Forgot your password?
typodupeerror
Science Technology

NIST Ytterbium Atomic Clocks Set Record For Stability 85

Posted by timothy
from the old-faithful-is-an-ambling-slacker dept.
New submitter bryanandaimee writes "An optical lattice clock like the one discussed earlier on Slashdot has broken the stability record. Comparing two OLC's using trapped atoms of Ytterbium, the stability of the clocks was measured to 2 parts per quintillion (10^18). While the previously reported OLC used strontium, these clocks, built by another group, use Ytterbium. Interestingly, while the stability of the clocks is now the best in the world, the accuracy has yet to be measured."
This discussion has been archived. No new comments can be posted.

NIST Ytterbium Atomic Clocks Set Record For Stability

Comments Filter:
  • by elfprince13 (1521333) on Saturday August 24, 2013 @12:40AM (#44662305) Homepage
    They measured the stability but not the accuracy? Aren't both essentially frequency measurements? Can someone explain what data you would collect that would allow you to determine one but not the other?
    • by vomitology (2780489) on Saturday August 24, 2013 @12:44AM (#44662317)
      Time. It's a big ball of wibbly-wobbly timey-wimey stuff.
    • by Anonymous Coward on Saturday August 24, 2013 @01:13AM (#44662389)

      I guess accuracy here refers to the cycle-to-cycle jitter. That is, how much the clock edge position varies when looking at single cycles.

      I guess stability here refers to wandering. That is, how much the single cycle error accumulates when looking at e.g. 1 million cycles.

      • by AmiMoJo (196126) * <mojo@NOspAm.world3.net> on Saturday August 24, 2013 @07:57AM (#44663177) Homepage

        I work with atomic clocks for a living. You are on the right track.

        Stability is the amount of variation over time. Clock frequency changes as parts age, temperature varies and so forth. Sometimes the variation is predictable and can be compensated for, sometimes not.

        Accuracy is how close the frequency is to the specified one. A cheap ebay atomic clock can do 1 second per century.

        There is also jitter which is the cycle to cycle variation of the output, which could be a square or sine wave. For some applications it matters.

        • by oldhack (1037484)
          What is the reference clock against which other clocks are measured?
          • by AmiMoJo (196126) * <mojo@NOspAm.world3.net> on Saturday August 24, 2013 @01:29PM (#44665051) Homepage

            What is the reference clock against which other clocks are measured?

            Good question. For things like drift you can compare a new and an old clock, or compare one in a cold environment to one in a cool environment. Actually, most atomic clocks have a heater built in that maintains a stable temperature internally.

            Even just comparing two clocks of the same age under the same conditions is useful, because they will never be at exactly the same frequency so will eventually drift apart.

            Beyond that it's a question of determining what the uncertainties are using physics. In other words it's a mathematical proof, rather than the result of an actual measurement. Obviously measurements are made to make sure the clock is close to other known accurate clocks, but the ultimate determination of frequency accuracy is a paper exercise.

            • by oldhack (1037484)

              Thanks for the reply.

              Beyond that it's a question of determining what the uncertainties are using physics. In other words it's a mathematical proof, rather than the result of an actual measurement.

              In an overly simplified nutshell, we are averaging out whole lot of stuff with circular dependencies, because, given the nature of the concept that is time, that has been the best we can do, and it seems to hold up reasonably well?

              • by AmiMoJo (196126) *

                The laws of physics that we think are probably correct state that certain things take a certain amount of time. We based our units of time off these things. Therefore we know that under perfect conditions a given clock should perform in a certain way. What we then have to do is determine how close to perfect the conditions are.

                Atomic clocks are basically an engineering problem. You have to build a machine that measures something difficult to measure and then turns that measurement into some kind of usable o

                • by oldhack (1037484)

                  The laws of physics that we think are probably correct state that certain things take a certain amount of time. We based our units of time off these things

                  And those laws of physics are, in turn, premised on the concept of time, which is non too simple or clearly defined, measured with clocks that are constructed based on the physics, which, in turn..

                  No, I'm not knocking it. In the end, we accept it (time, clock accuracy, the whole physics) as long as it works when thrown up against the wall called natu

                  • by AmiMoJo (196126) *

                    Your welcome. I'm not sure what you mean about the concept of time not being clearly defined. It is, conceptually by general relativity and physically by the things that atomic clocks try to measure.

