Tidbits About…Time [3/98]

Archives September 24, 2002 admin

Tidbits About…TIME

Posted by Tuan Ngo on March 27, 1998 at 6:03:40:

About milestones in the progress of timekeeping…

  • 3500BC : Egyptian obelisk
  • 1700BC : Mayan calendar
  • 1000BC : Aztec calendar
  • 1000 : Sung Su’s Chinese water clock (100S/day)
  • 1583 : Galileo free pendulum principle discovery (0.01S/day)
  • 1656 : Huygens pendulum clock (<10S/day)1736 : Harrison H1 chronometer tested at sea (1s/day)
  • 1918 : Quartz crystal oscillator developed (1uS/day, 1uS=1S/1000,000)
  • 1948 : First maser atomic clock
  • 1955 : First Cesium atomic clock (10nS/day, 1nS=1S/1000,000,000)
  • : Primary Cesium clock (current) (<100pS/day, 1pS=1/1000,000,000,000)1960 : Quartz frequency standard commercially available (1 S/year)
  • 1964 : Cesium clock commercially available (1s/10,000yrs)
  • 1978 : Global Positioning System 1st satellite launched
  • 1993 : GPS operational


About atomic clock…

  • The first atomic clock is based upon the oscillation of the Nitrogen atom about the plane formed by the 3 atoms of the ammoniac molecule (I believe it is the maser (Microwave Amplification of Stimulated Emission of Radiation) invented by Townes)
  • It was followed by the Cesium atomic beam clock based upon the wavelength of a transition between two hyperfine orbitals of the ground state of the Cesium-133 atom (9,192,631,770 periods)
  • The Rubidium gas cell atomic clock is another source with 10 times less the stability of the Cesium clock but 100 times the stability of TCXO quartz oscillator.


About Quartz based oscillator ….

  • Current time standard frequency is usually 5MHz or 10 MHz, although 1MHz to 100 MHz is also available
  • For wristwatch application they are usually 32768Hz although some “high beat” could reach the MHz range.
  • They are 3 types of quartz oscillator design (in order of increased stability)
  • The “uncompensated” quartz oscillator (1s/day), found on most quartz wristwatch
  • The Temperature Compensated Quartz Oscillator (TCXO,
    0.1S/day), found on some high accuracy quartz movement
  • The Oven controlled quartz oscillator (OCXO, 1S/Yr) and it’s variant the double OCXO where the crystal is in a highly regulated and stable temperature (+/-0.1
    Celsius), not available in a wristwatch for obvious reason.


About time standard….

  • GMT is astronomically based (earth rotation..)
  • UTC (Coordinated Universal Time) is based upon an aggregate of data from timing clocks laboratories throughout the world.
  • TAI (Temps Atomique International or International Atomic Time) is the weighted combination of 11 laboratory Cesium-beam based primary standards.
  • The length of a Second is the same for UTC and TAI.
  • UT1 (Earth time) is synchronous with UTC until 1972. Leap seconds are introduced in UT1 to compensate for earth spin rate slow-down (about 10**-10/Yr)
  • The current accuracy in determining time is +/- 3 nanosecond per day (1 Second in 10 Million years).

Because the way UTC/TAI are defined there is a need to synchronize all the clocks.

About clock synchronization’s…

  • Clocks before the advent of GPS were synchronized through
    LORAN-C, and or similar HF radio stations (WWWH..) or through specialized satellites broadcasting clock.
  • GPS consists of a set of 24 satellites (and 4 spares in orbit) launched by the US for the military purpose of positioning.
  • GPS satellites carry 2 on board Cesium clocks (redundancy) and sent data for the earth receiver to compute location and time.
  • All GPS satellites clocks are managed and synchronized from earth stations to provide accurate positioning
  • 4 satellites are required to compute time and coordinates (x,
    y, z, t). At any given time and place on earth at least 4 satellites can be seen, although up to 12 satellites could be in view under special circumstances.
  • Clock synchronization from GPS data is accurate to 100pS.
  • GLONASS is the Russian version of GPS

About the difficulty of clock synchronization:

  • Time could only be “coordinated” for “inertial frame of reference” i.e. for object in translation at constant speed (a postulate of Einstein’s relativity is that the speed of light is the same for all inertial frame)
  • The earth is NOT a inertial frame because it is rotating. Therefore the speed of light is not constant (Sagnac effect)!
  • A clock moving westward at a slow speed (no relativistic effect) and arriving at it starting point would retard by about 200nS compared to the same clock at rest!
  • Gravitational field affects time.
  • A Earth Centered Inertial (ECI) frame is used, and clocks at REST in that frame can be coordinated.
  • GPS satellites are in motion (eccentric orbit) in the gravitational
    field. Therefore they broadcast data so that GPS receivers can factor in clock correction (correction not required for GLONASS system)
  • GPS clocks readings must be corrected to account for motion and gravitation. GPS clocks are slowed to about 5×10**-10 before launch (gravitational effect)!

About future development…

  • Synchronization accuracy of 50pS
  • Clock accuracy of 10**-16 (1S in a billion years) to 10**-17

I have attempted to give a little insight and summarize some key problems on the science of timekeeping.

All comments and questions are welcome.