Why hotter clocks are more accurate


Research News


A new experiment shows that the more energy consumed by a clock, the more accurate its timekeeping.

Clocks pervade every aspect of life, from the atomic clocks that underlie satellite navigation to the cellular clocks inside our bodies. All of them consume energy and release heat. A kitchen clock, for example, does this by using up its battery. Generally the most accurate clocks require the most energy, which hints at a fundamental connection between energy consumption and accuracy. This is what an international team of scientists from Lancaster, Oxford, and Vienna set out to test.

To do this, they built a particularly simple clock, consisting of a vibrating ultra-thin membrane, tens of nanometers thick and 1.5 millimeters long, incorporated into an electronic circuit. Each oscillation of the membrane generated one electrical tick. The ingenious aspect of this design is that it is powered simply by heating the membrane, while the complete flow of energy through the clock can be measured electrically.

The scientists found that the more heat they supplied, the more accurately the clock ran. In fact, the accuracy was directly proportional to the heat released. To make the clock twice as accurate, they needed to supply twice as much heat.

The experimental team consisted of Dr Edward Laird at Lancaster University, Professor Marcus Huber at Atominstitut, TUWien, Dr Paul Erker and Dr Yelena Guryanova at the Institute for Quantum Optics and Quantum Information (IQOQI), and Dr Natalia Ares, Dr Anna Pearson and Professor Andrew Briggs from Oxford.

Their study, published in Physical Review X, is the first time that a measurement has been made of the entropy – or heat loss – generated by a minimal clock.

Understanding the thermodynamic cost involved in timekeeping is a central step along the way in the development of future technologies, and understanding and testing thermodynamics as systems approach the quantum realm.

It also shows a similarity between the operation of a clock and a steam engine. With a steam engine there is fundamental constraint on how much heat we must supply to do a desired amount of work. This constraint is the famous Second Law of Thermodynamics which is central to modern engineering. What this experiment suggests is that clocks, like engines, are constrained by the Second Law, with their output being accurate ticks instead of mechanical work.

Dr Edward Laird of Lancaster University said: “The subject of thermodynamics, which incorporates the most fundamental principles of nature, tells us that there are two types of machine that we cannot operate without releasing heat. One is the mechanical engine, which releases heat to do work, and the other is the computer memory, which releases heat when it rewrites itself. This experiment – in conjunction with other work – suggests that clocks are also limited by thermodynamics. It also poses an intriguing question: are all possible clocks limited in this way, or is it just a property of the ones we have studied?”

Interestingly, many everyday clocks have an efficiency that is close to what the scientists’ analysis predicts. For example, their formula predicts that a wristwatch whose accuracy per tick is one part in ten million must consume at least a microwatt of power. In fact, a basic wristwatch usually consumes only a few times this amount. The laws of thermodynamics, discovered in the nineteenth century, are still finding new applications today.


From EurekAlert!

4.7 11 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
Shanghai Dan
May 8, 2021 10:20 pm

So what you’re saying, is that Global Warming will make our clocks more accurate?

Well then, sign me up!

John Dueker
Reply to  Shanghai Dan
May 8, 2021 10:27 pm

No because it’s actually cooling.

Reply to  John Dueker
May 9, 2021 5:24 am

Yes , but follow the science. 97% of experts believe that global warming causes global cooling, so maybe it can do both and keep clocks more accurate while it’s on its way to causing some cooling?

Michael S. Kelly
May 8, 2021 10:33 pm

This makes perfect sense. And I don’t even have a joke to make about it.

Richard (the cynical one)
Reply to  Michael S. Kelly
May 8, 2021 10:45 pm

I’m warming up to the idea myself.

Bob boder
Reply to  Michael S. Kelly
May 9, 2021 6:02 am

So I have watch that runs fast, if I heat it up will it be more accurate?

Sal Minella
Reply to  Bob boder
May 9, 2021 12:41 pm
It will.  Hotter means more energetic so, it follows that there is less positional uncertainty, meaning lower amplitude quantum fluctuation in the ground state thus more .per tick accuracy.

