Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!caip!nike!ucbcad!ucbvax!XX.LCS.MIT.EDU!ELMO
From: ELMO@XX.LCS.MIT.EDU
Newsgroups: mod.telecom
Subject: (none)
Message-ID: <8607280012.AA15349@ucbvax.Berkeley.EDU>
Date: Sun, 27-Jul-86 20:12:47 EDT
Article-I.D.: ucbvax.8607280012.AA15349
Posted: Sun Jul 27 20:12:47 1986
Date-Received: Sun, 27-Jul-86 23:35:44 EDT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The ARPA Internet
Lines: 47
Approved: telecom@xx.lcs.mit.edu

2
Received: from IBM.COM by XX.LCS.MIT.EDU with TCP; Thu 24 Jul 86 09:42:04-EDT
Date: 24 July 1986, 09:12:22 EDT
FroTo:   telecom@mit-xx.arpa
Message-Id: <072486.091225.victor@ibm.com>
Subject: Do long cords cause errors?
ReSent-Date: Sun 27 Jul 86 19:59:54-EDT
ReSent-From: The Moderator <TELECOM-REQUEST@MIT-XX.ARPA>
ReSent-Sender: ELMO@XX.LCS.MIT.EDU.#Internet
ReSent-To: Telecom-Individual-Messages-List: ;
ReSent-Message-ID: <12226133997.15.ELMO@XX.LCS.MIT.EDU>

The following is an excerpt from a local bulletin board concerning
accessing our main computer from home.  Would anybody care to
comment on the assertion that long telephone cords/long modem cables
might cause the error rate on a line to go up drastically?

> I think I have more-or-less solved my line error problem.  While I was
> crawling under my PC desk disconnecting a defective monitor, I noticed
> that I had the remaining 5 feet of the 6-foot RS232 cable neatly coiled
> & tied with cable ties, and hanging down behind the desk.  Laying right
> on top of it was an equally neat coil of the remaining 5 feet of the
> 6-foot modem telephone cable!  I separated the two, and the error rate
> dropped somewhat.  Then I made up a new short telephone cable -- just
> long enough to reach from the modem to the telephone outlet, and the
> error rate dropped significantly.  I vaguely remember when I undid
> the rat's nest & coiled those cables -- it roughly corresponds to when
> the error problem became persistent.
>
> I have had the new arrangement for about two months now, & the error
> rate has remained reasonably low, so (knock wood) I feel safe in
> saying that my problem is solved.
>
> I don't know whether the mechanism causing the errors was coupling
> from the phone line to the RS232 cable or vice-versa.  One would
> normally expect the RS232 cable to be more noise-resistent, but one
> also wouldn't expect the RS232 cable to radiate noise when the terminal
> was idle, and I was indeed getting errors while idle.
>
> The connection still isn't rock-solid -- I think I am getting more
> errors than with the original rats-next arrangement.  Some day I will
> get around to making up a short RS232 cable and see if that improves
> things further.

                           Victor S. Miller
                            victor@ibm.com
                            IBM Research, Yorktown Heights