[time-nuts] Beginner's time reference

WB6BNQ wb6bnq at cox.net
Fri Dec 11 13:01:49 UTC 2009


Charl,

If it is autonomous operation you want, then your only choice is a Cesium Beam
frequency reference.  By definition it is absolute !  But that does not address
resolution.  For instance, if the Cs is old tube and noisy, while it may be
accurate, due to the noise you may not be able to resolve any better then, say, 1
part in 10-9 (not good).  If you need more absolute resolution, then your only
choice would be to have a new tube put in or buy another one.

However, all this says nothing about accurate time, because there is no such
thing !  Accurate time only means something with respect to a recognized
reference point.  In this case USNO for the USA or another county’s primary lab.

So, if you mean you want accurate time in addition to frequency you will need to
have a method of comparing or setting your time device to a selected external
time reference.  This is where the GPS comes into the picture.  It is the medium
by which you can track and adjust your local time device to an external
reference, USNO (for GPS), within a given time resolution (down to nanoseconds or
better).

Because time and frequency are related, having a time reference also gives you a
frequency reference.  Having a very good local oscillator for the short term
controlled by the GPS for the long term tracking is a good choice and relatively
inexpensive compromise between having nothing and spending a lot of money for
Cesium standard that may only last between 5 and 10 years if new.  Purchasing a
used Cs is a crap shot, may be good, but could be bad.

While a Rubidium oscillator looks good at first, you need to consider they are
not a primary reference because they drift, albeit generally slower then a Quartz
oscillator.  A really good Quartz oscillator will outperform the eBay run of the
mill Rubidium for short term measurements.  A good Rubidium will prevail for
medium term measurements and of course provide a much slower drift when in an out
of lock condition (called holdover) with the GPS tracking.

Another point to consider is the Rubidium has a short lifetime compared to a
Quartz oscillator.  Things do fail, but the Quartz oscillator likes to have
continuous operation and, actually, generally improves because of it.  There are
many Quartz oscillators that are still ticking after 30-40-50 years !  The
Rubidium, on the other hand, wears out faster when it is turned on with a life
span of 5 to 10 years if new.

So I would suggest doing some research like reading through the Timenut archives,
reading various publications available from NIST off the web, perusing TVB’s web
site at www.leapsecond.com and KO4BB’s site www.ko4bb.com under the timing and
manuals subpages.  There are a variety of manuals on frequency and time
equipment.  You should be able to find some hp app-notes on the subject as well.
There are other worthwhile sites too.

It doesn’t hurt to have a mix of items.  For instance a good oscillator (or two
or three) tracked by GPS, a Rubidium (or two) for good measure and after you get
your feet wet then decide if you really want to spend the money necessary to own,
operate and maintain a Cesium.  Oh, by the way, unless you are damn rich, the
wife will never agree to the Cesium when she hears the price.

Bill....WB6BNQ


Charl wrote:

> Dear Hal and others,
>
> Thank you for your suggestions. I suppose I should have emphasised that I
> was looking to have my own reference, i.e. something not dependent on GPS,
> LORAN, or other signals from the aether. Indeed I might get better accuracy
> for less money by tapping the GPS time signal, but to me that's not as much
> fun as building my own atomic reference.
> Hal, is there much difference in quality for these rubidium tubes? Will I
> get what I pay for? In any case, you make a good point about measuring the
> accuracy. I'm a university student, so perhaps I can pay the physics
> department a visit. Otherwise, maybe I could compare it to the GPS signal?
>
> Kind regards,
> Charl
>
> On Fri, Dec 11, 2009 at 9:30 AM, Hal Murray <hmurray at megapathdsl.net> wrote:
>
> >
> > > I'm hoping to build my own circuit around the device, which might cut
> > > down the costs somewhat. Some rubidium sources on eBay go for less
> > > than $100, but I'm not sure what quality to expect. Any advice or
> > > suggestions are appreciated!
> >
> > Short answer:  Sure, get one of the $100 rubidium boxes and see what you
> > can
> > do with it.
> >
> > Long answer:  You just stuck you toe into a huge tar pit.  Pick a corner
> > that
> > seems like fun and dive in.
> >
> > How are you planning to measure if whatever you build is any good?
> >
> >
> >
> > --
> > These are my opinions, not necessarily my employer's.  I hate spam.
> >
> >
> >
> >
> > _______________________________________________
> > 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.
> >
> _______________________________________________
> 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 mailing list