Feeds

back to article Time is on their side: NIST's new atomic clock accurate for 300 MEEELLION years

Some people are nuts about knowing the correct time, but the boffins at the US National Institute of Standards and Technology (NIST) have taken such habits further than anyone else with a new atomic clock that's accurate to 10 trillionths of a second per day. The clock, snappily dubbed NIST-F2, uses the frequency of a particular …

COMMENTS

This topic is closed for new posts.
Silver badge

Then we have leap seconds

To deal with. You see the earth is slowing down (ever so slightly) and days ARE getting longer. To counter this, we add leap seconds every once in a while. Yes, your GPS takes this into account when it finds the big X you painted on your foot (you are here!).

Leap years are another matter. They take into account that our trip around the sun isn't exactly 365 days. We add a day (almost) every 4 years to make the spring equinox be the same day every year.

Note to others: Please use the system's day-time routines. They are much more refined than the silly ones you might try to make up. They will work correctly as well. There are too many stories of date/time calculations going haywire.

Don't get me started on daylight savings (summer) time.

1
0
Bronze badge

Re: Then we have leap seconds

"...Yes, your GPS takes this into account..."

"Yes..."? Oh thank God. I'm so glad to learn that my GPS isn't confused by leap seconds. It'd be so annoying to be misdirected by 7,494,811.45 km (25 light-seconds) when I just wanted to pop down to the shops. It's amazing that they took such subtle details into account.

0
2
Silver badge

Re: Then we have leap seconds

Your GPS isn't confused by leap seconds because it doesn't use them.

Internally it uses GPS time (no leap seconds) your receiver adds the leap seconds when it display the time. The problem with leap seconds is that when somebody in Paris decides to add one, you cant guarantee that every satellite will get the message instantly.

The americans wanted to drop the whole silly idea - but the UK, among others, vetoed it.

3
0
Mushroom

slipped schedules

ok, so we all sync on this quantum quality transition, yet we must re-teach every new generation of children to tie their shoes before floundering aimlessly across the lawn to meet the 'maybe on time' school bus each day? did they chew their breakfast enough, so they don't get a tummy ache and lose a days worth of valuable school time? how is it we made it this far in our evolutionary transition from some pre-human form to now (relative to where you are) , and suddenly this time piece is sooooo important? can't we navigate into the future without it? do our fruits grow more precisely (without bruises and unusual bumps) now that it exists? do they care? as long as the sun keeps it's output steady enough to sustain our earth warm enough so we can wantonly spew noxious fumes and gaseous toxins freely and without thought, i'm happy..... ;-))

0
0
Silver badge

Our fruits do not grow more precisely.

But our planes do not get lost, and our Internets continue to anneal, and if I had one of these I'd paint it blue and put it on the mantelpiece to impress the neighbours, because it would be so much cooler than their cheap-ass illuminated betta tank.

1
0

"our planes do not get lost"

The Malaysian pilots must have forgot to wind their atomic clock.

0
0

Well done El Reg for doing this article.

It's about time...

14
0
Bronze badge

Synchronous Power Grids

I always understood that when you connected a new AC generator to a power grid, that if it was already almost in sync the grid would pull it into a fully synchronous state.

But what if the UK grid were running at 49.99 Hz and you wanted it ti be 50 Hz. How do you tune the frequency of the whole grid?

0
0
Silver badge

Re: Synchronous Power Grids

The clock gets messed up by loads

http://en.wikipedia.org/wiki/Utility_frequency#Frequency_and_load

0
0
Silver badge

Re: Synchronous Power Grids

"But what if the UK grid were running at 49.99 Hz and you wanted it to be 50 Hz. How do you tune the frequency of the whole grid?"

Well... I guess only a complete shutdown and restart with the first generator

being tuned to 50Hz and all the others turned on after that being synchonised

to the first would do the trick !

( I have seen pictures of what happens when an out-of-phase generator is

connected to the grid... Trust me, the carnage runs into the millions when

not only the generator but also the turbines get ripped to shreds ! )

8
0
Silver badge
Mushroom

Re: Synchronous Power Grids

The grid will pull a small generator into sync almost instantly, accompanied by a loud BANG as the rotor is yanked around and the coils try to leap off the stator.

It's a very, very bad thing to do.

3
0
Bronze badge

Re: Synchronous Power Grids

not a bad thing - if you're a fan of shrapnel!

many folks with small, self-installed, home emergency generators have learned the lesson that when reconnecting to the utility supply always disconnect the generator before closing the mains. there are such things as electrical codes and transfer switches for a reason.

otherwise, in the words of some government agency "the device undergoes rapid disassembly."

