🚀 go-pugleaf

In article <239@westmark.UUCP>, dave@westmark.UUCP (Dave Levenson) writes:
> In article <2799@ttrdc.UUCP>, levy@ttrdc.UUCP (Daniel R. Levy) writes:
> > In article <407880.880706.KFL@AI.AI.MIT.EDU>, KFL@AI.AI.MIT.EDU ("Keith F. Lynch") writes:
> > # > From: "Benjamin I. Goldfarb" <goldfarb%hcx9%ucf.CSNET@RELAY.CS.NET>
> > # 
> > # > OH YEAH?  More likely the "UUUUUUUUUUU" was an erroneously
> > # > initiated test mode in the modem in question.  Racal-Vadics are
> > # > notorious for this ...
> 
> The standard data-transfer mode for an AT&T 212-compatible 1200 bps
> full-duplex modem calls for scrambling (by the sending modem) and
> unscrambling (by the receiving modem) to prevent long strings of
> identical bits from being sent over the communication line.  The
The 212 scrambler is a linear feedback shift register designed to prevent
long runs of 1s or 0s on the channel and the attendant signal power density
problems on the telephone channel and modem synchronization problems.  One
of the test modes of the modem is remote digital loopback, where the remote
modem loops back its output and retransmits the signal for end-to-end
testing.  The way you put the remote modem into digital loopback is by
sending it a constant string (of 0 or 1, I forget) ON THE CHANNEL -
something that a self respecting scrambler would generate infrequently on
its own.  Thus, a modem could be fooled into thinking that it had been
asked to go into remote digital loopback if it heard a constant string of
channel symbols.  The symptoms of false remote digital loopback would be a
10101010 sequence emitted from the modem - which looks like UUUUUU.  This
falsing could occur (1) if the input to the sending modem happened to
scramble as a constant output or (2) if channel conditions caused a
constant string to be detected.

Bruce McNair
AT&T Bell Labs
Holmdel, NJ