johnk@opel.uu.net (John Kennedy) (11/13/89)
A coworker is experiencing a problem while using a 1200-baud dialup
modem, where he is seeing periodic "}" symbols, and characters 0x255
and 0x251.
When I was having a similar problem at another company, the phone
gurus there told me this was a phenomenon known as "bit slippage",
whereby two switches were losing sync on the digital connection
between them.
Several questions:
1) Does the problem I described seem like a fair diagnosis?
2) Where is this "slippage" problem occur - between the subscriber
and the switch, or between the two switches?
3) Can this problem be addressed by doing something with the subscriber's
interface to the local switch?
4) What can the subscriber do about it? Does he have the
right to have the problem corrected? Would ordering a so-called
"conditioned line" either correct the problem or give the subscriber
a right to have the problem corrected?
5) Are the local telco's sympathetic to residential customers' data
problems?
6) Is there someone to ask for in a local office, if the usual
customer assistance people are unable to help?
Thanks,
John Kennedy johnk@opel.uu.uunet
Second Source, Inc.
Annapolis, MDjohnl@uunet.uu.net (John Limpert) (11/17/89)
In article <telecom-v09i0509m03@vector.dallas.tx.us> johnk@opel.UUCP (John Kennedy) writes: >A coworker is experiencing a problem while using a 1200-baud dialup >modem, where he is seeing periodic "}" symbols, and characters 0x255 >and 0x251. >When I was having a similar problem at another company, the phone >gurus there told me this was a phenomenon known as "bit slippage", >whereby two switches were losing sync on the digital connection >between them. I have heard the same explanation of the periodic '}' problem. The '}' is produced when a T1 (1.5 mbps) link loses synchronization and loses/mangles a frame. Each end of the T1 link is supplied with its own clock source. If there is a discrepancy in the clock frequencies the link will lose synchronization periodically. The phone company has a hierarchal timing distribution system that is supposed to provide an accurate clock to the T1 equipment. I have been told that the clock frequency error may be caused by two problems. One, the local clock distribution system fails and the T1 equipment switches to an internal clock source when it detects the failure. It may be a long time before anyone notices the failure and corrects the problem. Alarms only work if someone checks them. Two, someone has installed some T1 equipment without providing an accurate clock source. This can be out of ignorance or unwillingness to pay for a quality timing distribution system. This information is what I have been told by others, please correct me if it is wrong. John Limpert johnl@gronk.UUCP uunet!n3dmc!gronk!johnl
rhg@uunet.uu.net (Richard H. Gumpertz) (11/17/89)
In article <telecom-v09i0509m03@vector.dallas.tx.us> johnk@opel.UUCP (John Kennedy) writes: >X-TELECOM-Digest: volume 9, issue 509, message 3 of 13 >A coworker is experiencing a problem while using a 1200-baud dialup >modem, where he is seeing periodic "}" symbols, and characters 0x255 >and 0x251. >When I was having a similar problem at another company, the phone >gurus there told me this was a phenomenon known as "bit slippage", >whereby two switches were losing sync on the digital connection >between them. I have also seen such problems when working with a PBX. The problem was that the trunk-card in the PBX was set for 600 ohm hybrid termination when 900 ohm was needed. Unfortunately, the manufacturer's testing sequence tested 900 first and then 600 so all trunk-cards were shipped in the 600 ohm position. The manufacturer's engineer said he had been trying to get the test sequence reversed because 900 is generally a better match. =============================================================================== | Richard H. Gumpertz rhg%cpsolv@uunet.uu.NET -or- ...uunet!amgraf!cpsolv!rhg | | Computer Problem Solving, 8905 Mohawk Lane, Leawood, Kansas 66206-1749 | ===============================================================================
bmk@mvuxi.att.com (Bernard Mckeever) (11/20/89)
[Moderator's Note: I am rather confused by this message which arrived here from Mr. McKeever with a different title, and was signed by John Kennedy of Annapolis, MD. I assume McKeever wrote to Kennedy, who then replied to him? But no introduction or cover note came with what follows below. PT] I hope this provides several answers. >Several questions: >1) Does the problem I described seem like a fair diagnosis? Very close. Most likely you are experiencing frame slips on the/a digital facility. A slip happens when both ends of a synchronous [as opposed to asynchronous] facility are not operating at the same average clock rate. If one terminal runs slow the buffer will overflow and a full frame of information will get tossed. If the same terminal is running fast not enough information is available so the last frame is repeated. >2) Where is this "slippage" problem occur - between the subscriber > and the switch, or between the two switches? Several possibilities here. If the trouble is between digital switchs both should have alarms. Slips do not have a noticeable affect on VF so the alarms are shut off. The slips can be caused by one switch free running [stratum 3] and the other timed from a standard. [stratum 1] If the trouble is between the station and the CO it could be a SLC system with a direct digital interface not being looped timed. [My first choice] Other equipment may also cause this problem if the Telco is sloppy with timing distribution. >3) Can this problem be addressed by doing something with the subscriber's > interface to the local switch? No. All you can do is install a modem with excellent error correction. >4) What can the subscriber do about it? Does he have the > right to have the problem corrected? Would ordering a so-called > "conditioned line" either correct the problem or give the subscriber > a right to have the problem corrected? In order, pray, no, yes. >5) Are the local telco's sympathetic to residential customers' data > problems? Not only no, but @$$# NO. Your best bet is to look around for somebody that knows what is going on, is in a position to do something about it THEN BEG. >6) Is there someone to ask for in a local office, if the usual > customer assistance people are unable to help? Try to get the repair supervisor if you have SLC systems to the CO, or the CO analyzer if you don't. The trouble can be any place in the network. Thanks, John Kennedy johnk@opel.uu.uunet Second Source, Inc. Annapolis, MD