Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!philabs!cmcl2!harvard!caip!lll-crg!lll-lcc!portman
From: portman@lll-lcc.UUcp (Roland F. Portman)
Newsgroups: net.lan,net.dcom
Subject: Re: Do you ground your Ethernet cable?
Message-ID: <306@lll-lcc.UUcp>
Date: Wed, 21-May-86 13:52:17 EDT
Article-I.D.: lll-lcc.306
Posted: Wed May 21 13:52:17 1986
Date-Received: Sat, 24-May-86 02:51:45 EDT
References: <261@pedsgo.UUCP> <7584@glacier.ARPA>
Reply-To: portman@lll-lcc.UUCP (Roland F. Portman)
Distribution: net
Organization: WG/USD, Lawrence Livermore Nat'l Lab.
Lines: 23
Xref: linus net.lan:1322 net.dcom:1679

In article <7584@glacier.ARPA> jbn@glacier.UUCP writes:
>
>       The Ethernet I spec called for the outer conductor of the Ethernet
>cable to be ungrounded.  Because the connectors are uninsulated, this
>created a safety hazard (a short to the braid due to wear between two
>rubbing cables, one Ethernet and one power cable would
>electrify all the exposed Ethernet connectors but wouldn't blow any breakers) 
>and violates the U.S. National Electrical Code.  So the Ethernet II spec 
>calls for grounding the Ethernet at exactly one point.
>       Back at Ford Aerospace, we grounded our Ethernet with a copper
>strap brazed to one of the connector barrels; the copper strap connected
>to the building ground system (3/8" x 1.5" copper busbars under the
>raised floor tied via heavy copper cable to the grounding rods out back
>of the building and cross-connected to power system ground at the wye
>of the main transformers; overkill, but a place that builds antennas
>takes a dim view of RF noise.)
>
>						John Nagle

One additional note, just be sure to ground only ONE end otherwise
nasty ground loop currents could creep up and cause major problems.

Roland