[time-nuts] hm H Maser
kb8tq at n1k.org
Wed Jan 11 19:08:13 EST 2017
> On Jan 11, 2017, at 6:07 PM, Angus <not.again at btinternet.com> wrote:
> On Tue, 10 Jan 2017 21:43:07 -0500, you wrote:
>> This does get back to state of the art Rb and what that means. In my suggested case thats measured in terms of ADEV for Tau = 1 to 1,000,000
>> seconds. If you wanted an Rb with (only) state of the art phase noise at 1 MHz offset
> thats a different thing. State of the art for
>> power consumption and size is also not what Im suggesting in this case. Why the choice of spec?
> this is TimeNuts.
> Some discussions on the performance that might be practically achieved
> with different designs may be a useful start - as long as it's done in
> the context of a practical unit that could actually get built, rather
> than just a theoretical wish list.
The biggest issue with doing this is that *if* there is a magic formula that tells you
“these parts give you that ADEV” …. it’s in the same vault as the formula for Coke.
That sort of thing (if it even exists) would be the gold standard of corporate IP for
a company making atomic clocks.
About all you can really say is “they did this and the units the shipped did that”. There
are some obvious thing like “bigger cells work better”. Coming up with an equation that
correctly predicts a cell of this odd geometry functioning at these dimensions for 10,000
second ADEV ….not so much.
> It would also be good to have some idea of the cost of any special
> parts like cells too. Without that info, it's hard to know how
> practical particular designs would be.
The most likely course would be to cut out a major chunk of cost and find somebody
who is willing to make up a couple hundred sets of cells. There has already been
a proposal to do this floated on the list. I don’t recall the exact numbers, but $2,000
is what comes to mind. Apologies if this is a bogus number.
> Looking at export/technology controls might be useful early on too,
> since we're going for high performance.
Indeed, if you get to crazy you could get in trouble. My guess is that a standard the
size of a 5065 or larger is unlikely to set off alarm bells.
> I've often wondered how a 21st century version of a 5065 would
> perform, so it's great to see that I'm not completely alone in my
Which obviously is an itch many of us share. The gotcha of course is that each of
us has (likely incompatible) ideas about how to do it. We may even have
incompatible goals in terms of “what’s good”. Based on many decades of designing
things like this, feature creep and elastic goals will kill a project dead (usually after
a lot of money has been spent).
Lots of Fun
> time-nuts mailing list -- time-nuts at febo.com
> 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