Date: Fri, 03 Oct 1997 13:28:15 +0000 From: =?iso-8859-1?Q?=DEor=F0ur?= Ivarsson <totii@est.is> To: jbryant@tfs.net Cc: Mike Newell <mnewell@newell.arlington.va.us>, freebsd-hackers@FreeBSD.ORG Subject: Re: thickwire<->thinwire Message-ID: <3434F2EF.41C67EA6@est.is> References: <199710030000.TAA03036@argus.tfs.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Jim Bryant wrote: > > In reply: > > On Thu, 2 Oct 1997, Wilko Bulte wrote: > > > > wilko> Not completely. The collision detect uses a diode connection to the central > > wilko> conductor on the thinwire tranceiver, and is directly connected for > > wilko> thickwire (or the other way around, I forgot that detail). This is at least > > wilko> the difference in the NS8392 based implementation I used when I built > > wilko> a couple of tranceivers myself (years ago, when they were rare and $ were > > wilko> few ;-) > > > > For short distances we had good success using an "N" to BNC adaptor (a few > > bucks in most "real" electronics stores). Just pull the terminator off > > one end of the thick, screw the adaptor on, twist on the thin, and put the > > standard BNC type terminator at the end of the thin. Wouldn't recommend > > it for cables near their max length, but for short ones... > > my memory fails me... the wire frequency is 20MHz, correct? > > the max length is limited by distributed capacitance [assuming a long, > random, non-resonant length of transmission line], correct? > > applying conventional transmission line theory, could you not use > longer lengths that are multiples of one wavelength of the base > frequency in use? this should apply maximum in-phase signal to the > downstream connection. > > or am i completely misunderstanding ethernet here? > > i would think that the only problems that would occur through the > N<->BNC adaptor would be a small insertion loss [avg for a good > connector around 1.05-1.1 dB], and slight increase in SWR. > I was told by network specialist and engineer at Novell that the time factor is the ruling factor of the length of cable. Rules was as following 1. Velocity factor (Size and dielectric of cable) 2. Number of clients (changes the SWR) 3. SWR of cable (damages cables, connectors, terminators) 4. External interference (Improper earthed computer) You might have seen cards from "Western Digital" that support extra long thin wire cables, just move jumper. Company here in Iceland had very long thin ethernet cable and lot of those cards but when we rushed in and put up our new computer with NE2000 from Novell nothing seemd to work. Thordur Ivarsson thivars@est.is
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3434F2EF.41C67EA6>