Why is it that historically the clocks have gotten more accurate, but the trains have been getting later and later?
Nuke clock incapable of losing time chimes with boffins
The force that binds neutrons to an atom's nucleus could be used to create clocks that are 100 times more accurate than today's best atomic clocks, say physicists at the University of New South Wales (UNSW). The nuclear clock outlined in a paper accepted for publication in Physics Letters Review would neither lose nor gain 1/ …
-
-
Friday 9th March 2012 11:56 GMT Still Water
Re: How?
We *define* the second by using atomic clocks, ergo, an average reading of many clocks can never be "slow" or "fast". If you do something odd with a single atomic clock like take it up a mountain, launch it in a satellite, etc, then it will drift, but that's relativity for you.
Why use atomic clock definition? The definition of a second came about because pendulums, the earth's rotation rate and the tropical year cannot be measured with anything like the same accuracy. A physical principle that nothing can ever be exactly measured.
NB: The measurement of metre is now defined by the speed of light (a defined constant) and the definition of a second rather than by using a platinum bar, or measurements of spectral lines, for much the same reasons - we can measure time far more accurately.
-
Friday 9th March 2012 11:58 GMT Anonymous Coward
Re: How?
> How on earth did we, way back when, define such a unit as a second if we couldn't even measure it accurately?
The same way we defined every unit of measurement we had at the time, as best we could.
> How do we know it's 2012 if we've never had an accurate timepiece?
Because 430 years ago Pope Gregory XIII decided that that year should be called 1582 and everybody agreed with him. Since then the Earth has orbited the Sun 430 times making this year 2012. It doesn't matter how accurate your second is, it is easy to count years which is how we know it is 2012.
> How do we know existing clocks are fast or slow? What do we use to measure the inaccuracy?
Up until 1967 a second was defined in terms of the Earth revolving around the Sun. Since then it has been define as:
"The duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom."
-
Friday 9th March 2012 15:02 GMT AndrueC
Re: How?
>Because 430 years ago Pope Gregory XIII decided that that year should be called 1582 and everybody agreed with him.
Not quite. That's only true if you live in Catholic country. The UK in common with most protestant countries didn't adopt his idea until 1752. The Russians didn't adopt it until the early 20th century.
http://en.wikipedia.org/wiki/Gregorian_calendar#Timeline
I read up on this stuff back at the turn of the millennium. A very interesting book - when it comes down to it it's all very arbitrary. Roman emperors for instance sometimes removed a month in order to bring tax collections in a bit early. Maybe that's an idea the ConDem coalition could look as an alternative to spending cuts :)
-
This post has been deleted by its author
-
-
-
Friday 9th March 2012 12:37 GMT dotdavid
So what?
"The nuclear clock outlined in a paper accepted for publication in Physics Letters Review would neither lose nor gain 1/20th of a second in 14 billion years, the age of the universe."
That's nothing, my wristwatch also neither loses nor gains 1/20th of a second in 14 billion years - it tends to lose or gain considerably more than that.
Now if they'd said their new fancy clock neither lost nor gained *more than* 1/20th of a second in 14 billion years, that would have been impressive ;-)
-
-
-
Friday 23rd March 2012 01:41 GMT An0n C0w4rd
Re: Excuse my ignorance, but...
The trick with high bandwidth stuff isn't knowing what the time is, its to do with recognising the pulses on the line. If you are generating a signal on a fibre optic line at N GHz, you need a clock source on the receiving end which has EXACTLY the same frequency to interpret the signals so you know if the bit is on or off. If one or the other drifts you'll introduce errors into the signals. Even a small drift can be bad as it upsets not only the bit in flow at the time, but also all subsequent bits which will probably overcome the ECC until the receiver figures out something is wrong and finds the next frame/packet boundary. Its one reason why, even though its mostly hidden, AFAIK high speed telco grade optical networks still use framing. (the other being that they're still based on the concept of carrying voice channels)
You can use a 2nd fibre to get around this (i.e. put the clock pulses down the 2nd fibre to let you interpret the signals on the first) but (a) that gets more difficult the higher the clocks go and (b) telco's really don't like that as it uses up a lot of fibre they could use to sell overpriced bandwidth to people.
