X3T9.2/90-104 Date: July 5, 1990 To: X3T9.2 Members From: John Lohmeyer, X3T9.2 Chairman Subject: High-Density Interconnection Working Group Meeting I chaired a working group meeting June 19, 1990 from 10:00 a.m. until 12:00 noon to investigate high-density interconnection issues for SCSI-2. Almost all of the meeting was consumed in presentations from the following people: Company Name/phone Foils H/W --------------------- ---------------- ------ --- NCR John Lohmeyer no yes (316) 636-8703 Hewlett Packard Ed Silva yes yes (916) 785-4204 Icontec Alan Haig yes yes (408) 945-7766 Context Elect./Mold-Con Roger Brickley yes yes Bob Krzynowek (413) 736-0371 CMS Wayne Douglas yes ? (508) 580-0111 Connective Technologies Fred Hengelhaupt yes yes (802) 257-5242 Black Box Corp. Bob Stevens (Observe only) (412) 746-6829 Amphenol Bill Sopchak no yes (607) 786-4370 AMP Bob Whiteman yes yes (717) 780-7481 Honda Connector David McFadden yes yes (708) 913-9566 Fujitsu Joel Urban yes yes (408) 562-1722 3M Bob Herron yes yes (512) 984-6807 Hirose Electric Oscar Kornblum yes ? (805) 522-7958 JAE David Shaff yes yes (714) 753-2600 I have attached copies of the presentations which were made available to me. Some issues that were identified are: 1. NCR has experienced difficulties in finding cost-effective solutions for mixing high-density shielded connectors on cabinets with nonshielded low- density connectors inside cabinets. They are especially concerned about finding good solutions for the upcoming P cable. 2. HP has experienced bent pin problems with the high-density connector when mounted horizontally with cables long enough to cause the connector to sag (droop). They are considering using jackscrews instead of the clips shown in the SCSI-2 figures. The cable droop does not cause loss of electrical continuity, but can deform the pins. 3. Several cable assembly vendors pointed out that all of their customers for SCSI cables have unique requirements. There is no such thing as "the" SCSI cable. 4. One cable assembly vendor strongly recommended against using 2-56 jackscrews presumably because they are too weak and break when over tightened. (Later in the week, I learned that the HIPPI standard will call out 2-56 jackscrews with a minimum torque specification that can be met with stainless steel screws. The 4-40 screws have interference problems.) 5. Some customers are frustrated by slow delivery times on the high-density connectors. 6. Bob Whiteman (AMP) showed a prototype 68-position daisy-chain high-density connector which uses 0.025 inch centerline ribbon cable. No follow-up meeting is planned, however the HP issue concerning the retention mechanism will be included on the August X3T9.2 agenda. Attachments: