Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
What is Time?
0 Plays2 seconds ago
In this episode of Breaking Math, Autumn and Andrew Novick delve into the intricate world of timekeeping, exploring the significance of precise time measurement in modern technology. They discuss the evolution from traditional atomic clocks to cutting-edge optical clocks, the critical role of time in various industries, and the implications of time on fundamental physics, including Einstein's theories. The conversation also touches on the quirky concept of leap seconds and the future advancements in timekeeping technology, emphasizing the relative nature of time and its perception.
You can learn more about Time at time.gov and NIST at nist.gov.
All opinions are of the individual scientist and do not reflect the opinions of NIST or the federal Government.
Subscribe to Breaking Math wherever you get your podcasts.
Become a patron of Breaking Math for as little as a buck a month
Follow Breaking Math on Twitter, Instagram, LinkedIn, Website, YouTube, TikTok
Follow Autumn on Twitter and Instagram
Follow Gabe on Twitter.
Become a guest here
email: breakingmathpodcast@gmail.com
Recommended
Transcript
Introduction to Precision Timekeeping
00:00:00
Speaker
Welcome back to Breaking Math, the podcast where we break down mathematics that shape our everyday world and explore the fascinating applications of technology and science. I'm your host, Autumn Feneff, and today we are going to take you on a journey into one of the most precise and fundamental aspects of our modern lives, timekeeping. From atomic clocks that keep our world synchronized to revolutionary advances in optical clocks, in this episode...
00:00:29
Speaker
It promises to be both mind expanding and practical. Time may seem like an unchanging constant, but measuring time with the utmost precision is one of the most significant challenges in modern science.
00:00:42
Speaker
Whether it's GPS signals guiding your navigation or the timestamp that keep global finance markets in order. Having precise time is critical for the function of nearly every system we rely on. But how do we measure time with such accuracy? And what is the technology behind public time services like Time.gov?
Andrew Novick's Role at NIST
00:01:03
Speaker
We're thrilled to have a very special guest with us today, Andrew Novick.
00:01:07
Speaker
Andrew has worked in the Time and Frequency Division of the National Institute of Standards and Technology, or most commonly known as NIST, since 1998, where he's been at the forefront of precision timekeeping technology.
00:01:23
Speaker
He's an electrical engineer with extensive experience in working on electronics for atomic clocks and remote measurement systems, making him one of the key figures in this development and management of advanced timekeeping services. Andrew's expertise doesn't stop there.
00:01:40
Speaker
He's also responsible for managing the UTC, NIST remote clock and oscillator calibration services, as well as the leading efforts on satellite timing systems and network time protocol or NTP measurements.
00:01:55
Speaker
Most impressively, Andrew is the creator and manager of the national web clock services, Time.gov, which millions of people rely on every day for accurate timekeeping.
00:02:09
Speaker
He's also authored and co-authored more than 40 papers related to time and frequency metrology, making him... one of the true experts in this field.
00:02:20
Speaker
So we're excited to have Andrew with us today to discuss the cutting edge technology behind atomic and optical clocks, as well as the real world implications of precise timekeeping.
00:02:32
Speaker
Whether it's been from GPS systems to satellite communications, the stakes are high when it comes to keeping time accurate down to the nanosecond. And Andrew's here to help us break this all down for you.
00:02:45
Speaker
But before we dive into detail, let's start with the basic question. Why is timekeeping such a crucial part of our technological infrastructure? You might think the clock on your phone or the microwave is good enough, but when it comes to things like synchronizing global data centers, guiding spacecrafts, or testing fundamental theories in physics, precision is everything.
00:03:12
Speaker
So today we're going to explore the science and engineering that make it possible.
Advancements in Optical Clocks
00:03:17
Speaker
This is where atomic clocks come into play. For decades, atomic clocks have been the standard bearers of timekeeping accuracy.
00:03:25
Speaker
These clocks work by measuring vibrations for atoms like cesium or rubidium to keep time with incredible precision. But now there's even more advanced technology on the horizon, optical clocks.
00:03:42
Speaker
These clocks promise to take more timekeeping to a whole new level, using oscillators from at optical frequencies, which are far higher than those traditional atomic clocks.
00:03:56
Speaker
Now, optical clocks are not only more accurate, but they also have the potential to revolutionize the way we measure time on a fundamental level. They are so precise that they can detect minute differences in time caused by the Earth's gravitational field, a phenomenon predicted by Einstein's theory of general relativity.
00:04:18
Speaker
We'll explore... even how these clocks could even help scientists uncover some new aspects of the universe, as well as how they impact the technologies we use.
00:04:30
Speaker
And of course, all of this cutting-edge technology has real-world applications, and that's where services like Time.gov, the Internet Time Service, ITS, and the Time Measurement and Analysis Service,
00:04:45
Speaker
TMOS comes in. So time.gov is a public facing service that provides accurate time sourced from the most precise atomic clocks at NIST to your internet enabled devices. And today, Andrew is going to give us an inside look at how time.gov works and how it's synchronized with atomic clocks and other precision timekeeping services that NIST provides us for industries and individuals so likes.
00:05:14
Speaker
So whether you're a physics enthusiast, a techie, or just someone curious about how our world stays so meticulously in sync, this episode is for you.
00:05:25
Speaker
We're going to take you behind the scenes of one of the most and important and yet often overlooked technologies that keep modern society ticking, literally. So let's get started and join us in this episode of Breaking Math.
Andrew's Journey in Timekeeping
00:05:49
Speaker
Hi, Andrew. How are you doing? I'm doing great. How are you? I'm doing wonderful. Thanks for coming on Breaking Math. Tell us a little bit about your background. Yeah, so I'm an electrical engineer ah at NIST, and NIST was my student job.
00:06:05
Speaker
So I went to CU Boulder, University of Colorado, for electrical engineering, and it's just like three quarters of a mile from NIST. And at the time, it was National Bureau Standards. And so I had a work-study grant, so I found a job doing electronics at NBS and...
00:06:24
Speaker
worked here all through school and eventually became NIST from NBS and eventually got hired on. And so this has been my one career, one job in my whole career. So it's awesome.
00:06:38
Speaker
Kind of a rare thing to hear nowadays. So consider yourself lucky with that. Yeah, very rare. ye So what got you interested in this work? And along with that, what is timed?
00:06:53
Speaker
Yeah. So I think, I mean, it's interesting because I i came here in like junior high if for a tour. And so maybe I knew a little bit about NIST, but really what got me interested was working here.
00:07:07
Speaker
You know, i I was working here. We're doing the work study and like I learned all about um you know time and frequency measurements and calibrations and metrology and standards. And so it was really just being here is how I got interested and how I've stayed interested all this time because it is just a really fascinating subject.
00:07:30
Speaker
And the subject of time, i mean, we we're kind of lucky with time because it's something that a lot of people understand to some degree. Like if you talk about like the kilogram or the meter, you know, how you measure a meter, like you lose people pretty quickly.
00:07:46
Speaker
But time affects people's lives, you know, all over the place. And really, you know, the at least time of day or the duration of how long it takes for something to happen. And something like, say, speed of light, right? Like people...
00:07:59
Speaker
are kind of confounded by that, like, you know, speed of light and, but it's also something that can kind of grasp it. And so it's, it's in our common vernacular. So I feel like time, but we're pretty lucky. And, you know, if you want to think of it, like if you want to define time as physics,
00:08:15
Speaker
You know, you have the three dimensions and then you have time. You could debate whether you want to call it a dimension or not, the fourth dimension. But without time, like none of these things in physics work. Like, you know, you drop something and like, how long does it take? Or you want to talk about mass or velocity. All of these things have a time component. So I feel like we're pretty lucky.
