[time-nuts] Re: Creating a D.I.Y Rubidium Atomic Clock

[Leo Ahluwalia]

Hi Stephen, I appreciate the advice, and while glass-blowing isn't
something I've had much experience with, making the glass package
definitely sounds like a fun challenge.

Thanks, Leo

On Wed, Jun 7, 2023 at 8:23 AM Stephen C. Menasian via time-nuts <
time-nuts at lists.febo.com> wrote:

> Leo,
>
> This is a complex, but interesting, project. The electronics/microwave
> part of the project is fairly straightforward and is best left for later
> phases of the project. The interesting part of the project is the
> "physics" package, which I refer to as "optical pumping". My experience
> in this area is a bit dated (5-6 decades ago), when my undergraduate and
> PhD work was centered on optical pumping of Rb, Na and Cs atoms as well as
> H2+ and Hg+ ions. A lot has changed since then; in particular, many
> refinements have been made to create practical miniaturized Rb based
> systems suitable for commercial time standards; I am not current on these
> changes.
>
> Your first task would be to establish and demonstrate optical pumping of
> Rb87 atoms. For this, you would need, at a miniumum:
>
> 1) some pure Rb87
> 2) a glass blowing setup (and some skill)
> 3) A selection of inert gasses (at least including Ar)
> 4) Light source (Rb discharge was used then with a circular polarizer;
>    intensity pumping via Rb89 (I believe) is often used now.
>    Infrared lasers look very interesting to me.
> 5) A low noise light detector (I used photomultipliers - many more modern
>    solid state possibilities exist now.
> 6) a uniform magnetic field (Helmholtz coils or solenoid) and magnetic
>    shielding.
> 7) A means of controlling cell temperature (Rb vapor pressure)
> 8) A bit of electronics, including RF power for the lamp. Note that a
>    noisy lamp discharge will mask the desired signal.
>
> You should start with the Zeeman transitions, which are much easier to
> observe and don't require microwave hardware. If you can't see a Zeeman
> transition, there is no hope for the hyperfine transition. Getting
> collisional disorientation down to a low enough level might involve
> experimentation with various buffer gas formulations and pressures and
> cell wall coatings (we used a very thin teflon film).
>
> If you get to the point where you can see the Zeeman transitions, you
> will have accomplished and learned a lot. With a little more effort, you
> should be able to see the hyperfine transition.
>
> Good Luck!!!
>
> Stephen
>
>
>
>
>
> On Tue, 6 Jun 2023 15:25:47 -0400
> Leo Ahluwalia via time-nuts <time-nuts at lists.febo.com> wrote:
>
> > Hello, my name is Leo, and I was wondering about the feasibility of
> > creating a simple rubidium atomic clock at home, and good places to
> > source parts for this sort of project. I was also wondering what
> > instruments and specialized knowledge would be required depending on
> > how pre-built the parts are, for example, the prerequisite knowledge
> > for creating PLL(s) for both frequency division to provide an output
> > and also to modulate the input of the microwave oscillator. Any good
> > resources or PDFs would be greatly appreciated, as I am very new to
> > frequency analysis and phase manipulation in general. I was also hoping
> > to keep the cost of the project soft-line below 500$, though that would
> > be assuming a decent level of precision and already possessing basic
> > equipment (oscilloscope, multimeter, etc…). Safety is also a major
> > concern, as I know while the intensity of the EM waves involved is low,
> > the power source may or may not be. This is all of course assuming a
> > basic design, where my assumption is that it will involve the
> > microwaves being shot at a rubidium vapor cell which is directed
> > towards a photodetector. This would then be connected up to the PLL and
> > circuitry to both provide a stable standard while also modulating the
> > microwave oscillator to offset any external environmental factors. I
> > don't have a clear vision of what would truly be the best design for
> > both simplicity and cost, though this is what I've found to be a common
> > design in what I've looked at online.
> >
> >
> > Thanks, Leo
> > _______________________________________________
> > time-nuts mailing list -- time-nuts at lists.febo.com
> > To unsubscribe send an email to time-nuts-leave at lists.febo.com
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe send an email to time-nuts-leave at lists.febo.com