Jul 2, 2017 - Greenwich closed and the site became part of the National Maritime Museum. Finally, we should not forget that clocks can be synchronised ...
Volume 22 Number 2
July 2017
CLOCK SYNCHRONISATION - PART 2 by Ray Essen In the first part of this article I described some early attempts at synchronising clocks using a number of different methods ranging from underwater cables to radio signals. My story ended in 1960 when Britain and the United States (US) started to coordinate their radio time services and a new system of standard time was introduced. Part 2 contains three more synchronisation experiments that have been largely forgotten but which played a key role in the establishment of today’s international system of coordinated time. Time and Frequency Before returning to the story of clock synchronisation I’d like to clarify what I mean by a ‘radio time signal’ in the context of this article, i.e., before the advent of satellite navigation systems. The BBC ‘six pips’ time signal is just one of many time services available by radio. Historically, time signals have been broadcast on a number of different frequencies that can be received and decoded by equipment ranging from radio-controlled watches costing a few pounds to expensive ‘time servers’ which synchronise time across large computer networks. In the early years, radio time signals were primarily intended to make Greenwich time available to ships at sea and the measurement of frequency was of secondary importance. For this reason, standards of time and frequency were developed independently with different users in mind. Time signals for navigation were the responsibility of Greenwich Observatory and standards of frequency - which satisfied the broader requirements of the radio and electronics industries – were the responsibility of the National Physical Laboratory, Teddington (NPL). NPL has always been an open establishment whereas Greenwich Observatory was under the control of the Admiralty from 1818 until 1965. After that, NPL gradually took responsibility for all time and frequency standards in the UK and, in 1998, the Royal Observatory at Greenwich closed and the site became part of the National Maritime Museum. Finally, we should not forget that clocks can be synchronised without necessarily telling the correct time and many electronic systems with embedded ‘clocks’ have little concern for accurate dates and calendars but they still depend on extremely precise synchronisation.
Omega In Part 1, I mentioned the transatlantic time synchronisation experiments performed by Jack Pierce and Louis Essen using very low frequency (vlf) radio waves broadcast by Rugby Radio Station, operated by the General Post Office (GPO). That research led, via a great many intermediate steps, to the development of the Omega Navigation System for the US Navy. Omega was the first all-weather radio aid to navigation to provide continuous worldwide coverage for aircraft, ships and submarines. It became operational in 1968 and its global coverage made Omega the forerunner of the current generation of sat nav systems, such as GPS (United States), GLONASS (Russia) and Galileo (the European system with full operational capability expected in 2019). So, what has time synchronisation to do with Omega? The answer is that Omega depended on the synchronisation of atomic clocks in each of its eight widely spaced radio transmitters and no other type of clock could have achieved the precision required for a hyperbolic navigation system with global coverage. Time differences (as small as one-tenth of one microsecond) between the arrival-times of signals from two or more Omega transmitters enabled a ship’s position to be determined with an accuracy of one nautical mile anywhere in the world. I have not been able to discover the full extent of Britain’s involvement in the development of Omega beyond the early experiments on clock synchronisation but the following information is taken from a report commissioned by the Royal Society of New Zealand in 1968 when there was considerable discussion on the desirability, or otherwise, of siting an Omega Station in New Zealand. The Royal Aircraft Establishment (RAE) designed and developed receiving equipment in the years before the US Navy authorised full-scale implementation of Omega in 1968 and these receivers were test flown extensively over Europe, the North Atlantic and the Caribbean Sea. Information released by the US Department of State in the 1960s acknowledged that RAE ‘have made and are making significant contributions to the solution of aircraft engineering problems.’ The Royal Navy was also very active in the development and testing of submarinereceiving equipment and operated a vlf transmitter (jointly with RAE) and several receivers in a programme which validated the system concept from a scientific standpoint. Some British companies also showed interest in making Omega receivers but British commercial interests in the mid 1960s decided that there was a need for only one global vlf navigation system and refrained from further development of similar systems until Omega had been thoroughly tested. However, a similar Soviet vlf navaid called Alpha/RSDN-20 became operational in 1968 (but without global reach). The Decca Navigator Company undertook a lot of development work on low frequency hyperbolic navigational aids and one such example was Delrac (DEcca Long Range Area Coverage). The technical similarity between DELRAC and Omega led Decca to successfully sue the US Government in 1976 for infringement of patents. Omega remained in service until 1997 but, long before then, scientists were looking at alternative ways of synchronising clocks using radio waves at far higher frequencies. In the late 1950s, Essen and his colleagues at NPL started to think about using microwaves for clock synchronisation. The Space Race Microwaves are a form of electromagnetic radiation similar to radio waves but with frequencies in the range 0.3 to 300GHz. These ultra-high frequency waves are capable of
carrying large amounts of information and can travel over long distances in a straight line via a chain of line-of-sight relay towers. This makes them very useful for point-to-point communications. However, bridging the Atlantic by this method would mean building a 760km high, microwave relay tower in the middle of the ocean. A solution to this problem was not particularly hard to find because in 1945 the science-fiction writer, Arthur C Clarke, had proposed a satellite communication system in which an artificial satellite performed the role of a microwave relay link in space. Clarke’s dream of communicating by satellite came a step nearer with the dawn of the Space Age on 4 October 1957 when the Soviet Union launched Sputnik 1 into earth orbit. The successful launch shocked and embarrassed the American public, who were convinced that Russian technology lagged well behind that of America, and it forced President Eisenhower to bring forward the planned launch of the first US satellite. In the event, it took almost four months before Explorer Figure 1. Echo 1 postage stamps. Credit: US Postal Service 1 was put into orbit. Explorer’s shape (a spinning cylinder 2m long and 15cm diameter) made it a poor candidate for a microwave reflector but, on 12 August 1960, America launched a pioneer communications satellite, Echo 1, and with it came the possibility for an entirely new approach to time synchronisation, Figure 1. The Anti-Aircraft Gun Echo 1 was large enough to be visible to the naked eye and thousands of people would have seen it move across the night-sky. It was a 30m diameter balloon coated with a thin metallic film and was the world's first inflatable satellite or ‘satelloon.’ The idea behind Echo was simple: put a giant mirror into orbit high above the earth’s surface then bounce microwave signals off its surface from one point on earth to another. A similar experiment had been performed by the US Army Signal Corps in 1946. Using modified wartime radar equipment, Project Diana showed that microwaves could be bounced off the moon and the reflected signals received back on earth. The round-trip took 2.5s. This was too long to be a practical method of synchronising clocks but it provided the inspiration for later space communication techniques. Echo was a passive satellite in the sense that it contained no active electronics (except a radio-location transmitter) but, despite this limitation, Echo was used for a variety of Figure 2. Denis Sutcliffe experiments during its thousands of orbits around the earth. Its success as a simple communications satellite is widely documented but an attempt to use Echo for time synchronisation has been forgotten until I heard a chance remark made by Denis Sutcliffe, a retired scientist who was working at NPL at that time, Figure 2.
Denis told me a story involving a colleague of his, James McA (‘Mac’) Steele, who had a long and distinguished career with the NPL. Steele was born in Scotland on 15 May 1924 and received a degree in electrical engineering from the University of Glasgow in 1945, Figure 3. After a year with the Ministry of Aircraft Production he joined NPL where he was involved in a number of areas of electrical measurements including radio time signals and the development of methods of precise time transfer by satellite. Steele worked alongside Louis Essen and was responsible for NPL radio time signals for many years. Denis picks up the story: ‘Soon after the United States launched the Echo 1 satellite, Steele said, “Well, all we’ve got to do is to get a laser or microwave generator and shoot a pulse of radiation towards the satellite and get it reflected off the balloon’s surface Figure 3. James McA ('Mac') Steele. Credit: NPL and receive it in the States – and vice versa. We need some means of tracking the moving satellite – such as an anti-aircraft gun which is either programmable or some bloke can turn a handle – and follow it across the sky as if it’s an aircraft that you want to shoot down.” Steele wrote a few figures on the back of an envelope and worked out that the energy in the pulse would be enough to receive it in the States.’ Fortunately, NPL had several contacts at Woolwich Arsenal so Steele phoned one of them and explained that he wanted to borrow an anti-aircraft gun for this project. A few weeks went by and Steele forgot all about it until the gate-house phoned to say that an army lorry had arrived ‘pulling a whacking great gun’ and where did he want it? The NPL standard quartz clocks were located in the basement of Bushy House so the driver was directed to take the gun to the front entrance of Bushy House where he reversed the gun to face the building, put the jacks down, unhitched it and went off. Denis explains what happened next: ‘This was in the morning. And at five past two Steele’s phone rang again and it was the Director’s secretary saying, “Do you mind coming over and seeing the Director?” Of course, he knew what it was about so he went over and explained that this had come much quicker than anticipated and if we did this experiment it would be good for everybody. The Director said, “Oh well, okay, but would you mind turning it round so instead of facing my front door it faces the sports-field.” Steele hastily agreed then made the necessary arrangements to move the gun and connect it up to a 3-phase electricity supply. They were getting this organised and, around tea-time in the afternoon, Steele and a couple of others came in all as white as sheets. They’d connected up the 3-phase supply and, as soon as they pressed the On switch, the gun barrel swung around horizontally through 180 degrees and almost decapitated a member of the team. It could have been our worse accident.’ Sutcliffe concluded: ‘The experiment did not get any further and was not pursued. After a year, the gun went back to Woolwich and Steele was very relieved to get rid of it.’ But Steele’s thoughts quickly returned to the rocket men from NASA (America’s National Aeronautics and Space Administration). The day was 10 July 1962 and he has just heard NASA announce that Telstar 1, the first active communication satellite, was finally in orbit.
Telstar Following the success of Echo, Bell Telephone Laboratories developed Telstar 1, an experimental satellite designed to send voice, data and live television signals around the world, Figure 4. But Telstar 1 also made history for another reason: it was used for the first two-way method of satellite time-transfer across the Atlantic. Within a few days of the launch, Steele and Essen had the idea of using Telstar to synchronise the UK and US national time signals. The satellite would provide a direct line-of-sight route for the microwaves and avoid errors caused by ionospheric reflection. Steele sent a cablegram to William Markowitz, head of the Time Department at the US Naval Observatory (USNO) in Washington DC, suggesting the idea. Markowitz recorded what happened next: ‘I received the cable on 25 July 1962. Within one day I obtained the approval of the USNO Superintendent, the Navy Department, NASA Figure 4. Telstar 1 experimental communications and AT&T. I called James the next day and told him that satellite we were ready to start. I was surprised when he told me that he had not yet consulted the GPO (who operated the UK earth station used for communicating with Telstar). This was done, however, and we did conduct the experiment.’ On 25 August, exactly one month after Markowitz received the cable from Steele, the first attempt was made to transmit time signals back-and-forth between the US earth station at Andover, Maine and the UK earth station at Goonhilly Downs, Cornwall, via Telstar, Figure 5. James Steele and Bob Donaldson controlled the time transfer at Goonhilly whilst Denis Sutcliffe and Louis Essen oversaw the operation at the NPL where the UK master atomic clock was located. The initial attempt was not successful but, with help from GPO staff at Goonhilly and by burning some midnight oil, it worked out in the end and the experiment was repeated successfully on 27 August. The results showed that it was possible to synchronise atomic clocks in the UK and US with an accuracy of ±1μs (microsecond) compared with accuracies of 2,000μs previously attainable using vlf radio signals. Time synchronisation between the ground station clocks was extended to the reference clocks at the Figure 5. Goonhilly, Cornwall UK, satellite earth station USNO, Washington DC, and the Royal Greenwich (1962) Observatory, Herstmonceux, by means of local, ground-wave radio transmissions. A comparison of reception times showed that on 27 August the UK time standard was ahead of the US by 2,234μs. This small difference was resolved by mutual agreement. Sutcliffe described the discussions as, ‘A great compromise – there was no politics – we agreed and it was adjusted with nobody noticing. No, it didn’t go before any government committees – the whole lot