Reply to  Sal Minella
May 10, 2021 7:57 am


Sal Minella
Reply to  bob boder
May 10, 2021 3:08 pm

I;m glad that you enjoyed my impersonation of S. Hawking.

May 8, 2021 10:37 pm

My quartz crystal watch would run slightly fast on my wrist and slower when not worn.

Reply to  tygrus
May 8, 2021 10:52 pm

Renditions are not only running longer, but seeming longer –

Reply to  gringojay
May 9, 2021 1:36 am

Are You Sitting Comfortably? Then I will begin.

Joseph Zorzin
Reply to  gringojay
May 9, 2021 6:02 am

never heard of interesetcional feminism- looked it up: https://en.wikipedia.org/wiki/Intersectionality

Now I know why I hate Time Magazine because that’s all it is!

Reply to  Joseph Zorzin
May 9, 2021 11:17 am

Gave up Time over 40 years ago in my early 20s, then News Week, but kept getting US News and World Report for a coupe of more years until it started to turn into an opinion rag like the other 2. I was done with all “news” magazines before I turned 30.

It was mostly the leftist change in their articles, not my maturation.

Reply to  Drake
May 15, 2021 12:17 am

I basically did the same with computer magazines. They just became total advertising rags. Subscribed to several and they all went the same way and that was before the internet became so pervasive. I used to love those CD’s that came with them so I could explore new things.

Edit : I should stress that when I said advertising rags I wasn’t meaning the usual advertising but the articles themselves.

Last edited 30 days ago by mikebartnz
Reply to  tygrus
May 9, 2021 9:55 am

That has to do with the quartz crystal and the way it is cut relative to the crystal axes — and is not related to the above topic.

A quartz crystal’s frequency is temperature dependent, and watch manufacturers assume the watch is taken off at night so they assume an average temperature based on so many hours on the wrist and so many hours off. Then they tweak the frequency of the crystal so that the average of the two frequencies works out to be (ideally) perfect.

Reply to  Observer
May 15, 2021 12:25 am

One thing I love about this site is that there is always someone that knows about a topic. Thank you.

Gary Pearse
May 8, 2021 10:38 pm

Look a little harder and you’ll see that the same holds true for electricity generators – the hotter, the better.

Reply to  Gary Pearse
May 8, 2021 10:54 pm

Also, rockets.

David A
Reply to  Joe
May 8, 2021 11:26 pm

…and women yes.

Alexy Scherbakoff
Reply to  David A
May 8, 2021 11:38 pm

Unless they are crazy.

Paul C
Reply to  Alexy Scherbakoff
May 9, 2021 1:33 am

The rightness constant of women is relative. If a woman at the front of a train, and a woman at the back of a train disagree with each other, they are both right in the frame of reference of men, though the hotter and faster woman is more right. A stationary male observer must move fast or the right woman will leave on another train. A stationary woman can still be hot and fast, but may be drunk – but is still right.

Perry Smith
Reply to  Paul C
May 9, 2021 5:56 am

More research on this subject.

Reply to  Alexy Scherbakoff
May 9, 2021 6:49 pm

Then they’re just a hot mess.

Bob boder
Reply to  Gary Pearse
May 9, 2021 6:03 am

Until it gets too hot

Joseph Zorzin
Reply to  Gary Pearse
May 9, 2021 6:04 am

So the productivity of solar and wind “farms” will improve as the climate heats up!

Corky the cat
May 9, 2021 12:13 am

This is a joke, right? Otherwise I despair for the human race, all those poor souls who don’t have a clue how anything works.
What these dumbos are trying to describe is a quartz crystal oscillator. Quartz is piezoelectric, set into sustained mechanical resonance with simple electronic feedback. It is electrically powered, not thermal! Resonance frequency of wristwatch crystals is a weak quadratic function of temperature, optimized for temperatures close to human body, slightly low (i.e. slow) at room temperature, also when too hot. They are made in billions and used all over, including many clocks. There is no 1uW limit of electrical power. Decent clocks use a better cut of crystal. Why is computer memory is singled out? Is there anything that moves that doesn’t take energy to operate?

