A hundred years of Electronics
We hear so much about
the information age these days that one is apt to think that it has always been with us.
Most people take mobile phones for granted and will tap into their home
computer or visit a cyber cafe to keep in touch with their nearest and dearest.
However, electronics, the key to this information age, started only 100 years
ago and the pathway to our present position has been a long and tortuous but extremely
exciting adventure.
1 Thermionic devices
After Thomas Edison
had invented the light bulb in 1878, John Ambrose Fleming, an English professor
at University College, London, carried out further investigations into metal
filaments in glass bulbs. Eventually he patented (in 1904) a device with two
electrodes in the same glass envelope called a diode and this was to result in
one of the most momentous journeys that mankind has ever made: it was the
evolution of electronics.
In the above diagram
the normal filament of an Edison bulb is labelled as F and the new electrode is
the cylinder, C, which encases the filament. It is possible to connect a wire
to this latter electrode by the terminal, T. As in the case of the Edison light
bulb, the glass bulb, B, is evacuated. The remarkable property of this device
was that, although electrons could flow from the Filament to the terminal T
they could not flow in the opposite direction. This was therefore what is known
as a non-Ohmic device and the term ‘valve’ was soon introduced as an analogy to
valves for controlling water flow. (Note. In USA these thermionic devices are
referred to as ‘tubes’)
An illustration of the
device is shown below.
So, why was this
fairly primitive device such a timely arrival into the 20-century?
In 1896, Marconi first
demonstrated the transmission and detection of radio waves as a means of
communication without wires, i.e. ‘wireless’. The weak link in his system was
the so-called coherer, which had to detect minute electrical signals. Now, the
Fleming diode was perfectly suited to detecting oscillating signals such as
those of a radio signal. The positive part of the wave flows across the device
and the negative part doesn’t ( one can imagine a sine wave with the negative
portions chopped off) and thus the
ensuing direct output could readily be detected with sensitive galvanometers.
With this increase in sensitivity the distance of wireless communication was
vastly increased and the communications era was born. It was not long after
Fleming had constructed the diode that a third electrode was added to the valve
and this produced the so called triode (Lee de Forest, 1906 ). This device was
very versatile in that it could amplify electrical signals and so became the
work-horse of early electronics. Later refinements give rise to the tetrode,
three electrodes and the filament ( Hull 1919) and yet later the pentode (Jobst
and Telegen 1926; a fuller account of all these so called thermionic devices is
given on the web site http://www.ginko.de). It
is fair to say that in the late 1950s, i.e. more than half way through the 100-year
period, the valve reigned supreme. There had been massive developments of
electronic components during the World War II years and quality sound
reproduction, radio, and TV was enjoyed by most of the population in the
post-war years.
An illustration of a
quality stereo amplifier from the late 50s is given below:
Despite the many
successes of thermionic valves there was one field of studies where problems
were arising. In the construction of digital computers a prodigious number of
valves were necessary to provide storage and data manipulation. The ENIAC
computer (initials stand for Electronic Numerical Integrator and Computer and
the computer was unveiled in USA in 1946) used 18,000 valves and these valves
failed at a rate of approximately one every 10 minutes making successful
operation somewhat intermittent. The Mark I computer developed by the Ferranti
Company in UK had similar problems so the development of such computers into a
Home Computer was a distant dream.
To pick up the story
of computer development it is necessary to back track to the turn of the
century once more.
2 Discrete solid-state devices
In the early days of
broadcasting a point metal contact pressed onto a semiconductor crystal, lead
sulphide, was used as a detector for wireless waves. A fine wire was required
to make the metal contact and it was likened to a ‘cat’s whisker’ so a cat’s
whisker radio was a gadget to be had at the time. The crystal detector was more
sensitive then the coherer but much less reliable and, therefore, when the
Fleming diode became available this detector (and the coherer) was replaced by
the thermionic diode. Whereas the basic theory of electron flow in evacuated
valves was well understood at the turn of the 20th century, there
was no adequate theory to show how electrons moved in solids and, in
particular, semiconductors. Thus the cat’s whisker diode development could not
proceed with confidence and, in retrospect, it has meant that a large part of
the 100 years of electronics has been devoted to the technology of valves which
is largely irrelevant in modern day electronics.
Eventually, the
Quantum Theory of solids was developed in the 1930s and this now opened the
route to the development of what is known as solid-state electronics. The point
contact transistor was announced to the world by Bardeen and Brattain in 1947
and is shown diagrammatically below:
An illustration of the
above device is given below;
The two critical
points about the transistor were that it was light and robust and
that it did not require a heated filament to produce electrons and therefore consumed
much less power then valves. The point contact device, shown above, was
a little reminiscent of the cat’s whisker diode and was not 100% reliable.
However, the technology of diffusing impurity atoms into semiconductors was
soon developed and a third advantage could be added to characterise transistors
– reliability. Planar diffusion processes, in which diffusion
occurs at only one surface, helped to create many different forms of the
transistor.
( Incidentally,
transistor circuits operative with low voltage supplies ~5 to10V compared the
valve circuits which require ~150 to 250V).
Using many of the
design ideas introduced in the valve days, transistor circuits using discrete
components (i.e. single resistors, capacitors, diodes, transistors, etc)
developed at a rapid pace and, radio receivers, once bulky and heavy, could be
manufactured as lightweight portable items. This dramatic reduction in weight
was partly due to the compactness of the circuit layout on printed circuit
board as compared to the aluminium ‘chassis’ (illustrated as the base of the
stereo amplifier given above) used for mounting valves and also due to the low
power requirement of the transistor circuit, typically, less than one Watt
which could be supplied by, say, four size AA, 1.5 Volt batteries.
