============================================================================== IBIS INTERCONNECT TASK GROUP Mailing list: ibis-interconnect@freelists.org ================================================================================ Attendees from June 4, 2025 Meeting (* means attended at least using audio) ANSYS Curtis Clark, Wei-hsing Huang, Juliano Mologni Arista Networks Jim Antonellis Broadcom James Church Intel Corp. Michael Mirmak*, Xiaoning Ye Keysight Technologies Ming Yan Marvell Steve Parker MathWorks Walter Katz* Micron Technology Justin Butterfield Siemens EDA Weston Beal, Arpad Muranyi*, Randy Wolff* Simberian Yuriy Shlepnev ST Microelectronics Aurora Sanna Synopsys Ted Mido, Edna Moreno University of Illinois Jose Schutt-Aine Michael Mirmak convened the meeting. No patents were declared. Michael reviewed the minutes of the May 28 meeting. He noted two issues 1) The end of an early paragraph was apparently unfinished. "parameters not M - Example 9 [AR]; reviewed examples" This should end as "parameters not mentioned." 2) The word "point" is repeated in a sentence. Randy Wolff moved to approve the minutes with corrections to the above; Walter Katz seconded. The minutes were approved without objection. Michael noted that all open ARs have been closed. Michael mentioned that he was reviewing a popular source for S-parameter information, "Microwave Transistor Amplifiers: Analysis and Design" by Guillermo Gonzalez (the publisher, Prentice Hall, is now defunct, though the book should still be protected by copyright. Second-hand editions are widely available for purchase. The following link is provided for reference only: https://archive.org/details/microwavetransis0000gonz ). The Gonzalez book rarely uses the ground symbol for referencing, except for physical measurements. Michael also mentioned the existence of an HP (later Agilent) application note on scattering parameters where ground never appears and all circuits are presented as loops. Several versions are available online, but the original has been preserved as a scan: https://www.hpmemoryproject.org/an/pdf/an_95-a.pdf Arpad Muranyi noted that, in one of the port mapping examples, the common base of a transistor was treated as the negative terminal of other two ports; the three terminal approach may be wrong. Michael asked whether this is common practice, or is the implication that other methods are wrong? Michael reviewed draft 14 of the Port Mapping proposal. He provided an overview of changes, including a new statement on referencing assumptions cautioning model makers and system designers regarding combining models from different sources. On referencing, Walter endorsed the statement, and Randy also agreed. Arpad stated that he needed more time to review the text. Walter suggested using the Open Forum for technical updates. Michael mentioned making example 8, 9 changes to remove redundancies and show different Data_usage options. Arpad asked whether proposed updates from online discussions were not included. Michael confirmed that this was correct. Walter noted that there were big confusions from the offline discussion involving Walter's fundamental assumption of a "ground" associated with all the signals in the Touchstone file. One could say these are examples of a 4-pin coaxial cable, and the ground return is well-understood. We just assume this. Should we say we are assuming this? Example 1 could include a statement regarding the application as a co-axial cable. Arpad added that Example 1 could also be a shielded cable. Randy suggested the example simply documents 4 signals of a cable, period. On the transistor example, Arpad stated that he would like to see three examples (2a, 2b, 2c) for Example 2, involving common emitter, common base, and common collector. Michael will send representative Gonzalez images for this case. [AR] Michael asked whether each example is a closed system, where all ports connect to all other ports. Arpad noted that Randy found VGA and HDMI (8-port VNA) characterization data, where the latter was mixed-mode. A shield was present, but between the signals. Arpad asked whether, if you plot differential data, you care where the ground is. For Example 3a, 3b Michael asked if they are describing the same thing twice. Walter replied that a co-planar waveguide is a single trace with ground on both sides; one needs a special probe for measurement (one for signal trace at the end, and the other two are grounds at both sides). We should be documenting at port 2 that this is the way measurements were done; the example may be nonsensical. Arpad asked whether this is a U-shaped ground plane with trace in opening. Walter stated that there are two different ways of measuring this structure (the reference moves between the cases). Randy stated that it is odd to have 3a using a pin for one reference and a group for the other reference. Walter replied that two ground traces may connect at a pin. Randy suggested that the text somehow describe this in notes. Michael accepted this AR. [AR] Walter noted that he thought Example 6 is good. Randy suggested to move Example 6 earlier in the documentation, as it is simpler. Arpad suggested making it Example 3. Randy pointed out that, in Example 7, the numbers are DIMM pins. In Example 8, the team suggested adding logical references to show connectivity (PadN, PadP, DieN, DieP); keep Physical notation. Michael accepted the AR to check the IBIS specification language for language around Ts4file port ordering. [AR] The team will resume editing Example 9 next time. Arpad moved to adjourn; Walter & Randy seconded. The meeting adjourned without objection.