Re: [time-nuts] Tbolt issues
Yes Charles
I was not part of or following the original Tbolt discussions. My setup at
the time was Loran C with Austron 2110, backed up with Cs and 60 Khz using
a Tracor 599H. My work was focused on FRK FRS Rb’s using Shera with very
good results. At the same time Juerg was using a PRS10 controlled by a Tbolt
and feeding a OSA FO3/8600. With the demise of Loran C I switched to Tbolt
feeding the 2110 because I did see the frequency jumps. Juerg’s PRS10 died
a year ago and he replaced it with his Tbolt, expecting the FO3 to clean
up the Tbolt. After disappointing results on some of our work urged him to
do a cross matrix on all his sources and found out the FO3 did nothing to
the Tbolt He bought his first one through Tarp. I bought 1 for him and 2 for
me all US sellers.
Off list Tom has looked at Juergs data and we and he are convinced that
antenna and equipment is good. We have used auto tune and variable settings.
On our to do list was temperature control and clean up loop. In order to do
an analog clean up we need short interval changes and that is why I went
on the list since we have not been able to do it and looking at past posting
have not found data that will get us there. For our work frequency has to
be better than 1 E-11 one second, he had it with the PRS 10 setup. Unless
some one can help us with settings that make analog loop clean up possible we
will not revisit Tbolt. Juerg will use it as an excellent 1pps source and
I will use it with the 2110 which is digital and has a 100 second loop.We
are already refocusing on work on a FRK/M100 GPSDO plan on having it up and
running with in a month, waiting for boards. Last two years distracted by
clean up of the Shera code before releasing it, controller for FE 5680 5650
and FE 405.
In a message dated 9/10/2016 5:48:49 P.M. Eastern Daylight Time,
csteinmetz@yandex.com writes:
Bert wrote:
Looks like we are not the only ones trying to improve frequency
performance and hopefully some one will share settings.
You are coming very late to the Tbolt party. There was a veritable
blizzard of posts about optimizing Tbolt performance, which began maybe
10 or 11 years ago (??) and lasted for several years, and a steady
trickle since then. Everything that is possible to be said was posted,
often several times. Go back in the list archives and read this
material -- it will answer all of your questions (and lots more that you
haven't asked yet).
Best regards,
Charles
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Bert wrote:
On our to do list was temperature control and clean up loop. In order to do
an analog clean up we need short interval changes and that is why I went
on the list since we have not been able to do it and looking at past posting
have not found data that will get us there.
To do any kind of "cleanup" at tau=1S, you need a source that is better
at that tau than the one you are trying to clean up. But if you have
that, why use it to clean up something else? Just use it as your
standard. (One might respond that the point would be to use the Tbolt
to discipline the better oscillator at longer tau to correct drift, but
you would need a very long time constant -- thousands of seconds --
which you cannot achieve with an analog loop.) I do not think there is
any realistic possibility of doing the kind of cleanup you propose in
the analog domain.
Are you absolutely certain you tried the Tbolt with the damping set to
10 seconds or more? Did you let it settle for several weeks before
deciding it wasn't doing what you need?
Perhaps you can try one last time:
-
Do a hard reset back to factory settings -
Change TC to 500 seconds and damping to 10-12 -
Set recovery jam synch to "ON" with a threshold of 55-65 nS -
Set recovery max offset frequency to >=1000 (this is in ppb) -
Put it in an undisturbed location, inside a box of some kind -
Let it sit for a month -
Now, measure it
If this is better, but not quite good enough, try damping settings of 30
and 100.
Alternatively, for a 10MHz reference you can just forget about and use,
buy another PRS10 and discipline it with the PPS from the Tbolt. I'd
recommend buying a new one from SRS so you know that it has the PPS
discipline feature and you could get help from SRS to optimize the
discipline for your needs (the discipline is very adjustable and it is
easy to get confused).
Best regards,
Charles