[time-nuts] Disciplining Rubidium
df6jb at ulrich-bangert.de
Fri Apr 25 11:35:07 EDT 2008
I have been an admirer to Stanford Research since many years. That is
why I thougth: Hey man, if they are that friendly to explain their PLL
in THAT detail in the PRS10 handbook I take them as my school masters
and make the regulation in my DIY GPSDO pretty much the same. The PRS10
features a (switchable) pre-filter for phase data that automatically
gets 1/3 the time constant of what the main loop filter is set to. This
pre-filter is thought to de-noise the data before it even enters the
loop. Since I have access to all stages of the data processing in my
GPSDO I attach a graph showing
1) Red line: Raw phase data, as delivered from phase comparator (RX =
2) Blue line: Sawtooth corrected phase data
3) Yellow line: Phase data behind the pre-filter
May everybody judge for himself which of the lines is more suitable as
the input of a regulation loop.
Unfortunately the original PRS10 does not feature the facility to
process sowtooth data and considered in detail this limitation even
makes sense: In order to make things easy for the user they would have
to make their microcontroller talk to the receiver in its specific
language which is by no means standarized but is not only different
between manufacturers but also between different generations of
receivers (at least concerned Motorola).
The alternative would have been to include a command in their own
PRS10-syntax which would need an external online translation process
from receiver language into PRS10 language and who of you owns a device
that would accomplish this task on the fly WITHOUT writing a special
program for it? There we got it! As far as I can see the PRS10 is one of
the very seldom cases that would benefit from generating a new sawtooth
corrected pps signal in hardware.
> -----Ursprungliche Nachricht-----
> Von: time-nuts-bounces at febo.com
> [mailto:time-nuts-bounces at febo.com] Im Auftrag von christopher hoover
> Gesendet: Freitag, 25. April 2008 05:36
> An: time-nuts at febo.com
> Betreff: Re: [time-nuts] Disciplining Rubidium
> Said wrote:
> > For such a large phase error I would have expected the
> PRS10 to just
> > reset
> > the 1PPS rather than drift it.
> The PRS10 uses a straightforward second order PLL to lock
> itself to an external 1PPS, so you will see a frequency error
> on phase jumps as it has to slew the frequency to get back on phase.
> There's a quite detailed description of the PLL in the manual.
> Incidentally, it is my experience that the PRS10 behaves
> fairly poorly if you simply feed it 1PPS without sawtooth
> correction. I'm in the midst of a move so I can't relate
> numbers at the moment.
> (Someone once mentioned of a firmware upgrade or a magic
> command or something to allow you to feed the sawtooth
> corrections from the receiver to the PRS10's time tagging
> circuit, but I've yet to find any documentation on
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-> bin/mailman/listinfo/time-nuts
> follow the instructions there.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 13077 bytes
Desc: not available
Url : http://www.febo.com/pipermail/time-nuts/attachments/20080425/2d9a0260/attachment-0001.pdf
More information about the time-nuts