Come on people, where I am it’s known as a leg-pull, but keeps those research funds coming.

Mike Ballantine
Reply to  Corky the cat
May 9, 2021 1:23 am

My take on their experiment is that doubling the heat simply brought their oscillator into it’s proper operating range.

Reply to  Mike Ballantine
May 9, 2021 4:06 am

It appears someone wrote the article to appeal to quite ignorant people or it was written by quite stupid people. Energy is input, heat is a by product of energy consumption. It is generally easier to control something when adequate energy can be used. This will generate more waste heat, that’s the way this universe works, but the waste heat is not relevant to the functioning; it is just an inevitable byproduct. The writing suggests accuracy will be improved by putting a bunsen burner under a clock.

Bob boder
Reply to  AndyHce
May 9, 2021 6:15 am

Yep, plug the thermostat in your car engine and see how much more efficient it gets the hotter it gets.

Reply to  Bob boder
May 9, 2021 6:55 pm

The basic statement is true, the hotter a reaction, the more efficient. But then we run into materials, we have to actually contain this reaction within a mechanical device, the reciprocating engine. Because of the limitations of the materials, we must limit the ultimate temperature of the reaction, and thus we limit its efficiency, but in engineering, everything is a trade-off. I can design you the perfect system, but do you my customer have the money to pay to have it built? Or even to pay me for my design? Or if I’m in manufacturing, can my client build the product that I designed at a price the public is willing to pay? So even if I have designed the perfect system, that does not make me the perfect engineer. If my product cannot be made at a price the customers are willing to pay, I am a failed engineer, and likely looking for another job. Clearly, no climate scientist (nor activist) has ever been der and done dat.

Reply to  Red94ViperRT10
May 10, 2021 8:00 am

The point is, if engines are more efficient when they are hotter than why do we need to cool them? in a frictionless world maybe they are and maybe they aren’t but in the real world hotter is only better to a degree then its not.

Reply to  bob boder
May 14, 2021 3:01 pm

Maybe I was too verbose. The maximum temperature of an engine is limited by the materials used to make it. Some 40 years ago I read an article in Scientific American, back when it was still an interesting publication, about the future of automobile engines is in ceramics. The details did not get hammered out, but the main point was, ceramic pieces can withstand temperatures up to 10,000° F without failure. Sounded like a great idea, but where has that revolution gone? I see no signs of it. Possibly because, one of my first thoughts, if you build an engine that operates at 10,000° F, where are you going to put it? If you use the same old engine compartment we’re used to seeing, within minutes you would be melting and/or incinerating the fenders, tires, suspension pieces, even the frame and your windshield.

I guess it’s important to note, the statement, “a hotter reaction is more efficient” applies only to the reaction! Where you would house this reaction, or how you turn that released energy into useful work are knock-on effects. One problem at a time, Kiles.

Reply to  Corky the cat
May 9, 2021 9:26 am

My first thought was how had they compensated for expansion of the framework holding their strip of plastic? For the effect of heat on the plastic itself (expansion, rigidity)? For the gas surrounding the apparatus? For the changes in resistivity of the wiring?

And I probably forgot a few.

Vincent Causey
May 9, 2021 12:24 am

Because information can replace energy in the second law. This was discovered when the famous “Maxwell’s Demon” thought experiment was solved. The demon uses information to select the molecules which consumes energy, if my memory serves correctly. Fascinating, as Mr Spock would say.

Reply to  Vincent Causey
May 9, 2021 1:39 am

What’s in a name?
Everything is in a name.
We need to stop talking about “Greenhouse Gases” and the “Greenhouse Effect”.
Instead, we must use a new phrase and call this the “Maxwell Mass Effect” (MME) in honour of a real scientist.

May 9, 2021 12:35 am

So, the 19th and early 20th century thermometers, on which thermodynamics is founded, were so inaccurate that climate can only be (properly) understood by adjusting the temperatures that they recorded?

Bob boder
Reply to  decnine
May 9, 2021 6:05 am


May 9, 2021 1:31 am

More research needed…..