          • The ultimate reference is the SI second standard that's based on the difference in energy levels associated with a particular ground-state hyperfine transition in the caesium-133 atom, approximately 9.192 631 770 GHz.

            A clock is 'accurate' to the extent that its long-term stability can be traced to the SI second. It is 'stable' if its stability over a given time interval (which must be specified for the term 'stability' to have any meaning) is consistent between intervals.

            • by oldhack (1037484)

              Which leads me to a question:

              Does time define cessium electron's frequency, or does cessium electron defines what time is?

              • I'm not qualified to explain the details, but it comes down to the fact that we had to pick an easily-reproducible definition of the second, and at the time, the ground-state Cs-133 hyperfine transition was the one that made sense.

                Today there are more stable clock designs based on other quantum transitions but the caesium standard serves well enough for almost all purposes that it's not worth the trouble to change it.

                There is no such thing as "perfect time," or perfect knowledge of just what time it is, but

      • But that leaves my original question: how do you collect data in a way that gives one but not the other? Both require you to record the frequency over time, do they not?
        • by msauve (701917)
          If you have a clock that's exactly, and consistently, 1 minute per day slow, you have a precise, but inaccurate clock. If you have a clock that might be off +- 1 minute per day, but the average makes it never deviate more than 1 minute from correct time, you have an accurate, but imprecise one.
          • Still missing the question here, but you hit it with the standards-related info in your other response. I'm not at all fuzzy on the difference between accuracy and precision. My question was based on the observation that both quantities require you to have sampled the waveform as a time-series in order to calculate them. The accuracy requires you to measure the average number of peaks per unit time, and the stability requires you to measure the variance in the time between peaks. Both require you to know *w
            • by ceoyoyo (59147)

              Except it's highly unlikely they collect a time series. The summary suggests they have two clocks and they're comparing the two of them. Relative drift between the two indicates instability, but it doesn't say anything about how accurate they are.

              • To compare drift at all, you have to be collecting a time series of some sort. If I draw two lines on the bank of a river, and say "this line is data showing that a stick floated from this point to the far end of the line, and the other is data showing how far another stick floated", there really are no meaningful comparisons you can make without knowing something about the respective times over which those processes occurred.
                • No, you don't. The easiest way is just to sum peaks, or rising edges, or whatever. Come back tomorrow and compare the totals for a bunch of clocks. There's your stability measurement with one number per clock. No time series.

                  Collecting a time series is hard, and probably impossible in this case. You have to be able to sample fast enough, and your sampling must be more consistent and more accurate than what you're measuring.

                  • Sounds like you're talking amplitude, not frequency. Stability (and accuracy) are both tied to duration. To quote TFA, "Stability can be thought of as how precisely the duration of each tick matches every other tick." Summing the peaks doesn't tell you how much time passed between them.
                    • by ceoyoyo (59147)

                      I'm not sure what you mean by amplitude. I doubt very much I'm talking about it.

                      Think about how you're going to measure the time between peaks without a timer that is more accurate than the one you're testing. You can't. One solution is to measure the relative difference between the time indicated by one clock, and the time indicated by another. You could do this simply by counting peaks for a period of time, or by doing something fancier such as looking at the relative phase of the two outputs. Both m

                    • Ah, you're using "summing" as a synonym for counting? That helps me understand your previous post a little better, even if it still strikes me as an odd word choice. And yes, the quote appears to be correct (though imprecise as might be expected from a press release), because a little bit more research brings up Allan-variance [slashdot.org]. You may also be interested in this reply from another well-qualified poster [slashdot.org].
                    • by ceoyoyo (59147)

                      Counting is a special case of summing when every entry is a one. The mathematical operation is summation. To count a bunch of digital clock pulses you'd use an adder circuit. To do the equivalent in the analog world you use an integrator.

