[time-nuts] PRS-10 Warm-up Time, Calibrating/Adjusting, and long-term poweron

[Attila Kinali]

Moin,

I'm sending this anwer to time-nuts as well (with permission of Taka)
as this might be generally interesting.

On Sat, 9 Mar 2019 04:23:27 +0000 (UTC)
Taka Kamiya <tkamiya9 at yahoo.com> wrote:

> I am experiencing what you are talking about "live".
> I have an experimental setup where PRS-10 is being steered by Trimble
> Resolution-T.  PRS-10 must be way off as it has been 3 hours and still has
> not phase locked. (physics lock was established long ago)  On LadyHeather, I
> can see a graph with steady decline, and as I understand it, it represents
> delta between the two.  Rate of decline is starting to slow down, so it must
> be approaching locking point.  Once I see it locks, I will have to make few
> hardware changes and restart it.  

I don't know what the PRS10 uses internally, but I wouldn't be surprised
if they would just do a long time-constant PID loop. As most people will
run a PRS10 continuously, having to wait a day or two for phase lock with
a GPS source is not that much of an issue.

 
> I am not in hurry to get it to lock.  What I really want is a stable standard 
> once it is locked. I am actually testing 4 time standards at the same time:
> one for each channel of scope. (Cs, Rb, GPSDRb, GPSDO)  The winner will
> become my house standard.  

I think you are making here a beginers mistake. You assume that there is
one standard that fits all needs. But truth is, each standard has its
uses and at different taus different standards will be the best one.
If you are in the confortable position of running all of these standards
at the same time, then I would do just that. Then you can use the one
standard that is best for what you are doing when you need it.

If you want to do it a bit more fancy, there are ways to build ensembles
out of different style standards and to combine them in such a way, that
always the best dominates the (in-)stability at a certain tau. 
E.g. [1] describes a decent way how to achieve this.


> This is just a test run though.  The real test
> will begin after few weeks of warm up.  I will have to come up with some kind
> of recording system.  This will be a challenge as none of the standards have
> GPIB.  I'm thinking of possibly taking a photograph of scope every hour for
> days, so I can see the drift.

If you have access to a TICC then you can use Timelab for it.
Alternatively, anything that can produce timestamps or time differences
of some kind is enough. You can then feed this to Timelab or Stable32
for further processing.

> I have shot my foot so many times, it's full of holes now.  

Each hole is the proof of experience :-)

			Attila Kinali

[1] "Composite clock: a new algorithm for servoing a VCO firstly to a hydrogen
maser clock and then to a caesium clock", by Mbaye, Makdissi, Plantard, Vernotte,
2008, 
https://doi.org/10.1088/0026-1394/45/6/S12
 
-- 
It is upon moral qualities that a society is ultimately founded. All 
the prosperity and technological sophistication in the world is of no 
use without that foundation.
                 -- Miss Matheson, The Diamond Age, Neal Stephenson