Wikipedia

User:Retired electrician/x2

David Theodore Nelson Williamson
FRS
Born(1923-02-15)15 February 1923
Edinburgh
Died10 May 1992(1992-05-10) (aged 69)
CitizenshipBritish
EducationGeorge Heriot's School
Alma materUniversity of Edinburgh
Employer(s)Ferranti
Molins plc
Known forWilliamson amplifier
Ferranti positioning system
System 24 FMS
SpouseAlexandra J. S. Neilson
Scientific career
FieldsPrecision metalworking machines
Flexible manufacturing systems
High fidelity audio

David Theodore Nelson Williamson (better known as D. T. N. Williamson; 15 February 1923 – 10 May 1992) was a British mechanical engineer, designer of precision metalworking machines, numerical control machines and the pioneer of automated flexible manufacturing systems. Williamson was the author of more than a hundred of patented inventions, including System 24, Britains' first flexible manufacturing system. He held the distinction of being elected Fellow of the Royal Society for engineering work alone, having never being engaged in scientific research, and having no university degree at all.

As an electronics amateur, Williamson is remembered as the creator of the classic Williamson amplifier and, along with Peter Walker, of the first production full-range electrostatic loudspeaker - the Quad ESL.

Early years edit

 
Gilmore Place in Edinburgh. Williamson lived most of his life in a former manse at 65, Gilmore Place[1]

Theo Williamson was born in raised in downtown Edinburgh, to affluent parents of Irish descent. Having contracted tuberculosis in early childhood, Williamson was left physically weakened for the rest of his life. The disease isolated him from his peers in childhood, saved him from conscription during World War II, and later forced premature retirement and emigration to Italy[1].

After completing secondary education at George Heriot's School, Williamson became an undergraduate in Engineering at the University of Edinburgh.[2] Engineering program has not yet been split into separate mechanical, electrical and civil branches; while Theo leaned to electrical and electronic technology, the University historically had a strong bias towards civil engineering.[2] Later, Williamson rated University of Edinburgh education very high, especially practical courses in hydraulics, strength of materials and heat engines.[2] However, in 1943, in the middle of the war, Williamson failed mandatory mathematics exam and dropped out of the University[2]. He would work without a higher education degree until 1971, when Heriot-Watt University awarded him an honorary Doctor of Science degree.[2]

Immediately upon discharge military authorities drafted Williamson, who was not fit for military service, for civilian work at Marconi-Osram Valve (MOV).[2] Williamson initially worked as valve tester at valve development laboratory, but eventually moved to more promising apllications laboratory developing new generation of electronic circuits.[2] There, he developed and tested a side DIY project that became known as the Williamson amplifier.[2] Towards the end of the war, Williamson briefly worked at Clifford Patterson's laboratory of the General Electric Company, which would later become his model for running industrial research.[3] After the war Williamson declined job offers from the Decca Company, returned to Edinburgh and joined Ferranti as a Development Engineer in Kenyon Taylor's Applications Laboratory.[4]

Machine design edit

Ferranti CNC machines edit

 
Ferranti positioning system. Drawing from Williamson's 1953 patent

Post-war Ferranti was a diversified company, shifting from its former industrial market in textile industry to high technology, high added value defense contracts.[5] In the beginning, Theo worked on ultrasonic airspeed indicators and magnetic amplifiers for aircraft catapults[6]. In 1950 he took the lead of a project aiming at improving Ferranti's own manufacturing technology.[6] By this time the company had reached a point when manually-operated precision milling of complex radar waveguides became a critical bottleneck, throttling down whole plant operations.[6] The only solution was replace humans with computer numerical control (CNC). No one in Britain had done it before; experimental technology developed by Gordon S. Brown at the MIT fell short of Ferranti expectation.[6] The company needed precision of no more than 50 μm (0.0002 in), the best American prototypes were an order of magnitude less precise.[6]