    • by AdamHaun (43173) on Saturday August 24, 2013 @01:27AM (#44662415) Journal

      Accuracy measures how close the frequency is to the target, on average. Stability measures how the frequency drifts over time (and temperature, etc.). Accuracy is more of an absolute measurement while stability is more of a relative measurement. From the article:

      The ticks of any atomic clock must be averaged for some period to provide the best results. One key benefit of the very high stability of the ytterbium clocks is that precise results can be achieved very quickly. For example, the current U.S. civilian time standard, the NIST-F1 cesium fountain clock, must be averaged for about 400,000 seconds (about five days) to achieve its best performance. The new ytterbium clocks achieve that same result in about one second of averaging time.

      and

      [U.S. civilian standard cesium reference clock] NIST-F1's performance is described in terms of accuracy, which refers to how closely the clock realizes the cesium atom's known frequency, or natural vibration rate. Accuracy is crucial for time measurements that must be traced to a primary standard. NIST scientists plan to measure the accuracy of the ytterbium clocks in the near future, and the accuracy of other high performance optical atomic clocks is under study at NIST and JILA.

      So it sounds like accuracy is defined in terms of how well the clock reproduces the ideal frequency of the physical process it's based on. Hopefully there's a physicist or two around who can give us the exciting details.

      • This is helpful, but was what I'd already pulled up via Google, and doesn't answer my question, per se. Both sound as if they are properties of the frequency distribution, which is why I was asking what data would give you one but not the other (I have a physics degree, so I actually "get" how the clock works. I'm interested in the data analysis).
        • by msauve (701917)
          Caesium clocks, upon which the second is defined, differ in frequency, due to gravity, electromagnetic fields, etc. (the definition of the second is based on a Cs atom at 0K, at rest, which isn't realizable in practice). Because of this, official time is only known after the fact, based on a weighted average of about 400 Cs clocks [bipm.org].

          If changes in the environment have less effect on a Y clock, or the physical processes upon which it's based have less random variation, then in might be more stable. Stability c
          • So in other words, they did take the correct set of measurements for calculating both stability and accuracy, but the number crunching on the latter takes longer because it has to be compared against reference time?
            • by msauve (701917)
              Yep. If you go to the source [nist.gov], they say "NIST scientists plan to measure the accuracy of the ytterbium clocks in the near future..."
              • So it seems then that this conversation has more to do with lay misuse of the term "measured" than with the experimental process. They *measured* the waveform. They're *calculating* both the accuracy and the stability based on those measurements.
                • by sFurbo (1361249)
                  They probably *measured* a voltage, and then *calculated* the waveform, from which they *calculated* the stability.

                  To discuss what is *really* being measured gets absurd pretty quickly with modern instrumentation.
    • by maxwell demon (590494) on Saturday August 24, 2013 @02:04AM (#44662523) Journal

      Imagine a mechanical clock that has such a heavy minute hand that it goes much faster down than up. But it returns after exactly one hour, and even after ten years, it shows the full hour accurate to the second. This is a clock with very low stability, but quite high accuracy (for a mechanical clock, for an atomic clock that would of course still be terrible accuracy).

      On the other hand, imagine a mechanic clock which doesn't have this stability problem, but the pendulum is not compensated for temperature changes. That is, when it gets warmer, the pendulum gets slightly longer and the clock goes slightly slower. Now temperature doesn't change very fast, so you'll not notice the effect in a short time span. However over time, the clock will drift away from the correct time, unless you manually correct it. This is clock with good stability, but not so good accuracy.

      • Thanks for the explanation, but that's the part I'd more or less already figured out. It still leaves my question unanswered, what *data* would you collect that allows you to determine one but not the other? A couple of the other posts had the same issue.
        • Well, in one case you let it run a relatively short time and see how well it matches another clock, in the other case you let it run a very long time. compensate for short-time errors, and again check how well it matches another clock.

          Clearly the accuracy measurement needs longer than the stability measurement (and moreover you need to know the stability in order to compensate for your accuracy measurement).

