Path: utzoo!utgpu!water!watmath!clyde!att!rutgers!gatech!ncar!ames!mailrus!purdue!decwrl!labrea!glacier!jbn
From: jbn@glacier.STANFORD.EDU (John B. Nagle)
Newsgroups: comp.sys.misc
Subject: Re: Protecting computer equipment
Message-ID: <17524@glacier.STANFORD.EDU>
Date: 3 Jul 88 20:41:55 GMT
References: <1188@gmu90x.UUCP>
Reply-To: jbn@glacier.UUCP (John B. Nagle)
Organization: Stanford University
Lines: 31


      Whenever someone discusses lightning protection and computers, they
seem to conclude that full protection against a direct lightning strike on
a power or phone line is hopeless.  This is simply not so.  Antenna
towers are frequently struck by lightning; for some installations, it's
a normal event, occuring hundreds of times a year.  The problem can be solved.

      A serious lightning arrestor consists of several stages.  The first
is an air gap to ground, with big contacts spaced a short distance apart.
Dime-sized silver contact faces are not uncommon.  The connection to ground
must be very heavy copper connected to a small farm of ground rods, which
may also be tied to the ground of the electrical distribution system.

      The second stage is an inductor, to block the lightning surge and
force it across the air gap to ground.  This coil is typically just a
few turns of copper busbar.  In some cases, it's just a few loops of 
the normal power or coax cable.  But tight turns of busbar inside a
grounded metal can are better.

      The big energy having been dumped to ground, now we're ready for
the usual "surge suppressor" electronics, the MOV devices and such.  

      Heathkit used to sell a soup-can sized unit that contained stages
1 and 2 above, equipped with coax connectors on each end.  But I don't
know a good source for such things today.  Try ham radio suppliers.
This is exactly what is needed where an Ethernet cable enters a building.

      The ARRL Handbook, section 37-4, contains some useful information
on lightning protection.

					John Nagle