[time-nuts] Cs tube pics

Michael Wouters michaeljwouters at gmail.com
Wed Nov 2 16:42:13 EDT 2016

I worked on a trapped ion frequency standard 20 years ago, a 12.6Ghz Yb
clock. It's still in the lab across from me and looking at it, and the
electronics, I think it is the sort of thing that a physicist might
contemplate building in his/her garage but ...

Building it took about 10 man years of concentrated effort so the time you
would need for a project like this would be the killer, assuming you could
buy the spectroscopic lasers and UHV equipment you needed at a price that
didn't require refinancing your mortgage.

One other vital component is the flywheel oscillator you need to take
advantage of the fabulous stability you now have at hand.We had a cryogenic
sapphire oscillator for our (microwave) clock.

For an optical clock, you're also going to need a frequency comb to get
back into the RF domain.


On Wednesday, 2 November 2016, Attila Kinali <attila at kinali.ch> wrote:

> On Tue, 1 Nov 2016 09:10:25 -0700
> "Tom Van Baak" <tvb at LeapSecond.com> wrote:
> > > I really would like to do that. But they are a tad bit expensive.
> > > Especially on this side of the big pond. If anyone is willing
> > > to part with a Cs standard and want to have it a good home, feel
> > > free to contact me :-)
> >
> > It's true that a cheap GPS receiver is more accurate long-term than
> > a surplus cesium standard, but you're right about the "just because"
> part.
> > To me at least, this time nut hobby is not so much about the pursuit of
> > accuracy as it is an appreciation for the variety, ingenuity and
> complexity
> > of timekeeping. In some cases "how it works" is far more interesting that
> > "does it work".
> I think, most of us are in it for the "how it works" and
> "how far can I push it". :-)
> > A used but known working cesium standard can be expensive, but like most
> of
> > you almost all my gear comes from eBay via automated daily keyword
> searches.
> > Many of my mil- or telecom-surplus FTS 4050 and HP 5061 were obtained for
> > just a couple hundred dollars. You may search for months or even years,
> but
> > amazing bargains show up. Not all of them work, of course, but the op/svc
> > manuals are superb, the design / construction is very repair-friendly,
> and
> > there's a weird group called time nuts with helpful advice.
> It still is prohibitevly expensive in Europe. There are much fewer
> Cs standards going around than in the US in the first place, and there
> is also less a tinkering mentality. Ie a lot of companies just say
> "it's broken, it's no use for anyone anymore, let's just scrap it" and
> thus a lot of stuff ends up in recycling instead of on ebay. Heck, a couple
> of years ago i got an 3458 because the company wanted to throw it away.
> Mind you, it was in full working condition, only the NVRAM batteries were
> low.
> I still would like to try to build my own atomic clock at some point,
> even if it would be a quite costly, and a many years project.
> IMHO the easiest to build would be an Rb or Cs vapor cell using either
> dual resonance or coherent population trapping. Cells can be bought
> for 300-500€. For a bit more you can get them made to spec. Machining
> a resonant cavity from aluminium is pretty easy and cheap these days,
> if one wants to go for the dual resonance. The biggest issue for both
> types would be the laser system. Either getting the laser diodes selected
> (makes them expensive) or build an external cavity for them (creates
> the need of a complex control system and not so simple mechanically).
> Putting all toghether would probably cost something in the order of
> 1000€ to 5000€.
> One up in difficulty would be a passive hydrogen maser. This requires a
> vacuum system and things like platinum leaks to generate atomar hydrogen
> and state selection magnets. If one knows glass blowing, part of it can
> be made using pyrex tubes, which simplifies some stuff (like keeping the
> state selection magnets outside the vacuum system). Also, the cavity
> needed would be quite big. A normaly used TE011 cavity is huge. One can
> load it with aluminia and get it down to 15-20cm diameter, but this
> requires
> crystaline aluminia to maintain low loss. Maybe one could use other
> resonating
> structures that are smaller, TE111 or loop-gap resonators have been
> proposed.
> The biggest cost here is definitely the needed vacuum system. Although
> the rough pumps are rather cheap (around 500-1000€ if one does not need
> fast pumping) and some of these actually end up on ebay without being
> destryoed by "testing", the high vacuum pumps (ion pumps, turbo pumps etc)
> are not cheap and need to bought new (the stuff you see on ebay are either
> systems that were removed from labs because they don't work anymore, or
> were destroyed by "testing" them in free air).
> An active hydrogen maser should not be that much more difficult. It mostly
> involves a low loss, correctly tuned cavity and low noise detection
> electronics.
> The input stream of hydrogen atoms needs to be more precisely controlled as
> well...
> Another step up in difficulty would be a system using a magneto optical
> trap.
> Glass cavities with flanges for vacuum system are readily available and
> also cheap. Again, the vacuum system is one difficulty, though probably
> simpler than for the hydrogen maser (less parts that need to be custom
> made).
> But the requirements for the vacuum are a bit higher. The laser system
> poses
> a similar difficult as with the CPT system above, but now there are more
> lasers and all need to be locked to eachother. The traping laser also need
> to be directed at the cavity from 6 sides, all meeting in the center of the
> cavity, which makes alignment problematic. This is also the first system
> that offers to be a primary standard, although it probably does not get to
> the stability of a 5071, as the trapping lasers will induce a light shift
> that is not so easy to control unless one goes for the expensive laboratory
> grade lasers.
> Next on the list would be an Hg ion trap. But there are so many parts
> in such a system that are not easily bought and require a very good
> understanding of the physics involved to design custom parts, that
> it becomes almost unrealistic that an amateur could be build one at home.
> Same goes for ion and neutral atom optical clocks. Yes, they can be build,
> yes the principle is simple, but getting it actually to work needs a lot
> of understanding and tinkering.
> Ah... I'm dreaming again :-)
>                                 Attila Kinali
> --
> It is upon moral qualities that a society is ultimately founded. All
> the prosperity and technological sophistication in the world is of no
> use without that foundation.
>                  -- Miss Matheson, The Diamond Age, Neil Stephenson
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com <javascript:;>
> To unsubscribe, go to https://www.febo.com/cgi-bin/
> mailman/listinfo/time-nuts
> and follow the instructions there.

More information about the time-nuts mailing list