        • The data you gather to determine accuracy is tedious, thrilling, gob-smacking, humbling, and generally fun. Here's what we measured to determine the accuracy (uncertainty) of NIST-7. http://tf.boulder.nist.gov/general/pdf/1497.pdf [nist.gov]
          • *That's* what I was looking for! Thanks for the good reading material :)
            • My pleasure. The dominant uncertainties are summarized in the last couple pages. If you would like to know the main errors in ytterbium let me know. I can ask Dr. Oates next week.
              • I definitely appreciate the offer, but the article in your previous post is enough to sate my curiosity for now :) Not looking to plunge headfirst into time/frequency metrology so much as I wanted at least one good article that wasn't dumbed down for lay audiences.
    • Exactly, all we have now is a clock that is very stable though 5 minutes ahead :-S
    • by unixisc (2429386)
      How do you measure the accuracy of something against a presumably less accurate standard? If the Ytterbium clock shows anything different from the standard NIST clock that uses, what, Caesium?, wouldn't it indicate that Ytterbium is less accurate, not more?
    • by c0lo (1497653)

      They measured the stability but not the accuracy? Aren't both essentially frequency measurements? Can someone explain what data you would collect that would allow you to determine one but not the other?

      Preliminaries [wikipedia.org]: stability (reproducibility) determines precision - in this case, how repeatable is the result obtained by measuring a defined duration. Accuracy would be how close is the measured value to the real one.
      As long as the the stability is high, one can organise different experiments/measurements to determine the accuracy.

      Assume you measure the difference between the time required for a laser beam to bounce from a mirror on the Moon. If your timer is not accurate, you can't be sure on how close th

    • by Chemisor (97276)

      A stopped clock is perfectly stable, but is only accurate twice a day.

    • by sFurbo (1361249)
      Without having read TFA, stability has to do with the frequency relative to the frequency of the same clock earlier, while accuracy has to do with the frequency relative to an external reference. Stability compares two instances of (nearly) the same frequency, which is a hell of a lot easier than comparing vastly different frequencies.
    • Hi, I work at NIST with this stuff. Stability and accuracy are often misunderstood. The rate of any clock is always measured against another clock. Suppose you measure the time difference between two clocks over some interval. The frequency is the slope of those data. The STABILITY is measured by how straight the lines are, i.e. how predictable the clocks are. ACCURACY is more difficult to measure. A clock is first a device that produces a frequency that realizes a principle of physics, e.g. inertia
  • by Anonymous Coward

    has great stability.

  • by Gravis Zero (934156) on Saturday August 24, 2013 @01:55AM (#44662495)

    so... it has come to this.

  • by fahrbot-bot (874524) on Saturday August 24, 2013 @02:30AM (#44662571)

    From the following newspaper article: http://www.latimes.com/science/sciencenow/la-sci-sn-atomic-clock-stability-nist-20130822,0,6785801.story [latimes.com]

    The ytterbium optical lattice clocks at the [ NIST ], achieved a so-called stability of one part in 10^18. In plain English, that means that "if a clock had been running since the Big Bang, by now it would only be off by one second,” said Vladan Vuletic, a physicist at MIT who was not involved in the work.

    [these clocks] could help industry build GPS systems that can rapidly pinpoint locations with sub-centimeter-scale precision.

    In addition, I heard a report on NPR that said researchers studying Einstein's theory of general relativity could make use of this clock to more precisely measure how time is different depending on the surrounding gravitational force - over a change in altitude of 1 inch.

    • by Anonymous Coward

      In addition, I heard a report on NPR that said researchers studying Einstein's theory of general relativity could make use of this clock to more precisely measure how time is different depending on the surrounding gravitational force - over a change in altitude of 1 inch

      Such advanced technology.....such archaic measurement units....

    • by yusing (216625)

      [these clocks] could help industry build GPS systems that can rapidly pinpoint locations with sub-centimeter-scale precision

      Now, if only I could get transceivers to strap to fruit flies, I might find out where they're all coming from.

    • This is the most interesting comment on this thread, yet the only replies it got were bashing the Imperial System, and a lame joke...

      Thanks for the article link!

  • Ytterbium, better than the prototype at least.

    Thanks, I'll be here all week. Sorry.

  • Hey, NIST, quit bogarting all the ytterbium! We need it to make Hilbert Drives [duntemann.com]!
    • Read the 23 page excerpt, bought a hard copy from Amazon. Will have to wait 5-9 business days to read the rest...

      Thank you.

The reason that every major university maintains a department of mathematics is that it's cheaper than institutionalizing all those people.

Working...