00:08:37
Speaker
But also time is pretty relative, you know, at least how we perceive time. I mean, what we do, what we do at NIST is we define time. very a very specific duration that's not relative, doesn't change, but how people...
00:08:51
Speaker
perceive of time like oh wow that 20 minutes went by really fast or well this last five years like was really fast or like man this you know i have a test on friday and there's no time to to do it like it's time either goes fast or slow depending on what you're doing what you're perceiving and so those things are relative which is cool i mean if you want to talk about time dilation and relativity time is also relative in physics but that's a whole separate discussion i Absolutely.
Functioning of Time.gov and Atomic Clocks
00:09:20
Speaker
So to kick things off, really, could you provide an overview of what Time.gov and Internet Time Service and the roles they play in delivering precise time to the public?
00:09:33
Speaker
Sure. We have here what we call the timescale, so NIST timescale, and that's a bunch of atomic clocks, which I'm sure we'll talk about later, but we define the duration of one second for America. So we're the federal lab that defines the second, and then we define what the time of day is, and we're also the resource or the reference for frequency.
00:09:56
Speaker
ah So with this timescale, there' a bunch of atomic clocks all averaged together, and we could talk about timescales all day, but let's assume for a minute that we know exactly what time it is and we can define the second and we can define frequency.
00:10:09
Speaker
So how do we get that out to the world? One of the biggest ways we do that is time.gov. So it's our federal government web clock. And the other way is NTP, Network Time Protocol. So that's our internet time service.
00:10:23
Speaker
And these are both free service. They're called services, but they're free. And time.gov basically from any browser, from any device, you go to www.time.gov and that browser sends a request packet to our server.
00:10:38
Speaker
And our server is on time, right? It's linked to our time scale. And so we get this exact time packet and we send it back to that browser and here's your time, right? Except for, we know about latency, right? There's latency in that. So it took a while for the for that request to get to our server and it took a while for that exact time packet to get back.
00:11:03
Speaker
And that latency can be milliseconds or hundreds of milliseconds. And so like a thousand milliseconds is a second. So if we're already at a hundred milliseconds, like you're, you're, you're coming up on like something that actually could matter. So what we do is we use the,
00:11:20
Speaker
whatever the local oscillator say in your phone or in your computer. And when you send that request, it actually makes a timestamp from that oscillator, i'm like your computer clock or your phone's clock to microseconds. So like, you know, a lot smaller ah designation of a fraction of a second. And then the network transfer goes to our server, comes back,
00:11:42
Speaker
And then when you get those time packets back, you look at that clock again and you say, oh, 342.009 milliseconds transpired in round trip. so... milliseconds transpired in that round trip and so We don't know how long it took for the packet to get to our server versus how long it came to get back, but we know that round trip.
00:12:03
Speaker
So if we can estimate or assume that half of the round trip was there and half the round trip was back, which is probably never going to be exactly half, but we can estimate that and we can have a pretty good way of taking out that delay from when our server set the packets to when it gets to the user.
00:12:22
Speaker
And In the case of time.gov, it doesn't totally matter because it's really just a visual clock that's running. So you can see this clock. You're not really using it to measure something. really You're really just getting a visual clock.
00:12:34
Speaker
And then we do have different time zones. And so those things change sometimes. Time zone rules or time zone borders change. So we have a map that shows you and you can look at the time in different time zones.
00:12:47
Speaker
So NTP actually sets a clock. And so there's nothing visual about NTP. You can actually run it from a command line. it's It's, or you can write a program that in the background goes, but it does the exact same thing. It goes to a NIST server, gets a time packet, comes back, uses that round trip delay, takes out the half the delay as the estimate.
00:13:09
Speaker
And NTP is, um it's, so it's a software clock. You get these packets back as numbers and NTP, The good thing about that is you can set a clock with it.
00:13:20
Speaker
So if you're running it on a device, say a computer or a stock market timestamper thing, or even like a punch clock, like ah you at at your job where you're like clocking in, I don't know people do that anymore, but it's probably somewhere. There's quite a few places that I know that still do. Yeah.
00:13:36
Speaker
And it might be a software, you know you're you're clocking, so you want to make sure your computer's right. So NTP can set a clock, but because of the network latency in that estimate, we don't really know Exactly. But it is ah ah on the order of milliseconds, probably single digit milliseconds. And so and with time.gov, we get about 20 million hits a day on during the week. It's less on the weekends.
00:14:00
Speaker
I guess people are less on their on their devices on the weekends, which is kind of nice. And then NTP, our Internet time service, is over 100 billion per day. Wow.
00:14:12
Speaker
So it's like, I don't even know how many computers there are in the world, but it's probably less than that. yeah so So there's computers that are checking quite often, especially server computers um that want to keep yeah're very close. They set themselves like multiple times a day, maybe multiple ah times an hour.
00:14:31
Speaker
Absolutely. Now, could you explain to us why precise time is so critical in everyday activities and how these network time services ensure this level of accuracy for us?
00:14:46
Speaker
Yeah, it's questionable. like how you know Something like time.gov or like your cell phone clock or whatever, it's going to ensure that you get to the meeting on time or you're late, but you know you know you're late because you can see what time is. Yes.
00:15:02
Speaker
You know, it used to be that people would set their alarm clock by like seven minutes ahead or nine minutes ahead so that their alarm goes off and they trick themselves into thinking that they're late so they get up.
00:15:13
Speaker
um But nowadays, like a lot of clocks are set automatically. so You really can't do that now. You have to get like an analog clock to to trick yourself. But most people's like cell phones, you can't set your clock ahead. You've got to, you can't.
00:15:25
Speaker
really trick yourself now. But NTP is used, I mean, it's synchronizing networks, network to synchronizing computers, it's synchronizing data. So data transfer is a huge thing. I mean, what what do we do all day, right? We're we're on the internet, we we're on our phones, phone calls, and we're we're you know listening to something, internet radio. So those things seem to be synchronized um you know, at least milliseconds, if not microseconds. So synchronization is important for timestamping, for just communications or power grids or things. Sending either data signals or power signals, you have to synchronize both sides so that they don't, you know, cross paths the wrong way.
00:16:08
Speaker
Yeah. So synchronization gets more and more important. And especially with the amount of bandwidth that we we shoved through the the airwaves and the internet, waves now. now we're We're trying to put so much information through. so The better synchronization and the better time you have on both sides of that synchronization, the more bandwidth you can you could do. so I'd say on these kind of levels, you know milliseconds, microseconds, that's the kind of things that you're that you're seeing. Other things, you know stock market transactions is a big one. High-speed trading, they want they really want nanoseconds. Let me know when you're ready to talk about nanoseconds, but right now,
00:16:47
Speaker
You know, NTP is this microsecond timing and it's it's most of the clocks, you know, most of the clocks that It's not just time of day to get to a meeting, but something better than that is probably synchronized through a network.
00:17:01
Speaker
Definitely. Now, if we're going to dive a little bit deeper into how we're looking at networks and how even NIST synchronizes its devices to atomic clocks, can you explain that for us a little bit and how that process works?
Principles of Atomic and Quartz Clocks
00:17:19
Speaker
Yeah, so an atomic clock is, free you want to if you want to delve into that, yes atomic clocks use the naturally occurring frequency from atoms. That's what makes it atomic.
