====================================================================== IBIS EDITORIAL TASK GROUP http://www.ibis.org/editorial_wip/ Mailing list: ibis-editorial@freelists.org Archives at http://www.freelists.org/archive/ibis-editorial/ ====================================================================== Attendees from March 4 Meeting (* means attended at least using audio) ANSYS Curtis Clark* Intel Corp. Michael Mirmak* Keysight Technologies Radek Biernacki* Mentor Graphics Arpad Muranyi* SAE ITC Maureen Lemankiewicz, Logen Johnson Signal Integrity Software Walter Katz*, Mike LaBonte* Teraspeed Labs Bob Ross* Michael Mirmak convened the meeting. No patents were declared. Michael summarized the stated objective of the meeting as resolving the use of “GND” and related terms in the IBIS specification in a consistent manner. Bob Ross stated that the team needed to agree on a process for meetings, including ground rules. Radek Biernacki agreed, noting that he had produced a slide set an e-mail on the topic from some months ago. Walter noted that he had also produced a slide set, plus a collection of historical presentations. He added that voltage statements in the specification must be followed with statements regarding “in reference to what”. Bob added that the reference must be defined with respect to technology, and that’s part of the clean- up of the document. Mike LaBonte suggested that the focus is development of IBIS 6.2, and believes this will involve the normal process (BIRDs). The alternative is to issue specification by fiat. Walter did not necessarily agree. A full IBIS 6.2 can be covered by a single BIRD. Mike stated he would accept this if the changes are clear. Bob suggested a single BIRD is possible if we minimize the changes. Technical content might be brought out separately. Radek suggested that the work be done in parallel, with changes made to the specification, extracting that material for editing a BIRD. Michael asked whether the changes required an opening statement only, or page-by-page revisions. Walter suggested both. Anything that requires parser changes needs close consideration. He agrees with Radek – the team may go through the entire document and extract pages into a separate BIRD. Walter added that no parser changes are expected. Radek suggested otherwise. Walter replied that this assumes Pin Mapping may require detailed changes. Bob suggested holding off that discussion until the terminology is reviewed; that’s the key issue. The team should stabilize the document as is. Michael asked four questions 1) Is a particular IBIS section in need of particular focus, such as [Model]? 2) Should the BIRDs be submitted directly from this Task Group? 3) Would there be any interconnect proposal impact? 4) Are any other documents affected? Bob, on question (1), noted as an example that the GND symbol is used in a diagram. In addition, [Pullup_Reference] may be connected to a terminal or some other connection. [Voltage Range] and similar terms in models need to be clarified. On question (2), Mike and Radek agree, unless the Open Forum throws them to a different Task Group for discussion. Arpad also agrees, but ATM can bring any issue in as needed. On question (3), Radek stated Interconnect agreed to let GND be resolved elsewhere. On question (4), Radek noted that we allow IBIS-ISS to use node 0; this needs clarification. Mike asked who shall take minutes in future. He suggested that, in the next meeting, the team start a numbered task list or list of recorded ideas, e.g., change “x” to “y” in every place it appears. The team needs a working document to maintain, that forms our institutional memory. Bob replied that a working document would be separate, and that a review page-by-page would still be required. We might limit scope to exclude the EMI section, for example. Mike replied that grounding could be particularly important there. Radek stated that we may go page by page, or in groups of pages; we can search the document for different things rather than sequentially. On a task list, he sent an e-mail months ago suggesting a number of tasks. Mike agreed. Walter asked that Mike upload to one location all his relevant reference documents. Radek reviewed his slides. Key points: - There’s a silent assumption of global reference in [Model] - Global node zero is OK to use as a reference - IBIS-ISS subcircuits support use of global node zero Two total pages of topics included: global node 0, nodes for C_comp connection, impact of different pulldown and GND clamp reference, non-zero pulldown/GND references and signal I/O reference node with or without pin mapping? Pulldown_ref bus declarations with pin mapping? Bob suggested that Pin Mapping not be touched by this discussion. Walter agreed, noting that several issues were resolved in separate presentations/discussions. Mike accepted an AR to produce the task list for review at next meeting. Walter moved to meet twice out of every 3 weeks, when the IBIS Open Forum is not meeting. Arpad asked about duration. Walter stated that 1 hour, meetings to exclude holidays. Radek seconded the motion. No objections were raised. Walter presented his slides on referencing, including on VDD- relative and Vcc-relative connections. He suggested that Vcc- relative means that voltages are referenced to the Vcc pin, as is stated in the [POWER Clamp Reference] keyword. The standard equation involves Vtable = Vcc – Voutput. Similar language is proposed for “GND pin” under [GND Clamp Reference]. Test fixture concepts are involved here, so the proposed text adds Device in Action and Device Under Test concepts. For next time, Vinl and Vinh will be addressed. Bob objected to adding the new concept of a VCC pin. Walter moved to adjourn. Curtis Clark seconded. The meeting adjourned.