Integrated Circuit History
- The history of the development of the integrated circuit - how it was developed, the main names of Noyce, Kilby, Project Tinkertoy, etc.
The history of the integrated circuit is one of the most important stories within the electronics arena.
The history of the integrated circuit shows that the IC developed as a result of the need for very small electronic assemblies.
The transistor had shown the way, now history shows that the direction had been set: engineers and scientists saw the possibilities of much greater levels of miniaturisation.
IC history beginnings
With the transistor well established, people soon started to wonder if several components could be placed on the same piece of semiconductor. If this could be accomplished then considerable improvements in performance and reliability would be obtained in addition to reductions in size.
One of the main driving forces in the history of the integrated circuit, IC came out of the need for improved military equipment. The Second World War had conclusively proved the value of electronics beyond all doubt. Radar had been an outstanding success, and many other new uses had been found for electronic equipment.
One of these was an early computer called Colossus which was developed by the British to help decipher German encrypted messages. It contained over 1500 valves and generated a phenomenal amount of heat. It was the most complicated piece of electronic equipment at the time and it proved to be very successful although somewhat unreliable.
As electronic equipment became more sophisticated and complicated a number of problems arose. Firstly the physical size grew. This was a particular disadvantage for aircraft where size and weight were very important. As a result it limited the complexity of equipment which could be carried in aircraft. The second disadvantage was even more important. As the complexity of the circuitry grew, so the reliability fell. It often fell to a point where it was being maintained for longer than it was in use. This was particularly true of some of the early valve based computers.
Some of these problems were solved to a degree by the use of new construction techniques. Smaller valves enabled the size of equipment to be reduced a little, as did the introduction of printed circuit boards. However the main advantage brought about by the introduction of printed circuit boards was an increase in reliability.
Despite these improvements the basic problems were not solved. Reliability was still too low, and the equipment too large. Then in 1948 the Soviet Union exploded its first nuclear bomb. The U.S.A. saw this as a great threat. It meant that the Soviet Union could easily launch an atomic attack on the U.S.A.. With existing technology the U.S.A. would not be able to detect this until it was too late. Better methods of detecting possible threats were needed, and this required more complicated electronics.
Tinkertoy & IC history
The integrated circuit history shows that one of the first major attempts to solve the problems of size and reliability was started in 1951 when the U.S. Government funded a study. Code named Tinkertoy, it investigated a number of possibilities, many of which are in standard use today.
Within Tinkertoy, double sided and even multi-layer boards were developed, as well as the techniques for making plated through holes on a board. Whilst the transistor may have seemed an obvious candidate for inclusion in the project, it was not used because the technology was very new and unreliable at the time.
Other developments and ideas that were key within the integrated circuit history were beginning to surface. Across the Atlantic in England, Dr G Drummer from the Royal Radar Establishment proposed the idea of building a circuit as a solid block without any interconnecting wires. However this was more of a vision of the future because there were no practical ideas to support it. Nevertheless it was a remarkably accurate prediction of what the future might hold.
A year later in May 1953 the first patent for an integrated circuit was filed by H Johnson working for the radio Corporation of America (RCA). He proposed that all the components for a phase shift oscillator could be contained on a single chip of silicon. He detailed how the individual components could be made, but as the first p-n junction transistors had only just been made the technology did not exist to be able to manufacture it.
IC history moves on a-pace
Meanwhile back in the U.K., Drummer kept working on his idea. In 1957 he placed an order with the research wing of Plessey to investigate methods which could be used to manufacture an IC. This was a key development within the integrated circuit history.
It took some time for work on the project to start properly. In fact it was not until 1959 that work was really under way. By this time it was too late because wok was progressing far more swiftly in the U.S.A..
The key elements were now in place within integrated circuit history for the IC itself to come to fruition.
With the basic foundations in place, the story of the IC is able to move on to the actual invention of the integrated circuit itself.
the invention of the IC required many elements to be in place: the transistor, silicon processing technology, ideas about printed circuits, and many more.
Tinkertoy moves on
By 1957 transistors were becoming far more widespread. They were beginning to find their way into more equipment. Even military equipment which tended to use tried and tested technologies was starting to become transistorised in some areas.
With this change it soon became apparent that transistors could give significant improvements in reliability and size reduction. This caused the U.S. Government to update their Tinkertoy project to include various aspects of semiconductor technology.
Their work was split up so that a number of different companies progressed separately, but followed similar lines of research. One of the companies which had been awarded a contract was Texas Instruments, the first company to produce a silicon transistor, and one of the first to produce field effect transistors. It was to this company that a very gifted young engineer named Jack Kilby came about a year after the project had started.
First integrated circuit invented
It was as a result of a quirk in the company bureaucracy that major advances were made. When Jack Kilby joined Texas he had very little leave entitlement. When the annual company shutdown occurred he offered to work there on his own. This gave him the opportunity to follow many of his own ideas through.
Kilby started by making a number of phase shift oscillators on a single chip of germanium. The circuit was simple but quite sufficient to be able to prove the feasibility of the technology. During the shut-down Kilby made tremendous progress, first deciding the pattern to be made on the germanium substrate and then transferring these onto the semiconductor substrate. Then on 12th September 1958 he succeeded in making the first of his circuits work. Following on from this success he made a further batch to prove the repeatability of the process. Again he was successful and produced a high yield from the circuits he had made.