00:17:32
Speaker
We're using frequency, so it's called resonance frequency. And so all atoms have different properties of their orbital levels and energy levels. So if you think back to chemistry class, you have an atom, we have a nucleus and you have these orbitals, right? And there's electrons on those those outer layers of orbitals.
00:17:52
Speaker
And so when I say different atoms have different resonant frequencies, A resonant frequency is something that happens when an electron jumps from one level to the other.
00:18:03
Speaker
And so if an atom goes into a lower energy state, like it decays into a lower energy state, we're not talking about like nuclear or I mean, ah like radioactive decay, but just an atom is decaying into a lower energy state.
00:18:17
Speaker
It might give off a photon because that energy goes somewhere, right? Yes. And so when an atom changed states, that's significant and it's related to frequency, either that photon or if we can hit an atom with a right frequency that it's it's resonant frequency for that particular state of atom, we can actually make that electron jump to a different orbital or...
00:18:39
Speaker
There's a a spin state. there's ah There's an orbital spin state of an electron and a frequency can make that spin go from one way to the other way. So there's all these different properties of atoms and it's related to frequency. So it's an energy level and spin relationship with frequency.
00:18:58
Speaker
So that means we have this really valuable thing that frequency can be used as a naturally occurring frequency. If you have a cesium atom, of a certain in a certain state and I have a cesium atom in that same state and we hit it with a frequency to make it change to another state and we do that to from one state to the same other state, you and I have the same frequency.
00:19:21
Speaker
So we can somewhat synchronize frequency without ever comparing or being in the same place. And so atomic clocks exploit this fact, this natural frequency that you can get in or out of atoms. And we actually use it as we generate a frequency and we hit atoms with the frequency.
00:19:38
Speaker
And did they change state? We move that frequency back and forth until we get that peak of where the most atoms change state. And we lock onto that. And we just keep generating a frequency and keep steering it till we get the most state changes. And how that moves and how that steer goes back and forth, that's really the stability. um So we just kind of try to fine tune it.
00:19:59
Speaker
And so if you turn off, you know if you get the right frequency, okay, we're on, we got the right frequency. If you turn off the atoms, then that frequency will be drifting. So you're constantly adjusting the frequency. it has to be on all the time because you're generating that frequency from, i mean, in a lot of atomic clocks, you're generating from a quartz crystal still. So so we lock to frequency is a naturally occurring thing from atoms.
00:20:23
Speaker
And so you have an atomic clock. And the the second is defined by cesium. So cesium atomic clocks are the very particular 9 gigahertz frequency that will make cesium atoms change from the F3 to the F state.
00:20:36
Speaker
And not that those states matter, except for pointing out that there's very specific state change that you're looking for. And so two atomic clocks that use a cesium atom are going to be very close to each other in frequency.
00:20:51
Speaker
And so there's rubidium and that's a different frequency. Rubidium atoms will change with a different frequency and those it's a little easier, a little cheaper, a little bit less power. So you can buy a rubidium oscillator for maybe a thousand dollars or two thousand or three thousand dollars.
00:21:07
Speaker
Whereas a cesium oscillator is going to cost upwards of 80,000 or a hundred thousand dollars. So. Much different. So not everyone can afford to see the cesium, which is why you have those, the rubidiums. It just depends on what you need and how often are you going to steer it or synchronize it.
00:21:23
Speaker
But once you have like the best and and really to get better than cesium, you can get a bunch of cesiums, right? And you average them together. And so then any of the instabilities might average out or cancel each other out.
00:21:37
Speaker
And so when I mentioned the time scale earlier, that's a bunch of atomic clocks and they're averaged together in a weighted average. So you know your math, right? a weighted average is a statistical average.
00:21:49
Speaker
And so in this case, what we do is the the clocks that are closest to the average get the most mathematical weight. And so that way, if a clock starts drifting off, say it's ah some temperature thermal runaway or it's running out of cesium, it's getting worse somehow or some components are failing.
00:22:05
Speaker
As it gets away from the average, it gets less mathematical weight. So it's not pulling the whole average with it. And so that's what a time scale is. So we have a bunch of atomic clocks and...
00:22:17
Speaker
and Different countries around the world have multiple atomic clocks in their time scale. And we all contribute our clock data to the Bureau of International Measurements, the BIPM in France. It's an international territory and within France. And they're the international metrology lab that takes all the data in for all the standards.
00:22:40
Speaker
So we all contribute clock data. And then they take the hundreds of clocks around the world and they do a weighted average. And they come up with UTC, the Coordinated Universal Time.
00:22:50
Speaker
Then they publish a document called the circular T that says, okay, we came up with the average. It was a while back, but what we did, we came up with this average. And here's how close your lab or your clocks were to that average.
00:23:02
Speaker
And so everyone can kind of gauge or steer to the international average. So we just get better and better over time. But I did want I don't want to point out something. When I was talking about atomic clocks and frequencies from atoms, I wasn't mentioning time at all.
00:23:17
Speaker
That was frequency. So atomic clocks are not really clocks. They are sorry to burst the bubble on that. But tell us more. Atomic clocks are oscillators. They're getting their frequency locked to resonance frequencies.
00:23:33
Speaker
So if I have a cesium oscillator, it's running at 5 megahertz or 10 megahertz output. It's very, very, very stable. That's a very well-defined 10 megahertz, right?
00:23:45
Speaker
But 10 megahertz is not time, it's frequency. Frequency is one over time. So that's 10 million cycles per second, 10 megahertz, right? So that's cycles. A cycle is a sine wave.
00:23:56
Speaker
That sine wave can be anywhere in phase, right? And it's perpetually oscillating. Yeah, it's propentially oscillating. And so it's very stable, but we have to decide what's on time. Like when that zero crossing happens, you could mark that and then you count 10 million later, 10 million cycles later, and then you have one second.
00:24:20
Speaker
So now you've converted by dividing down, you've divided down this 10 megahertz by 10 million. And now you have one cycle per second, which is one second. And so when I when i explain like what clocks and ah how we tell time ah to to anybody of any age, young or old, I define a clock as something periodic and something that's counting.
00:24:47
Speaker
And so if you think about like the really early clocks, like the sun, right? You could count a day because every time the sun is directly overhead, And then 24 hours later, this the sun's directly overhead. Like that's a day, right? And the sun is periodic.
00:25:03
Speaker
It's not human controlled. It's just happening over and over again. And so you count it. And you can break down that sun going overhead with a sundial, right? Where you're now you're taking that one day thing and you're making us using a shadow and some demarcations to break it down to smaller amounts.
00:25:21
Speaker
If you have a grandfather clock is another good example. That's a mechanical human made thing. So the the period of the oscillations is based on the the length of the arm.
00:25:34
Speaker
And so you can tune it by changing the length of that arm, but it's going to swing back and forth. And so that's your oscillator. And it's not perpetual oscillation, right? You have some kind of a weight or a spring that you you've pulled to add a little bit of energy to that to that arm each time so that it keeps swinging.
00:25:53
Speaker
But the rate of which is of which it swings is based on the length of the arm. So you can calibrate that. But then you got to count it, right? So you could someone can be there like one, two, three, you know, 59, one minute, 58, 59, two minutes.
00:26:08
Speaker
fifty eight fifty nine two minutes Right. But ah that kind of a clock, a pendulum clock or a grandfather clock, there's gears in there. Right. So every time that goes back and forth, another gear turns in the same direction.
