Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!iuvax!purdue!haven!decuac!netsys!vector!telecom-gateway From: gentry@kcdev.uucp (Art Gentry) Newsgroups: comp.dcom.telecom Subject: Re: AT&T Long Lines Message-ID: Date: 9 Oct 89 20:32:56 GMT Organization: AT&T Kansas City, MO. Lines: 62 Approved: telecom-request@vector.dallas.tx.us In article , gabe@sirius.ctr. columbia.edu (Gabe Wiener) writes: > I was just thinking about the AT&T Long Lines that have been used in > this country for decades. I'm sure all of these questions have > ridiculously simple answers, but here goes anyway. > 1. Over the _really_ long runs, such as through the Rocky Mountains, > or through the deserts of the southwest, how do they prevent > line resistance from degrading the signal to a point where it would > become undetectable? We use repeaters, spaced anywhere from 1 to 30 miles, depending on the type of carrier technology. These repeaters are powered by DC current carried down the same pairs of the transmission. Each direction takes one pair, one for the E->W and the other for W->E (no, I'm sorry, but AT&T does not go N->S or S->N ) Actually, all repeaters will show an East/West transmission, just to keep things easy. On the older technology based repeaters, they were vacumn tube amplifiers; later came solid state amps; and now with digital transmission, they are not repeaters anymore, but rather regenerators. They actually regenerate the signal at each repeater location, rather than just amplify it. This has the immediate advantage of not amplifying the noise base along with the signal. > 2. When one of those lines is damaged out in the middle of nowhere, > and the damage is _inside_ the cable, how do they locate it? > Moreover, how do they splice in a new piece of cable? In other > words, how do they connect up those hundreds of individual lines? > It would be like trying to rewire a spinal cord. Ahhhh, back in the good-ol-days....:-} All the wires within a cable are color coded, in pairs. In larger cables, pairs were grouped into bunches, which in turn, were color coded themselves. So while tedious, it was not overly difficult to match pairs in a splice. As to locating the fault, it's really pretty simple, you just measure the resistance out to the fault, use a formula (which has long since escaped me) to figure out the distance from the test board to the problem. Careful measurments and calculations would usually drop the splicer right on top of the problem. > 3. Are the long-lines used today by AT&T digital or analog? Sprint > obviously is touting their fiber-optics, but what is AT&T doing? > Do they still use the analog long-lines that they've been using for > years? Or do they send the signals over them via a digital encoder? The answers are: we are rapidly converting our entire network over to digital transmission, where we are in that project, I can only guess. I've been out of the toll end of the business for about 10 years now; yes, we still have some analog facilities on-line, but they are being replaced; and yes, we do use digital encoders, crude way to call it, but in fact that's basicly what they do. > Thanks... You're welcome... | R. Arthur Gentry AT&T Communications Kansas City, MO 64106 | | Email: attctc!kcdev!gentry ATTMail: attmail!kc4rtm!gentry | | The UNIX BBS: 816-221-0475 The Bedroom BBS: 816-637-4183 | | $include {std_disclaimer.h} "I will make a quess" - Spock - STIV |