On metronomes

Reply to  fretslider
May 9, 2021 2:53 am

They already did research on Yo-Yos and results were frightening …

Reply to  John Shewchuk
May 9, 2021 5:47 am

Results like that tend to go up and down….

Tombstone Gabby
Reply to  John Shewchuk
May 9, 2021 12:25 pm

G’Day John,

I note: it’s from the “Barrier Miner”, Broken Hill, 23 Jan 1933. It was either a slow-news day, or someone wanted a laugh. I have a hunch it was probably printed as a ‘strange’ humor piece. Working at Broken Hill in that era was no picnic.

Peta of Newark
May 9, 2021 2:51 am

Quote:”One is the mechanical engine, which releases heat to do work

No. You put Heat Energy into an engine to get mechanical work out of it

Yes heat engines ‘release heat’ – in their exhausts.
They have to. They have to have a thermal gradient, from input to output.
The steepness of that gradient defines the efficiency of the Heat>Mechanical work engine, via Carnot’s Law = an amazingly subtle/simple beguiling little equation if ever there was.

Now then, take that to the Green House Gas Effect – where ‘energy’ radiated from the Green House Gases in the (cold) atmosphere has a heating effect on the (warm) surface.
Put the numbers into Carnot’s Law and you find you’ve built a Heat Engine with negative efficiency.

Patent nonsense, unless you are describing a Cooling Effect.

Yes, a warming atmosphere will exhiit a decrease in the thermal gradient between the surface any any altitude you choose – slowing conductive and convective cooling of the surface.
The very things that define ‘weather’ – ultimately ‘climate’

Yet Climate Science says The Exact Opposite will occur.

How wrong about anything is it possible to be?
Silly question, go back to the top of this comment….

A Warming Atmosphere is the very last thing any critter living on the surface needs or wants.
Primarily because all life depends on bacteria – and as anyone who owns or uses a refridgerator knows, bacteria (quite abruptly) stop working at 4 Celsius and less.

Strangely coinciding with the temperature where water has its greatest density.

Take care
Gaia is watching you……..

BTW: What do they mean by ‘twice as accurate’
Does it keep better time compared to other clocks or does it simply, as their own description implies, run twice as fast.
Thus its ‘time resolution’ (is that what they mean) is better, not accuracy as most folks would read it.
To my mind, that is all it can possibly do.

They’ve built a very simple little engine that, when you ‘give it more gas’, it goes faster. Even kindergarten kids know that.

Last edited 1 month ago by Peta of Newark
Rud Istvan
Reply to  Peta of Newark
May 9, 2021 7:56 am

Peta, your ‘example’ misunderstands how the GHE works. The backradiation does not warm the surface. Rather, it is indicative of the IR scattering caused by GHG that retards IR escaping to space, thus cooling. It is reduction in cooling, not an increase in warming, that’s the GHE essence. All the warming comes directly from insolation

Reply to  Rud Istvan
May 9, 2021 8:49 am

The climate scientists’ over-simplified models treat co2 insulation as a separate heat source however, independent of the initial value of the Sun’s insolation, and none of them consider that the heat coming from the core of the Earth, from radioactive decay and possibly other mechanisms, should also be taken into account. And on a side note, the alarmists’ conceptions all assume that co2 sinks and sources were all in perfect magical synchronization, even though their colleges in the geology department like about the huge changes the Earth has undergone.

Reply to  PCman999
May 15, 2021 1:06 am

Recently they discovered I think it was 90 volcanoes under the West Antarctica shelf that they didn’t know about before. As the world is around 70 percent oceans with most of it being unexplored how many do we not know about.
Iceland at the moment must be adding an incredible amount of heat and Hawaii before it. I have been told that that heat is negligible but I have never come across any data what so ever about it.
You might get some data about individual volcanoes like St Helens but nothing about the total world data and they truly don’t know the underwater data.
All we need is Yellowstone parks caldera to erupt and it will make the Taupo one look like a puppy.
Christchurch is 611K from Taupo, NZ and yet there is evidence there of that eruption.
It is scary if Yellowstone decides to truly erupt. It wouldn’t just be the devastation within the USA but worldwide.
I live by the motto “Live fast die young and have a good looking corpse” but I think that I have left it a little late.

