Minutes from the 31 Mar 2020 IBIS Quality Task Group meeting are attached.
Mike
Minutes, IBIS Quality Task Group
31 March 2020
11:00-12:00 EST (08:00-09:00 PST)
ROLL CALL
Intel Corp. Michael Mirmak
Mentor, a Siemens Business Arpad Muranyi
Micron Technology * Randy Wolff
Signal Integrity Software: * Mike LaBonte
SPISim Wei-hsing Huang
Teraspeed Labs: * Bob Ross
Zuken USA: Lance Wang
Everyone in attendance marked by *
NOTE: "AR" = Action Required.
-----------------------MINUTES ---------------------------
Mike LaBonte conducted the meeting.
Call for IBIS related patent disclosures:
- None
Call for opens:
- None
Review of previous meeting minutes:
Minutes from the March 24 meeting were reviewed. Bob Ross moved to accept.
Randy Wolff seconded. Without objection the minutes were approved.
ARs:
- Bob Ross to find more up to date TSCHK2 source code license document
- Done, but an updated version could not be found. We have one under review
anyway.
- Mike LaBonte to update IBISCHK "orphan" message sections 3, 4, and 6.
- Done.
- Bob Ross to reduce over-extrapolation bug testcase size and post it
- Done, but will be reviewed before posting.
New IBISCHK bugs:
Bob Ross showed BUG211 with updates. I-V curves had been reduced to 5 points
each,
and precision reduced. Mike had researched the history of the decision to
check for
excessive currents 1.5V outside the operating range. It was part of EGG10. The
1.5V
and 10A limit values were found to be outside the range seen in models at the
time.
Bob said there were no checks for [Pulldown] or [Pullup] tables. It could have
been
changed to an In model, but it wasn't. Bob had changed the data to make all 3
corners
give the warning message.
Bob suggested making it a caution message, but not removing the test. Another
question
was whether to make it apply to all [IBIS Ver]. Randy Wolff suggested fixing
the
extrapolation to use a valid voltage. Mike had suggested a code change for
that.
Bob noted that I-V curves can go beyond what is required. Mike LaBonte felt
that
extrapolation was not needed. We could simply find the nearest table point to
1.5V.
Bob noted that it would also interpolate to 1.5V if the table extends beyond.
Mike said
it should be OK to interpolate, but never extrapolate. He was assuming the
last points
would have the highest current. Bob said the last point could be the highest,
but maybe
[Voltage Range] would be better. Randy agreed.
Bob said the typical corner voltage should be used for all checks. Randy asked
if
the positive side was checked. Bob said voltages on both sides of the
operating range
were checked. Randy said power clamps would sometimes have a different voltage
than
voltage range.
Mike showed his suggested code changes. Bob said all reference voltages were
required
if any was given. Randy said all four were required if voltage range was
missing.
Either [Power Clamp Reference] or [Pullup Reference] were required for power
clamps.
Bob felt this issue was holding up BIRD181.2
Mike said the point was to forward bias the diode and check the current. The
voltage
range was only a loose indication of the clamp forward bias voltage. Bob said
the
extent of table points should give a good indication. Mike suggested starting
at
1.5V, interpolating as needed, but not extrapolating. Randy said diodes
connected
to references other than [Voltage Range] existed in the past but were rare
today.
Those reference voltages were all beyond 1.5V.
Bob said he would introduce the bug at the Open Forum meeting Friday. Randy
and Bob
felt it should be a caution message. Mike said the original intent seemed to
be to
guard against forgetting to put "m" for milliamps.
TSCHK2 license agreement:
No discussion.
Better IBISCHK error messages for BUG210:
Bob said the bug report needed to be amended.
AR: Mike LaBonte to update BUG210 with message update suggestions.
IBIS-ISS parser:
No discussion.
Randy Wolff moved to adjourn. Bob Ross seconded. Without objection the meeting
ended.
Meeting ended: 12:08 ET
Next meeting March 31, 2020