Development of the first Ferranti CNC machine lasted over five years. First, Williamson had to create a precise and reliable measurement instrument, suitable for interfacing with a digital computer[6]. Leadscrews of traditional milling machines were not for the purpose, and Williamson arrived at the idea of using diffraction gratings which evolved into the Ferranti positioning system, patented by Williamson[7]. Gratings fabricated by the National Physical Laboratory allowed measuring displacement up to 30 cm (12 in) and could be stacked for displacements up to 1.5 m (5 ft); even longer objects could be measured using stainless steel tape carrier[7]. As of 1995, the Ferranti system remained the industry standard for mearuring displacements with 1 μm precision.[6] Having measurement instruments in place, Williamson tackled positioning problem.[7] Traditional leadscrews, again, were to coarse for the job. Instead, Williamson used ball screw actuators, recently introduced in the United States, and refined it to reduce backlash.[7] General-purpose computers of the period like the Ferranti Mark 1 were too large and too expensive for manufacturing plants, so the Williamson team employed a custom-built, mixed analogue and digital computer with magnetic tape storage that had sufficient speed to run several machine tools simultaneously.[8] Ferranti CNC machines were a market success, not in the least due to Williamson's fame among radio enthusiasts[9]. According to Joseph McGeough, "Ferranti used to send him out to try and sell these machines and people would say ‘are you the Williamson of the Williamson amplifier?’ Once they knew that they would do business with him – his name and reputation was so high. He brought a lot of business."[9].

 
The Trident's S-duct was milled from a single piece of alloy on the Fairey-Ferranti CNC machine[10]

Williamson' own favorite job at Ferranti was the Fairey-Ferranti CNC machine - a giant vertical milling machine for manufacturing jet aircraft spars up to 3 × 10 m in size.[11] For the first time in the world, all slideways were remotely operated via hydraulic drives.[11] Cutting head operated at 8000 revolutions per minute, so that most of heat was removed with swarf, thus evading undesirable thermal expansion of work-in-progress.[10] The machine was completed in 1957 and immediately employed to make tail engine cowlings for the Trident jetliner.[10]

Molins high-speed CNC machines edit

In 1960 Williamson accepted a job offer from Molins plc, a worldwide monopolist maker of cigarette-rolling machinery that had "ran out of technology" and desperately needed first-class engineering expertise.[10] Williamson had no sentiment about tobacco industry, but Molins offered him almost unchecked executive powers, a seat on the board, and a fourfold increase in pay[10] (to more than the salary of the Prime Minister at the time[9]). In ten years at Molins Williamson and his team, recruited from the former staff of D. Napier & Son, increased output of Molins machines from 1800 to 4000 cigarettes a minute, redesigned corporate supply chain, invented and patented quality control instruments.[12] However, most of his tenure at Molins was dedicated to designing and building CNC metalworking machines.[12]

Traditional batch production of steel parts for Molins cigarette machines was time-consuming and expensive.[12] Williamson suggested switching from steel to light alloys that enabled almost unlimited theoretical speeds of metal removal.[12] A numerically controlled, hydraulically driven and liquid cooled machine cutting alloy at 30000 could make desired shapes exceptionally fast.[12] Given small size of typical parts in Molins inventory, it could also be very compact and thus inexpensive.[12] First two experimental machines designed by Williamson cost Molins only 50000 pounds and where deemed a success.[12] Then, Molins built a batch of fifteen such machines for sale; two of these machines, purchased by Aérospatiale, would later make parts for the Concorde.[12] Williamson persuaded the company to diversify from tobacco industry to building metalworking machines, a new market with significantly higher profit margins.[13] Building state-of-the-art machines required exceptional skills and cleanliness on the work floor, so Williamson settled to build a brand new, fully automatic plant.[13] This course of action culminated in his main achievement - System 24 flexible manufacturing system.[13]

System 24 edit

 
Layout of System 24 (from US patent 4621410). Leftmost row: milling machines, rightmost row: manual workbenches, with a workpiece rack and automatic storage and retrieval unit in the center
 
Flexible manufacturing system of similar layout (East Germany, late 1980s). Left: Workpiece rack, center: automatic storage and retrieval unit, right: milling machines

System 24 was a compact, fully automatic metalworking shop designed for non-stop day and night operation.[13][14] After four years of trial and error Williamson formulated optimal three-line configuration of the FMS: a row of CNC machines, a parallel row of manual workbenches for handling workpieces, finished products and tooling, and a linear automatic storage and retrieval unit (AS/RS) in between.[13][15] Internal storage capacity was sufficient for 18 hours operation,[13] thus System 24 could operate 24 hours a day although only one of three daily shifts was fully manned.[13][14] Two other shifts were supervised by a single maintenance operator[13][14] (building the system was easier than persuading British workforce to work night shifts[14]).

