From owner-freebsd-hardware Thu Sep 18 18:07:45 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id SAA24154 for hardware-outgoing; Thu, 18 Sep 1997 18:07:45 -0700 (PDT) Received: from math.berkeley.edu (math.Berkeley.EDU [128.32.183.94]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id SAA24148 for ; Thu, 18 Sep 1997 18:07:40 -0700 (PDT) Received: (from dan@localhost) by math.berkeley.edu (8.8.7/8.8.7) id SAA26720; Thu, 18 Sep 1997 18:07:37 -0700 (PDT) Date: Thu, 18 Sep 1997 18:07:37 -0700 (PDT) From: dan@math.berkeley.edu (Dan Strick) Message-Id: <199709190107.SAA26720@math.berkeley.edu> To: FreeBSD-hardware@FreeBSD.ORG Subject: Re: SCSI3 cables Cc: dan@math.berkeley.edu Sender: owner-freebsd-hardware@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk The "high quality" SCSI cables sold by Granite Digital and other companies seem rather pricey. I think they are taking excessive advantage of a SCSI FUD factor resulting from SCSI errors caused by end-user ignorance and cable vendor behavior verging on outright fraud. For example, many end-users think nothing about connecting "fast-20" devices together with 6 foot cables and wouldn't even know to wonder about the termination of the extra signals at a wide/narrow SCSI cable junction. Then there are the cable manufactures/vendors who shave pennies off the cost of cables by using thin wire and omitting some of the grounds. Avoiding these kinds of really gross configuration errors is essential. You should pay attention to the rules for maximum cable and stub lengths. You should know exactly where your terminators are. Care is especially required with "fast" and "ultra" SCSI. However, using gold plated connector shells with embedded LEDs does not make your electrons move any faster. Gold plated contacts and teflon insulation can make the physical cable and the connections a little more robust, but they aren't worth the extra cost in most applications. You don't need silver wire. The manufacturer should have used an adequate wire size (28 awg of better) and the cable should have a "characteristic impedance" similar to the termination resistance (usually about 115 ohms these days; was about 132 ohms before "active" terminators). It is not easy to find out exactly how your cables are made, especially when they have molded high density connectors. The vendor probably doesn't even know. I feel ripped off at both ends. Dan Strick dan@math.berkeley.edu