CuriousMarc - 2022-10-31
We adjust our atomic clocks with the Zeeman effect, and dive even deeper into the quantum world. It's complicated. How an Atomic Clock Really Works, Round 1: https://youtu.be/eOti3kKWX-c HP 10638A Degausser: https://youtu.be/hvxX5A1nTIU Cathode Ray Tubes and the discovery of the electron: https://youtu.be/zqUlfMmhJQs Our sponsors - PCBWay: fast turn PCBs, https://www.pcbway.com - Electro-Rent: https://www.electrorent.com - Keysight: test instruments: https://www.keysight.com - Samtec: connectors: https://www.samtec.com - R&D Microwaves: https://rdmicrowaves.com Support the team on Patreon: https://www.patreon.com/curiousmarc Buy our merch on Teespring: https://teespring.com/stores/curiousmarcs-store Get more info on the companion site: https://www.curiousmarc.com Contact info: https://www.youtube.com/curiousmarc/about Music Credit: Crinoline Dreams by Kevin McLeod
I never thought that adjusting a clock could be so interesting!
Hands down the most interesting I've ever en count ered.
Now to put one on the roof and one in the basement.
Would the relative gravitational force at that scale be significant enough to affect the precision of these clocks?
@@tristshapez Yes, if you waited long enough to observe the cumulative drift in timekeeping. The challenge here is to keep both units running continuously, on frequency, without any hiccups, so that the elevated unit can systematically gain a small integer number of total cycles over the duration of the exercise.
For reference, the article cited below says that a clock elevated one kilometer above its mate will gain 31 microseconds in ten years. Let's scale that down to 100 meters and one year. Then the elevated clock will gain 0.31 microseconds or 310 nanoseconds. (Note that the speed of light is about one foot per nanosecond. Oddly enough, if you were on the ground using binoculars to peer at the clock 100 meters above you, it would take 333 nanoseconds to observe the display so that, on the one-year anniversary, the observer on the ground would declare the two clocks reported the same time from his vantage point.)
https://www.washingtonpost.com/national/health-science/an-atomic-clock-is-used-to-measure-not-time-but-the-height-of-mountains/2018/02/23/5a845166-11c3-11e8-9570-29c9830535e5_story.html
There was an experiment done with optical atomic clocks that could detect the lab bench being jacked up a few inches!
Funny! :)
You know that’s coming lol
Excellent discussion! I was happy to see the 5334B Frequency counter in the video as I was the Service Engineer for that product and designed the self test and service strategy for it.
Awesome! I love that counter, it’s my primary instrument below 1.5 GHz. I have the high stability oven oscillator in it and calibrate it once a year with the Cesium. It stays pretty much on target.
@@CuriousMarc Hopefully the service manual for that counter is dramatically simpler than the 5061. I went to great pains to have the unit troubleshoot itself from known good kernal out to the edge circuits. Most of the testing (assuming the microprocessor is alive and the power supplies work) is just hooking it's outputs up to it's inputs and running it through it's various switching paths. It doesn't require a lot of external test equipment to test and calibrate it. This also greatly simplified the amount of production test equipment needed on the manufacturing line to test it out. I got a nice internal award for that. This was way back in 1980 IIRC so I'm glad you are getting good use out of it.
@@brianbeasley7270 Just out of interest, do you know any of the people that worked on the HP53131/2A? I've always wondered what the story behind the genuinely awful standard oscillator in those counters was - I suspect it has to be accountant related, since I can't see the engineers being exactly happy with shipping a product with a borderline unusable internal timebase.
@@TrimeshSZ Sorry, I didn't know any of them. That was after I left the division. It was most likely a low cost oscillator for the time to lower the price. Sorry it wasn't up to standards!
@@brianbeasley7270 Well, it just meant you had to use an external ref or get the (excellent) high or ultra-high stability options - it was just that a 5 part in 10^-6 oscillator was not what I expected to find in a 10/12 digit/s HP counter!
I knew several of the words used in this video.
18:00 Actually, that was very clear. Best ever. I found myself wishing you had chosen not to skip some of the details e.g. how the C-field disallows certain transitions. Very, very satisfying learning. Thanks.
I would be very up for a follow up/bonus video about this too yes.
I had dropped the relativity course in university due to getting lost and it being an optional course of luxury. I appreciate the review and explanation!
Agree. Excellent explanation.
