Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!uwm.edu!zaphod.mps.ohio-state.edu!usc!jarthur!elroy.jpl.nasa.gov!ames!pacbell!att!dptg!mtunh!dls
From: dls@mtunh.ATT.COM (Dinah Sloan)
Newsgroups: comp.dcom.lans
Subject: Re: 10BaseT on 26 AWG UTP?
Summary: clarification, media specs
Message-ID: <2881@mtunh.ATT.COM>
Date: 20 Mar 90 19:10:05 GMT
References: <8564@lindy.Stanford.EDU> <1990Mar20.163348.6033@agate.berkeley.edu>
Distribution: usa
Organization: AT&T Bell Laboratories - Middletown, NJ
Lines: 26

In article <1990Mar20.163348.6033@agate.berkeley.edu>, macklin@garnet.berkeley.edu (Macklin Burnham) writes:
> In article <8564@lindy.Stanford.EDU> romeo@lindy.UUCP (Patrick Goebel) writes:
> > <question about using 26 AWG UTP wire vs. 22 or 24.>
> >Does anyone know if the current form of the 10BaseT "standard" specifies
> >only a 24 AWG medium?
> >
> The standard specifies 22-26 AWG wire, with no distance penalty.The equipment
> is supposed to drive 100 meters of wire of any of those sizes.

Actually Draft 10 of the proposed standard, section 14.4 is relevant 
to this question.  It states requirements for insertion loss, delay,
differential characteristic impedance, jitter, NEXT (near-end
crosstalk), etc.  It also states in 14.4.2 that "These characteristics
are generally met by 100m of unshielded twisted-pair cable composed of
24 AWG twisted pairs."  Past experience has led me to answer the
question "will this cable work" with a request for a copy of the
characteristics of the cable.  Choosing cable of a particular guage
only tells you a small part of the story.  If you need more specifics,
get a hold of the draft standard.  (Of course, if that fails, you can
always try contacting me and I'll try to answer your questions.)

Dinah Sloan
att!mtunh!dls
AT&T Bell Laboratories
200 Laurel Ave.
Middletown, NJ 07748