The FMS employed two type of twin-spindle, three-axis Molins CNC machines - one for general-purpose milling, another for drilling and boring, and a single-spindle, six-axis machine for milling complex parts.[14] Williamson deliberately sacrificed flexibility for maximum productivity rates.[16] Contrary to existing CNC designs, aluminium workpieces were mounted on compact vertical pallets in front of the machine column - to facilitate swarf removal.[17] Each machine carried a magazine of 14 interchangeable tools, and four similar magazines in reserve and accessible on demand.[17] Tool change was done directly by the spindle swapping tools within the magazine.[17] Condition of tools was checked during each tool change, and during milling via continuous torque monitoring.[18]

Williamson secured support for his project from IBM and the Minister of Technology, obtained 500,000 pound government subsidy, and began building his model System 24 workshop in Detford.[19] The new three-level building had noisy hydraulics and swarf compactors in the basement, machine hall in the ground floor, and the control computer and supervisor stations in the mezzanine.[13] The yet untested shop was presented to general public in 1967 to much acclaim, and then Williamson's fortunes rolled downhill.[20] A change in top management led to a major restructuring; System 24 was earmarked for disposal to clean up Molins balance sheet in anticipation of a public offering.[20] Williamson's health failed; in September 1969 he suffered a esophageal hemorrhage and went through a surgery that, according to Bob Feilden, "proved to be a two-week cliffhanger".[19] While he was convalescing, the board of Molins made the decision to kill System 24.[19] Williamson remained with the company until 1973, when the management decided to quit machine building altogether.[19][21]

At the same time, advent of inexpensive and compact minicomputers allowed Kearney & Trecker, Milacron and White Consolidated to build and sell their own, less costly flexible systems.[21] Japanese competitors followed in 1977 and soon surpassed the Americans.[21] Williamson's patent rights to FMS layout were contested in United States courts, but eventually Molins won. In 1983-1985 the United States Patent and Trademark Office finally issued patents in Williamson's name,[22] with all royalties payable to Molins.[19]

Public activities edit

In 1968 Williamson was elected Fellow of the Royal Society.[23] Promotion of an engineer, who was never engaged in scientific research, never lectured at the university, and had no university degree at all, was an unprecedented event; it reflected expectations of a new industrial revolution that apparently materialized in System 24[9]. Williamson joined the Industrial Activities Committee (IAC) chaired by James Lighthill and in 1972 began campaigning for greater representation of engineers in the Society.[24] He argued that the Society was unnecessarily biased towards academics and fundamental science; only a quarter of the Fellows worked in applied science and engineering, and only 6% worked in the industry.[24] By 1975 the Society agreed to increase the number of fellows, but refused to lower the bar of academic achievement.[24] The opinion that the engineering community will be better served by its own academy prevailed, leading to the creation of the Fellowship of Engineers.[25] When Victor Rothschild and Frederick Dainton suggested a reform of national research financing, Williamson spoke against specific steps proposed by Rothschild, but supported its ultimate goal of "coupling science more effectively to the social and economic needs of the community"[26]. He advised President of the Royal Society Alan Hodgkin "to be constructive about that objective", and Hodgkin did exactly that, agreeing with ulimate goals and strongly defending fundamental university research.[27]

After the shutdown of System 24 Williamson concentrated on the analysis of national economy.[28] He predicted inevitable decline of British manufacturing.[28] The government, wrote Williamson in 1971, was artificially supporting trade balance with stopgap actions and did not address fundamental social and technological flaws. Britain, according to Williamson, had only twenty years left before losing the edge in high technology trade.[29] The economy in general was lagging behind the competition; the industry needed to concentrate on product design and development stages, rather than research; .[29] American or Japanese success stories relied on free competition of fast-growing innovative businesses, but these models did not work in Britain.[29] British banks were not willing to finance businesses that they did not control, British market for industrial services was almost non-existent, and finally there were no national technological development programs that could benefit the industry like the Apollo program in the USA[30]. Thus, argued Williamson, the only remedy left was redistributing investments into products with maximum added value.[30][28]

Williamson was a strong critic of national higher education system: "Engineering education in Britain today teaches only fundamentals, although not always in sufficient depth for modern needs. What it does not teach is how to apply this knowledge to the synthesis of new products or to develop the type of iconoclastic yet creative outlook ... majority of engineering teachers are not creative people, and have little or no understanding of the design process ... they turn out back-room specialists in their own image".[30] To remedy the flaws of higher education, Manufacturing Technology Committee chaired by Williamson proposed institution of Teaching Company Schemes - a series of government-backed partnerships between universities and private corporations.[28] The proposal became one of the most successful partnership models in the UK, with more than 500 Teaching Company Schemes in operation by 1995.[28]