I believe the other transitions destructively interfere under the influence of the C-field’s frequency :)
@@kaitlyn__L Ahh, thank you. Make sense.
I, for one, get excited when the elevator music starts. Great stuff! I look forward to seeing how you use your clocks.
My electrical engineering internship was in the Time and Frequency Division at NIST, Phase Noise Measurement group, '93-97, just down the hall from the NIST-7 cesium beam standard. 25 years later, after working in RF communication systems and test equipment, I recently started working at ColdQuanta on quantum computers based on laser cooled neutral Cesium atoms.
Your explanation of Zeeman splitting, and demonstration of the tuning, really helped my understanding of clocks and MOTs. I know many of these concepts as absorbed facts I can regurgitate, without much intuition. It also gave some more insight into the march from parts in 1e14, to 5e-15 of NIST-7, to 1or2e-16 of NIST-F2. And, it's one thing to see an Allen Deviation chart of an atomic clock (or a wandering OCXO tone on a SPA), but watching atomic clock drift on an o-scope made it come alive.
And fun fact, the gravitational redshift at Earth’s surface corresponds to a fractional frequency gradient of -9.8(2.3)×10-20/mm (at that particular lab in Boulder). With Strontium optical lattice clocks at the 1e-21 level, geopotential uncertainty is the dominate error term. It's fascinating to see all the metrology/sensing that atomic clocks play a huge role: magnetic fields, gravity, inertial navigation, Josephson voltage standards, atomic physics, astronomy, electric fields, mass.....
https://arxiv.org/ftp/arxiv/papers/2109/2109.12238.pdf
I love those clocks! Great video, and I'm really interested to see your time dialation experiment.
Ben, wanna do some Zeeman experiments? I just scored two large mid-century spectrometers. Should be able to resolve the lines (ahem, once restored).
@@CuriousMarc I'm definitely waiting for that!
@@CuriousMarc Mid 17th century? A slightly famous physicist did some ground breaking spectroscopy work that century.
@@johndododoe1411 No, nice instruments from the 1960’s. But good optics don’t get obsolete nearly as fast as electronics. Still very good instruments today.
When I studied engineering the quality of the lecturer absolutely steered my appreciation and interest in a particular subject, you have that gift, I dont know what you do professionally but if you dont lecture or teach its a waste.
I opened this channel and looked at the picture. OMG this guy found a pair of HP Cesium clocks. When I was working with HP, I sold two of them. Both went to Lockheed. After delivery I never saw them again. In those days they were accurate to within one second in 200,000 years. Better accuracy now.
Here I am learning about NMR spectroscopy in ochem 2, deciding to take a break with an exciting new CuriousMarc video, and getting most of the lecture theory explained back to me (plus a lot more.) No complaints though. Fascinating video. Thank you all for what you do and for taking the time to document it
Back in the early 60's NMR was the latest thing never realized how that experimental lab instrument would change chemistry. I had a medical MRI a few years back and got curious how they worked, most fascinating.
What a fantastic explanation, you make a wonderful teacher Marc, Thank you.
Waking up and finding 37 minutes of Marc discussing atomic clocks and Zeeman alignment is the best way to start the day.
My exact thought this morning as well.
Well, you just blew the old saying: A guy with one watch can tell you the time. A guy with two watches is never quite sure.. BRAVO !!
At this level even having 3 clocks or (horrors) any odd number of clocks you're still not sure what time it is. You just maybe, statistically, get a better guess.
{^_-}
The only thing better than this episode is the fact that I understood exactly what you were doing and explaining! I wish I had some of that test gear, and one of those beautiful clocks!
Thank you Marc and Team for another amazing experience! I can't wait for the time dilation clock check!
The HP 5518A heterodyne laser Interferometer also uses Zeeman effect to split HeNe laser light into 2 opposite polarized beams, with a few MHz difference in frequency, so that the receiver can tell, if the measurement mirror is moving towards or away from receiver.
Another instrument with lots of HP magic inside, can measure absolute distances in nm range or flatness of surface plate very accurately.
Wow, I definitely have to look that one up!
HP engineering sure was great.
magic is the proper word...or perhaps reverse alien enginering....
me: forgets about daylight savings time and doesn't even notice for a day or two
marc: my cesium clock is 0.0000000001Hz out of spec and we need to calibrate it.
