Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!apple!voder!pyramid!csg
From: csg@able (Carl S. Gutekunst)
Newsgroups: comp.dcom.modems
Subject: Re: 19200 bidirectional extension to V.32 bis.
Message-ID: <129998@pyramid.pyramid.com>
Date: 11 Oct 90 17:05:04 GMT
Sender: daemon@pyramid.pyramid.com
Reply-To: csg@able.pyramid.com (Carl S. Gutekunst)
Organization: Pyramid Technology Corp., Mountain View, CA
Lines: 21

Neither twisted-pair nor shielded is a Good Thing if you are going to try to
run RS-232 at high speeds (beyond 20K). Your big enemy is capacitance, both
between adjacent wires (cross talk) and to ground. Nice straight wires are
best. The worst thing in the world that you can do is run twisted pairs with 2
and 3 in a pair, and 15 and 17 in a pair. (I've seen it done.) Black Box sells
some nice low-capacitance cables that, with modern line drivers (not the old
Motorola 1488/1489 pair), you can run up to 48Kbps into 250 feet of cable and
remain within spec. But they are unshielded. 

This is one reason it's not a good idea to run RS-232C beyond its rated speed.
You really do want shielding (especially in the Midwest, where lightning is a
very real problem), but it does generally increase cable capacitance.

Ribbon cable is right out. :-)

I think Lon mentioned that V.35 implementations tend to be marginal. So true.
It's very easy to build a working, reliable RS-232 interface, yet a *lot* of
them get done wrong. V.35 is much more difficult. When done right, it's very
good up to T1 speeds. When done wrong, you might as well go back to RS-232.

<csg>