EEVblog Electronics Community Forum

EEVblog => EEVblog Specific => Topic started by: EEVblog on February 12, 2013, 11:30:23 am

Title: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: EEVblog on February 12, 2013, 11:30:23 am
A look at one of the original HP 5061A Cesium Beam atomic clock frequency standards used in the "flying clocks" experiment in the 1960's, that showed (not definitely at the time) that time dilation as part of Einstein's relatively theory was correct. The effect was not confirmed until a more precise run in 1971, again with the same clocks.

EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard (https://www.youtube.com/watch?v=3yTL06ZU5SU#ws)

Dave.
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: lewis on February 12, 2013, 01:15:58 pm
Cooooooooooooooooooool! I want one!
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: Ed.Kloonk on February 12, 2013, 01:30:12 pm
What was the price tag back in the day?

Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: olsenn on February 12, 2013, 02:17:02 pm
Quote
What was the price tag back in the day?


What would be the price tag today? How long does Cesium last?
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: lewis on February 12, 2013, 04:17:52 pm
There's one on the 'bay: http://www.ebay.co.uk/itm/Refurbished-HP-Cesium-Beam-Frequency-Standard-5061A-Battery-Backup-Atomic-Clock-/160917714428?pt=LH_DefaultDomain_0&hash=item25777175fc&_uhb=1 (http://www.ebay.co.uk/itm/Refurbished-HP-Cesium-Beam-Frequency-Standard-5061A-Battery-Backup-Atomic-Clock-/160917714428?pt=LH_DefaultDomain_0&hash=item25777175fc&_uhb=1)
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: free_electron on February 12, 2013, 04:45:32 pm
The tubes last about 15 to 20 years. After that ot's game over.

Agilent sold the atomic clock line to Symmetricom. You can still buy agilent designed systems from them , and they do service them.
I had a gps disciplined clock from agilent. Kinda looked like a 34401 multimeter. Same package.
Came with a huge battery pack and charger. You needed to put a cone antenna on the roof of the building , lightning arresters, and run a big fat coax to the machine. It had a local 10MHz oven that was disciplined by the gps.

58503a with display option. There was a special distribution amplifier that would send the 10MHz out.
I had long bnc coax cables under the raised floor of the lab. If we needed precise timing measurements you would plug thos in the back of the machine and tell the scope/ generator/ counter or logic/spectrum/network analyzer to use the 'external clock'.
You could then do a phase compensation ( due to the long bnc cables ) and have all machinery work in sync. Was used when doing I/Q modulation work on ADSL . The internal oscillators have phase jitter and are outof sync with each other. By linking them up this all went away and we could see how good our electronics was.

I have pictures somewhere of it... Need to look

The little mechanical clock on the front is kinda hilarious. Well, it ticks extremely accurate , but good luck seeing where the hands point.... Measure with a micrometer , mark with chalk , chop with an axe... Bwahaaaa

http://www.symmetricom.com/products/frequency-references/chip-scale-atomic-clock-csac/SA.45s-CSAC/ (http://www.symmetricom.com/products/frequency-references/chip-scale-atomic-clock-csac/SA.45s-CSAC/)


Herr's the 5071a they talk about. Still in production

http://www.symmetricom.com/products/frequency-references/cesium-frequency-standard/5071A/ (http://www.symmetricom.com/products/frequency-references/cesium-frequency-standard/5071A/)

That company (symmetricom) is really time-nuts. They got it all. Cesium beam standards, rubidium vapor standards, hydrogen masers. You want how many digits ? Come this way ...
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: free_electron on February 12, 2013, 05:00:00 pm
More i teresting reads :

http://www.wired.com/science/discoveries/news/2007/12/time_hackers?currentPage=all (http://www.wired.com/science/discoveries/news/2007/12/time_hackers?currentPage=all)
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: mikeselectricstuff on February 12, 2013, 05:14:57 pm
The little mechanical clock on the front is kinda hilarious. Well, it ticks extremely accurate , but good luck seeing where the hands point.... Measure with a micrometer , mark with chalk , chop with an axe... Bwahaaaa