Thank you for that reminder. I thought I was just having strange experiences in the kitchen over the weekend.. but my clock in there needed changing
Here I am on the last steps of restoring and adjusting a french turn of the century mantel clock to + - 10 minutes PER DAY and super happy with that, and these guys adjusting to the billionth decimal a cesium clock.
Turn of 19th century I take it!
I should adjust my best crystal clock, in a HP 5360A frequency counter. The manual suggests adjusting it against WWV. Only about 5 powers of ten worse then this cesium clock.
That was an excellent and very good explanation of the C-field and why it is required. Thank you for another outstanding video.
I've never understood why this channel sits under 150k subscribers, it's content and quality is outstanding every time.
It is because you have to be of a certain level to understand what's going on. The chin droolers that make up 90% of youtube can't keep up.
@@stargazer7644 Yes, and I still find it interesting in spite of drooling. Wish others could dream as I do, but you know....
I’ll stay tuned (hope I don’t drift too much!) 👍
This atomic clock is a genuine condensate of Nobel prizes.
Every time a video is posted, I am amazed at what is presented. I would have bet money that 2 HP atomic clocks were not operable at this late date. Those tubes are not replaceable anymore.
I can't wait to see what experiments you're able to perform with two highly calibrated cesium clocks.
The main thing to demonstrate is that timekeeping is affected by gravity. If one of the clocks is elevated above the other by a few meters, it should be able to observe that it gradually gains time on the one positioned below it. Even with just one working cesium clock, the GPS reference time can play the role of the baseline clock.
Now they just need couple of business jets to fly around the world.
Or just mountain top to Death Valley gravity field.
@@VainoOskariAstala81 Scott Manley is currently getting his pilot license. And he is living in the area.
Keeping one clock at sea level, and the other up a mountain, should produce a measurable difference.
Of course, that means finding a way to make one of the clocks portable, but I think HP was thinking of that when they built these things.
I remember when I was a kid in 70s, I saw a TV science program which was about relativity, time and speed of light. They used two atomic clocks like those here. They put one in a NASA airplane and another one on the ground. After the airplane did a couple spins the atomic clocks showed some difference about a few nanoseconds. Then the funny guy placed his twin brother in an spaceship which traveled at the speed of light. After a couple minutes, he came back to Earth and he found his brother pretty old. They said the story wasn't a science fiction at all and it was very amazing for me. Perhaps the best science program I ever saw on TV in those years.
Atomic physicist here - great explanation overall! Here's some additional information about selection rules:
The 'selection rules' you mentioned are determined by the polarization of the light - the applied magnetic field matters, but only indirectly. If you (arbitrarily) choose any fixed direction in space, you can define your m_F levels to have fixed total angular momentum projection along that direction. If the polarization of the interrogation laser is aligned with this direction, then we call it 'pi-polarized' and it can only drive transitions that do not change m_F. Note that the definition of the m_F levels depends on your choice of quantization axis.
The clock would still work perfectly in exactly zero B. At finite residual B, however, population will begin to be transferred between different m_F levels, unless the B-field is perfectly aligned along your preferred axis. The role of the C-field is to reduce the effect that stray fields have on the angle of the total B-field with respect to this axis. (Simple picture: moving the end of a meterstick by 1 cm changes its orientation much less than moving the end of a 30cm ruler by 1 cm.)
An alternate view (equivalent in any nonzero B field) is to choose the direction of the B-field as your quantization axis. Then it's the impurity of the laser polarization projection (along the B-field axis) that causes the unwanted transitions to occur. This has the advantage of ensuring that the corresponding m_F levels are so-called 'energy eigenstates' of the full Hamiltonian, including the Zeeman interaction - but at the expense of more complex selection rules.
But either way you do the computation, you get the same result - the component of the B field perpendicular to the light polarization must be zero, or you will get unwanted frequency components.
I learned quite a lot. One of the most entertaining classes I ever attended.
That's a very nice demonstration and explanation of the Zeeman effect and the magnetic alignment of the HP5061. I've never seen that before in such a detail. Anyhow, I suggest to do frequency comparisons (Cs vs Cs or vs GPS) by time-nuts methods. That is a phase comparison method by means of a high resolution T.I. counter, like your HP5334A. By using John Miles free TimeLab program, which might directly support your counter over GPIB, you can make comfortable stability (Allan Deviation) and uncertainty measurements down into the 10^-13 region, over many hours. That is required to overcome short / medium termed jitter especially of the GPS reference. Easy to accomplish, and maybe making a nice next video.
