Minutes, IBIS Quality Committee

16 Dec 2008

11-12 AM EST (8-9 AM PST)

ROLL CALL
  Adam Tambone
* Anders Ekholm, Ericsson
  Barry Katz, SiSoft
  Benny Lazer
  Benjamin P Silva
  Bob Cox, Micron
* Bob Ross, Teraspeed Consulting Group
  Brian Arsenault
* David Banas, Xilinx
* Eckhard Lenski, Nokia Siemens Networks
  Eric Brock
* Guan Tao, Huawei Technologies
  Gregory R Edlund
  Hazem Hegazy
  Huang Chunxing, Huawei Technologies
  John Figueroa
  John Angulo, Mentor Graphics
  Katja Koller, Nokia Siemens Networks
  Kevin Fisher
  Kim Helliwell, LSI Logic
  Lance Wang, IOMethodology
  Lynne Green
* Mike LaBonte, Cisco Systems
  Mike Mayer, SiSoft
* Moshiul Haque, Micron Technology
* Pavani Jella, TI
  Peter LaFlamme
  Randy Wolff, Micron Technology
  Radovan Vuletic, Qimonda
  Robert Haller, Enterasys
  Roy Leventhal, Leventhal Design & Communications
  Sherif Hammad, Mentor Graphics
  Todd Westerhoff, SiSoft
  Tom Dagostino, Teraspeed Consulting Group
  Kazuyoshi Shoji, Hitachi
  Sadahiro Nonoyama

Everyone in attendance marked by *

NOTE: "AR" = Action Required.

-----------------------MINUTES ---------------------------
Mike LaBonte conducted the meeting.

Call for patent disclosure:

- No one declared a patent.

AR Review:

- Mike propose IBISCHK bug for 5.5.6
  - This has been drafted.
  - Still not sure if it should be proposed.

New items:

- Eckhard emailed a document "Possible Errors-revised-16dec08.doc"
  - This topic (Section 6) will be discussed in this meeting.

- In a previous meeting David had proposed a discussion of correlation
  - This is section 7 and will be discussed after section 6.

Review of Eckhard's document:

6.1.	{LEVEL 2} Typ/min/max order of parameters correct
- We have been filtering out checks that call only for observance of the IBIS 
  specification.
- We deleted this check

6.2.	{LEVEL 3} C_comp checked in both input and output mode
- Mike: There has been some controversy on this topic
- David: Some structures produce a very wide C_comp range
  - Lower pin capacitance is a competitive issue
  - One idea is to release IBIS files with:
    - C_comp = C_comp_off
    - C_comp_on value commented out
  - Bob: This might be a good cookbook suggestion
- David: There should be a more elegant way to do this
  - Mike: The ibis-atm group is considering using [External Circuit]s
    - This would allow C_comp to depend on voltage, frequency, and state
  - David: Should there be separate state keywords?
    - There are 9 possible C_comp state values:
      (hi, lo, tri) x (typ, min, max)
  - Moshiul: In most cases tri gives good correlation
  - David: Having the right value gives better rise/fall time correlation
- David: We should agree in the IQ spec what the best practice is
  - We are not yet certain what the "mission" of section 6 is
  - Should IBIS files get bonus points for meeting these?
- Bob: The lack of C_comp typ/min/max correlation is a problem
  - C_comp is the only IBIS keyword that does not follow typ/slow/fast corners
  - New C_comp keywords could address this
  - A 4th dimension is the actual pad voltage, not the driven state
  - Methods to determine C_comp:
    - Using ramp sweeps to calculate C covers a wider voltage range
    - Can also AC sweep the bias point
    - There is a way to sweep frequency
    - A TDR can be used to see the C_comp blip
  - All of these methods give different results
- David: Section 6 should be for guidance, not affecting IQ scores
  - Mike: We could have an IQ score letter to signify extra checks
- Section 6 could be called "Consistency Checks"
- There are two separate topics here:
  - How to represent C_comp
  - Detecting C_comp values that are guessed or incorrectly determined
- Mike: Will IC vendors have trouble with a C_comp consistency check?
  - Moshiul: No Problem
  - Pavani: Agree: We do correlation checks.
- Moshiul; Sometimes devices are used in different modes
  - This results in changing C_comp
  - Mike: We could have a 5.x "driving and non-driving" mode
    - This coule be check 5.1.4
    - David: We may be adding confusion with that
  - Bob: We should insert 5.1.5 as a conditional check:
    - For higher speed applications consider a check that calls for
      more detailed C_comp
    - Separate driving and non-driving values, for example

AR: Moshiul draft new IQ check for reasonable C_comp for high speed
AR: Mike post 1.1ak IQ spec with deletions processed and renumbered

Discussion of meeting dates
- The next 2 Tuesdays might be poorly attended due to holidays
- We decided to meet 6 Jan 2009

Next meeting:

6 Jan 2009 11-12 AM EST (8-9 AM PST)

Meeting ended at 12:08 PM Eastern Time.