Someone in Marketing probably insisted that as it was called an atomic clock it needed a dial.
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: N2IXK on February 12, 2013, 05:20:42 pm
Probably there so when they bring the suits on a tour through the lab, they can all set their watches and feel all "technical" about it... :-//
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: free_electron on February 12, 2013, 05:55:16 pm
Bwahahaa. I can visualize a bunch of military brass on a tour of their high end atomic clock facility for rockets in the 60's... All setting their analog watches to this thing and being all smug about having a wristwatch that was set to atomic time....
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: billclay on February 12, 2013, 09:53:41 pm
More i teresting reads :

http://www.wired.com/science/discoveries/news/2007/12/time_hackers?currentPage=all (http://www.wired.com/science/discoveries/news/2007/12/time_hackers?currentPage=all)

There's more of Tom Van Baak's recent general relativity experiment with a few HP 5071A here:

http://www.leapsecond.com/great2005/ (http://www.leapsecond.com/great2005/)
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: G7PSK on February 12, 2013, 09:54:37 pm
So what was the time keeping like then, in 1971 when I started as a horologist i was told by a rep for one of the Swiss watch company's (Louis Patek I think)that atomic clocks were good for a second in a million years but on Dave,s video they were talking about 30 odd thousand years for a second
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: Slothie on February 12, 2013, 09:58:52 pm
I wonder how the accuracy of this ceasium clock stacks up against those Rubidium frequency standards available on eBay that Dave reviewed ages ago?

Did you use it for anything Dave, or is that still on the "to do" list?

Ian
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: Dave on February 13, 2013, 01:10:23 am
The second is defined as 9,192,631,770 (yes, I do know the number by heart) oscillations of the radiation emitted by Caesium 133 while shifting between two fine ground states. How is that guy's coworker able to know the frequency to 14 digits, if the thing is defined only by 10 digits? :-//
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: amspire on February 13, 2013, 02:26:12 am
The second is defined as 9,192,631,770 (yes, I do know the number by heart) oscillations of the radiation emitted by Caesium 133 while shifting between two fine ground states. How is that guy's coworker able to know the frequency to 14 digits, if the thing is defined only by 10 digits? :-//
That is when the Caesium atom is at zero degrees Kelvin. If it is at a higher temperature - like in the atomic clock - I gather you have to correct for the difference, so that could be where the 14 digit number comes in.

Alternatively when that clock was built, the second was defined as 1/31,556,925.9747 of the tropical year for 1900 January 0 at 12 hours ephemeris time. In 1967 after the clock was built, it was changed to the Caesium definition. There is a good chance the clock frequency was never altered from the time it was built for exactly the reasons that Dave has explained in his calibration blogs.  There was a very slight difference between the 1960 second and the 1967 second, and so the clock frequency may require 14 digits of resolution when it is based on the 1960 standard second.
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: ejeffrey on February 13, 2013, 07:27:29 am
The second is defined as 9,192,631,770 (yes, I do know the number by heart) oscillations of the radiation emitted by Caesium 133 while shifting between two fine ground states. How is that guy's coworker able to know the frequency to 14 digits, if the thing is defined only by 10 digits? :-//

It isn't defined to 10 digits, that is an exact number.  i.e., 9,192,631,770.000000000 ...  In reality, cesium is not the best choice.  The big problem is that the transition is not perfectly sharp, but has some width.  Furthermore, as I recall, it isn't a single transition, but a spectrum of nearly degenerate spectral lines.  The second is actually defined as the center of the spectral line, but the width and structure limit the resolution of measurements.  Newer atomic clock designs have considerably more precision than a cesium standard.  The aluminum ion quantum logic clock at NIST has enough resolution to measure the time dilation due to a 30 cm altitude change.  Of course the accuracy can't actually be better than cesium clocks until the second is redefined.
Title: Re: EEVblog #423 - HP5061A Atomic Clock Cesium Beam Frequency Standard
Post by: dakiller on February 13, 2013, 11:18:40 am
I actually have one of these at home myself. I bought it off ebay a few years ago for a few hundred bucks.

Completely non-functioning but completely intact unit. Weights a friggin tonne, probably near the 30kg mark. I'm pretty sure it'll power up but wont sync to the field because the caesium tube is gone.