would have been out of sync again before anyone came to a decision.’ As a result of this experiment, time signals in the US were advanced by 1ms and time signals in the UK were retarded by 1ms, thus bringing the two time-scales into alignment. As other countries began to use atomic standards they adopted the same time and an international synchronised time service was established. Telstar 1 had a productive but short life of only a few months and went out of service in November 1962. Ironically, the satellite failed after the US Department of Defence conducted a high-altitude nuclear test (known as Starfish Prime) below the inner Van Allen radiation belt. The explosion caused an electromagnetic pulse which was far larger than expected and which may have affected Telstar’s control transistors. Even so, the dead satellite still orbits the earth and its presence is a constant reminder of its contribution to the establishment of an international synchronised time service. Flying Clocks The Telstar time-transfer experiment showed that it was possible to achieve accuracies in time comparisons of about 1μs between the UK and the US. Most other countries, however, had to rely on short-range high-frequency radio signals for the comparison of distant clocks. And this meant that their time comparisons were about a thousand times less accurate. This gap disappeared over the next few years because of an enterprising programme carried out by Hewlett-Packard (HP) between 1964 and 1967. HP launched its first transportable atomic clock (HP 5060A) in early 1964. At the time, it was one of the most expensive development projects ever undertaken by HP and its success was essential to the Company’s reputation and profitability. A few months later, HP embarked on an ambitious programme known as the ‘Flying Clock Experiment’ which, in the words of HP, ‘was undertaken as a service in the spirit of international scientific cooperation and as a further demonstration of HP’s portable caesium chronometers.’ HP atomic clocks were flown to astronomical observatories and government standards’ laboratories around the world where they were used to successfully correlate international time to within one millionth of a second, Figure 6. The travelling clocks were mains operated with a standby battery power supply so could be kept running for the duration of each tour. In 1964, two HP portable atomic clocks were taken to locations in America and Switzerland but a more ambitious programme in the following year meant that time-scales in eleven different countries were correlated to within 1μs. On 25 February 1965, the Figure 6. Louis Essen (r) with Hewlett-Packard HP clock arrived at the NPL where it was correlated Travelling Clock Team (1965) with Essen’s long-beam caesium atomic clock. To mark the occasion, participants in the experiment were presented with a commemorative paperweight mounted with a gold-plated quartzwafer resonator similar to the one inside the HP travelling clock, Figure 7. By 1966, the tour extended over forty-one days and the travelling clocks covered a total of about 100,000km. The programme culminated in 1967 when the clock comparisons included
East Europe, Africa and Australia and HP used its latest portable clock (HP 5061A) to correlate time between fifty-three places in eighteen countries to within 0.1μs. The Flying Clock Experiment was a clever publicity stunt that guaranteed the worldwide success of HP’s range of portable atomic clocks but it should also be seen as a serious piece of research that proved portable clocks could be transported over thousands of miles by air and road with no loss of accuracy. It also meant that timescales maintained at major time-keeping centres around the world could be synchronised with microsecond accuracy. Finally, in a follow-up Flying Clock Experiment performed in October 1971, four HP atomic clocks were flown around the world twice – once eastward and once westward – and the results were claimed to confirm Einstein’s relativity theory. Figure 7. International Time Comparison (1965). Quartz Paperweight presented to L Essen Postscript I have tried to pick out some of the landmarks that led to the creation of a worldwide system of synchronised time. Progress was particularly rapid in the early 1950s when NPL and Greenwich Observatory in Britain established good working relationships with the National Bureau of Standards and the Naval Observatory in the United States. The ensuing collaboration meant that international time synchronisation went from millisecond to microsecond accuracy within a few years and the foundations for our current system of Coordinated Universal Time (UTC) were established. Better clock synchronisation is just as important today as it was fifty years ago. Optical lasers had just been invented in 1960 when the ill-fated Echo time-transfer experiment was being planned; nowadays, laser technology is at the heart of the next generation of optical atomic clocks. A super-fast fibre-optic communications network across Europe is also under construction. When finished, it will connect together a number of major timekeeping laboratories, enabling clocks to be synchronised with much higher accuracy. Finally, the current definition of the second is based on atomic clocks operating at microwave frequencies. A number of laboratory atomic clocks already work at much higher frequencies in the visible part of the electromagnetic spectrum. They promise significant advances in both pure and applied science, which may lead to a future re-definition of the unit of time. References Ray Essen, The Birth of Omega, Transmission Lines Vol. 17, No. 3, December 2012. Denis Sutcliffe, interview conducted by the author, 1 May 1999. Wm Markowitz, private letter to Denis Sutcliffe, 5 January 1979. The Flying Clock Experiment, brochure published by Hewlett-Packard, 1966.
