[comp.dcom.telecom] Northern Telecom History

Tony Harminc <TONY@mcgill1.bitnet> (06/12/91)

gaarder@anarres.ithaca.ny.us wrote about United Telephone of 
Pannsylvania:

> While I was in college in Carlisle, Pa., United Telephone replaced
> their Kellogg K-60 switch with a North (not Northern) Electric NX-1E
> mega-whizz-bang.  A recent posting mentioned that North has cut its
> business back to just power systems.  The NX-1E may be part of the
> reason why.  This was -- brace yourself -- a computer-controlled
> crossbar switch.  That's right, one foot in the future and one in the
> past.  I wouldn't be surprised to find out that North took an existing
> crossbar design and glommed on computers so they could have a
> computerized switch quick.   ....
 
Interesting.  Northern Telecom also made a computer controlled
crossbar switch called the SP1.  They installed a large number of them
between about 1972 and 1979.  There are still many thousands of lines
of SP1 in Toronto and other Canadian cities.  They also made a four
wire toll version (SP1/4W) and a TOPS option for toll operator (TSPS)
service.  The first TOPS switch went into service in Alaska in about
1974.
 
A little Northern Telecom history is necessary to understand why this
strange beast was ever made.  In the late 1960s, NT was still
essentially doing no R&D of its own.  License agreements with the
Bell System in the USA provided the technology, which was then
manufactured under licence (or occasionally just imported).  NT
manufactured 1ESS systems this way, 100% unchanged from the Western
Electric version as far as I know.  There are about half a dozen large
1ESS switches in Toronto for example - all installed between 1969 and
1972 or 1973.  But the 1ESS was too large - or rather was not cost
effective outside large cities - and there are only a few large cities
in Bell Canada's (parent of NT) operating territory.  Also - and more
important - the legal goings on in the USA made the continuation of
the cosy licensing arrangements with the Bell System impossible.  So
NT decided to develop a switch that would make economic sense in
configurations from small-town to big city sizes.
 
(Now here I speculate; I should make it clear that I am not an NT
insider and all this information comes from published material): the
digital writing was already on the wall.  I am sure that the R&D that
eventually became the SL/1 and then the DMS series was in progress by
1970 or so.  But something was needed that could be delivered soon.
So tried and true crossbar switches were married to a computer.  In
those days everyone designed their own CPU; there were no off-the-shelf 
CPU chips with ready-to-go C compilers.  The SP1 CPU was typical for
its time, wide words (36 bits? - I forget) all made of TTL and
programmed in assembler.  But it worked.
 
Various stuff was added to the SP1 over the few years it lasted;
centrex, hotel and hospital versions, etc. and of course the four wire
version mentioned above.  Although the DMS switches caught up on the
SP1 quite quickly, the time it was being installed was a time of major
replacement of SxS gear and expansion in general, so a large number of
lines (and switches) went into service.  One entire new building in
downtown Toronto (Simcoe Street -- the first new CO building in many
years) was filled with nothing but SP1.
 
(I can't resist an aside on the Simcoe Street building.  This was
built in the early 1970s, and is a very tall, modern looking building
*with no windows* right in the middle of downtown.  If I am conducting
out-of-country visitors around town, I often contrive to pass by the
building and say in an off-hand manner: "that's the headquarters of
the Secret Police".  The building looks forbidding enough that it
sometimes works with gullible visitors.)
 
One amusing thing that gives an idea of the strengths and weaknesses
of the SP1 is that in the late 1970s, when the DMS switches were just
starting to take off, NT discovered that the entire SP1 CPU could be
replaced with the DMS100 CPU running a software emulator which ran
substantially faster than the original!  I think most if not all were
eventually replaced.  This increased the call processing capacity
significantly, and also did wonders for reliability.
 
Former phone phreaks with good ears will easily recognize the crossbar
sounds when calling into an SP1 - clicky sounds quite different from
the silent DMS or the ESS wire-spring stuff.  The call progress tones
are nice, analogue, PTP too.
 
Caveat: all this is from memory.  The dates are roughly right, but in
the very unlikely event that anyone needs correct dates, "recourse
should be had to the literature" :-)
 

Tony H.