Fry got laughed at for suggesting that packet switching networks need to know what the time is. I've yet to meet a router (or other packet processing device, other than some mobile phone standards which use clocks for encryption related things) that won't run if it doesn't know what time it is. Likewise I've never seen a SONET mux with a time source. But they all have clock sources built in.
(Note I haven't worked in the telco field for a few years, but while the speeds have increased I don't think the fundamentals have changed)
-
-
-
Saturday 10th March 2012 09:06 GMT Richard 12
Re: Chimes are one thing
Fundamentally impossible unfortunately, DAB simply cannot do real-time transmission or reception.
FM only suffers speed-of-light delay, DAB adds huge compression and decompression delays, and as each 'block' is compressed separately it's impossible to reduce the delay to less than one block of time, even if the processors at the BBC and in your radio could do it instantly.
Which it can't - and your radio is probably really slow.
-
-
Friday 9th March 2012 14:12 GMT Ian Michael Gumby
I want one....
Sorry,
But while a lot of the commentards are ranting and raving about the accuracy...
Think about the technology required to actually build one, and then build one that would be practical for any of the mentioned uses. (Except trying to keep the trains on time. I don't think its possible.)
Yeah I'm one of those guys who wears an automatic watch because I think its cool to have a timepiece made up of hundreds of mechanical moving parts costing $$$ and being hand made, when an electric watch costing $ is probably more accurate.
Sometimes the journey is more important than the destination itself.
-
Friday 9th March 2012 18:33 GMT Nigel 11
The real problem
The real problem is that time itself in a mystery. How do we know that ten seconds measured today (accurate to 14 decimal places) is the same amount of whatever time might be, as ten seconds measured yesterday? Or last century? Or in the age of the dinosaurs? Or "seconds" after the start of the universe?
We don't. We know only that multiple clocks involving different vibrating entities that agreed on a number yesterday, also do so today, within their individual limits of accuracy (whatever that means). We can't take today's clock back to double-check yesterday's measurement. Time goes forwards only.
IT angle: in a computer or other clocked logic, the exact frequency or regularity of the clock is fairly unimportant. The presence or absence of skew between the clock here and the clock there is critical to the correct operation of the whole thing. A wild speculation: the universe ends when it expands so far that "clock skew" prevents it from working.
-
-
Friday 9th March 2012 21:22 GMT Vic
Re: Re Chimes are one thing
> Now teletext isn't even sure what the second is, it only gives h/m and that disagrees (runs late)
Digital teletext is based on the MHEG-5 standard. That doesn't split out into individual lines, like analogue teletext does, so getting the time right isn't really viable.
It also means that fast-clearing isn't all that easy either, which is why the teletext offerings we get at present have a shrunk video pane in them. That means precious little space left over for teletext - particularly when you've got advertising to display...
All in all, I see digital teletext as an enormous retrograde step compared to analogue.
Vic.
-
-
-
Sunday 11th March 2012 23:54 GMT Gavin King
Re: Of course, if the speed of light isn't constant all bets are off
Even if it isn't constant, it is still likely that a we could find a function to fit to the changes i.e. rather than c = c, have c = c(t).
Then would come the simple task of recalibrating everything from the last few hundred years.
-
-
Tuesday 20th March 2012 08:04 GMT elderlybloke
Time passes - like wind and water
And regarding the passing speed in the new allegedly level of accuracy,
Who is going to require / make us of it?
Time passes at different speed/velocity depending on the strength of the gravitational field you are in and the velocity that you are traveling.
so time will be different for every thing and everybody if you are going to go in for the new order of time keeping.
As was said by the troops in WW2- It all sounds like Bullshit to me.