(attaching a text version of the minutes for ease of archiving)
======================================================================
IBIS EDITORIAL TASK GROUP
http://www.ibis.org/editorial_wip/
Mailing list: ibis-editorial@xxxxxxxxxxxxx<mailto:ibis-editorial@xxxxxxxxxxxxx>
Archives at //www.freelists.org/archive/ibis-editorial/
======================================================================
Attendees from March 25 Meeting (* means attended at least using audio)
ANSYS Curtis
Clark*
Cadence Design Systems Bradley Brim
Cisco David
Siadat
Intel Corp. Michael
Mirmak*
Keysight Technologies Radek Biernacki*
Mentor Graphics Arpad Muranyi*
Micron Technology Justin Butterfield,
Randy Wolff*
SAE ITC Maureen
Lemankiewicz, Logen Johnson
Signal Integrity Software Walter Katz*, Mike
LaBonte*
Teraspeed Labs Bob Ross*
University of Aveiro in Portugal Wael Dghais
Michael Mirmak convened the meeting. No patents were declared.
Walter Katz moved to approve the minutes of the March 11 meeting. Radek
Biernacki seconded. No objections were raised and the minutes were approved.
During the discussion of opens, Bob Ross noted he had a version of the
spreadsheet table showing terminology usage in the IBIS document. Walter added
that he had sent out an e-mail regarding figure 16 (per an earlier AR) and in
response to Bob's recent IBIS-ATM task group presentation.
Walter reviewed his e-mail, to provide an overview of the work before the
Editorial Task Group. The e-mail can be seen at
//www.freelists.org/post/ibis-editorial/Conclusions-from-Figure-16-presentation
Rails are kept constant in a DUT (device-under-test) with respect to a test
fixture reference node. Voltages are provided to rails specified by
Pullup_Reference etc. Bob raised concerns about packaging being included as
part of the DUT test fixture or results. Two sections would be provided in an
updated IBIS document: one to describe data for a DUT (which is essentially
today's specification) and the other for the DIA or device-in-action.
Radek noted that references are needed for I/O signals. This is the only item
still missing. Walter replied that GND is the test fixture reference node.
Bob expressed generally agreement with the e-mail, though he has some issue
with the idea of a "best guess" about C_comp referencing by the EDA tool in the
absence of C_comp_* subparameters. EDA tools may apply C_comp differently in
their tools. It's not clear whether separate sections for DUT and DIA are
needed.
Walter replied that the DUT approach assumes non-varying rails/reference. This
may not be true of the DIA usage. Bob responded that K-tables have terminals
connected to the I-V tables, meaning that the K-tables have references.
Radek stated that the document needs a measurement that can be used
unambiguously. IBIS should discuss DIA vs. DUT; on this, he is perfectly
agreed with Walter.
Michael asked about a universal ambiguous-reference statement - text near the
beginning of the document that states C_comp and similar keywords today have
ambiguous references if C_comp_* or other, future subparameters specifying the
references are not included. Walter replied that this is not needed.
Bob reviewed his IBIS model terminal names document. He noted that a pulldown
resistor or pulldown is used for ECL/PECL referencing, an idiosyncrasy of IBIS.
He pointed out an issue with [Ramp], where the conditions are embedded in
IBISCHK as assumptions. A single-waveform model can apparently be created with
just the [Ramp] data, and pass the parser.
Walter replied that he concludes that the names in Bob's document Column B are
connections to local or global ground. Column F should use the same names as
Column C. He proposed editing the interconnect draft document to use the [Pin
Mapping] names from Column C. A standard name is still needed for the I/O pin
or pad. He suggested going through graphics in similar detail. Bob replied
that the team would have to be careful about expressing voltage drops vs.
"absolute voltages".
Michael asked whether column B names get changed for the proposed DIA section
in IBIS. Walter replied that the voltage in the I/O would be specified as a
voltage rail in column C and a number for column A.
For next time, the team will review Mike LaBonte's task list.