3 Integrated circuits
In little more than
ten years after the transistor had been announced, Kilby (1959) filed a patent
for the production of integrated circuits (ICs). Using the planar process, all
manner of components (diodes, resistors, transistors, etc) could be fabricated
adjacent to each other on the same piece of semiconductor with masks to define
the areas to be diffused. Then, by interconnection through a metallization
procedure, a complete circuit could be fabricated on a semiconductor wafer as
shown in the following illustration.
Each small element on
the semiconductor slice (usually silicon – hence the term silicon chip) is a
complete circuit and after cutting and mounting on a suitable terminal base is
ready for use.
The dramatic size
reduction, which accompanied transistor circuits, was dwarfed by the IC
revolution and whole instruments were reduced to a single chip. For instance,
the author used a 10 MHz frequency counter (Marconi Type TF1345), which was a
product from the ‘valve days’, and the complete functionality of this Marconi
instrument was replaced by a 28 pin counter chip, type IC 7216.
In the early days of
IC development it was noted by Gordon E. Moore, director at Fairchild
Incorporated, that the complexity of IC’s doubled every year. His predictions,
now blessed with the name Moore’s Law, have proved to be remarkably correct
over the decades which have followed.
So, electronics in its
integrated form was now in fast forward mode and all branches of the subject
were to benefit from this rapidly changing technology. In particular, the
manufacture of computers made rapid advances. In 1960 the Minicomputer PDP1 was
introduced by Digital Equipment Incorporated and it was to be followed rapidly
by other PDP models. The PDP8 was the first minicomputer to sell for less than
£10,000 and therefore a computer for the home was beginning to come into sight.
The first microprocessor, which was basically a whole computer on a chip,
(model number 4004), was launched by Intel in 1969 and it is the microprocessor
that has brought computers into our homes.
From the 1970s to the
present date one has witnessed remarkable developments in electronics with
scientific, military and commercial interests driving the industry forward. One
cannot possibly describe all the facets of this huge industry or determine the
extent to which the silicon chip has changed people’s lives so only a brief
examination will be made into one topic area, namely, that of home computers.
Perhaps the division
of the above thirty-year period into decades will help in our telling of the
home computer story.
The 1970s were very
much involved with developments of the microprocessor families. The Intel 4004
had a 4-bit instruction and data word length and therefore had limitations in
data manipulation and memory addressing. By 1974 Intel had introduced the 8-bit
processor 8008 but there were other companies in the semiconductor processing
business; Fairchild, Texas, RCA, National Semiconductors, Zilog and Mostek who
were all producing microprocessor chips. In the late1970s these chips were
finding their way into computers, for example, Apple, TRS80, Commodore PET,
Sinclair ZX80/81 and Acorn BBC computers to name but a few.
Features of the
earlier transistor technology, such as small size, low power, good reliability,
were translated to integrated circuit technology but advantages of low cost
became evident, also. Since the fabrication of an IC chip is completely
automated, very complex circuits could be produced at low cost provided the
quality control of processing was of a suitable standard.
In the 1980’s IBM
joined in the race. IBM, the world’s largest computer company of the day,
entered the desk top computer market in a somewhat half-hearted way, possibly,
believing that it would remain only a hobbyist’s fad. The IBM Personal Computer
(PC) was launched in 1981 and was based on the Intel 8086 processor. Using a
‘ready-made’ processor was a serious departure from IBM’s normal strategy in
that it would normally develop a critical component, such as a processor, in
house so that design security would be maintained. Not only did IBM ‘farm out’
manufacture of the processor but it commissioned Bill Gates of Microsoft to
develop an operating system called MS-DOS ( Microsoft Disk Operating System)
for the PC. A year later in 1982 IBM launched an improved model called the IBM
Personal Computer-AT (Advanced Technology) and published many details of this
new design. It was only a matter of time before IBM clones appeared since any
company could build the computer with the circuit details provided by IBM and
purchase the processor chip and software directly from Intel and Microsoft,
respectively.
The benefit to the
consumer was the availability of a sophisticated cloned PC computer at a
fraction of the cost which IBM would have charged. The PC home computer was to
dominate the world from this moment in history.
In the 1990’s the
story of home computing has broken all limits of excellence. Data links with
the Universal Serial Bus (USB) and modems has enabled the home computer to be
connected to the world in a manner impossible to envisage 100 years ago when
Fleming was developing his diode. Indeed, the decade has given rise to
exceptional developments in processor capability and software provision and,
unquestionably, the story of the PC, as it links into other media areas, has
much further go.
There is no doubt that
the modern PC has impressed today’s generation with its versatility, speed and
low cost but another device will be mentioned to show what a varied subject electronics
has become.
Zarlink Semiconductors
has recently introduced their M2A capsule endoscope. This pill-sized capsule,
which passes naturally through the digestive tract, incorporates a miniature
camera, LED-based flash, Radio Frequency transmitter chip and two batteries.
During the course of its passage, the camera takes two high-resolution colour
images per second of the digestive tract. These are relayed to a data recorder
in a belt worn by the patient who can carry out his/her normal daily routine.
The M2A is, very much,
an example of extreme circuit miniaturization operating at such small power
levels and will be of great benefit to medicine.
It is, however, just
one device amongst thousands that are being manufactured in the electronics
industry of today. With such a picture of the recent past, the future can only
lead to even more exciting times, so,
‘Here’s to the next 100 years’
Frank
Thompson, November 04
A fuller version of this article
is published in the journal Physics Education, Volume 40, Number 3, Page 252–256.