Williamson vocally rejected academic management science that insisted on centralization of product design.[30] He believed that most effective design is created by "family groups or cells" - compact, closely-knit and self-sufficient workgroups.[30] To be effective, these groups should be as independent and removed from staff management as possible.[30] "The more bureaucratic organization and regimentation is applied to R&D facility, the more costly to run it will be and lower will be the output and quality... under no circumstances should design-draughting be provided as a service. The drawing office is an anachronism that cultivates the worst abuses in design."[30]

Amateur electronics edit

Williamson inherited his passion for electronics from his father, a capable engineer and an avid radio amateur. He learnt the skills as his father was building John Scott-Taggart's broadcast receiver designs, and soon began designing and refining his own shortwaver receivers and then illegal transmitters.[1] In 1938-1989 he built his first high-power, three-stage valve amplifier with global negative feedback, achieving tenfold (20 dB) reduction in distortion - a feature not common in commercial products until the 1950s.[31] In 1939-1943 Williamson experimented with then novel concept of dynamic range expansion, leading to a series of letters and a featured article published in Wireless World.[31] All his subsequent work in electronics engineering focused on high fidelity audio reproduction.

In 1944, while working at MOV Applications Laboratory, Williamson obtained access to first-class test and measurement instruments, and settled on building from scratch a complete audio system - a lightweight custom-made phonograph pickup, a low-distortion 20 Watt amplifier and a matching speaker.[2] With support from Marconi management and experts from Decca Records, Williamson managed to attain the threshold of high fidelity,[2] previously materialized only in pre-war experiments at RCA and Western Electric. The Williamson amplifier, which produced 20 Watts with less than 0.1% total harmonic distortion, proved that a well-matched combination of a quality output transformer and global feedback loop is a prerequisite for high fidelity reproduction.[32][33][34] It was not the first amplifier in this class, but only Williamson dared to freely reveal the secrets of the trade.[35][36] Instead of selling his design to corporations, in 1947 he published a detailed description in Wireless World, addressing the DIY community.[37][36] A second series of articles on the improved, revised Williamson amplifier followed up in 1949–1950. A collection of articles from Wireless World was reissued three times in the 1950-s, and once in the 1990-s.[38] Williamson's articles educated DIYers and consumers and set the guidelines to the industry; this framework, outlining the vague subject of high fidelity, remains in force in the 21st century.[34] Richard Sequerra, designer of Marantz 10B and Day-Sequerra tuners, said that "There are very few people who have ever done it really well. I think Frank McIntosh made a wonderful amplifier. Williamson, in England, made a wonderful amplifier. Even the first Leaks were wonderful. The object of these people was not just money, but to reproduce music for the home. The problem today is that profit is the design objective. Music is incidental to it."[39].

The Williamson amplifier was an instant success, copied by thousands of amateurs and businesses all over the world.[40][41] In the early 1950-s it briefly dominated the preassembled hi-fi amplifier market, but soon lost to cheaper, simpler and less demanding pentode-based designs[40][41]. Exact number of units built is unknown, but it was not less than hundreds of thousands[23]. Williamson, still in his twenties, obtained lasting worldwide fame for being the Williamson of the Williamson amplifier[42], although it did not bring any instant material gains. The design, based on the pre-war Quality amplifier by W. T. Cocking, was not patentable.[43][44][32] Williamson and Peter Walker tried manufacturing and selling "authorized versions" of the Williamson amplifier, but lost to cheaper competition.

Another wartime project, a novel magnetic cartridge design, was duly patented and was manufactured since 1948 as the Ferranti ribbon pickup.[45][46]. Less known is the fact that Williamson was Walker's co-author[47] and business partner in the design of the world's first commercially viable electrostatic loudspeaker - the original Quad ESL, manufactured since 1957.[48] At this point in time Williamson firmly believed that it is the speaker, not the amplifier, that is the weakest link in any audio chain, and that the designers should concentrate on speaker technology, rather than inventing yet another "best" amplifier and chasing "merits that the ear can't hear".[49] The ESL survived into the 21st century with mimimal alterations and still excels in low coloration, low distortion, transient response and overall subjective sound quality.[48] According to Newell and Holland, "what Walker and Williamson did ... was to take a step forward that has rarely been equalled in the world of sound reproduction, especially so considering all the technical difficulties which they had to overcome."[48]

