[time-nuts] Homebrew H Maser
jmiles at pop.net
Sun Aug 29 23:39:42 UTC 2010
> After keeping an old EFOS 2 H maser running the last couple or three
> years here are a few bits of advice.
> -The triple magnetic shields are VERY important. The first time I removed
> the top
> shield to access the RF section the Maser stopped oscillating! Replacing
> the shield
> restored oscillations!
As I understand it, not only the intensity of the ambient H-field but its
orientation is critical. Sounds like you either pulled the 1,0 -> 0,0
transition frequency too far for the control loop to compensate, or the
static field was no longer oriented properly with respect to the cavity's TE
The impression I got from my reading on the subjecct is that tuning an
H-maser isn't something you can do incrementally. It's not a conventional
RF tank circuit -- there's a list of factors as long as your arm that have
to be just right, or you will get nothing at all for your trouble. Getting
those factors right seems to require a graduate-level understanding of both
the materials and the math. Ars longa, vita brevis.
> -this maser has two vacuum sections, an outer that enclosed the cavity
> and thermally insulates it. This allows the cavity temperature to be
> regulated to within .001 degree C. The inner chamber keeps the cavity
> evacuated so all you have in it is the H atoms.
I imagine the cavity has to be pretty stout not to either collapse from
barometric pressure or flex excessively.
> I'll dig them up and see if anyone could host them on a website. (Files
> are quite large!)
It would be great if you could upload these to the Manuals page at
www.ko4bb.com. I've looked over the Symmetricom maser's manual but it's
>> "Building Scientific Apparatus" by Moore, Davies, and Coplan
> It's Davis, as in C.C. Davis of the University of Maryland, my former
> MS thesis adviser.
> Sorry about that... yes..
> Its a great book, but very expensive.
> worth every penny if you're doing this sort of thing..
The current edition is $70 at Amazon, not too bad by textbook standards.
It's a good book, but often more valuable as a pointer for further reading
than as a practical, up-to-date handbook. Seems like a good overview of
The book you really want to start with is Major's "The Quantum Beat," IMHO.
>>In thinking about it, it would be a terrific project to run with LabView!
>Rubbish, LabView would _never_ be able to do that.
(Shrug) PLLs are PLLs. I don't see a role for a PC in the cavity tuning or
oscillator disciplining loops, but that's a very small part of the overall
control picture. Most of the actual software work would involve UI design
for monitoring (and ideally graphing) the dozens of operating parameters
over time. It would probably make the most sense to use either analog or
microcontroller-based controls for the realtime (RF) loops, and use Labview
or another instrumentation package to monitor everything.
There are also various high-latency thermal loops that could be controlled
as well as monitored by Labview-like software on the PC.
-- john, KE5FX
More information about the time-nuts