I didn't catch the first message from Mr. Jones, but it is completely
wrong. Ethernet runs on a Carrier-Sense Multiple Access with Collision
Detection (CMSA-CD) scheme, and the problem with long cables is that the
number of packet collisions starts to increase dramatically over about
100 meters. It has nothing to do with the speed of "1"s and "0"s
traveling through the wire. See
<http://userpages.umbc.edu/~jack/ifsm498d/tcpip-intro.html> for an
introduction to the concepts of TCP/IP and packet switching, etc.
This has nothing whatsoever to do with the problems of interference with
USB cables, which carry serial data streams, not packets, and are
subject to external electromagnetic interference.
Mike
Rich Schinnell wrote:
At 12:00 AM 3/4/2008, you wrote:
Date: Sun, 2 Mar 2008 15:58:43 -0500
From: Brian Jones <[EMAIL PROTECTED]>
Subject: Re: USB cables
Well, it does make a difference. With Ethernet, the faster the
transfer rate, the more twists you must have in the cable (per foot)
to minimize crosstalk. That is why you have different grades of
Ethernet cable (5, 5E 6). And length is a factor as well. Typically,
(but not always), higher data rates dictate shorter distances with a
given technology. The problem is that the digital '1' and digital '0'
do not travel at the same speed though the wire... over short
distances it's not an issue, but over longer distances, or at higher
frequencies, a '1 bit can catch up with the '0' ahead of it, thus
corrupting both. Using higher quality cables (less resistance) can
obtain for you a little more distance.
Is this the 1st of April? I really need a reality check with this "1" bit
catching up to a "0" bit.
<GGGGGGGGGGGGGGGGGGGG>
*************************************************************************
** List info, subscription management, list rules, archives, privacy **
** policy, calmness, a member map, and more at http://www.cguys.org/ **
*************************************************************************