Williamson, in a jokeful 1953 interview, admitted that "my ears are catholic ... I like concertos and dance music... I can listen to about two Dixielands a night. But may be that's because I have but two".[50] Like his peers Walker, Ray Dolby or Gerald Briggs of Wharfedale, Williamson was raised and schooled exclusively on classical music. This generation of audio designers, wrote John Linsley Hood, "was almost exclusively concerned with the reproduction, as accurately as possible, of classical music".[51] They sought "ways which helped to enhance the perceived fidelity in the reproduction of classical music, and the accuracy in the rendition of the tone of orchestral instruments".[51] Williamson believed that lifelike reproduction of solo instruments, maybe even small ensembles, was techhically feasible,[52], but insisted that the designer should not chase the folly of recreating a large orchestra in a small room: "I would like to see more common sense brought to bear on sound reproduction at home...".[50]

Retirement. Private life edit

Notes edit

References edit

  1. ^ a b c Feilden 1995, p. 518.
  2. ^ a b c d e f g h i j Feilden 1995, p. 520.
  3. ^ Feilden 1995, pp. 520–521.
  4. ^ Feilden 1995, p. 521.
  5. ^ Feilden 1995, p. 521–522.
  6. ^ a b c d e f g Feilden 1995, p. 522.
  7. ^ a b c d Feilden 1995, p. 523.
  8. ^ Feilden 1995, pp. 523–524.
  9. ^ a b c d Davidson, G. (2011). "Turning up the interest in city's unsung visionary". The Scotsman (19th August).
  10. ^ a b c d e Feilden 1995, p. 525.
  11. ^ a b Feilden 1995, p. 524.
  12. ^ a b c d e f g h Feilden 1995, p. 526.
  13. ^ a b c d e f g h i Feilden 1995, p. 527.
  14. ^ a b c d e Talvage 1987, p. 48.
  15. ^ Talvage 1987, pp. 46–48.
  16. ^ Talvage 1987, pp. 48–49.
  17. ^ a b c Talvage 1987, p. 47.
  18. ^ Talvage 1987, p. 49.
  19. ^ a b c d e Feilden 1995, p. 529.
  20. ^ a b Feilden 1995, p. 528.
  21. ^ a b c Forester 1987, p. 181.
  22. ^ Feilden 1995, p. 532.
  23. ^ a b Feilden 1995, p. 517.
  24. ^ a b c Collins 2015, p. 97.
  25. ^ Collins 2015, pp. 97–100.
  26. ^ Collins 2015, p. 116.
  27. ^ Collins 2015, pp. 116–117.
  28. ^ a b c d e Feilden 1995, p. 530.
  29. ^ a b c "Quality remains key to engineering exports". New Scientist (April 15): 152. 1971.
  30. ^ a b c d e f g Williamson, D. T. N. (1971). "Carving out an engineering niche for Britain". New Scientist (10 June): 618–621.
  31. ^ a b Stinson 2015, p. 14.
  32. ^ a b Electronics Australia 1990, pp. 4–5.
  33. ^ Hood 2006, pp. 148, 163.
  34. ^ a b Stinson 2015, p. 37.
  35. ^ Frankland 1996, pp. 115–116.
  36. ^ a b Stinson 2015, pp. 22, 36.
  37. ^ Crabbe, J.; Atkinson, J. (2009). "John Crabbe: Firebrand". Stereophile (July 14): 4.
  38. ^ Williamson, D. T. N. (1994). The Williamson Amplifier (PDF). Audio Amateur Publications. p. 38. ISBN 9780962419188.
  39. ^ Lander, David; Sequerra, Richard (2009). "Richard Sequerra: Tuning In". Stereophile (April 28).
  40. ^ a b Jones 2003, p. 425.
  41. ^ a b Frankland 1996, p. 115, 117, 119.
  42. ^ Feilden 1995, pp. 517, 520.
  43. ^ Frankland 1996, p. 115.
  44. ^ Stinson 2015, p. 16.
  45. ^ "The RCMF Exhibition" (PDF). Electronic Engineering (April): 115–116. 1948.
  46. ^ "Sound Reproduction. Gramophone equipment" (PDF). Wireless World (April): 135–136. 1948.
  47. ^ GB patent 815978, Peter James Walker & David Theodore Nelson Williamson, "Improvements relating to electrostatic loudspeakers", published 1959-07-08 
  48. ^ a b c Newell & Holland 2012, pp. 378–379.
  49. ^ Wallace & Williamson 1953, p. 32.
  50. ^ a b Wallace & Williamson 1953, p. 106.
  51. ^ a b Hood 2006, p. 229.
  52. ^ Wallace & Williamson 1953, p. 104.

Sources edit

Retrieved from "https://en.wikipedia.org/w/index.php?title=User:Retired_electrician/x2&oldid=1092348690"