IBIS Accuracy Sub-Committee Minutes

IBIS Accuracy Sub-Committee Minutes

Thursday, March 26, 1998
Held at Digital Equipment's PKO3 building, 129 Parker St., Maynard, MA

Attendees

Greg Edlund, Digital Equipment (chair)
Fawn Engelmann, EMC
Bob Haller, Digital Equipment
Bruce Heilbrunn, Stratus Computer
Peter LaFlamme, Fairchild Semiconductor
Harvey Stiegler, Texas Instruments (via conference call)

Business

The chairman will be responsible for circulating the agenda the Monday
prior to the meeting and posting the minutes to the IBIS reflector and
IBIS user's group mailing list within one week. The sub-committee
members will be responsible for circulating their homework among the
group on or before the Monday prior to the meeting.

Bob Haller has begun writing a rough draft. He will fold other's work
into the document and circulate it among the group. When the document
gets closer to a first release, it will go through a more formal editing
and formatting procedure.

When it becomes apparent that the group cannot agree on an issue, we
have agreed to table the issue until the next meeting to prevent
grid-lock. In the mean time, parties will have the opportunity to
discuss the issue off-line and build a case, which will be presented
briefly at the next meeting and brought up for a vote.

Correspondence

The chairman received much positive and constructive feedback from
publishing the "Proposal for an IBIS Accuracy Specification," and we
reviewed this feedback at the meeting. We also received some dissenting
opinions which in general ran along the lines that the work of the IBIS
Accuracy Sub-Committee did not need to carry the full weight of a
specification. (In our discussions we maintained anonymity for those
who had negative opinions so as to avoid any flame wars.) In addition,
there seemed to be some points of confusion in the proposal; it may need
to be reworded in some areas. The IBIS Accuracy Sub-Committee would
like to thank all who took the time to respond. We hope that people
will continue in the future to be liberal with their critical input.

Milestones

The group approved the setting of three milestones, to which we will add
as we work out the details.

May 15, 1998: Post the IBIS Model Test Board to the web.
September 18, 1998: Distribute the first draft of the IBIS Accuracy
Specification
February 1999: Present the IBIS Accuracy Specification at DesignCon99

Editing

1. Scope

Peter Laflamme successfully completed his research into IBIS
"parameters" from version 1.1 through 3.0 which might be relevant to
accuracy. In summary, these parameters include I/O buffer capacitance,
package RLC, IV curves, and ramp rate. He also presented a list of
other parameters which might be relevant in the future.

The discussion continued around how much of IBIS we should attempt to
cover in the first release of the IBIS Accuracy Specification. Bruce
Heilbrunn, supported by a former colleague from Stratus, argued that the
first draft of the IBIS Accuracy Specification should cover the entirety
of IBIS 3.0. Greg Edlund and Bob Haller argued that IBIS has evolved to
cover various new I/O buffer designs which may require different test
conditions than a generic IBIS 1.1 push-pull or open-drain buffer. Greg
and Bob's position is that it would be too much work to attempt to cover
all possible I/O buffer designs currently covered by IBIS 3.0 in the
first draft. The sub-committee tabled the issue for a vote at the next
meeting.

2. Test Loads

Bob Haller documented a first cut at a set of test loads in his rough
draft. These include standard load, transmission line terminated into
its characteristic impedance (or resistive load if board layout
permits), open-ended transmission line, and output buffer driving an
input buffer on the same component.

Bob also proposed two other "optional" net topologies, but the group has
yet to decide whether these belong in the IBIS Accuracy Specification or
the IBIS Cookbook.

3. Measurement Techniques

Bob's rough draft also included an outline of measurement techniques,
including oscilloscope bandwidth, test environment characterization, and
test conditions.

4. Metrics

We did not discuss comparison metrics at this meeting.

IBIS Model Test Board

Peter Laflamme, Fawn Engelmann, and Greg Edlund have successfully
debugged everything on the test board except for the capacitance
measurement. If anyone has experience measuring pin capacitance on a
powered-up DUT using an HP4175 LCR meter, please drop us a line! Fawn
and Greg will be making modifications to the schematics and layout and
submitting the design for fabrication and assembly in the next two
weeks. Peter Laflamme has graciously offered the services of
Fairchild's PC board shop to build this new revision of the board.

IBIS Developers Tool Kit

Following the lead of Hyperlynx's, Bruce Heilbrunn has drafted a letter
to the other EDA vendors requesting a similar tool be made available to
the IBIS community in the coming year. It is our feeling that a
reduced-feature simulation engine would be a great aid in the model
development process. Furthermore, if modeling engineers have several
simulation engines available, they can assess the magnitude of the
differences among them for various test cases. For the purposed of the
IBIS Accuracy Sub-Committee, this may allow us to avoid any pitfalls
arguing over third-order effects.

IBIS Cookbook

Bob Haller and Stephen Peters have agreed to work together to complete
the remaining two chapters of the IBIS Cookbook related to measurement
and verification. Clearly, there is much overlap between the work of
the IBIS Accuracy Sub-Committee and the Cookbook.

Homework

Greg Edlund - finish schematics and layout of the IBIS Model Test Board.
Build a strong case for why different driver families may require
different test loads and why the first version of the IBIS Accuracy
Specification needs to be constrained to IBIS 1.1; discuss off-line

Fawn Engelmann - finish schematics and layout of the IBIS Model Test
Board.

Peter LaFlamme - list all IBIS 1.1 keywords that are relevant to the
IBIS Accuracy Specification. Ask Mike Ward if he is willing to provide
input to the Measurement Techniques section.

Bob Haller - continue developing the Measurement Techniques section.
Discuss with Stephen Peters an outline of IBIS Cookbook chapters 5 and
6.

Bruce Heilbrunn - add a list of suggested features to the IBIS
Developer's Toolkit letter. Build a strong case for how the IBIS
Accuracy Specification can cover all I/O buffers covered by IBIS 3.0
within the constraints of our milestones; discuss off-line.

Next meeting

Thursday, April 30, 1998 from 3:00 to 5:00 pm at Digital Equipment's
PKO3 building, 129 Parker St., Maynard, MA.

----------
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
Received on Thu Apr 2 12:28:34 1998