00:26:21
Speaker
And so that gear is the seconds. And then there's a there's a different gear on that, that once that 60 teeth of that gear goes around, then that one clicks once. That's the minute. Right.
00:26:33
Speaker
Yeah. And then that one has 60 teeth or however, it might be more smooth than that and not just go every minute. but There's gears for the minute hand, gears for the hour hand. And so that kind of a clock is something periodic, the swinging of the arm and the counter, which is the gears.
00:26:53
Speaker
In the case of a quartz crystal, you know early nineteen hundreds quartz crystals, you know, we found that this piezoelectric quality of of quartz, where you can you can take a thin sliver of quartz and bend it, and it actually gives off a voltage.
00:27:08
Speaker
Or if you have a thin sliver of quartz and you give it a voltage, then it will bend. And so you can use this by making an electronic circuit that makes some piece of quartz bend.
00:27:19
Speaker
Then when it bends back, it gives off a voltage. You feed that back into itself and it makes it bend again and then it unbends. And so you can create this oscillation by this bending back and forth. And properties of the electronic circuit and the properties of the quartz and how it was cut and the shape and the size can dictate that frequency.
00:27:38
Speaker
So you can actually tune tune in that oscillator. And so there again, a quartz watch or quartz oscillator is just the oscillator part.
00:27:49
Speaker
And then you have to ah you have to count it. And so you have some 32,768 hertz oscillator. seven hundred and sixty eight hertz oscillator Inside your wristwatch.
00:28:03
Speaker
um And you have to count that down by 32,768 to get to one second. Do you know what the magic of that number 32,768? No. me. two seventh. you know what the magic of that number is thirty two seven sixty eight no help me it's two to the seven So in in early digital electronics, it's very easy to divide by two, right? So if you have 32,768 and you just send it through where you, a sine wave, you count every other cycle.
00:28:33
Speaker
Now you have a frequency at half. And then you you count that every other cycle or like a digital electronics, maybe it's just measuring pulses. And it says like, okay, just look at every other pulse and you cant you can look at every other pulse and divide it by two.
00:28:49
Speaker
And so you divide by two, divide by two, divide by two, divide by two, and then you get to one second. And so because of quartz crystal, there's no quartz crystal that won't run that slowly at one second.
00:29:00
Speaker
That would be, know, it has to be in this kind of self oscillation. So you have to get a higher frequency and really the higher frequency, the better. So if you have like a 10 gigahertz quartz crystal,
00:29:12
Speaker
Then you divide it by 10 million. And if it's it's certain if it's stable at 10 megahertz, you're dividing down that stability. And so you you could, if you're very stable at 10 megahertz, you're much even more stable at one hertz.
00:29:29
Speaker
And so the higher frequency, the better. so the So the benefit of like atomic clocks, they might be ah at a gigahertz or nine gigahertz in the case of cesium. So you're really stable at nine gigahertz. When you divide that down to a second, you're it's yeah that clock will not be off by a second in millions of years.
00:29:46
Speaker
And so um you have this huge advantage of higher frequencies and you're stable at a high frequency and you divide that down to a low frequency to get your your time your time of day or your seconds.
00:29:59
Speaker
I think, what was the original question? Did I answer it? You're talking about how do you get time from atomic clocks? I think you did. And as you're talking about that, I just had the floating thought of, we talk about time, everyone talks about crystals, crystal vibrations.
00:30:14
Speaker
And I just had the silly little joke in my head. Is it time? Is it magic? Or is it witchcraft? Yeah. Witchcraft. We have a big quartz crystal in our lobby.
00:30:26
Speaker
Yep. That's like a big natural quartz crystal. It's really cool. And so when I tell people how quartz, you know, what that this, how quartz clocks work, I talk about this crystal and I, and I, and I talk about, cause lot of people have heard about like the magic,
00:30:40
Speaker
magic why magic magic of magic of quartz you know the quartz has vibrations so if someone's holding it and they're feeling of vibrations and then it's kind of magic and like I don't have much to say about that kind of magic or witchcraft side of it but it's true that quartz does have this Really interesting, maybe magic quality of this piezoelectric, which is physical, right? It's right not magic.
00:31:07
Speaker
It's physics. But the fact that the fact that if you take a rock or you this crystal and bend it and it gives off a voltage, it it really is storing some energy in some way. And if it bends back, if you bend it and it gives off a voltage...
00:31:23
Speaker
And then if you give it a voltage and it bends, I mean, that's a pretty cool property. Call it what you will, but it's pretty neat. So precise timekeeping impacts many different sectors. So how industries such as telecommunications, financial services and IT t rely on NIST for their operations?
00:31:43
Speaker
And can you find some examples of how these critical systems or services depend on accuracy of time by the services that are provided by Nest? Yeah, so it's a little hard to explain like something like the power grid, but power grid is time and frequency, right? You're sending off voltage in a three-phase power down the lines and then you're combining those back for some voltage at 60 Hertz, for instance, like why you plug something in the wall that's 60 Hertz.
00:32:17
Speaker
right But from power station to power station, they're exchanging energy back and forth with each other. So they have to be in phase. So they the one way is if you they have some way to see the phase coming in and match that. Right.
00:32:34
Speaker
But two different sides coming in, it's handy or helpful if they are in phase to begin with. So then when you're sending power back and forth and it's already in phase.
00:32:47
Speaker
So how you do that is you have to synchronize the stations to each other so that when they send power back and forth, they they come there in phase when they get there. So power grid is a great example.
00:32:59
Speaker
I mentioned a little bit about Internet and ah stock market is a good example of you have millions, maybe billions of transactions happening at a time, like every second, you say a billion in a second.
00:33:14
Speaker
And so which one came first, right? If you don't have an accurate clock, ah you say, okay, well, a billion happened in this second. And, you know they used to, it used to be, i think, eight seconds, seven or eight seconds.
00:33:28
Speaker
So all the timestamps on stock transactions, you know, in those, in movies where they're like passing back and forth, like pieces of paper and they're buying, selling. The timestamps were something like seven seconds.
00:33:42
Speaker
So it means every trade that happened within seven seconds happened at the same time. And then it went down. So like, so who gets to decide? Whoever's like stacking up those papers. I'm going to put this one first and then this one or whatever. Or then it was down to one second.
00:33:57
Speaker
Right now, the United States stock market designation for timestamps is 50 milliseconds. And that's huge, right? you they're You're still having like almost a billion transactions happen in there.
00:34:11
Speaker
Right. So somebody... One teeny tiny click, two people clicking at the same time is not the same time. So you need to decide because if someone bought it and the other one person sold it, or if you sold a lot of them,
00:34:26
Speaker
the price is gonna maybe go down or depending on if there's offers, price is gonna go up. And so every trade changes the price. So you really have to tell who got in their trade first and second and third, because the price is changing between those. So high-speed traders can exploit this fact because they say, I see a trade that's happening and I'm going to get in on that before that trade because they're faster or they're seeing something or they see something, they're reacting faster because they have better clocks or better synchronization.
00:35:01
Speaker
And so... I'm not sure if they need nanoseconds, but in this in the markets, and the high-speed trading, they want nanoseconds. So whether or not they can use nanoseconds or tens of nanoseconds, they all want it. They want to be as close as they can so the latency is is small and they want to have exact synchronization so that they can plan and ahead by microseconds or nanoseconds.