Mark - Helsinki
May 9, 2021 3:06 am

“with their output being accurate ticks instead of mechanical work.”

an oscillating membrane is still part of the “mechanical work” of the clock, ticks here is not a real thing. We use “tick” because the machine is used for time keeping

“Tick” cannot replace mechanical output in this context, efficiency and accuracy are irrelevant.

Nit picking over 🙂

Last edited 1 month ago by Mark - Helsinki
Reply to  Mark - Helsinki
May 9, 2021 7:08 am

If the membrane they mention is not in a vacuum then it must do mechanical work to move the air on either side.

May 9, 2021 3:40 am

Do we really know what time it is? Their research seems to prove that we don’t. Maybe it’s too late or not yet time to start. If global warming makes us early will we too late with global cooling?

Why would anyone care if we don’t even know what time it really is? Wait …. give me a second to think about this. Humm … maybe it’s time to stop all the foolishness but I can’t be sure.

Reply to  eyesonu
May 9, 2021 7:53 am

Hey, that was only 0.99999999999999999 seconds. Gotta get that phone in a warm pocket dude.

Tom Johnson
May 9, 2021 3:59 am

“The scientists found that the more heat they supplied, the more accurately the clock ran.”

What we have here is a total failure to define terms. What is a “clock”? What is “accuracy”? For that matter, what is “heat”, or even what is a “scientist”.

It is well known that almost all mechanical systems change significantly with temperature. Some go “up”, some go “down”, and some do both, having an optimum where they are most “efficient”. Have these “scientists” only now discovered this, and have they not yet even reached the most “efficient” point?

Bob boder
Reply to  Tom Johnson
May 9, 2021 6:07 am

Yep most engines run better the warmer they get, until they don’t.

Reply to  Bob boder
May 15, 2021 1:14 am

I have been told that it takes an ICE engine about twenty two miles to come to its optimum temperature and around here very few would do that. Most of them are running their brats to school instead of them cycling.

May 9, 2021 5:03 am

Even a broken clock is right twice a day. I can’t recall the name of the Law right off, but that’s the well-known minimum accuracy for clocks.
Ah……. so f I continue to put more heat into a broken clock, eventually all I get is a flaming clock, not more accuracy.

That’s probably in the “Further study is needed” section of the paper.

Last edited 1 month ago by H.R.
Reply to  H.R.
May 9, 2021 5:47 am

This research is correct: a sundial is more accurate at noon when it’s hot than at midnight when it’s cold….

Last edited 1 month ago by Yooper
Steve Richards
Reply to  Yooper
May 9, 2021 6:00 am

I never realised that!

Bob boder
Reply to  Yooper
May 9, 2021 6:09 am


Reply to  Yooper
May 9, 2021 7:57 am

Indeed, we do live in strange times, when a sundial is more accurate than a computer model. Solution: we need a bigger computer ….. tempus fugit

Last edited 1 month ago by philincalifornia
Bob boder
Reply to  H.R.
May 9, 2021 6:08 am

Actually a slow moving clock can be right less then twice a day

Reply to  Bob boder
May 9, 2021 6:39 am

Well then, Bob, apply more heat until it’s a flaming clock. Then take the ashes and combine them with resin to make a sundial.

Eventually, more heat can make a more accurate clock, as Yooper points out above.

The researchers are on to something, or perhaps as Peta pointed out earlier, maybe on something. In any case, more research (and funding) is needed.

Bob boder
Reply to  H.R.
May 9, 2021 11:45 am


Corky the cat
Reply to  H.R.
May 9, 2021 12:23 pm

A broken clock is not right twice a day. It’s broken.

May 9, 2021 5:59 am

How do you test the accuracy of a clock?

Reply to  AARGH63
May 9, 2021 6:42 am

1) A man with one watch knows what time it is. A man with two watches is never quite sure.

B) It’s always 5 o’clock somewhere.

