Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site fortune.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxt!houxm!ihnp4!fortune!olson
From: olson@fortune.UUCP (Dave Olson)
Newsgroups: net.dcom
Subject: Re: lightning, PACXs and computers
Message-ID: <5445@fortune.UUCP>
Date: Mon, 5-Aug-85 17:55:17 EDT
Article-I.D.: fortune.5445
Posted: Mon Aug  5 17:55:17 1985
Date-Received: Tue, 6-Aug-85 12:43:31 EDT
References: <1397@islenet.UUCP> <646@mcnc.mcnc.UUCP> <1420@islenet.UUCP>
Reply-To: olson@fortune.UUCP (Dave Olson)
Distribution: net
Organization: Fortune Systems, Redwood City, CA
Lines: 28

I used to work at a company in Mt. View, CA, that regularly had
problems with terminals and DZ-11 interface boards (on 11/60's and
11/70's) being destroyed by lightning.  Because of the product we
manufactured, we had lots of UARTs, and just kept replacing the
terminals' UARTs, and getting the occasional DZ-11 replaced.

Finally, in one particularly bad storm, we lost 12 of 15 DZ boards,
and ~40 terminals.  About 80% of the terminal damage was restricted to
the UARTs, the rest had more extensive damage.  All the systems went
down about 2-3 minutes later, due to a related power outage, however
there was no other damage to the computers.)  After that, we built a
number of optical isolators.  We never lost any boards or terminals
after that.  Some of our RS232 lines ran 900 feet, most were 80-200 ft.

All connections were unshielded 4-wire to a plug-board, and std 25-pair
telephone wire fanning out to various locations from there.  Most
of the destroyed terminals were in the same ferro-cement building as
the computers.  (Side note:  all of them were used at 9600 baud, even
those 900 feet away.  Just goes to show how far you can push the
standard...)

We also had related problems because some of the terminals were in a
building with a different power supply; we measured substantial
differences in the ground potential (which no doubt fluctuated
radically with time).  The isolators (and making sure signal and frame
grounds were NOT connected) solved that problem too.

	Dave Olson, Fortune Systems