X3T9.2/92-127 R0 To: Membership of X3T9.2 From: D.W. Bill Spence & Lawrence J. Lamers Date: May 21, 1992 Subject: Minutes of X3T9.2 May SPI Working Group The SPI working group is the umbrella for all contact, connector, cable, transceiver issues related to SCSI-2 Parallel Interface (SPI). Bill Spence chairs this working group which is chartered with developing a set of recommendations for the SPI standard that will get the physical plant to a reliable state. Agenda -- X3T9.2 May SPI Working Group Harrisburg, PA -- May 19, 1992 1. Opening Remarks 2. Attendance and Membership 3. Cable Issues 3.1 Review of cable impedance parameters 3.1.1 Length and Magnitude of Zo longitudinal variations - Ham 3.1.2 Zo distribution across a round cable 3.2 Review of cable prop delay parameters 3.3 Minimum bend radius --- near connector and away from connector 3.4 Applicability of SPI to A cables 4. Terminator Issues 4.1 New terminator concept from TI -- Mark Granahan 4.2 Review of terminator current-source limits 5. Miscellaneous 5.1 Zo and crosstalk, differential vs s/e cable usage 5.2 Further input on skew budget parameters -- Kurt Chan 5.3 Filter mod proposal -- Asami, Steele, Spence 5.4 Differential driver output voltage specs -- TI 5.5 Single-ended receiver input current specs -- TI 5.6 Which lines should have active negation? 5.7 Annex C - Spence 5.8 Bus Length Requirements - Chan 6. Other items which may be presented 7. SPI Working Group Schedule 8. Adjournment Results of Meeting 1. Opening Remarks Bill Spence called the meeting to order at 1:00 p.m., Tuesday May 19, 1992. He thanked Chuck Brill of AMP for hosting the meeting. As is customary, the people attending introduced themselves. A copy of the attendance list was circulated for attendance and corrections. It was stated that this is an authorized X3T9.2 SPI Working Group meeting. The meeting was conducted under Roberts Rules of Order. Since this is a working group meeting no final actions are taken. The working group prepares a set of recommendations for approval by the X3T9.2 Plenary Committee on Lower Level Interfaces. The minutes of this meeting will be posted to the SCSI BBS and the SCSI Reflector within 14 calendar days. 2. Attendance and Membership Attendance at working group meetings does not count toward meeting the attendance requirements for X3T9.2 membership. The general working group meetings are open to any person or company to attend and express their opinion on the subjects being discussed. The voting rules for the meeting are those of the parent committee, X3T9.2. These rules are: one vote per company; and any participating company member may vote. The following people attended the meeting: Meeting Attendees Name Status Organization Phone Number ------------------------- ------ ------------------------------ -------------- Mr. Robert C. Herron A 3M Company (512) 984-6807 Mr. Ken Wolfswinkel V 3M Company (512) 984-6719 Mr. Thomas Newman A Adaptec, Inc. (408) 945-8600 Mr. Charles Brill P AMP, Inc. (717) 561-6198 Mr. Bob Whiteman A AMP, Inc. (717) 780-7481 Mr. Wills Xu R AMP, Inc. (717) 428-1775 Mr. Russell D. Moser V AMP, Inc. (717) 986-5022 Mr. Jeff Rosa P Amphenol Interconnect (607) 786-4222 Mr. Michael Wingard A Amphenol Interconnect (607) 786-4241 Mr. Richard M. Ross O Amphenol/Spectra-strip (203) 281-3200 Mr. Scott Smyers P Apple Computer (408) 974-7057 Mr. Rich Mizia O C&M Corp. (203) 774-4812 Mr. Bob Gannon C&M Corporation (203) 774-4812 Mr. John Geldman A Cirrus Logic Inc. (510) 226-2368 Mr. Wayne Sanderson P Control Data Corp. (612) 482-2712 Mr. Peter M. Blackford O Cooper Industries (317) 983-5200 Mr. Michael Smith V Dallas Semiconductor (214) 450-0457 Mr. Greg McSorley O Data General Corp. (508) 870-8829 Dr. William Ham A Digital Equipment Corp. (508) 841-2629 Mr. James A. Somerville O DuPont (302) 992-5869 Mr. Skip Jones A Emulex Corp. (714) 668-5058 Mr. D. W. Spence A ENDL Associates (512) 255-0339 Mr. Joel Urban O Fujitsu Microelectronics Inc (408) 922-8928 Mr. Jeffrey L. Williams P Hewlett Packard Co. (208) 323-5030 Mr. Kurt Chan A Hewlett Packard Co. (916) 785-5621 Mr. Howard Wang O Hitachi Computer Products (408) 986-9770 Mr. Tom Kulesza O Honda Connector (708) 913-9566 Mr. George Penokie P IBM Corp. (507) 253-5208 Mr. Claude Mosley S IBM Corp. (507) 253-4064 Ms. Gricell Co V IBM Corp. (512) 838-3664 Mr. Kevin R. Pokorney O Intellistor, Inc. (303) 682-6649 Mr. Lawrence J. Lamers P Maxtor Corp. (408) 432-3889 Mr. Bill Kutsche O Murata Erie N.A. (814) 237-1431 Mr. Bruce DeBree R NEK/Helix Cable (CBM) (516) 567-5000 