Similar IC inventions
As the U.S. Government had a number of similar contracts with several companies, it was hardly surprising to find that they were reaching the same conclusions.
Robert Noyce, working for Fairchild reasoned that it was foolish to make a large number of transistors on a chip which was then cut up to give the individual devices.
In the manufacture of equipment these devices were then assembled together. Instead he thought it would be more sense to remove the splitting and reassembly stages. Using this reasoning, Noyce applied his knowledge of transistor production technology to lay the foundations of much of today's IC technology. In view of their complementary work, Kilby and Noyce are jointly credited with the invention of the IC.
More impetus for IC invention
Like many revolutionary ideals the IC was not an immediate success. The idea caught the imagination of many engineers and scientists but the reality of their high cost limited their use to a very small number of specialised applications.
It was not until 1961 that the first ICs were marketed. Even then only two companies: Texas and Fairchild were producing any, and at $120 for a typical IC, it was hardly surprising their use was limited.
Then in 1961 President Kennedy announced his vision for space research saying that America would place a man on the moon by the end of the decade. For this to be achieved vast amounts of money had to be made available to develop the new technology needed. One of the prime areas for research was in electronics. Size weight and reliability were some of the prime requirements. As a result of this new impetus more ICs soon became available, although their cost was still very high.
With the basic IC now in existence, the next stage of the story was in the integrated circuit development.
It had to develop from being a high cost laboratory experiment available for a few niche applications to one where it was available at low cost and for all the electronics arenas.
The development of the integrated circuit to its current state of widespread use took many years and much development.
However costs gradually fell, and usage steadily increased as more products were developed to take advantage of IC technology.
Early development
The early progress in the development of the IC was not easy. The high cost gave an indication of the difficulties which were being encountered. Yield was a major problem. Only a limited amount of accuracy was available with the processes available at the time, and this meant that only a small proportion of the chips worked correctly. The more complicated the chip, the smaller the chance of it working. Even circuits with a few tens of components gave yields of about only 10%.
Most of the IC development in the 1960s was devoted towards increasing the yield. It was recognised that the key to success in this field lay in being able to manufacture ICs economically. This could only be achieved if the percentage of working circuits in a wafer could be significantly increased.
Most of the development and advances were made in the USA because of the amount of money which was available for space research.
Despite this other countries made a number of significant advances. Europe was well up with the field. In the UK a lot of preparatory work had been undertaken by Plessey for the Royal Radar Establishment. Other companies including Ferranti, Standard Telephones and Cables (S.T.C.) and Mullard (now part of Philips, which in turn has changed to NXP) all joined the IC club. Other countries in Europe saw similar interest in these new devices.
Japan, which fast becoming a very major force in world economics saw the significance of semiconductor technology. In most areas of research from the first production transistors to IC technology itself they were only about two years behind the U.S.A.. One of the first Japanese companies to produce ICs was the Nippon Electric Company, NEC which brought its first products to market in 1965.
Realising the vast amounts of research which would be needed to gain a world leadership, five of the largest Japanese IC manufacturers cooperated on a joint research venture with the Government in 1975. This scheme paid enormous dividends placing some of these companies right at the top of the tables for IC sales.
New IC technologies developed
All the early work on IC technology had been undertaken using bipolar technology. Very soon it was found that heat dissipation was the greatest factor limiting the development of the size and complexity of ICs. With the number of components on an IC being packed into a very small area heat problems were many orders of magnitude worse than if the circuit had been built up using discrete components.
Initially, the work was concentrated on finding more efficient ways of removing the heat, but this only gave limited success. It soon became obvious that a more revolutionary approach was needed if integration levels were to rise.
The answer to the way forwards for the integrated circuit development came in the form of a new transistor technology. First manufactured in 1963 the field effect transistor had great advantages in that the gate consumed virtually no current. Also the channel had a relatively low "on" resistance and a high "off" resistance. This made it ideal for digital applications where the current consumption could be reduced by many orders of magnitude.
Texas Instruments were again leading the way and they were the first company to launch an MOS device onto the market in 1966. Their first device was a binary to decimal converter, but many others followed shortly afterwards.
Further integration levels
As MOS technology had largely conquered the problem of heat dissipation, the way lay open for the development of much higher levels of integration.
Progress in this area of integrated circuit development was very rapid. Only a year after Texas launched their first device, Fairchild took the lead by manufacturing a device with over a thousand transistors. The chip was a 256 bit RAM and it was the first major attempt at conquering the dominance of the magnetic core memory which was used in computers at this time.
While it was a milestone in semiconductor technology the device was not a commercial success. The chip was about twice as expensive as the traditional core memory and it did not sell. However it showed the way which semiconductor technology was to progress. Only when 1 kbit RAMs were launched did semiconductor devices start to show an advantage.
As the 1970s progressed MOS technology became the dominant format for ICs. Although linear ICs were gaining in popularity and chips like the famous 741 operational amplifier were introduced, it was MOS technology which dominated the market. Integration levels continued to increase and new ideas started to develop in the minds of IC designers.