00:35:27
Speaker
And so we have basically SNS-steered clocks and time services inside of data centers all over the world. And at first you're thinking, wait, you're just helping the high-speed traders, are you? Like, great, you know, thanks a lot, NIST.
00:35:43
Speaker
But what happens is if we're in the data center and people are synchronizing and someone wants an advantage by being within 10 nanoseconds, but then the other person gets that same advantage because they're getting the same service.
00:35:57
Speaker
Now you've equalized them. They both have the advantage, which means no one has the advantage. Right. So what we've essentially done is we've made the stock market accountable.
00:36:10
Speaker
Because now you have a lot better clocks and a lot better. You can time those transactions better, even though the legal limit is still 50 milliseconds, but people are doing it a lot faster. And so you're helping order those transactions and cut down on those advantages by homogenizing the advantage.
00:36:28
Speaker
And so this has become very popular. So we're in lots of data centers. NPL, the national lab in England, has services for their data centers too.
00:36:39
Speaker
And in Europe, it's 100 microseconds. So it's it's a more stringent rule than here, and but it's 100 microseconds to UTC. So in America, the stock market rule is 50 milliseconds to NIST. So we're you kind of, we're part of that legal traceability.
00:36:55
Speaker
In Europe, it's 100 microseconds to UTC, which means if we have our system at a data center in London, people can use UTC NIST to get to UTC, just like they could use NPL to get to UTC.
00:37:10
Speaker
So we have this advantage that you're able to use UTC NIST in any data centers, pretty much, I think, around the world because we have them in a lot of places. Whereas in America, you couldn't really use NPL because the legal traceability is to NIST.
00:37:24
Speaker
Now... Tell you a little bit more about how this works also for power grids and internet services. Yeah. So power grids, I'm not, I'm not, I don't think I could explain any better with power grids, except for they need to be in phase.
00:37:41
Speaker
If you had two, if you had two say 60 Hertz signals coming in, just as an example, I mean, three phase power is all different, but, and you try to just put those in the same wire. If they're not in phase,
00:37:52
Speaker
they're going to explode or they're going to cause power going back the other way. Just being in phase to some degree better than what microsecond time synchronization could do for that phase determination is good enough. I think it's something like 10 or 12 microseconds in time to keep that phase in for power grids. Internet is more about...
00:38:17
Speaker
yeah You know, you send a JPEG somewhere, an email, right? It's all ones and zeros. You know, everyone says ones and zeros. Here's where it actually becomes ones and zeros because that JPEG is data. And it's not just ones and zeros. It's a bunch of ASCII characters all encoded and compressed.
00:38:34
Speaker
And you send that. And some of the data might go, you know, from you to Chicago to California. Some of the data might go from you to Florida to California. They're just picking whatever is the the fastest way to get packets through.
00:38:47
Speaker
And so that all of those packets from that one JPEG are not going the same route. But when they get to where they're going, it's getting all these packets in. You know the catchers, they are catching with this catcher's mitt. And you're catching all these packets in. Like some are coming from here, some are coming from there.
00:39:03
Speaker
But you know you need to put all these packets back together. And so in that sense, the timing is not that critical, except for... If you're the catcher and you're catching millions of packets from all over, you need to have a good oscillator to be able to just accept those packets or send those packets.
00:39:25
Speaker
And so the more stable the oscillator, the more bandwidth you can handle. And so it's very hard to explain. One other sector that I wanted to talk about because you were saying, you know what sectors are important.
00:39:36
Speaker
GPS. is kind of ubiquitous, right? Global Positioning Service or Global Positioning Satellites Service. And so the only way the only way that GPS works is because of atomic clocks.
00:39:52
Speaker
So let's explain that a little bit because it's fascinating.
Impact of Precise Timekeeping on Technology
00:39:56
Speaker
Yes. So if a satellite's sending a signal to you and it it says, hey, here's where I am, its position and in the orbit, semi-hemispheric orbit, it's sending you a signal that hey, here's what time it is.
00:40:11
Speaker
Here's where I am. And you get that information. You're like, OK, great. here's what time it was when they sent it, not what time it was when I received it. Same as problem that with that time.gov problem, right?
00:40:24
Speaker
It sends it, but there's some latency to when you get it, right? And the satellite says where it is. And then a second later, it says, here's another signal ah ah what time it is and where I am. But now I'm over here. I'm, cause it's moving in orbit, right?
00:40:39
Speaker
So it's, you're getting all these packets from a satellite, where it is and what time it is and lots of other stuff. But, Break it down into that. So if you know roughly where you are on the globe, you could say, OK, well, I got this and the satellite told me where it is. And so I might know kind of where I am already.
00:40:59
Speaker
So if it took and I might know kind of what time it is. Right. So. Or you could say, okay, well, the first satellite that comes in, I'm just going to synchronize to that, even though I know it's late, but I'm going to synchronize.
00:41:11
Speaker
ah But then I get a signal from another satellite at the same time or almost the same time, and it's much farther away. So you get that pulse a little bit later from the one that's farther away, even though they both satellites sent them at the same time, but one of them you got so quicker than the other one because you're you're closer to that.
00:41:30
Speaker
So imagine, you know, they say about triangulation. I was just going to say that. Yeah. So if if you someone someone yells and three people hear it or three people yell and one people one person hears it and you hear it slightly different, you start to say, oh, I'm closer to that person and farther from that person.
00:41:48
Speaker
And this person's about in the middle. So you can start walking towards the one, the closest one or walking towards the farthest one and see if the closer one gets farther or not because it's a triangle.
00:41:59
Speaker
So you could get to the some point in that triangle where they all get to the sound all gets to you at the same time. So now if you know where those people are, now you know where you are because you've you've kind of centered yourself amongst the three signals.
00:42:15
Speaker
So imagine if you have eight signals. of those signals or 12 of those signals or 100 of those signals, however many satellites you can see, and they're all sending information. They're all synchronized to within nanoseconds or a nanosecond to each other.
00:42:31
Speaker
um And that's a whole different part of that. But for now, let's just assume that they all are synchronized to each other and they're all sending you signals, but you're getting those at slightly different times.
00:42:42
Speaker
So you know where they are and you know how long relatively they took. So you can start to figure out your position and your time. And the more satellites you have in view, the quicker and better position you can get.
00:42:56
Speaker
And so the only way that works is because... it's It's radio propagation, which through free space or not really totally free space, but free space. which So that's roughly the same for one satellite to you as the other satellite to you. And they're synchronized.
00:43:14
Speaker
And so you're getting the signals at different times, which means you know your relative distance between you and all those satellites and you know where they are. So you can start to average into where you are. When I say free space, not really.
00:43:26
Speaker
Like, they're all going through the ionosphere, the troposphere, the magnetosphere, right? So the signals actually get refracted through there, and they don't come exactly...
00:43:37
Speaker
at the same rate because of that refraction. Or if there's a solar storm, you know we've had a lot of solar storms in the last year, and that actually affects the signals of GPS or other satellites coming through that magnetosphere.
00:43:52
Speaker
And so GPS is noisier during a but high magnetic, solar magnetic period. and so But nevertheless, Because of atomic clocks and because of radio propagation from a satellite to you, it's pretty amazing.
00:44:05
Speaker
Almost, you could almost call it magic. It's incredible how you could determine your time within nanoseconds and also your position within, you know, meters or even centimeters with the right kind of math around it.