Mike moved to adjourn. Radek seconded the motion. The meeting adjourned.
======================================================================
IBIS EDITORIAL TASK GROUP
http://www.ibis.org/editorial_wip/ ;
Mailing list: ibis-editorial@xxxxxxxxxxxxx
Archives at //www.freelists.org/archive/ibis-editorial/ ;
======================================================================
Attendees from March 25 Meeting (* means attended at least using audio)
ANSYS Curtis Clark*
Cadence Design Systems Bradley Brim
Cisco David Siadat
Intel Corp. Michael Mirmak*
Keysight Technologies Radek Biernacki*
Mentor Graphics Arpad Muranyi*
Micron Technology Justin Butterfield, Randy Wolff*
SAE ITC Maureen Lemankiewicz, Logen Johnson
Signal Integrity Software Walter Katz*, Mike LaBonte*
Teraspeed Labs Bob Ross*
University of Aveiro in Portugal Wael Dghais
Michael Mirmak convened the meeting. No patents were declared.
Walter Katz moved to approve the minutes of the March 11 meeting.
Radek Biernacki seconded. No objections were raised and the minutes
were approved.
During the discussion of opens, Bob Ross noted he had a version of the
spreadsheet table showing terminology usage in the IBIS document.
Walter added that he had sent out an e-mail regarding figure 16 (per
an earlier AR) and in response to Bobs recent IBIS-ATM task group
presentation.
Walter reviewed his e-mail, to provide an overview of the work before
the Editorial Task Group. The e-mail can be seen at
//www.freelists.org/post/ibis-editorial/Conclusions-from-Figure-16-presentation
Rails are kept constant in a DUT (device-under-test) with respect to
a test fixture reference node. Voltages are provided to rails
specified by Pullup_Reference etc. Bob raised concerns about
packaging being included as part of the DUT test fixture or results.
Two sections would be provided in an updated IBIS document: one to
describe data for a DUT (which is essentially todays specification)
and the other for the DIA or device-in-action.
Radek noted that references are needed for I/O signals. This is the
only item still missing. Walter replied that GND is the test
fixture reference node.
Bob expressed generally agreement with the e-mail, though he has some
issue with the idea of a best guess about C_comp referencing by the
EDA tool in the absence of C_comp_* subparameters. EDA tools may
apply C_comp differently in their tools. Its not clear whether
separate sections for DUT and DIA are needed.
Walter replied that the DUT approach assumes non-varying rails/reference.
This may not be true of the DIA usage. Bob responded that K-tables
have terminals connected to the I-V tables, meaning that the K-tables
have references.
Radek stated that the document needs a measurement that can be used
unambiguously. IBIS should discuss DIA vs. DUT; on this, he is
perfectly agreed with Walter.
Michael asked about a universal ambiguous-reference statement text
near the beginning of the document that states C_comp and similar
keywords today have ambiguous references if C_comp_* or other, future
subparameters specifying the references are not included. Walter
replied that this is not needed.
Bob reviewed his IBIS model terminal names document. He noted that a
pulldown resistor or pulldown is used for ECL/PECL referencing, an
idiosyncrasy of IBIS. He pointed out an issue with [Ramp], where the
conditions are embedded in IBISCHK as assumptions. A single-waveform
model can apparently be created with just the [Ramp] data, and pass
the parser.
Walter replied that he concludes that the names in Bobs document Column
B are connections to local or global ground. Column F should use the
same names as Column C. He proposed editing the interconnect draft
document to use the [Pin Mapping] names from Column C. A standard name
is still needed for the I/O pin or pad. He suggested going through
graphics in similar detail. Bob replied that the team would have to be
careful about expressing voltage drops vs. absolute voltages.
Michael asked whether column B names get changed for the proposed DIA
section in IBIS. Walter replied that the voltage in the I/O would be
specified as a voltage rail in column C and a number for column A.
For next time, the team will review Mike LaBontes task list.
Mike moved to adjourn. Radek seconded the motion. The meeting
adjourned.