The answer to your question lies somewhere in between those two points. Higher order math is probably involved.

Last edited 1 month ago by H.R.
Reply to  H.R.
May 9, 2021 8:01 am

B) It’s always 5 o’clock somewhere.”

Thanks for reminding me. Mimosas with breakfast on Mother’s Day. YAY.

Reply to  H.R.
May 15, 2021 1:17 am

Quote : B) It’s always 5 o’clock somewhere.
Time for a beer then.

May 9, 2021 8:48 am

Perhaps this is why the battery in my Timex Expedition watch last so long, it never leaves my wrist and only gets cooler in the shower!

Reply to  2hotel9
May 15, 2021 1:18 am

Do you like cold showers?

Reply to  mikebartnz
May 15, 2021 9:16 am

Around body temp is good. Even warm water has a cooling effect on objects as it evaps.

May 9, 2021 10:17 am

They are not even measuring “accuracy”. They are measuring close-in phase noise (aka Allan variance), and calling it “accuracy”. What a pile of rubbish. I would love to see the feedback they would have gotten if they tried to publish this in an IEEE journal.

Danley Wolfe
May 9, 2021 10:37 am


Reply to  Danley Wolfe
May 9, 2021 1:48 pm

Perhaps that someone at Lancaster can, indeed do verifying experiments in thermodynamics? But the news release hacks still can’t write coherently about it.

May 9, 2021 3:17 pm

So it takes a minimum amount of power for a clock to reach certain level of accuracy. But that doesn’t mean that all clocks using that amount of power are as accurate as it would allow. One reason those cheap mechanical quartz clocks they sell everywhere are so inaccurate is that they are only as accurate as the crystals installed in them. Meanwhile, most clocks with digital displays are measured against an atomic standard at the factory to determine how long it takes for each sample to loose or gain a second, then that time period is set in a nonvolatile location (FLASH, ROM, etc.) so that the clock chip will set itself ahead or back each time said interval has passed. So if you want a wall clock that doesn’t go off every week, a digital one is usually a better bet.

Jim G.
May 9, 2021 7:45 pm

I read a joke one occasion, it goes:

Most things expand as they are heated.
Therefore, I am not fat, I’m just hot.

Larry Butler metrologist since 1966
May 9, 2021 7:47 pm

Er,ah, you people need to stick with your profession. time.gov is accurate to nanoseconds in thousands of years. AGW has no effect on the cryogenic cesium atoms at NIST. The clocks on the webpage are compensated for your internet or direct radio propagation delays. Metrology has nothing to do with weather.

Mark L. Gilbert
May 11, 2021 8:52 am

At the risk of confirming my own stupidity, the study would seem to make perfect sense.

Since energy can only be subdivided so far (to minimum unit quanta), the more ENERGY per division, the more finely it can be divided (accuracy) by measuring the quanta.

I am not understanding why this is a surprise to anyone.

May 13, 2021 3:27 pm

“To do this, they built a particularly simple clock, consisting of a vibrating ultra-thin membrane, tens of nanometers thick and 1.5 millimeters long, incorporated into an electronic circuit.

Each oscillation of the membrane generated one electrical tick. The ingenious aspect of this design is that it is powered simply by heating the membrane, while the complete flow of energy through the clock can be measured electrically.

The scientists found that the more heat they supplied, the more accurately the clock ran. In fact, the accuracy was directly proportional to the heat released. To make the clock twice as accurate, they needed to supply twice as much heat.”

Appears to me that they’re measuring the accuracy of physics, not a clock.

Gene Selkov
May 15, 2021 1:15 pm

I think thermodynamics has nothing to do with energy dependence. Clocks are feedback devices. Their precision depends on the accuracy of the feedback mechanism and the switching speed. The less time the clock spends switching from one state to another, the more accurate it is. Fast switching requires more energy. Maintaining a steady state between reversals also requires energy. Driving the output mechanism (e.g., hands in a wristwatch) is another energy sink. An extremely low-energy clock will be affected by all of those sinks to a greater degree than an excessively energized one.

%d bloggers like this: