[time-nuts] You can build a fountain from the things you find at home...

Attila Kinali attila at kinali.ch
Sat Oct 1 04:41:26 EDT 2011

On Tue, 27 Sep 2011 15:39:14 -0400
brent evers <brent.evers at gmail.com> wrote:

> So at its most basic, I'm wondering what type of clock would make the
> most sense to consider - cesium fountain, or hydrogen maser?  I
> haven't done enough research in either to know which modes they
> oscillate the cavity in, but presumably, the cavity for the cesium is
> going to be a lot smaller just based on the oscillation frequency
> alone, so there is some cost benefit there.

Depends on the build up of the cavity. IIRC the unloaded hydrogen maser
cavity is about 50x25cm. AFAIK a caesium fountain has a smaller diameter
but a longer length.

The lower frequency of Cs has a big gotcha: The electronics get much more
difficult. While H clocks are build around a ~1.4GHz resonance, Cs clocks
are build around a 9GHz resonance. Although 1GHz isnt that easy to handle
for someone who is not trained in electronics, it's doable. Also quite a bit
of electronics are available up to 2.5GHz, thanks to the abundance of WLAN & Co.

Another matter is being able to measure what you are doing. A oscilloscope
that is capable of showing 1GHz is already hard to get (and expensive!).
And you have to know exactly how to measure stuff w/o too much disturbance
of the signal. Getting a 10GHz oscilloscope is for the avarage amateur nearly
impossible and measuring stuff at that frequencies requires a lot of knowledge.
Oh.. and just in case you are not good at soldering. Forget trough-hole
parts in that frequency ranges. You will have so many troubles to get it
working, that it will be a lot easier to learn how to solder 0.5mm pitch ICs.

> What about material availability?  I don't know what form of cesium or
> hydrogen are needed, 

Both have to be of high purity. But H has the advantage of being a leaky
gas by itself. I.e. passing it trough a leak purifies it to a high degree.
Cs has the big disadvantage (like all alkaly metals) of a very high reactivity
with oxygen and water. Thus handling is very difficult if you dont have good
chemistry lab equipment.

> but I'm assuming that hydrogen is more available
> with less hassle?

Yes. But getting Cs isnt that much of a trouble either. Just ask your local
chemistry lab supplier.

> It seems like there are a few simple things that could be done to
> cheaply gain on the stability curve.  I live in an area with a high
> water table (right next the the chesapeake bay).  Why not drill a
> shallow well - 20ft?? to  submerge the cavity and gain the thermal
> stability of the well?  I'm from the subsea world, so putting
> something 20 feet down is nothing (most of my stuff goes to 6000m).

You are operating in vacuum anyways. Build a resonant cavity within
the vacuum cavity and you get a temperature stability that is a lot
higher than if you'd put a single cavity desing into a well.
> Last - I've got a good friend that works at Jefferson Lab, involved in
> niobium deposition to get their cavity Q's up.  What's the feasibility
> of an amateur doing something like that to improve cavity Q.

I wouldnt worry about this. The Q factor of an unloaded cavity is high
enough that you need to be able to tune it eletronically, otherwise you'll
be so far off that you will not detect the resonance you want.

> Again - I apologize for my complete lack of knowledge on the subject,
> but these are just some of the ideas I've had bouncing around my head
> for how this might be done to make a pretty good clock on a pretty
> tight budget.

You want to have a look at the papers written in the 60s and 70s about
H masers and Cs fountain clocks. These explain in great detail how these
devices work and what the big stability issues are.

			Attila Kinali
The trouble with you, Shev, is you don't say anything until you've saved
up a whole truckload of damned heavy brick arguments and then you dump
them all out and never look at the bleeding body mangled beneath the heap
		-- Tirin, The Dispossessed, U. Le Guin

More information about the time-nuts mailing list