(217)   Mon 4 Nov 91  2:15
By: John Nagle
To: All
Re: Re: Early ESSs?
St:
------------------------------------------------------------------------------
@PTH 1:340/201.0@Fidonet
From: nagle@netcom.com (John Nagle)
Newsgroups: comp.dcom.telecom
Subject: Re: Early ESSs?
Organization: Netcom - Online Communication Services  (408 241-9760 guest)

The Moderator questions:

> [Moderator's Note: Wait a minute. What about Morris, IL? I thought
> they were first or nearly first back in the sixties.   PAT]

      There were a number of experimental electronic switching systems
before the 1ESS.  Interestingly, an early Bell System experimental
system, in 1953, involved pole-mounted concentrators remotely
controlled by a #5 crossbar at the central office.  Field trials were
conducted in LaGrange, IL, Englewood, NJ, and Freeport, LI.  Further
experimental work was done on distributed switching, but the notion of
active components in outside plant was premature; components were not
yet reliable enough.  So the 1ESS was, like its electromagnetic
predecessors, designed with all the active components in the central
office.

      The system installed in Morris, IL in 1958 was in some ways "more
electronic" than the 1ESS system.  Unlike the 1ESS, which uses reed relay
type devices for the actual call switching, the Morris system used
cold-cathode gas tubes.  So, unlike 1ESS, Morris had no moving parts.
But Morris required special telephone sets, with active components,
because the gas-discharge tubes couldn't handle the usual 86V ringing signal.
Coin lines, PBX lines, and loop testing, all of which use nonstandard
voltages, were not supported at all.

      While electronic, Morris was only partially solid-state.  The
test system had over 2000 vacuum tubes, plus 30,000 gas-discharge tubes.
The test system served only 400 customers, so a big Morris-type system
would have had rather large numbers of vacuum tubes.

      Pictures of the Morris switch show a truly strange-looking system.
The gas-discharge tubes had to be illuminated by fluorescent lamps to
provide enough free electrons so that the tubes would ionize quickly.
So the banks of tubes sat in brightly lit racks with built-in fluorescent
tubes running vertically down the racks.

      After Morris came the 101 ESS, which was a PBX in the 200 line
range.  This, interestingly, used a time-division bus front-ended by
ferreed relays.  But the bus was analog, not digital, using
"pulse-amplitude modulation", intermittently connecting a
sample-and-hold circuit to the analog bus when the time slot went by.

      But after these forays into truly electronic switching, Bell
Labs decided to build the 1ESS around what are essentially big arrays
of reed relays controlled by a computer.  Semiconductors just weren't
ready for the job of physically switching telephone-line levels.  So
the first 1ESS, at Succasunna, (dedicated May 27, 1965), still
switched calls with moving contacts.  And so did all the other 1/1AESS
switches.  (Was some kind of electronic retrofit developed, or are the
1/1AESS switches running today still using fereed relays?)

      It's interesting to think of how things might have developed.
If the distributed concentrator concept had been pursued, the phone
system might look very different today, with much more intelligence in
the outside plant.  We might have ended up with a phone system that
looked more like the ARPANET or Datakit, rather than the central CPU
operating a dumb crosspoint architecture we still see today.

      If the cable TV people start selling dial tone, we may see it
yet.


[Source: A History of Engineering in the Bell System, Switching
Technology, 1925-1975.]


John Nagle

@Path: softwords!news.UVic.CA!ubc-cs!uw-beaver!micro-heart-of-gold.mit.edu!wupost!spool.mu.edu!telecom-request
@Message-ID: <telecom11.885.7@eecs.nwu.edu>
@Date: 4 Nov 91 02:15:36 GMT
@PID: FredMail 1.8-BETA