3
0
Silver badge

Re: Synchronous Power Grids

I remember doing that exercise in the power lab at university (the correct way, bringing the system into sync and then throwing the switch, not the other way). The lecturer told us that in some parts of the world they would just throw the switch and let the grid fight it out even with a big generator, but hopefully in the intervening years there has been more enlightenment.

0
0

Re: Synchronous Power Grids

How did you know that the grid frequency was 49.99 Hertz? you had to have an accurate clock! If you want your power to be at your frequency then you should make your own power grid. In other words live "off the grid." Obviously the commercial power grid is not going to let you change it!

0
0

Serious question

How do they determine the accuracy of the most accurate clock ever built?

And how do they know it will have lost a second rather than gained one?

1
0
Bronze badge
Pint

Re: Serious question

The second is now *defined* as 9...770 of those Cs transitions. I guess that they can guesstimate the uncertainties and errors in their hardware. Things like residual magnetic fields and the like.

On my mantle is a hilarious but lovely $20 Made in China pure mechanical wind-up clock. Beside it is one of those "Atomic" radio clocks tuned to WWVB. I also have a rubidium brick that needs to be integrated into the household time system, somehow.

0
0
Silver badge

Re: Serious question

"And how do they know it will have lost a second rather than gained one?"

Pah! Some people have no common sense! Just phone the speaking clock in 300 million years time.

3
1
Silver badge

Re: Serious question @Stratman

With a simple question you have opened the twin gates of Horology and Metrology. It doesn't matter what color pill you take. I recommend taking them all, and just rolling with it. If anything starts to make sense take more pills and/or turn up your bourbon drip until confusion is restored. Don't worry about the yellow, they don't itch.

In all seriousness, your question is one of the most complex questions any Human has ever asked. Right now, or then, as the case might have been, incredibly intelligent people are engaged in developing proofs and models that have taken years to build all to definitively answer your question and some little smartass post doc has just presented a paper which will distract those incredibly intelligent people just long enough so that he can steal their sanity and ruin the lives of their families.

The sciences of time and measurement are separate entities, but are so intertwined that learning one is learning much of the other, like reading and writing. The stuff of Einstein, Heisenberg and Hawking is a subset of the study of time and measurement. It is really, really deep stuff and much of it is highly theoretical to the point that the correct answer to your question is now incorrect, however that is subject to change if you decide to measure it again and the original answer was correct, but only at that moment but it's on you to determine when that moment was.

The short answer to your question, and I don't think you're going to like it, is that the accuracy of the most accurate clock on Earth is verified to be the most accurate clock on Earth by a group of people who say it is the most accurate clock on Earth. Therefore, it is the most accurate clock on Earth until the aforementioned smartass post doc comes along and proves it isn't.

The Metrologist who runs our tool room is the most intelligent person I've ever met and he's also way up there on the list of the highest paid non-executive employees in the US. He is in charge of calibrating all of our tooling and inspection equipment and is the reason we can create rather large objects with tolerances across any given plane down to a few millionths of an inch. As part of his contract we fully fund his research which is currently an experiment to develop a method to accurately measure the movement of a 1mm tungsten sphere caused by accumulation of individual atoms of the materials that make up the vacuum chambers all these little balls live in. This is the fifth year of that experiment and he is exceedingly pleased that nothing has detectable has occurred. No family or pets. Just a bunch of tungsten balls in oddly shaped sealed containers and that's his thing. But they're all just like that! He goes off to the conferences where NIST type folks like in this story frequent and if the news reported that the entire conference center, after emitting a low humming noise had simply flashed out of existence to the accompaniment of an 19th century accordion ditty I wouldn't be remotely surprised.

All that to say, if you ask him what time it is he responds with 'when?' Just accept they've got a super accurate clock (don't call it a watch :)

4
0
Silver badge
Boffin

Re: Serious question @Stratman

You just *know* it's right.

More seriously, you know (or can measure) the tolerance of every component and thus can calculate the possible range of the total error.

In some cases this can even tell you the direction of the error - for example, car speedometers are (supposed to be) designed so they don't read lower than the actual speed. This does of course mean they nearly always read higher than the real speed.

The other way to tell is to build several and compare them, which was how it used to be done. That doesn't really work for very high accuracy clocks though because you have to wait too long before you see them diverge by a measurable amount.

(As opposed to other instantaneous physical quantity measurements, where it doesn't take too long to check.)