00:44:23
Speaker
So simply put, this These frequencies will actually affect the GPS systems and deep space navigation. Yeah, yeah. So the deep space navigation and VLBI, very long baseline interferometry, those things, you need to be very well synchronized in time and frequency, because if you're ah if you're off in frequency, you're not going to communicate with that long, very distant ship or transmitter. And so you need to have very accurate and stable oscillators and synchronization to communicate in, like you're saying, deep space. Okay.
00:45:02
Speaker
So from a purely theoretical perspective, what impact would these clocks have on fundamental physics, especially when testing Einstein's theory of relativity?
00:45:15
Speaker
Yeah, you know, it's amazing. A lot of things that Einstein did are amazing, but like right something like this, like, I mean, Einstein, how did Einstein know like the the subparticles of atoms without being able to observe them?
00:45:27
Speaker
How did he know like time dilation? I mean, it's it's wild that he came up with that, but he did. And eventually clocks caught up to that where it's like, okay, now we have cesium based atomic clocks and we're going to accelerate So you synchronize two clocks and you accelerate one, put it on the Concorde jet in the 80s, and you're going to leave the other one and you're going to fly around the world or something.
00:45:57
Speaker
And then you're going to decelerate back and you're going to compare those clocks and those clocks because the. they um because of the, the velocity, the G force and the, um, the relative motion of one clock to the other one clock went, was slower than the other for that period.
00:46:18
Speaker
And so you're like, wow, we can actually measure it now at the time of clocks, but fast forward to modern times in the last 10 or 15 years with optical atomic clocks, we've got to the point where even the, um,
00:46:35
Speaker
The relativity of a clock just on based on gravity, we can measure because the atomic clocks are so good now, with the atomic the optical atomic clock. So in other words, if you have some frequency on a laser table with an optical clock and you change the height of that laser table by several centimeters,
00:46:59
Speaker
that frequency, that ah changes, that relative frequency changes based on its effects due to gravity. And so you're talking about measuring like very, like very minute changes of clocks themselves, relativity using atomic clocks with just by moving it by like, you know, 10 centimeters or something.
00:47:24
Speaker
Okay, so let's take half a step back here because you just use switch to optical clocks. Now, let's explore these mechanics a little bit and how do optical, I believe it's optical lattice clocks work and what makes them more accurate than some of these previous generations of the atomic clocks.
00:47:49
Speaker
Yeah. So remember I said earlier that the higher the frequency, the better. Right. So an atomic clock is better a course oscillator at 10 megahertz because you're at this high frequency and you're stable at this high frequency.
00:48:08
Speaker
Absolutely. like Those kind of atomic clocks like cesium, rubidium, those are microwave clocks. Just like your microwave oven, it's microwave frequencies. So nine gigahertz is within that, that's the cesium frequency.
00:48:21
Speaker
So optical clocks are based on atomic transitions that happen in an optical field, like a laser. So instead of hitting an atom with a certain frequency and making it change states, you're hitting it with also a frequency, but a light frequency.
00:48:37
Speaker
And so now your frequency is not derived from quartz crystals multiplied up to nine gigahertz. It's derived from locked lasers at very high frequencies. and those So those atomic transitions are like in the terahertz.
00:48:52
Speaker
So now you're locking to atomic transitions at very, very, very high frequencies. So when you divide that down to divide down that instability, it's incredibly better. So like if a cesium clock is good to parts in 10 to the 14th, so that means like, you know, you have some frequency with 14 decimal places.
00:49:17
Speaker
of 14 zeros or whatever, like one hertz at 14 zeros or 10 megahertz at several zeros, right? um When you're talking about, so that so we say like parts in 10 to the 14th or 10, you know, like five times 10 to the minus 14th.
00:49:35
Speaker
same Same thing, different kind of nomenclature. The optical clocks are stable at like parts in 10 to the 17th or 18th. So many, many magnitudes better. And so because it's higher frequency, because you can determine optical, it's a higher frequency. So every zero crossing is you know much closer together.
00:49:55
Speaker
So it's just a much more fine tuning with those high frequencies. And you can lock in those optical frequencies so much better. Optical lattice clock is... So if you just say you have one atom and you hit it with a frequency and it changes state. So, OK, we know, okay their frequency is good, but you need a bunch of atoms.
00:50:13
Speaker
Right. And so with ion clocks, they would take it. They would take ions and put them in an electromagnetic field. So they line up charged particles because it's electromagnetic. They use charged particles and they can line those up with like putting forces in all directions. So you have a line.
00:50:30
Speaker
And then with a fine-tuned kind of um frequency, you can hit one atom with the frequency and make it change states where the other ones didn't. Or you go over and change another atom and change that frequency. So you're starting to have this quantum-based, so atomic states, subatomic particle, charge particle, ion clocks. You're able to change one and not the other.
00:50:54
Speaker
And so... You could really make a logic gate out of that. Like you're talking about ones and zeros. if this is it Did this change state or not? Is it in one of two states, one of three states? And so if you can control atoms, and you probably know where I'm going with this, but quantum computing, right?
00:51:12
Speaker
course. So some of our work, just all we're trying to do is make better clocks, right? So we're lining up atoms. We're turning on one atom and not the other. Because you're in ah in a cesium clock, you just got have a bunch of atoms kind of streaming past and you're you're hitting them with a frequency and you see how many change state on the other side.
00:51:30
Speaker
So that's kind of like the the shotgun approach, whereas right the ion clocks, you're actually taking ions, putting them in a row. And how fine-tuned can you just change one and not the other? So you're not like just blasting them all because they'll all change states. So you didn't really learn anything there.
00:51:45
Speaker
So if you have an optical clock and you have to line up these atoms, you would really have them in a line, right? But that's not efficient because if you wanted to have a billion atoms or 100 billion atoms or whatever, that's really getting to be a longer line.
00:52:00
Speaker
i mean, 100 billion atoms is not that long of a line, but the more atoms you have, it's a long line. So optical lattice says, OK, let's not put them in a line. Let's put them in a grid like a checkerboard.
00:52:11
Speaker
So you have atoms in every square in the checkerboard. So now you don't have to have this really long line. You have this this grid that's getting exponentially bigger by you know, if you just go out one more, you know, one more checkerboard grid, like a five by five grid,
00:52:29
Speaker
I was 25, right? If you just go one more level out, that's a six by six grid. You have 36. So an optical lattice is basically just putting those atoms in a well and you're spreading them out that way.
00:52:41
Speaker
And so you can you can interrogate those atoms with different frequencies and have in more data, basically more bandwidth, more data. It's all about getting more bandwidth, right? You have more atoms, but in this case, you can control them independently.
00:52:56
Speaker
And so that's that's where that lattice part comes in. it's it's the It's the checkerboard of atoms instead of like a long tube of atoms. Now you talked a little bit, you touched base it a little bit on ah quantum computing, not to foreshadow another episode that we're going to have in the next few weeks this season, but what advancement do researchers hope to achieve with optical clocks even over the next decade? And how could these advances open doors for new scientific discoveries or even drive new emerging technologies?
00:53:34
Speaker
Yeah, so we have this odd thing at NIST, and it's we're creating technologies or abilities with not really a total goal in mind.
00:53:46
Speaker
In other words, we're like, if we had better clocks, what could we do? Like, i have no idea. And then, you know, like GPS comes out. We're like, see they figured we we came up with these really good clocks starting in the 60s, the atomic clocks.
00:53:58
Speaker
GPS didn't come out until the late 70s or early 80s. So... so The atomic clock development wasn't because someone had an idea, i think, for GPS. It was like, let's just figure out how to make better clocks.
