[time-nuts] Re: ammonia, cesium, masers, etc.

[Andy Talbot]

It ought to be possible to find the resonance with off the shelf equipment
An ADF5355 synthesizer is capable of small frequency steps and will
generate up to 13.6GHz.  Use it's 2nd harmonic leakage (4th harm of the
VCO), there is a bit coming out at 24GHz.   If the synth  reference input
is derived from a 48 bit DDS like an AD9852, it can be tuned in really
small frequency steps, far below the 1Hz steps that ought to be enough to
spot the absorbtion.   Isn't Caesium about 1Hz wide ?

For detection, an amateur 24GHx converter is close enough to use.
Copper water pipe that has been hammered into an oval can support TE01
modes reliably enough if you need to suppress polarisation rotation.

Andy
www.g4jnt.com



On Mon, 20 Sept 2021 at 18:14, Andy Talbot <andy.g4jnt at gmail.com> wrote:

> What's the exact resonance frequency of ammonia?   To as many sig figures
> as it's been calculated.
> Googling just seems to throw up "approximately 23.8GHz"
>
> Andy
> www.g4jnt.com
>
>
>
> On Mon, 20 Sept 2021 at 18:10, paul swed <paulswedb at gmail.com> wrote:
>
>> Well at the very next MIT flea I will grab some 23 GHz waveguide. Maybe
>> Nearfest in October. Another trick I have actually used on 10 GHz id 3/4"
>> copper tube as waveguide. Its quite low loss but circularly polarized. So
>> to use it I needed to rotate the coupling in the shack for best return
>> loss. So perhaps 1/2" or 1/4" would be the same. At least in the earliest
>> Amonia system they simply seemed to have filled the waveguide with amonia.
>> Then launched 23 GHz down the length of it. Good thing about amonia if you
>> have a wave guide leak you would know it.
>> Indeed there is a ton of modern 23 GHz technology available. I was going
>> to
>> make some comments on system oscillators and deleted them. I haven't
>> looked
>> at what types of oscillators or multipliers are out there at a cheap cost.
>> All this being said its a curious possibility. But if its no better then a
>> Cesium not worth any effort. Its one serious advantage is you could refill
>> the amonia. Not really sure what would get used up. I think as mentioned
>> here the amonia H would disassociate.
>> Regards
>> Paul
>>
>> On Mon, Sep 20, 2021 at 11:07 AM Lux, Jim <jim at luxfamily.com> wrote:
>>
>> > On 9/20/21 7:17 AM, Poul-Henning Kamp wrote:
>> > > --------
>> > > paul swed writes:
>> > >
>> > >> Though honestly bending 23 GHz waveguide might be tricky. Or simply
>> > leave it
>> > >> straight after all it Amatuer grade.
>> > > I seem to recall that there was about 10-ish meter coiled up in the
>> > > one on the famous old picture.  That might be a bit unwieldy.  On the
>> > > other hand, that was probably not waveguide as much as just plain
>> > > copper-tubing ?
>> >
>> > That would depend on whether you want to control moding. (the picture
>> > says rectangular) Bending WG is pretty easy if the radius is large
>> > compared to the size of the WG - it's very soft copper, so you just need
>> > an appropriate form around which to bend it. The trick of filling it
>> > with sand also helps.  At 23 GHz, you probably want to use WR34 or WR42
>> > (Hmm, probably the bigger, so you're farther from cutoff, 17 and 14 GHz
>> > respectively).
>> >
>> > >
>> > >> Lastly they discussed an Amonia maser. Never knew. Wonder if it
>> actually
>> > >> vented amonia slowly. Sort of sounds that way.
>> > > I would suspect the main problem with an amonia maser is that the
>> > > Amonia disassociates and the hydrogen leaks out though the metal ?
>> > >
>> > I wonder, though, if you were "building your own atomic clock" whether
>> > ammonia might not be easier than Cs or Rb, for instance. These days a 23
>> > GHz oscillator and other parts are fairly easy to come by, and ammonia
>> > is somewhat easier to handle than Cs or Rb. I would normally think you
>> > can get any of them by just ordering them, but since ammonia is used in
>> > synthesis of methamphetamine, it may not be so easy to get a small
>> > bottle of ammonia gas.
>> > _______________________________________________
>> > time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe
>> send
>> > an email to time-nuts-leave at lists.febo.com
>> > To unsubscribe, go to and follow the instructions there.
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe
>> send an email to time-nuts-leave at lists.febo.com
>> To unsubscribe, go to and follow the instructions there.
>>
>