0
0
Silver badge

Re: Serious question @Stratman

Interestingly, do you know why car speedometers have variance in them? It's pretty cool, that variance is built in so that if the OEM tires on the vehicle are within 20% either way of their recommended inflation pressure range, or you buy aftermarket tires within that threshold the speedometer will still read within MPH of the actual speed traveled. I believe it has been that way in the US since 1937, which is when Pontiac began offering an easy to change speedometer gearing assembly inside the instrument instead of the transmission. The wide variety of tires at the time, with terrible standards compliance, meant speedometers were either somewhat useless, or completely useless.

Modern cars assess speed through the transmission/transaxle, the same Hall Effect sensors that provide information to the ABS system with any difference being split down the middle. With the adoption of OBD and OBD II (in 1996 for most light and medium duty vehicles) the speed of your vehicle is more accurately displayed on your speedometer than on the speed cameras of police cars. From 1998 on all GM cars actually run long term accuracy checks and automatically compensate for a pretty wide variety of tire changes. The accuracy of those systems is extremely high and far, far to precise for low resolution instruments like an in dash speedometer. If you think the speedometer in your modern car is reading incorrectly it is far more likely that your comparison data source is incorrect, not your speedometer.

In reference to the original question, my initial answer is still valid. The details are summed up in a variety of peer reviewed papers dealing with calculating the transition of a component of the decay rate of the cesium element within a stabilized environment and extrapolating the long term accuracy of the clock based on that. Just like calculating you pulse based a 10 or 15 second count.

It's an enormously complex mathematic proof that deals more with operator selection theory and chemistry than with timekeeping. It's boring as shit and I'm firmly convinced that being surrounded by people who enjoy it is what makes our Vice Presidents all have that same kind of drunk, stabby look in their eyes.

The NST guys and their ilk all hang out as the US Naval Observatory, which is where we make our Vice Presidents live. That used to seem really odd, because all of our Vice Presidents are always unhinged loons and having them live at one of our oldest scientific institutions was just weird. But then Cheney came along and I'm 100% certain he was kind of like a reverse Captain America and that's what they had been waiting on to test the research they had been doing since the end of WWII. That's why they had to build the new super bunker where they kept Cheney between feedings.

1
0
Gold badge
Thumb Up

Impressive.

Of course I wonder when they will build the MEMS sized one for my personal ded reckoning "personal navigator."

0
0
Bronze badge
Pint

Re: Impressive.

Google: Chip Scale Atomic Clock.

$1500 price class.

On my Bucket List.

1
0
Silver badge

Impressive.... but needed?

Yes you do need an accurate clock to be able to decode/send various signals. That does not necessarily mean you need to carry around an accurate clock source though.

In most settings you can re create the clock from the signal itself using some sort of phase locked loop (PLL) or similar trickery. This allows you to correct for all sorts of clock errors such as poor clock sources, Doppler and such.

For example, a GPS receiver needs to be within a few ppm to track the signal from a GPS satellite. You can get those tight tolerances by using very precise temperature controlled oscillators, or you can use a relatively cheap crystal oscillator and the GPS signal itself as a high precision clock source.

0
0
Bronze badge
Happy

I won't believe it until I can see it and check it for myself

0
0
Bronze badge
Angel

Re:- "I won't believe it until I can see it and check it for myself"

What will you be checking it against?

0
0
Silver badge

There is better

In theory there is some tech orders of magnitude more accurate.

"The estimated amount of uncertainty achieved corresponds to a Yb clock uncertainty of about one second over the lifetime of the universe so far, 15 billion years, according to scientists at the Joint Quantum Institute (JQI) and the University of Delaware in December 2012."

NIST in USA have at least two prototype clocks far more accurate than this one. But their standard time isn't based on them.

http://en.wikipedia.org/wiki/Atomic_clock

Also there some development of Optical Clocks instead of Microwave (as the one in article is Microwave). Would that be a Crystal Clock?

Why is the handle on my clockwork winder on the back where I can't reach it?

2
0
Bronze badge

you can trust NIST, they're the only US government agency with standards.

[maybe they should be in charge]

0
0
Silver badge

Only NIST?

What about ANSI then?

0
0

Re: Only NIST?

ANSI is Not Système International. NIST is Ingenuity Spearheading Technology.

Do you prefer Americans Never Stopping Inching or New and Improved Seconds Ticking?

0
0

Not sure...

you can trust NIST. They crippled the encryption standards at the request of the NSA. So, who knows?

0
0
Silver badge

I will not be satisfied

until we have a timepiece that can tell us the number of Planck-times elapsed since the Big Bang. Now that's an accurate clock!

0
0

Please tell me...

...that the scientists announcing the world's most accurate clock were late for the press conference.

1
0
This topic is closed for new posts.