Doc Number: X3T9.2/92-22
 Date: January 17, 1992
 Author: G. Penokie/C. Mosley
 Desc: SPI Connector Contact Working Group Minutes 1/15/92

 SCSI-3 Connector Contact Meeting Attendees

 Name                   Organization
 ---------------------- -------------------------
 Mr. Robert C. Herron   3M Company
 Mr. Jim Havener        3M Company
 Mr. Charles Brill      AMP, Inc.
 Mr. Bob Whiteman       AMP, Inc.
 Mr. Ken Schreder       AMP, Inc.
 Mr. Bill Garver        AMP, Inc.
 Mr. Edward Hrvatin     Burndy Corp.
 Mr. Geoffrey Hyman     Cables To Go
 Mr. George Zisk        CW Industries
 Dr. William Ham        Digital Equipment Corp.
 Mr. Mark Sausert       ELCO Corp.
 Mr. Tom Kulesza        Honda Connector
 Mr. George Penokie     IBM Corp.
 Mr. Claude Mosley      IBM Corp.
 Mr. Larry Grasso       IBM Corp.
 Mr. Tom Forrer         IBM Corp.
 Mr. William D'Andrea   J.S.T. Corporation
 Mr. Bob Masterson      Methode Electronics, Inc.
 Mr. Jim McGrath        Molex Inc.
 Mr. Max Bassler        Molex Inc.
 Mr. Dave Bowen         Molex Inc.
 Mr. Ram Battu          Seagate chnology, Inc.
 Mr. Vit Novak          Sun Microsystems, Inc.
 Mr. Dean Wallace       Texas Instruments
 Mr. Harvey Waltersdorf Thomas & Betts

 26 People Present

 Claude Mosley presented his SCSI-3 Connector proposal (see
 X3T9.2/92-16).  There was a great deal of interest in
 Claudes' presentation.  Many issues where discussed.  The
 major points where as follows:

 Questions were asked as to what criteria Claude used to
 determine a connector failure.  Claude indicated one criteria
 was that if the resistance increased by 100% or more after
 environmental testing the connector failed.  This was
 acceptable, but questions about what test does the connector
 get exposed to was not settled.  All players were asked to
 consider whether or not the EIA and IEC documents would be
 acceptable to define the these tests.

 In the discussions about surface finish Claude indicated
 that, although he calls out a surface finish of N5 or better,
 he realizes that a curved surface finish is difficult to
 measure.  He would rather measure the porosity and verify
 that the number of mating cycles does not cause wear thru of
 the noble plated surface.

 There was a long discussion on how many connect/unconnects
 should be allowed in a SCSI connector system.  It was finally
 agreed that 200 cycles with no wear thru of the noble plating
 was a reasonable number.

 The next topic was pin radius.  Claudes' picture which shows
 a pin digging into a contact was enough to convince the
 non-connector people in the room that a radius was needed.
 After some talk about how to specify the radius it was agreed
 to call out the same thickness as todays pin (0.0156") with a
 radius of 0.045 +/- 0.005 ".

 The issue of which connectors should the radius be
 incorporated on was addressed.  The general feeling was that
 it should at least apply to all the square pin style
 connectors which are in the standard today.

 The last major issue was wipe length.  Claude showed that the
 numbers in the SCSI standard (A14 and A18 in figure 9) were
 way too conservative and should be changed.  This caused
 quite a stir among the connector people but in the end it was
 agreed that the values would only be changed to values which
 did not cause any changes to existing connectors.  The
 connector vendors where given the task of determining what
 the values of their pins are and getting that information to
 Claude before or at the next plenary meeting in February.

 George Penokie and Claude Mosley agreed to write a proposal
 which would incorporate the above agreements.  This proposal
 will be distributed at the February SCSI meeting.  Another
 SWIG on connector will then be held at the March SCSI meeting
 to discuss and finalize the proposal.  The final proposal
 will then be presented to the plenary for a vote at the April
 SCSI meeting.