@1:00 as Alec from Technology Connections would say, "through the magic of buying two of them!"
What a fantastic video, Marc and the guys on top form. I love the amount of work that went into explaining the somewhat complicated (understatement of the year) quantum physics - Marc, you're a hero.
These videos are pretty awesome. Love the theoretical description next to the hands on demonstration.
I wonder if a vector voltmeter such as the HP 8508A or 8405A might not be a better instrument for detecting minute phase drifts between clock signals. They offer phase resolution down to 0.1deg, which would be difficult to resolve with the oscilloscope. They might also be less susceptible to drift or jitter in the triggering point as compared to an oscilloscope.
Brilliant explanation as ever. Took me right back to Electronic Theory of Matter lectures 48 years ago.
Awesome explanation! I feel smarter for watching, and that often happens when I watch your vids
I didn't understand a thing, and I still enjoyed it ;)
Marc, you guys should really get in touch with Tom Van Baak for your journey with cesium clocks and precision time keeping. His collection is probably second only to NIST itself.
On the day of the transition from DST to normal time. What perfect timing! 😁
I hate to tell you, but you're a week early.
Great explanation. I was actually able to follow you, and I'm not the brightest bulb in the electromagnetic spectrum.
Very cool!!! Avoid brain twisting explanation you ask, NEVER! Thank you so much for sharing all this with us! I grew up around engineers, listening to your explanations give me a warm feeling of being home again. Fascinating stuff. By far my favorite episode on this channel. What could be more mysterious then time? Liquid gold!
As a stained glass artist I highly approve of that window!!
Should you ever wonder again whether to include 18 minutes of theory, by all means, do include it. This was one of your best episodes ever.
I can't wait for the seemingly inevitable episode about the Ramsey effect!
I wish you had been my quantum mechanics teacher during my physics degree. You’re speaking directly to me when you express the theory intertwined with the experimentation and engineering utilization of the phenomenon. What an 18min!
It's amazing how many 'explanations' for the clock are 'it measures hyperfine transitions of cesium.' And are as long if not longer than this video. It's like telling you how a car works: 'it makes gasoline explode.' Excellent work. You've done a great job especially on elements that are hard to grasp in text. I don't think people really appreciate the scale of these measurements. The leap from theory to a practical machine is almost unreal.
Fascinating. I wish you’d been a lecturer whilst I was at university. You’re a natural! Very clear, and your enthusiasm for the topic is infectious.
This is a wonderful example of engineering as the application of science to a problem. Kudos to the HP engineers.
I think I understood more about this topic than the rest of my education in this one video. Very well done on the explanation and the restoration
I watched this video shortly after waking up and I still managed to follow your explanation of the hyperfine energy levels and Zeeman lines. Excellent job at communicating it clearly! You've now inspired me to dig into the literature on magnetic quantum numbers and selection rules, since that's a level of detail that I've only ever heard of in passing before. It's good to have a practical application to tie the knowledge back to.
If Dr. Barney M. Oliver were still around, he would love to see your workshop. So would I, if I still lived in the Bay Area!
Barney would be able to tell you exactly what to do to calibrate the clock, and why -- even if he had nothing to do with the clock product. He had the kind of curious mind that never let anything go by without understanding it, in a way, similar to you, Marc! Barney Oliver was a fascinating man, and a kind mentor to this young engineer, once upon a time, many years ago.
Barney Oliver was an amazing genius!
This video gave me a series of those headaches with pictures. You know, “an education!” That was awesome to see!
I wonder if Segal's Law ( "a man with a watch knows the time. A man with two does not" ) still applies when both the timepieces are Atomic?
Ha, that is why I got my second Rb oscillator. I could see a difference between my first Rb and the GPS disciplined oscillator but I didn't know which. With two Rb oscillators I could see the GPS was the issue. It needed a better antenna, small dropout on the GPS would slightly skew the PLL slightly.
Only until your accuracy reaches 5.39×10−44 s. 😜👍🏻
"... A man with two and an HP oscilloscope does not".
I could listen to you lecture about this kind of stuff all day!
@ReneKnuvers74rk - 2022-10-31
This really shows the marvel of HP being able to convert physics experiments to a useful product that can be calibrated with standard equipment and by people without a Nobel prize in their pockets.
Your explanation is super clear. You were able to explain enough of the details so us noobs can understand what is going on.
@hisham1269 - 2022-12-01
standard equipment ...