Mr. Gene Milligan A Seagate Technology (405) 324-3070 Mr. Gerald Houlder A Seagate Technology (612) 844-5869 Mr. Hale Landis A Seagate Technology (408) 439-2443 Mr. Robert L. Simpson P Sony Corp. of America (408) 944-4348 Mr. Vit Novak A Sun Microsystems, Inc. (415) 336-2455 Mr. Mark Granahan Texas Instruements (214) 997-5955 Mr. Ricardo Dominguez P Texas Instruments (512) 250-6204 Mr. Harvey Waltersdorf P Thomas & Betts (803) 676-2905 Mr. Mark Jordan S Unitrode Integrated Circuits (603) 429-8628 Mr. Shishir Shah A Western Digital (714) 932-7235 3. Cable Issues 3.1 Review of cable impedance parameters Bill Spence led a discussion on the cable impedance values since many of the cable folks had previously voiced concern over the numbers in the current SPI document. The cable impedance values were adjusted to 72/115 ohm min for S-E and Diff. The cable impedance values were adjusted to 160 ohm max for Diff. The cable impedance difference values were not changed. The new numbers should allow for use in differential mode of any cable which is within spec for single-ended mode. The question of measuring the difference arose. The test method still needs to be included in appendix. Vit Novak questioned the measurement technique since it could cause differences in the numbers. The cable folks seem to have reached some agreement on test methods so they could provide comparative data. This still needs to be included in the document. 3.1.1 Length and Magnitude of Zo longitudinal variations - Ham Bill Ham made a claim that a significant variation in impedance can occur locally without materially affecting the system integrity. He felt that this would solve many of the system integrator's concerns with bends in the cables, mis-match when changing cable types, stub lengths, and short runs of flat ribbon cable. The technical justification is primarily that the rise time of 5ns minimum consumes at least 3.8 feet of cable and impedance variations with that length should have little affect on the signal. Bill's proposed adding the following wording to the SPI document: "A 20% variation in local impedance is allowed over 1 foot." There was some discussion on this, in particular relating to the difficulty in meeting the 5ns minimum rise time. There was also concern about who gets the local variation, and how many are allowed within a system. The consensus was that an implementor's note for the annex on cables should be considered. 3.1.2 Zo distribution across a round cable Rich Mizia of C&M presented data on 28 AWG, 68-conductor round cable impedance measurements. The data showed that the middle and center conductors are close in impedance, but the outer conductors are lower. 3.2 Review of cable prop delay parameters The prop delay numbers were rounded to 5.4 ns/m max, and the difference was restored to 0.15. There was discussion on removing the minimum and it was agreed to remove it. The Project Editor will make these changes to the SPI document. 3.3 Minimum bend radius --- near connector and away from connector Rich Mizia presented test data that showed bend radius does not have a serious effect on the C&M cable tested. The bend radius used was 1/4- to 1/2-inch. The cable survived 9200 full 180-degree bends before failing. This is very comforting, and all but eliminates a concern that had been expressed in past meetings. 3.4 Applicability of SPI to A cables Bill Spence spoke on the A-cable issue. Bill would like a bit stronger wording relating the SCSI-3 cable parameters to A-cables. George Penokie stated that electrical characteristics are in the next section and so 5.2 should be moved to section 6. The Project Editor will move the section and work with Bill to improve the wording. 4. Terminator Issues 4.1 New terminator concept from TI -- Mark Granahan Mark Granahan of TI presented a new terminator concept. The TI TL2218-285 SCSI Terminator utilizes this concept. Each package has 9 signals. The service current is 1ma. The package size was very small. The waveforms presented showed great improvements over a Boulay-type terminator in configurations where one device in a data transfer is at a bus end. Bill Spence commented that while this terminator significantly improves data transfers with one of the devices at the end of the bus, it cannot improve the first step level experienced in transfers between two devices when neither is at a bus end. The first step in the middle of the bus is determined by the cable impedance and the line current. As long as the line current is limited to a "legal" 24ma, not much can be done to improve the first step signal. 4.2 Review of terminator current-source limits Bill Spence gave a little sermon on the "outlaw" FPT terminators. They have an important presence in the SCSI market, and the fact that all existing SCSI drivers have always sunk well over 48 ma for the first 20-40 ns of each assertion proves that exceeding the driver current spec limit is nothing new. He asked if the 24 ma limit should somehow be reconsidered. Gricell Co affirmed the need for FPT terminators to get systems to work. She would very much like to get the SPI document to address these real world needs. In order to do this the current would have to be raised to 60ma (30 ma per terminator). The silicon folks were adamant against changing the SPI document, but all indicated a willingness to qualify parts to a private qualification requiring 60ma. One possible compromise: do not change the 48/24 ma limits but ease the stringency of the limits. 5. Miscellaneous 5.1 Zo and crosstalk, differential vs s/e cable usage This topic was not discussed due to lack of time. 5.2 Further input on skew budget parameters -- Kurt Chan Kurt Chan showed a revised skew budget. He explained that the 6ns skew allowed for external is not sufficient given the temperature variations that can occur. Revised to 10ns this ripples through the skew budget leaving only 8ns for the cable. There was a ground swell of opposition to giving the transceiver folks 10 ns. There was not a clear consensus on making a change. What is needed is a consultation with the external transceiver folks (IEEE, SC6) (National, TI, AT&T, AMCC) to come to an agreement on the skew budget needed for those devices. 5.3 Filter mod proposal -- Asami, Steele, Spence The consensus of the silicon folks was that there is no need to limit the glitch rejection to 1 volt. The Project Editor will remove the 1 volt requirement from the SPI document. Tak Asami and David Steele were not present so this issue may get revisited. 5.4 Differential driver output voltage specs -- TI Bill Spence briefed the group on Kevin Gingerich's document regarding differential output characteristics. At this SPI working group meeting the differential transceiver folks were not present. These folks need to take a look at this document (92-123) and comment on it. 5.5 Single-ended receiver input current specs -- TI Bill Spence briefed the group on Steve Hunley's document (92-124) regarding single-ended input characteristics. John Geldman stated that we have gone to great lengths to recover the leakage allowed in SCSI-2. What Steve was offering is a doubtful justification to reverse the progress made. The power consumption needed to allow this bias is not acceptable today. The consensus is that allowing a +- 20 microamps tolerance might be acceptable, but that raising the limit was not. This topic was suggested to have a plenary vote. The SPI working group recommends allowing for bipolar leakage but maintaining the 20 ua limit. 5.6 Which lines should have active negation? This topic was not discussed due to lack of time. 5.7 Annex C - Spence Bill Spence brought in a proposal (92-122) for Annex C of the SPI document. This topic was not discussed due to lack of time. 5.8 Bus Length Requirements - Chan Bill Spence stated that this topic has already been discussed within the SPI working group and the present SPI provisions resulted. To reopen the topic there should be a wider audience. The document (92-115) was referred to the General Working Group for consideration. 6. Other items which may be presented Due to a lack of time during the SPI working group the following topic was discussed the next day during the general working group. Bill Ham - Single cable for SE and Diff SCSI: Does the current SPI document require REQ/ACK pair placement for Diff? In the opinion of the working group, the answer was no. However, the SPI document is a little ambiguous on this point. Bill Ham would like all P-cables be wired as single-ended cables. The project editor will move the paragraph relating to this topic in the SPI document from section 5.2 into section 5.2.1. 7. SPI Working Group Schedule The consensus was that another SPI Working Group meeting is needed. There is not a meeting scheduled at the next plenary in June. Bill Spence will request authorization for a SPI Working Group meeting during the next working group week in July, 1992. The meeting announcement will be posted to the SCSI BBS and the SCSI Reflector. 8. Adjournment The meeting was adjourned at 5:45 p.m. on Tuesday, May 19, 1992.