So what’s inside that Cs Beam Tube anyway?
Hi Tom,
On 11/02/2016 06:23 PM, Tom Van Baak wrote:
Hi Rick,
That's consistent with what I've seen and heard as well. Most of the 5061B / 5071A on eBay with dead tubes are opt 001 (high-performance) tubes. The surplus units with standard performance tubes are more likely to power up and lock. And if they don't, it's probably something simple rather than a dead Cs tube.
I've also heard that the choice of high- vs std-performance is made after the tube is built. The difference would be in things like the oven temp and detector gain, both of which are controlled outside the tube. My first hint of this was in a paper by Carroll Alley...
You know the famous 1971 Hafele-Keating experiment with four 5061 cesium clocks flying around the world. Several years later, a more precise measurement was made by Carroll Alley using better clocks. What I read is that he got Len Cutler to hack the 5061A so that its beam current would be higher, thereby improving performance for the duration of the experiment. Same tube, just different operating parameters.
See page 3 of:
http://www.hp.com/hpinfo/abouthp/histnfacts/publications/measure/pdf/1977_04.pdf
Nice find, except, that's not what it says there.
"I also modified the cesium standards, which already had the new
high-beam flux tubes, for further improve their performance."
This implicates that the high-beam flux tubes already existed, and that
he did other tweakings to the clocks.
I'm not disputing that beam-current changes might be completely
controlled by the resistors rather than being a physical aspect.
Cheers,
Magnus
Hi Magnus,
About the high-perf CBT, I wrote: "The difference would be in things like the oven temp and detector gain, both of which are controlled outside the tube. My first hint of this was in a paper by Carroll Alley."
Since you're curious, let me explain the hint in the paper by Alley [1][2]. He says:
"The clocks were modified in order to give the performance needed. Following suggestions by Len Cutler and others at Hewlett Packard, we increased the beam current by a factor of 2, we added an integrating loop in the crystal control, and there was a proprietary modification of the beam tube (now standard on all high performance tubes)."
"All in all, we could achieve stabilities over the 15 hours at a couple of parts in 10^14 with standard commercial clocks, as shown in Figure 29. We paid much attention to providing a stable environment for the clocks. Let us look at some pictures to show you the equipment and give you some feeling for the experiment."
I hope that answers your question.
/tvb
[1] C.O.Alley, "Proper Time Experiments in Gravitational Fields with Atomic Clocks, Aircraft, and Laser Light Pulses" in "Quantum Optics, Experimental Gravity, and Measurement Theory".
https://books.google.com/books?id=GLwGCAAAQBAJ
This article / book is behind paywalls. I can send you a PDF off-list. But...
[2] There's papers with about the same content. C.O.Alley, "Introduction to some fundamental concepts of general relativity and to their required use in some modern timekeeping systems", PTTI, 1981.
And this PDF for this is freely available, from all the usual suspects:
http://tycho.usno.navy.mil/ptti/1981papers/Vol%2013_37.pdf
https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19820012654.pdf
http://www.dtic.mil/dtic/tr/fulltext/u2/a494262.pdf
The quote above is from page 710.
Hi Tom,
That's more relevant text. Thanks.
Cheers,
Magnus
On 11/03/2016 08:34 AM, Tom Van Baak wrote:
Hi Magnus,
About the high-perf CBT, I wrote: "The difference would be in things like the oven temp and detector gain, both of which are controlled outside the tube. My first hint of this was in a paper by Carroll Alley."
Since you're curious, let me explain the hint in the paper by Alley [1][2]. He says:
"The clocks were modified in order to give the performance needed. Following suggestions by Len Cutler and others at Hewlett Packard, we increased the beam current by a factor of 2, we added an integrating loop in the crystal control, and there was a proprietary modification of the beam tube (now standard on all high performance tubes)."
"All in all, we could achieve stabilities over the 15 hours at a couple of parts in 10^14 with standard commercial clocks, as shown in Figure 29. We paid much attention to providing a stable environment for the clocks. Let us look at some pictures to show you the equipment and give you some feeling for the experiment."
I hope that answers your question.
/tvb
[1] C.O.Alley, "Proper Time Experiments in Gravitational Fields with Atomic Clocks, Aircraft, and Laser Light Pulses" in "Quantum Optics, Experimental Gravity, and Measurement Theory".
https://books.google.com/books?id=GLwGCAAAQBAJ
This article / book is behind paywalls. I can send you a PDF off-list. But...
[2] There's papers with about the same content. C.O.Alley, "Introduction to some fundamental concepts of general relativity and to their required use in some modern timekeeping systems", PTTI, 1981.
And this PDF for this is freely available, from all the usual suspects:
http://tycho.usno.navy.mil/ptti/1981papers/Vol%2013_37.pdf
https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19820012654.pdf
http://www.dtic.mil/dtic/tr/fulltext/u2/a494262.pdf
The quote above is from page 710.
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Never heard that story. You gotta love it: a hot rod atomic clock.
Rick
A well-written, very readable version of the story is in here:
http://www.hp.com/hpinfo/abouthp/histnfacts/publications/measure/pdf/1977_04.pdf
and the technical paper with all the details of the experiment is here:
https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19820012654.pdf
Another story about the 5071 is "turbo" mode which hp/Agilent briefly touted at PTTI 10 or 15 years ago. Do you remember that? It was a firmware upgrade that would boost stability even above the high-performance level. Of course it would burn Cs like there was no tomorrow, but for a brief time you'd get another couple of dB. Hard to say if that was a Jack idea or a Len idea. But it reminded me of "Spinal Tap". I wish I had kept a copy of the flyer or the ADEV plots. Maybe you know?
See, these are all the kinds of things Attila could play with and tell us about if he decided to build his own cesium clock ;-)
/tvb
Tom
I may ask for the paper offline. Though kind of wonder how much sense I
will make of it.
What I am reading is that increasing the temp of the oven in an end of life
tube can indeed increase the signal. Maybe it doesn't shift the spectrum or
resonance. That would tend to suggest my funny offset is the DC oven
controller. Seems hard to believe because its only the oven thats under DC
heating.
Thanks for teaching.
Regards
Paul
WB8TSL
On Thu, Nov 3, 2016 at 5:36 AM, Tom Van Baak tvb@leapsecond.com wrote:
Never heard that story. You gotta love it: a hot rod atomic clock.
Rick
A well-written, very readable version of the story is in here:
http://www.hp.com/hpinfo/abouthp/histnfacts/publications/measure/pdf/1977_
04.pdf
and the technical paper with all the details of the experiment is here:
https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19820012654.pdf
Another story about the 5071 is "turbo" mode which hp/Agilent briefly
touted at PTTI 10 or 15 years ago. Do you remember that? It was a firmware
upgrade that would boost stability even above the high-performance level.
Of course it would burn Cs like there was no tomorrow, but for a brief time
you'd get another couple of dB. Hard to say if that was a Jack idea or a
Len idea. But it reminded me of "Spinal Tap". I wish I had kept a copy of
the flyer or the ADEV plots. Maybe you know?
See, these are all the kinds of things Attila could play with and tell us
about if he decided to build his own cesium clock ;-)
/tvb
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/
mailman/listinfo/time-nuts
and follow the instructions there.