be drafting the BIRD and sending it
electronically to each member for input before we submit it to the IBIS
Committee. This first BIRD will be short and sweet, stating the
intentions of the IBIS Accuracy Specification and some rough content
related to implementation.

Discussion

Bruce Heilbrunn argued that the IBIS Accuracy Specification ought to
encompass the entire IBIS 3.0 specification. There was some confusion
among others in the group over what new features are available in IBIS
3.0 and what their implications might be regarding accuracy. Greg
Edlund argued that at the very least some of these new features might
require a wider variety of test loads to cover the circuit behavior they
represent. The committee decided to table the question of IBIS 3.0
coverage until we all have the chance to become more familiar with the
new revision.

There was considerable discussion about whether or not our committee
needs to address the question of discrepancies between simulation
engines. We all realize the political sensitivity of this issue. It
would be prudent and expedient for the work of the committee to steer
away from this question as much as possible. One would imagine that at
the most fundamental level, say IBIS 1.1, all simulators must give the
same answer for a simple push-pull output buffer. It would behoove us
to keep things as simple as possible until we establish a solid
foundation. The question becomes, at what level do we begin to
encounter differences between simulation engines. Kumar's presentation
at the IBIS Summit leads us to believe that features introduced in IBIS
2.1, namely VT curves, bring out these differences.

Mike Ward highlighted the need to include in the IBIS Accuracy
Specification material related to data-taking techniques. This material
would include topics such as accounting for probe parasitics, scope
bandwidth, junction temperature, etc. We went on to talk about which of
these items should be specified and which might be considered "cookbook"
material. This will become more clear when we release the BIRD.

Test Board

Greg Edlund, Fawn Engelmann, and Bob Haller worked on powering up the
test board. Even though some of the footprints were wrong, Fawn's
remarkable EC wiring resulting in verification of the board's
functionality. We will be changing the footprints and making some other
minor changes to accommodate a 16821 part so that other vendors
participate in testing. There has been great interest in this little
board. When we are done, we intend to publish the schematics, Gerber
files, parts list, and test plan on the IBIS web page for all to use.

Next meeting: Thursday, March 26 (location TBD)

Homework Assignments

Greg Edlund - draft BIRD and circulate
Fawn Engelmann - turn test board layout
Peter LaFlamme - report on IBIS keywords that are relevant to accuracy
Bob Haller - begin writing the section on data-taking techniques
Bruce Heilbrunn - contact simulation vendors about releasing a
developer's tool kit

----------
Greg Edlund, Principal Engineer
Server Product Development
Digital Equipment Corp.
129 Parker St. PKO3-1/20C
Maynard, MA 01754
(978) 493-4157 voice
(978) 493-0941 FAX
greg.edlund@digital.com