Voice
Communication in Business Volume 1
Essays on telecommunications,
1969-1980
PART II
Requests for Proposals, Features and Related Topics
Introduction
When interconnect went
into effect in January, 1969, the only competition for the
ubiquitous SXS PBX was a variety of small crossbar and reed-switch
systems (many imported from Japan) and Centrex. Wired logic control,
using relays or solid state circuits copying relays, dominated the
PBX field; No. 1 ESS, just a couple of years after its first
successful installation/field-trial at Succasunna, N.J., had not yet
done much to bring the wonders of stored program control to Centrex.
Although it may be
difficult to remember now, computers were still quite expensive at
the end of the 60s. One had to make very large telephone systems to
get the per-line cost of the control computer down to something
reasonable. When a diode cost $1.25 and a telephone relay, able to
switch 12 circuits simultaneously (12 transfers), cost $1.35, it
took vision, indeed, to inspire faith in stored program control.
Bell Labs came up with
the solution: No. 101 ESS, or EPBX as it had been called informally.
The idea was to have a large computer in a Central Office to control
a number of small PBXs on nearby customer premises via a data link
to each. Unfortunately, because its designers chose to use stored
program control to provide a number of features of questionable
value while omitting many standard features of SXS, the 101 could
hardly be considered a success in the field.
From a scientific point
of view, it also left much to be desired. It served to contrast TRL
logic with the LLL logic used in the No. 1 ESS computer, but the
test became meaningless as others developed Large Scale Integration
and wiped out all forms of logic using individual resistors,
transistors, diodes and the like. However, 101 did serve to
illustrate a very important scientific principle: the NIH factor.
NIH (Not Invented Here)
implies that there in no need to study history or the professional
literature; if we didn't do it here, nobody else could possibly have
invented it anywhere. Most R&D organizations operate on an NIH
basis, this being a direct result of current theories in engineering
education. Bell Labs, of course, was no exception. The switching
matrix in 101 used an idea called "resonant transfer," and many
man-years were invested in its perfection. Unfortunately for BTL,
Gunnar Svala of Ericsson and North Electric had patented the whole
idea, lock, stock and barrel, some years before (US Patent
2,718,621, Sept. 20, 1955.)
It is against this
background that I, as a customer, started looking at PBXs.
[
Top ] [ Next ] [
Table of Contents
] |
