================================================================================ 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 April 6, 2018 Meeting (* means attended at least using audio) ANSYS Curtis Clark Cadence Design Systems Bradley Brim Intel Corp. Michael Mirmak* Keysight Technologies Radek Biernacki* Mentor, A Siemens Business Arpad Muranyi* Micron Technology Justin Butterfield*, Randy Wolff* SiSoft Walter Katz*, Mike LaBonte* Teraspeed Labs Bob Ross* 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 April 4 meeting. Arpad Muranyi moved to approve the minutes. Walter Katz seconded. The minutes were approved without objection. Review of ARs: - Arpad to review the "Important" paragraphs on page 30 for inclusion in BIRD189 [AR]. - Arpad reported he sent out an email about this shortly before the meeting. He hoped to review this today. Opens: - Michael stated there is a DASC related meeting directly following this meeting, and we need end about 5 minutes early. - Bob noted we sometimes use node 0 and sometimes we use node "0" with quotes. Michael asked Bob's preference. Bob preferred to use node "0" in quotes. Michael stated we should add this to the editorial bin list. BIRD189.5_draft18_v6 Review: Arpad stated there are three "Important" notes in the Terminal line rules section. He had an AR focussing on the last two of these paragraphs, and he wanted to ensure after the A_gnd changes that this section is still needed. Arpad proposed to reduce the text to one sentence. His intent is to emphasize that the model maker should ensure a proper connection between the buffer and the interconnect models. Bob asked if this applies to File_ISS and File_TS. Arpad noted the text does not mention either and is intended to apply to both. Radek Biernacki agreed the new text is better. He stated for IBIS_ISS, the reference node may not be defined. Arpad noted the model maker should know, for IBIS_ISS models, how to connect the reference node. Radek thought there could be some room for improvement in the wording. Arpad asked how the sentence could be improved. Radek said he would have to think about it. Bob stated that although you may not know the reference node in for File_ISS, it may not be necessary. Michael asked if this can be resolved as part of an editorial pass. Arpad moved to replace the text as noted in his in email. Randy Wolff seconded. There were no objections. Michael replaced this text in the BIRD. Arpad asked if we can remove any comments related to this section. Michael replied there were none. Arpad asked about the other "Important" note paragraphs. Radek stated the second note says A_gnd becomes a node which spans the model, and it should become the reference node. We have to make sure the node 0 inside the IBIS-ISS subcircuit is an external connection. Arpad noted that we did decide that A_gnd is a global node. He has issue with the phrase "shall be treated as the A_gnd node" and asked what it is really referring to. Radek noted that node 0 is the A_gnd in the ISS node list. Randy asked if Radek's intention is to treat the node 0 inside the ISS subcircuit as a global node 0. Arpad asked why we need to do this if ISS already says this. Radek stated node 0 can be brought out as a node on the subcircuit definition. He would like to make sure that node 0 and A_gnd become the reference at the pin interface. Bob commented there are ISS subcircuits where node 0 is used, but it is not exposed as a terminal. Michael asked in the case where you have a two terminal model, but there is a node 0 inside, if this forces the pin side to use node 0 as the reference. And he asked if node 0 is truly global. Radek replied that a global node is equivalent to having that node exposed on the definition of every subcircuit. Walter thought we need to state in the BIRD that there are two types of models: those that reference node 0 and those that define a specific local reference. Michael asked Walter if he would write up some text on the two types of models [AR]. Walter agreed. Michael asked Radek if there should be a restriction that if a model has A_gnd or node 0, it has a new terminal that effectively appears at the pin interface of the device. Radek commented that the text can be simplified. Bob noted we would need to change figure XX1 to include the A_gnd terminal. We also have a rule that node 0 in a subcircuit should not be exposed. Michael asked if the model maker should have to add the node 0 terminal when the simulator can do this automatically. Randy noted that you cannot put node 0 on the subcircuit definition line in some tools. If node 0 exists inside the subcircuit, the tool has to find it and assign it. Walter had written a sentence stating if node 0 is inside the subcircuit, you cannot account for all currents in the subcircuit. Radek stated that it could be okay to include node 0. Michael stated he does not want a situation where models are adding the node 0 on the [Pin] list. Michael asked what else is left to address in the BIRD. Mike noted there were five comments remaining. Michael stated that we have now three comments left after replacing Arpad's text. Arpad asked about the first "Important" note on page 30. Radek stated this is a good note. Arpad asked what is an I/O signal port. Radek stated this is the IO pad. Arpad asked why we cannot include power rails in this. Mike suggested to just say ports. Radek said he will improve the language of the first and second "Important" notes on page 30 [AR]. Bob stated he is not sure we fully understand the impact of exposing A_gnd as a terminal. Next Meeting: The next meeting will be April 11. Bob moved to adjourn. Mike seconded. The meeting adjourned without objection. Task List BIRD189.5 editorial additions/changes to be completed: 1. Resolve inconsistency with using node 0 and node "0" with quotes.