🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: sci.electronics
Article #4754

Re: tapping into a fiber optic cable

#4754
From: Rob Warnock
Date: Sat, 02 Jun 1990 05:25
39 lines
2600 bytes
In article <3332@ash31.UUCP> svoboda@motcid.UUCP (David Svoboda) writes:
+---------------
| From article <61338@sgi.sgi.com>, by rpw3@rigden.wpd.sgi.com (Rob Warnock):
| > In fact, there are companies [not mine] out there working on using this
| > phenomenon to build practical fiber-optic "quick-connects"...
| Look in your handy-dandy Inmac catalog.  A year and a half ago I installed a 
| fiber optic "parallel cable" on an unruly laser printer.  The setup consisted 
| of an optic xcvr at either end (a black box the size of a null modem) and a 
| two "conductor" plastic optic cable.  Just strip, stuff, and twist, as you 
| said.  About 130 bucks for two xcvrs and a cable...
+---------------

You are talking about *plastic* fiber, at *low* speeds. That plastic stuff,
usually 1mm diameter, has enormous loss, typically .3-.5 dB/meter, and bad
dispersion, usually .5-1 ns/meter. That means at 10 Mb/s data rates ypu can
go at most 30-50 meters. (Yes, I have pushed Ethernet packets down el cheapo
Hewlitt-Packard plastic fiber.) This plastic stuff is nice, but it's the
"RS232" of fiber. And it's not "bent-fiber", but "easy termination" (with
an eXacto knife) of a more-or-less standard (but plastic and thus low-cost)
butt-connect (straight end-on). It's the large 1000 micron aperture, over
50 times the area of even the largest glass fibers, that lets this work in
the presence of sloppy alignement. And the 2-4 dB of the resulting crude
termination gets lost in the high loss of the fiber itself. So it's a win,
at very low speeds and very short distances.

But I was talking about being able to quick-connect, say, low-loss 62.5 micron
glass, the 500 MHz-km stuff you use for FDDI that carries the 100 Mb/s data
rate (125 Mbaud) for up to 2 km per hop. "Bent-fiber" connectors for such
glass don't exist on the open market yet. But when they do, they'll start
to give "10baseT" twisted-pair Ethernet some real competition...


-Rob

-----
Rob Warnock, MS-9U/510		rpw3@sgi.com		rpw3@pei.com
Silicon Graphics, Inc.		(415)335-1673		Protocol Engines, Inc.
2011 N. Shoreline Blvd.
Mountain View, CA  94039-7311
Technical Details
Message-ID: <61464@sgi.sgi.com>
Path: arc1-ca.pugleaf.net!utzoo!attcan!uunet!mailrus!ames!sgi!rpw3@rigden.wpd.sgi.com
References: <61338@sgi.sgi.com> <3332@ash31.UUCP>

Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!utzoo!attcan!uunet!mailrus!ames!sgi!rpw3@rigden.wpd.sgi.com
From: rpw3@rigden.wpd.sgi.com (Rob Warnock)
Newsgroups: sci.electronics
Subject: Re: tapping into a fiber optic cable
Message-ID: <61464@sgi.sgi.com>
Date: 2 Jun 90 05:25:43 GMT
References: <61338@sgi.sgi.com> <3332@ash31.UUCP>
Sender: rpw3@rigden.wpd.sgi.com
Reply-To: rpw3@sgi.com (Rob Warnock)
Organization: Silicon Graphics, Inc., Mountain View, CA
Lines: 39