00:54:12
Speaker
Who knows what we need them for, right? But so well once we come up with it, someone will come up with a use for it. So in that kind of pure research kind of... angle, I think that we we don't really even know what the possibilities are.
00:54:25
Speaker
i mean, we're we're the pure research side. It's like, we we just want more stable clocks. We've got high frequency. um We want them to be able to run for longer. So in other words, like, you know, a lot of optical clocks nowadays is just a laser table with a bunch of lasers and mirrors and, right you know,
00:54:42
Speaker
And it can run for five minutes or run for 24 hours or something, but it doesn't stay locked. there's two It's too much fine tuning. So we got it to a point where like here, we know it's possible.
00:54:56
Speaker
But then commercial companies take that idea. They read our papers. you know they We talk at conferences. they They learn that it's possible. They figure out how to miniaturize it and sell it.
00:55:08
Speaker
And so once they can sell it, then people can actually use it. So in other words, like we have for some part in 10 to the 18th, you know, stability that we can't get that out anywhere. There's no, there's no time.gov.
00:55:22
Speaker
There's no satellite. that There's nothing that you can get that from even across the street. Right. I guess across the street you could with an optical fiber, but, You can't get that time out.
00:55:33
Speaker
But what we did is we proved the technology and then a company can figure out how to make it smaller. And hopefully that stays running for a long time so that then you can have two really good clocks.
00:55:45
Speaker
So then if you have a really good clock, you can synchronize it one place and just carry it to another place and take that synchronization kind of and put it at the next place. And so it's hard for us to say what how this is even going to be used, except for um you know more data, faster, um you know more minute positioning, I guess, if ah if the atomic clocks in space are better.
00:56:09
Speaker
and quantum some computers, I mean, you know it started out by having 10 atoms, kind 10 transistors, right? like right build or You could build a radio with 10 transistors, but you can't build a computer. A computer's got a CPU. It's got like a billion...
00:56:25
Speaker
or billions of of transistors on it, billions of gates, right? So if transistors got smaller and smaller to where you could put a billion on a chip, what if you could put, you know, 10 billion or hundred billion, million billions on a chip?
00:56:40
Speaker
And so that if if a single atom is a gate, that's the fastest, smallest gate you could have. And so there's no quantum some computer that has a billion gates on it yet.
00:56:52
Speaker
But it's getting there, right? So if it does, then now you're talking about like processing, cryptography, you know, solving your computer, solving equations and stuff. So you can be able to encrypt, you know massive amounts of digits and things like that. So at NIST, we have not really ever been too much in the business of figuring out what to do with the clocks. We just try to make better clocks.
00:57:17
Speaker
Completely understandable. Out of curiosity, is there anything that i you've learned over the course of your career about time that you found extremely interesting or a fun fact or quirky little topic.
00:57:36
Speaker
Man, it's so esoteric that everything right about it's kind of quirky. of you know and It's all relative. It's relative. It's quirky. And and you think about like what you know now or say what someone knew 100 years ago, about so two things that happened at exactly the same exactly the same time.
00:57:55
Speaker
right right It was exactly the same time and what they could measure it with. in that era. Clocks get better and we're like, oh, we actually know that this one happened before this one, right? so even though, but these other ones are at the same exact time, but these other things we can measure. So then you find two other things that seemed like that they were the exact same time 20 years ago. and Now you can tell one happened before the other one.
00:58:18
Speaker
So that simultaneity It means that you just need better clocks and you can you can see that things aren't simultaneous. so So is anything simultaneous? I mean, just depends on what minute level you're talking about. So think what I've kind of learned is that the closer you look at something, the less you know, almost, you know, and that's the same. but So if you want to think on a grander scale, like about people, right? Like you meet someone for the first time or you get to know them,
00:58:48
Speaker
and you're like, oh, okay, I kind of know that person, you you spend a week with them or you something, or you spend your yeah know a year with them, or you get married, or you know someone for a long time. The more you know about someone, you realize,
00:59:00
Speaker
You keep learning more and more things. And you're like, I feel like I know less about them now than I did when I first met them because I know all i know a lot more things, which means that there's a lot more things that I realize I don't know.
00:59:12
Speaker
So it's come it's like that. So that metrology is like looking at things closer and closer to find out all the things that you don't know. And if you have the best clock in the world, how do you know?
00:59:24
Speaker
Right. There's nothing to compare it to. except something worse. And if it's something worse, then all you're measuring is how good the bad clock is. So you need a better clock to tell you how much your current, clock how good your current clock is.
00:59:39
Speaker
And so that's relative in a different way, right? It's like, so if you're going to characterize how good is the best clock, You have to characterize all of the things you oh the things that you got to that stability.
00:59:53
Speaker
So you're looking at you know temperature or solar radiation or black body radiation. You have to basically, to come up with the uncertainty analysis of a sum clock, you have to come up with the uncertainty, all of the uncertainties of all of the things you got to make that clock, and you come up with an uncertainty estimate.
01:00:12
Speaker
And so until you have better clocks, you don't know how good that clock was. so like So it's kind of like that, you you submit data to the circular T for all the countries, but you don't know how good you were until you got like several weeks later, you got that result back of that international average.
01:00:29
Speaker
So like, okay, here's how good you were. three weeks ago or four weeks ago, you know? So I feel like if you have the really good clock, you can't tell how good it is. You can make another one. And if they're both at the same rate, you know, yeah it's repeatable, but you still don't know how good they really are until you come up with better clock. And now you say, okay, well now we know how good those clocks were. so I think that kind of relative nature of things and looking at things closer and closer, I think that's something that I've i've learned and gotten interested in.
01:01:00
Speaker
But I also, you know I did a blog post um at at New Year's um coming from into 2023 2024. It's like on the NIST like on the nist vlog And it's it's more about the relative, how our our perception of time.
01:01:15
Speaker
And, you know, if you have a really busy day, it might go by really fast. Or if you're just watching a pot boil, it might seem like it takes forever or whatever. Or if you put if you're put in a dark room with no no light, no sound, no external influences.
01:01:31
Speaker
You sit there and if you said, okay, tell me when a minute's gone by tell me when five minutes has gone by. Maybe someone would say after 20 seconds, they think a minute's gone by or maybe after four minutes, they think a minute's gone by.
01:01:43
Speaker
We perceive of time based on all of these things happening around us or what we're doing. And so that's why time is so relative. Or like when you get older, you know, an 80 year old person thinks of a year as a really short time, whereas a 15 year old person thinks of ah a year as like a really big possibly. I mean, the theories are, you know, that's relative. That's the relative or the ratio of that one year in the 80 year old life is a small chunk.
01:02:13
Speaker
So that's that seems very fast to them and because they've lived for 80 years. And so I kind of tried to delve into that a little and, and you know research just people's perceptions of time. So I think it's that, you know, it's ah it's the psychology and the physics kind of coming together when it's something
Historical vs. Modern Timekeeping
01:02:32
Speaker
like time.
01:02:32
Speaker
Now, as we're coming up, we're just over the hour for right now for recording. Is there anything that you would like to add or have listeners take away from this conversation?
01:02:46
Speaker
I think leap seconds is a fun thing to talk about. Okay. Tell me more. So... imagine early timekeepers and like the ancient Sumerians, Mesopotamians. Remember I talked about the sun as a clock.
01:02:59
Speaker
Yeah. And the astronomers were like the the super smart folks, right? They said, Hey, look, We can tell where the stars are and you know this the sun is up for less time you know in this periodic. you know like If you see these things where the sun doesn't go all the way overhead, it goes over there. So it means like you're not on the equator.
01:03:21
Speaker
So the astronomers were seeing things that other people weren't seeing. And so they were the original, really the original timekeepers. They were the ones who were saying, look, things are cyclical and the earth is rotating on its own axis.
01:03:34
Speaker
you know, the sun's not going around the earth. The earth is rotating. And so it's the sun goes by every time the earth rotates. But... The Earth processes around the sun.
01:03:45
Speaker
So you have the seasons. So you start you start thinking about time and periodicity, not in one day, one minute, one hour, but in one year or multiple years. And so you're talking about like long term.
01:03:59
Speaker
But really what that all boils down to is Earth and its rotation is their clock. You know, it's that original clock. It's that periodicity of the Earth rotating.
01:04:11
Speaker
And so you can count the days and years with Earth, right? But when clocks got better, they're like, oh, wait, we got to adjust the clocks because they're getting off from what the Earth is, right?
01:04:25
Speaker
But when you get even better clocks, like atomic clocks, you're like, the atomic clock doesn't care where Earth is. It's just going at its own stable thing. It's much more. So what you realize is the Earth is not a very good clock, even it's been a clock for thousands of years or whatever, because you got a better clock.
01:04:41
Speaker
And so once we define time as based on atomic clocks, which happened with a treaty or agreement in 1968, but it started in 1972, we said, okay, we said okay We're going to base the definition of a second based on atomic clocks.
01:04:56
Speaker
We're not going to move it when we find that Earth is changing and we thought that we thought the clocks were changing, but it was really, really Earth changing. You know the Earth's axis is on a wobble.
01:05:07
Speaker
And then you and you have like something like an earthquake or whatever, and it wobbles and the the tides and things. Earth is not really that good of a clock. But we didn't know that until atomic clocks. And so now we said, OK, we're going to define time by atomic clocks, but we're going to keep that Earth time significant.
01:05:24
Speaker
So when Earth gets off by by about around a half a second from our atomic clocks, which we know are more stable, We're going to adjust just the time of day by a second. So that's called a leap second.
01:05:40
Speaker
And so basically you add an extra second, kind of like leap day where you add an extra day in the calendars because the Unfortunately, when the Earth goes around the sun, it's not exactly 365 days. So you have to have leap days every so often.
01:05:53
Speaker
And so this leap second lets Earth catch back up because the original definition of the second, and um the duration of the second was kind of based on.
01:06:06
Speaker
a long time's worth of data, 100 years of data. And so Earth was getting behind, but not behind at a steady rate. If it's the steady rate, was just like, oh, we should change the definition. Some some years you would have a leap second.
01:06:19
Speaker
Sometimes you wouldn't have a leap second for two years. And so that's Earth being inconsistent. And so when we get about a half second off, and this is based on long baseline interferometry, there's ah there's an agency called the International Earth Rotation Service.
01:06:36
Speaker
So there they're astronomers with good clocks, right? So they say, here's how far Earth has gotten off. And when Earth is behind by about half second, they say, let's do a leap second.
01:06:48
Speaker
Then that lets Earth catch up and get ahead by a half second. So then we were never off by more than 0.5 or 0.6 seconds because we went from off by 0.5 or 0.6 to 0.4 or 0.5 seconds. point four point five So that's just us trying to stay cool with the astronomers. They're like, look, like Earth is the original clock. Like, don't leave us behind.
01:07:08
Speaker
And so since 1972, there's been equivalent of 37 seconds.
01:07:12
Speaker
seconds And so that's how far Earth has gotten off, except Earth has changed its rate. And we haven't had a leap second since 2016. So Earth has actually caught up to us.
01:07:31
Speaker
Yeah, we're speeding up. Yeah. And so so that rate has caught up. So now we're about at zero. But if it keeps speeding up, we're going to have to take away a leap second.
01:07:43
Speaker
what they call a negative leap second. Because we would, so instead a leap second is an extra second. So you go like, it's like 58, 59, zero for the next minute. You go 58, 59, 60, then zero.
01:07:55
Speaker
then zero If there's a negative leap second, because Earth is ah ahead, then you're going to have to go 58, zero. You know? So, like... and And so far, that's never happened. So, any digital systems have a chip potential.
01:08:12
Speaker
Think of statistics, right? You're taking a data point every second. Now you have this this one minute that has 61 seconds in it. Like, it throws things off. Or clocks that synchronize are going to get freak out on their steering. And so, like...
01:08:26
Speaker
networks like Google and or maybe eBay, there's other ones that they've decided we're not doing this one second quantized second thing. We're going to distribute that second offset in six hours or 12 hours.
01:08:43
Speaker
So they actually change frequency in all of the network synchronization. They change frequency so that absorbs that second into they're changing the frequency before and after.
01:08:54
Speaker
So they've theyve they've they've they've kind of slid it in instead of this quantized second that can make a bunch of you know data have problems or whatever.
01:09:05
Speaker
So a negative leaf second has never been done. You guarantee that most systems are going to be able to handle that. That could be ah maybe catastrophic. I don't want to an alarmist or anything, but they're talking about getting rid of the leap second, not just for that, but just for systems that statistically have to deal with it.
01:09:24
Speaker
And so we could just say, forget leap seconds. We're going to go on these atomic clock frequencies. And... The IERS can still tell us how far Earth is off, and that's just going to get further and further off in one way or the other. it doesn't matter, but we don't have to change our our time of day. you know so It's kind of like um daylight saving time, right? like Everyone hates changing their clocks.
01:09:48
Speaker
It's so minute for some people that the leap second is almost trivial until it's a significant amount of time.
01:09:59
Speaker
Yeah, but in the case of networks, you're talking about bandwidth and data and all this stuff. As we were talking about, you need to divide that second by smaller and smaller. One second is a billion nanoseconds, right? So if you need nanoseconds and here you are with like slamming a whole nother second or taking away a second, that's billion nanoseconds. Like that's huge for some people. But, you know, if we had this, if you so if you got up in the morning and say, everyone set your clocks back by a second.
01:10:28
Speaker
That's not going to affect us, right? You might have been late to that meeting by five seconds, and now you're six seconds late or four seconds late. mean... In our daily lives, we don't think of a second.
01:10:40
Speaker
But that hour, everyone doesn't like that hour. that's Nobody likes that hour. that shit know That's the jet lag of daylight saving time. so But yeah, computers don't like that that second offset. So so they're talking about getting rid of the leap second. They've talked about it for 20 years. So it's probably getting closer.
01:10:57
Speaker
But that's another one of those kind of fascinating things that you might see on the news.
Conclusion: Future of Timekeeping
01:11:02
Speaker
And be like, what, what is this? You know, so now you kind of know it's another thing that time throws a wrench in the, in the works.
01:11:11
Speaker
Absolutely. Now, anything else before we wrap up? Um, man, I don't, I think we covered, we covered a lot of bases. Yeah. So, Andrew, thank you for coming on Breaking Math Podcast.
01:11:26
Speaker
It's been ah pleasure chatting with you about time. And for our listeners, or you can follow us on on our website at breakingmath.com and also Twitter or twitter or depending on what platform you like to call it for Breaking Math Pod. and now we're also on Blue Sky for Breaking Math.
01:11:48
Speaker
Just strictly breaking math. Thanks for joining us, folks.