Date: July 9, 1990 To: Cable Working Group Participants From: Lawrence J. Lamers Subject: July 9, 1990 Cable Working Group Meeting Larry Lamers called the meeting to order at 2:30 p.m., Monday, July 9, 1990. He thanked George Penokie of IBM for hosting the meeting. The attendance list was circulated. Participants were requested to add or verify their facsimile numbers so they could be contacted if needed. As is customary, the people attending introduced themselves. The final agenda was as follows: 1. SCSI Cables Observations & Proposals (90-100) [Spence] 2. Crosstalk Comparisons (90-106) [Chan] 3. NCR Cable specification (90-108) [Lohmeyer] 4. Cable Specification [Lamers] 5. Action Items 6. Next Meeting The following people attended the meeting: Name Organization ------------------------------ ------------------------------ Mr. Denis D. Springer 3M Company Mr. James Fiala 3M Company Mr. Jon Abilay Apple Computer Mr. Peter M. Blackford Astro Cable Company Mr. Thomas Debiec Belden Wire & Cable Mr. John Ellis Berk-Tek, Inc. Mr. I. Dal Allan ENDL Mr. Robert Liu Fujitsu America, Inc. Mr. Kenneth Post Future Domain Mr. Kurt Chan Hewlett Packard Co. Mr. Chuck Grant Madison Cable Corp. Mr. Lawrence J. Lamers Maxtor Corp. Mr. Richard Wagner Montrose Products Company Mr. Ron Schlenker Montrose Products Company Mr. John Lohmeyer NCR Corp. Mr. Thomas R. Marks NEK/Helix Cable, Inc. Mr. Robert N. Snively Sun Microsystems, Inc. Mr. D. W. Spence Texas Instruments Mr. Erik Jessen Western Digital 19 People Present The following new documents were distributed at the meeting: Document Doc Date Author Description of Document ------------- -------- --------------- --------------------------------------- X3T9.2/90-100 6/26/90 B. Spence SCSI Cables Observations and Proposals X3T9.2/90-106 7/6/90 K. Chan Best/Worst Case Cable Crosstalk Comparison X3T9.2/90-108 W. Stengel NCR Contribution to Cable Spec. WG (Cable, Shielded-25 Pr, #28AWG...) The results of the meeting 1. SCSI Cables Observations & Proposals (90-100) [Spence] Bill Spence gave an overview of his document and advocated that the cable working group not come up with "hard" specifications. He wants "soft" specifications that do not infer a specific construction. This would allow advances in cable making and alternative styles of cable to be used in the future. The group agreed with these aims, however several members felt that a practical cable assembly be written up in an appendix as a guide to users. 2. Crosstalk Comparisons (90-106) [Chan] Kurt Chan presented the results of his testing of cable crosstalk as it relates to signal location within the cable. No one was surprised with the results. A cable that is constructed with the REQ/ACK signals in the inner layer and the data signals in the outer layer showed the least crosstalk. Kurt also reported that the difference in impedance between the inner and outer layers of the was on around 2 ohms on the cable tested. The conclusion is that the placement of signal pairs is the overriding factor in the measured crosstalk. 3. NCR Cable specification (90-108) [Lohmeyer] John Lohmeyer brought in copies of the NCR cable specification for SCSI-2 cables as an example. 4. Cable Specification [Lamers] The main task of this meeting was to develop a specification for 50- conductor cable. Erik Jessen got out a foil and started the collecting of parameters to be specified. Then the real fun began when values were selected for the parameters. The following parameters and values were agreed to as a starting point for a recommendation to the X3T9.2 committee. o Propagation Delay < 5,41 ns/m o Signal Skew (delta Pd) < 5 ns measured over 6 meters o Rise time degradation <= 2 ns measured over 6 meters o Impedance 70 ohm min, SE (higher is better) o Capacitance to ground (max) 100 pF/m o Delta capacitance (max) 5 pF/m o Square wave attenuation 1V drop from 5V input @ 5MHz, with a rise time of 3ns on the input signal measured over 6 meters o Average pair to adj pair crosstalk 7% max (digital) The above parameters do not apply to TERMPWR lines. The TERMPWR lines will only have a resistance specification of DC Ohm <= 0,23 ohm/m. This requires 28 AWG wire for TERMPWR, while the other signals could be 28 or 30 AWG. The impedance discussion was the longest. However it is clear that cable has been designed for use in SCSI systems by at least two vendors that will meet an 80 ohm minimum specification. The two possible constructions are 28 AWG with foam insulation (80 ohm) and 30 AWG with pvc insulation (84 ohm). Bob Snively and Larry Lamers felt that the minimum specification should be 80 ohms. Bob Snively questioned whether the above values were realistic since they are interrelated and a cable could not be built with an arbitrary set of values. Jim Fiala maintains that if you specify three of the parameters the rest will be dictated by a nomograph and the practical considerations of building a cable. Jim stated that measuring the rise time over the maximum cable length is a good indicator of the quality of the cable as a transmission line. Jim plans to review the above parameters as part of his study on specifying cables. He promised to mail out a report on his work in August to members of the cable working group. Several members felt that test methods for measuring the parameters need to be specified. Jim Fiala stated that his report will address this issue. 5. Action Items 1) Jim Fiala to complete report and mail it to cable working group members. 2) Jim Fiala to re-check the cable parameter values against the samples. 3) John Lohmeyer to provide set of mailing labels to Jim Fiala. 4) Larry Lamers to write-up proposed cable specification. 6. Next Meeting The chairman felt that the group had accomplished its main objective and that if there as there were no further work items the group need not meet again separately. Bob Snively immediately requested that the group develop a recommended color coding scheme for an appendix. Several cable vendors volunteered to look into the color coding. The 68-pin cable also needs some work. Therefore the group will meet again. The next meeting is scheduled for Wednesday, September 5, 1990, at the Ramada Inn in Westminster, CO. Westminster is located midway between Denver and Boulder. The starting time is 1:00 p.m.