================================================================================ IBIS INTERCONNECT TASK GROUP http://www.ibis.org/interconnect_wip/ Mailing list: ibis-interconnect@freelists.org Archives at http://www.freelists.org/archive/ibis-interconn/ ================================================================================ Attendees from November 1 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, Ming Yan Mentor, A Siemens Business 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. Justin Butterfield took minutes. Review of Minutes: - Michael called for review of the minutes from the October 25 meeting. Mike LaBonte moved to approve the minutes. Walter Katz seconded. The minutes were approved without objection. Review of ARs: - Arpad Muranyi to work with Bob Ross and Walter on Sets and Groups language. - Arpad reported that this is started but not yet complete. Opens: - Per Mike's suggestion last time the unused port termination discussion item is now at the top of the agenda for this meeting. Summary of IBIS-ATM Discussion: - Arpad reported that the unused port termination issue was discussed. Bob had distributed some draft text for rules on the new Unused_port_termination parameter. Bob stated that Mike also had an interesting idea on Unused_port_termination rules. Unused ports termination: Michael asked Mike to summarize his proposal on unused port termination from ATM. Mike reported that they had discussed the Unused_port_termination parameter which includes everything you would want to include. Mike stated that he had created some examples which brought up some interesting questions. Michael asked if these examples were shared in ATM. Bob said that they were not. Mike shared the examples sent to the reflector. He took one of the existing examples in the BIRD and changed it from an .s2p to an .s6p with four of the ports as unused. The issue is when we have per port resistance. Mike looked at a couple of cases with the Unused_termination_list defined before or after used ports. In all cases, the parser has to look at all the ports. He commented that some rules about the order of the terminal lines might be necessary. Mike proposed to add the word "Unused" as a column entry with the resistance value in the next column. Walter liked the idea but with just the resistance value. Randy Wolff asked if we would still have the shortcut to make a selection for all the unused ports. Mike replied that the shortcut would still work. Arpad was concerned that this might open the door for conflicting termination settings. Michael asked if resistance would be the only choice or if we could have a reactance or something else. Mike commented that this could be a possibility as well as an option to use an ISS subcircuit for the termination. Bob stated that his concern is with the clarity of the specification. With Mike's proposal we would have to redo Table 41, and it changes the meaning of the terminal line columns. Bob also stated that it is redundant to say "Unused" and give the resistance value, as we could just list the termination impedance. He stated that we need to make a choice on how we define the columns. Bob's opinion is that the Unused_termination_list subparameter should separate the unused terminals. Mike asked in that case, if we would have rules of how to order the terminals list, as the Unused_port_termination_directive could be anywhere. Bob noted that we have these options as syntactical choices, and Mike's proposal deviates from some conventions we have followed in the past. Michael asked Randy if any of these choices would be more efficient for his purposes. Randy replied that for large models you would not want to use per port termination. Bob commented that this is very detailed information and most models would have a global termination impedance. Randy noted that one use case would be if you have a large model and only a few ports need per port termination. Mike commented that you could have the unused port termination directive, and the per port selections would override the selection for the model. Walter proposed to remove the per port termination feature due to it complicating the syntax. He noted that Radek may object and vote to keep it. Michael will take an AR to ask Radek about it offline. Walter stated that models can be wrapped in a subcircuit as a workaround to eliminate the unused ports. Bob agreed that enumerating all the unused ports complicates the syntax. Michael asked if removing the per port termination feature would break anything else in the syntax. Bob thought it would not break anything. Michael asked if we can remove this feature and be open to it as a separate BIRD. Bob agreed with this approach. Randy stated that he does not like the word "directive" in the proposed Unused_port_termination_directive subparameter. Bob agreed that the word "directive" is redundant. Arpad asked how it would work to add the per port termination later. Bob proposed that we could add per_port as option for the Unused_port_termination subparameter. Bob commented that we could add a third column to the Unused_port_termination subparameter for the resistance value rather than have a separate additional subparameter for the resistance value. Randy agreed that this would be simpler. Mike will take an AR to write an example using Bob's syntax with a third column for resistance. The next meeting is tentatively scheduled for Friday November 3. Although, Michael will not be able to host. Randy moved to adjourn. Bob seconded. The meeting adjourned without objection.