Frederic Calland Williams

principal papers

hardware

williams tube (CRT)

software

keywords

see also

related subjects

Biography

Pre-War Years at Manchester and Oxford

Professor Sir F.C. Williams FRS was always known to his friends and collaborators as Freddie Williams, or more often just "F.C.". He was born at Romiley, close to Stockport and just south of the city that he would help to make famous in the computer revolution.

He was educated at Stockport Grammar School and the University of Manchester, in the Department of Engineering, and gained BSc (1932) and MSc (1933) degrees. He then joined a two-year "College Apprentice" course with Metropolitan-Vickers, but left after a year because he was awarded the Ferranti Scholarship by the IEE to do two years research at Oxford University. At Oxford he was awarded a DPhil (1936) for work on circuit and valve noise.

Freddie Williams then took up the post of Assistant Lecturer in his old department at Manchester. During the next few years he made many outstanding contributions to research in electronics, publishing some 20 papers. Two of these were with Professor P.M.S. Blackett on an automatic curve follower for the Hartree Differential Analyser, a famous mechanical calculator constructed at the University of Manchester in the early thirties. In 1939 he left the University - after being distinguished with the award of a Doctor of Science (DSc) degree - to join the Telecommunications Research Establishment (TRE).

The War Years at TRE

At the outbreak of war Freddie Williams, already with an established reputation, had recently joined TRE, which moved to Malvern in 1942.

During the war years Freddie Williams made outstanding contributions to the electronics of radar and other military equipment, producing, amongst other things, the first "operational amplifier". He made a major contribution to the development of IFF systems (Identification Friend or Foe), which enabled radar operators to distinguish between friendly and enemy aircraft, and to AI (Airborne Interception) systems, which provided onboard equipment allowing aircraft to automatically track and intercept other aircraft. A late version, AI mark IX, used his "Velodyne" system, and was widely used. Another important system he made a significant contribution to was Oboe (a precision ground-controlled blind-bombing system).

At Malvern he was renowned for his emphasis on circuit "designability", the development of circuits whose operation can be predicted accurately before they are built. He saw this as a major weapon in attacking the problems of precision, reliability and producibility -- obviously crucial in the war years, especially for circuitry on board aircraft. A characteristic of his circuit design was that it was generally voltage based rather than current based.

Freddie Williams ran a small group (in a room in the cricket pavilion of Malvern College) the main purpose of which was to design electronic circuitry to solve problems great and small met by other groups both within and beyond TRE. He also had a reputation as a prolific generator of ideas, not all of which came to fruition! These would often be the subject of an intense burst of activity, often by a member of his group who had been invited to try it out. Freddie Williams also liked to give flamboyant names to his devices, like the "Phantastron".

The Cathode Ray Tube Storage System (1946-47)

In 1945, and again in 1946, Freddie Williams visited the Radiation Laboratory at the Massachusetts Institute of Technology (MIT) in the USA in connection with their massive 24-volume Radiation Laboratory Series on Electrical Engineering (informally known as the "five-foot-shelf"!); he was a co-editor and contributor to volumes 19, Waveforms, and 20, Electronic Time Measurement. Here he learned of attempts to use cathode ray tubes for data storage. In 1946 he also visited the Moore School of Engineering, home to the ENIAC, where research on cathode ray tubes was also under way. ENIAC was the first general-purpose electronic computer, but it did not have an effective electronic storage, and changing its "program" involved a painful process of reconfiguring the hardware, which could take days.

The general belief was that long-term storage (on a CRT) was difficult if not impossible. On returning to England, Freddie Williams made this the main subject of his research. He started active investigation into the storage of both analog and digital information on a CRT. Storage of analog information could help solve the problem of static objects cluttering the dynamic picture on a radar screen (because). Storage of digital information could solve the problem holding up the development of computers worldwide, i.e. lack of an effective storage mechanism that would work at electronic speeds.

He started by repeating the experiment he had seen for analog storage, which involved trying to maintain a permanent record of analog signals by copying them backwards and forwards between two CRTs. However he found the storage of binary digits more tractable, and by November of 1946 he had demonstrated the use of a (single) CRT to store a single binary digit.

Freddie Williams was appointed to the Chair of Electro-technics at the University of Manchester in December 1946, and so returned to Manchester. TRE were still interested in the digital CRT storage and continued to fund and support his research. They seconded Tom Kilburn, who had worked in his group since 1942, and Arthur Marsh, and allowed them to transfer the equipment for the single bit storage, and continue drawing stores. (Arthur Marsh left after a few months and was replaced in June 1947 by Geoff Tootill. The Electro-technics department on Williams' arrival (soon to be called the Electrical Engineering department) was strongly biased towards heavy engineering, and possessed very few modern valves. So the contribution of electrical components by TRE over the next couple of years was crucial.

The research into CRT storage continued throughout 1947, with Tom Kilburn, who was able to work full time on it, now making a major contribution. They tried a number of different storage mechanisms over the year, and by the Autumn were able to store 2048 bits over a period of a few hours. This general method of CRT storage of binary information became known as the "Williams Tube". It had the advantage over other storage mechanisms being designed at the time (notably the Mercury Acoustic Delay Line) that it was made out of simple standard components, it was compact, and it did not require temperature control or accurately controlled power supplies. Most importantly, it was a true random-access storage device. These properties were of course highly significant in it being the first working storage system.

The Mark 1 (1948-51)

The next step was to build a computer around one or more Williams Tubes, both so that the speed and reliability of the Tube could be tested, and so that (if it proved satisfactory) the aim of the world's computer researchers for a computer with an effective electronic storage for both program and data could be realised. So with Kilburn as the main driving force behind the computer design, the Small-Scale Experimental Machine, the "Baby", was designed and built in 1947 and 1948, proving the effectiveness of both the Williams Tube design and the stored-program computer. It ran its first program in June 1948.

By the autumn of 1948 an expanded team had been set up under Freddie Williams to design and build a usable computer, the Manchester Mark 1, and the government had awarded a contract to Ferranti Ltd. to manufacture a commercial computer based on it. The Manchester Mark 1 was fully operational by around October 1949, and included many new features, including a magnetic drum backing store. By this time the detailed design was already being handed over to Ferranti, to produce a machine with a number of enhancements and improved engineering, the Ferranti Mark 1. The first machine came off the production line early in 1951, the world's first commercially available computer.

Freddie Williams and his team repeated the achievment of 1947, of devising and building the first working Random Access Memory in a computer, by designing and building the first working magnetic backing store attached to a computer, the Manchester Mark 1 drum. The idea had been thought of, in particular by A.D. Booth of Birkbeck College, London, who provided useful input. The gathering of the raw material for the first drum showed Freddie Williams in typical practical free-thinking form. As Tommy Thomas recalls: the drum itself was scrounged from discarded equipment from another department, a motor attached in the department, and the magnetic surface applied by the motorcycle electroplating shop across the road from his office! Freddie Williams took a particular interest in the development of the drum, carrying out with J.C. (Cliff) West the early experimentation on servo-mechanisms to synchronise the drum's rotation with the refresh cycle of the CRT stores.

Patents

Freddie Williams filed the first patent associated with the CRT store in December 1946, a provisional patent for the single-bit store. There were a succession of later patents on the CRT storage (now all with Tom Kilburn's name on as well) and the further patents associated with the Manchester Mark 1.

These patents were managed originally by the Board of Trade and then by the National Research Development Corporation (NRDC) when it was set up in 1949. This was the first major project the NRDC had to deal with, and it laid the foundation for a long and important involvement by the NRDC with universities and industry, and set the pattern for much of its activities in other fields. The NRDC negotiated rights on the Williams Tube with IBM, and defended the invention when rival claims were made on the tube in the USA. The Manchester connection got the NRDC off to a good start, and resulted in substantial revenues in its early years.

But in connection with patents Freddie Williams had two important political battles to fight. Firstly, within the Universities, there was a strong feeling in conservative quarters (especially the Arts?) that private profit from patents was inconsistent with academic learning, and new knowledge should be made freely available. The response to free availability was that in areas of national interest there was a danger without patents that the nation would end up having to spend large amounts of valuable foreign currency to more successful exploiters of the inventions from other countries. Freddie Williams countered the distaste of private profit by politely but firmly drawing a comparison with royalties from books.

The second battle Freddie Williams fought was that universities and their researchers should be entitled to the financial rewards of their inventions, rather than the government just because the government funded the Universities. In practice there were many beneficiaries from the patents of Williams and Kilburn, and the later patents, with the main intent being to channel as much as possible into the further development of computers in the department.

Electrical Engineering Full Time (1952-77)

Freddie Williams was a highly inventive electrical engineer with a wide range of interests. His whole professional life had been devoted to solutions to other people's problems as well as his own interests. The Williams Tube, and its key contribution to early computers worldwide, was just another case of solving other people's problems. Having developed the CRT storage with Tom Kilburn, and having managed the Mark 1 operation (while still making significant contributions), he was happy to leave Tom Kilburn to lead the drive to progress the development of computers beyond the Mark 1, as head of the Computer Group within his Electrical Engineering department.

By the early Fifties his focus had returned fully to general electrical engineering problems. He was however always happy to give assistance to solve circuit and motor problems for the Laboratory. A lot of his effort in the next 25 years was devoted to electric motors, especially variable-speed induction motors and linear induction motors (including much collaboration with E.R. Laithwaite, on his staff from 1950 to 1964). He also developed an automatic transmission for motor vehicles, an experimental version of which he installed in his own car!

When the time came for the Computer Group to break away from the Electrical Engineering department to form the new Computer Science department in 1964 under Tom Kilburn, Freddie Williams was very helpful and supportive (even though he was losing a third of his staff).

Freddie Williams was head of his department until he died in 1977. During his time in charge the department had grown from 5 staff, with a graduate output of 17 students a year, to 23 staff, with 50 students a year graduating. And this of course despite losing the Computer Group to form a new department, now as big as his own. He also made his contribution to the University as a whole with more than one spell as a Faculty Dean, and as a Pro-Vice-Chancellor. And he was always a stout defender of the Engineer as a worthy profession, asking "Why is it laudable and proper to show that a thing can be done, but quite improper to do it?".

Chronology

Honors and awards

Freddie Williams received numerous honors in his life, most notably professionally Fellow of the Royal Society (1950), the Faraday Medal of the I.E.E. (1972) and the Pioneer Award of the I.E.E.E. (1972), and nationally the O.B.E. (1945), the C.B.E. (1961) and a Knight Bachelor (1976).