[time-nuts] Newbie question: GPSDO filtering principles

Peter Vince pvince at theiet.org
Tue Apr 24 07:31:42 EDT 2007


Dr Bruce Griffiths <bruce.griffiths at xtra.co.nz> wrote:

>A more prosaic explanation springs to mind.
>What are the specs for the Disciplined oscillator package?

I don't have the manual in front of me tonight, but I do remember 
that they say it should stay within 300ns of UTC.

>It is possible that the frequency setting resolution is a few times
>1E-11 and the frequency steps you see correspond to 1-2 bits of the
>frequency setting DAC.

No, it's much finer than that - see below.


"Tom Van Baak" <tvb at LeapSecond.com>

>What make/model GPSDO is this?

I'm not really comfortable publicly "naming and shaming", but it 
isn't one of the big names that I've seen discussed in this forum.

>Is there a setup menu
>where you can control the integration period or is it fixed
>at 8192 always?

It is automatic.  Apparently it uses shorter time constants when 
first switched on, but lengthens them once it has settled down.  The 
8192 period I just discovered by examining the output log data.  It 
can also be switched into manual mode, whereby the DAC control value 
is set from the console.  The DAC is fed with a 16-bit control word - 
although I don't know if the DAC uses all sixteen bits. 
Experimenting with that this weekend, the step-size appears to be 
about 6.67 x 10^-14.


>Can you send a few lines of the log so we
>can see the granularity of the DAC?

See below: I have extracted the DAC change data, and annotated each 
line with the the change from the previous reading, the period since 
the previous change, and the frequency change that equates to 
(assuming the step-size determined above).


>When you say the
>phase drifts, is this something the GPSDO reports by itself
>or is this an external measurement made by you with some
>counter and some other standard time reference?

I have measured it with a 53132A counter, initially just comparing 10 
MHz outputs from two identical oscillators, but subsequently also 
comparing against free-running crystal and rubidium oscillators.  The 
1 pps can also be output from these units, and I now have a second 
counter comparing that against my free-running rubidium so that I can 
indirectly compare the disciplined rubidium against GPS, and cancel 
out any wanderings by the free-running ruby.

	Thanks again for your help,

		Peter



Delta frequency assumes 6.66666667E-14 Hz per step

5893 at 15:00:18 02/12/2006  Change:   0 after     0 seconds, dF =  0 Hz
58B0 at 17:16:50 02/12/2006  Change:  29 after  8192 seconds, dF = 
1.933E-12 Hz
58B2 at 20:00:04 02/12/2006  Change:   2 after  9794 seconds, dF = 
1.333E-13 Hz
58BA at 22:16:36 02/12/2006  Change:   8 after  8192 seconds, dF = 
5.333E-13 Hz
58A5 at 02:09:12 03/12/2006  Change: -21 after 13956 seconds, dF = 
-1.400E-12 Hz
5870 at 05:29:35 03/12/2006  Change: -53 after 12023 seconds, dF = 
-3.533E-12 Hz
5853 at 08:02:10 03/12/2006  Change: -29 after  9155 seconds, dF = 
-1.933E-12 Hz
585E at 10:18:42 03/12/2006  Change:  11 after  8192 seconds, dF = 
7.333E-13 Hz
5810 at 12:35:14 03/12/2006  Change: -78 after  8192 seconds, dF = 
-5.200E-12 Hz
5898 at 14:51:46 03/12/2006  Change: 136 after  8192 seconds, dF = 
9.067E-12 Hz
5926 at 17:08:18 03/12/2006  Change: 142 after  8192 seconds, dF = 
9.467E-12 Hz
5934 at 19:57:43 03/12/2006  Change:  14 after 10165 seconds, dF = 
9.333E-13 Hz
593C at 22:14:15 03/12/2006  Change:   8 after  8192 seconds, dF = 
5.333E-13 Hz
58E7 at 00:30:47 04/12/2006  Change: -85 after  8192 seconds, dF = 
-5.667E-12 Hz
58BF at 02:47:19 04/12/2006  Change: -40 after  8192 seconds, dF = 
-2.667E-12 Hz
586F at 05:03:51 04/12/2006  Change: -80 after  8192 seconds, dF = 
-5.333E-12 Hz
57FA at 07:56:38 04/12/2006  Change:-117 after 10367 seconds, dF = 
-7.800E-12 Hz
580E at 10:13:10 04/12/2006  Change:  20 after  8192 seconds, dF = 
1.333E-12 Hz
57CD at 12:29:42 04/12/2006  Change: -65 after  8192 seconds, dF = 
-4.333E-12 Hz
5879 at 14:46:14 04/12/2006  Change: 172 after  8192 seconds, dF = 
1.147E-11 Hz
592C at 17:02:46 04/12/2006  Change: 179 after  8192 seconds, dF = 
1.193E-11 Hz
5961 at 19:49:07 04/12/2006  Change:  53 after  9981 seconds, dF = 
3.533E-12 Hz
5969 at 22:05:39 04/12/2006  Change:   8 after  8192 seconds, dF = 
5.333E-13 Hz
58F8 at 00:22:11 05/12/2006  Change:-113 after  8192 seconds, dF = 
-7.533E-12 Hz
58C1 at 02:38:43 05/12/2006  Change: -55 after  8192 seconds, dF = 
-3.667E-12 Hz
5852 at 04:55:15 05/12/2006  Change:-111 after  8192 seconds, dF = 
-7.400E-12 Hz
57D2 at 07:55:47 05/12/2006  Change:-128 after 10832 seconds, dF = 
-8.533E-12 Hz
5806 at 10:12:19 05/12/2006  Change:  52 after  8192 seconds, dF = 
3.467E-12 Hz
589B at 13:22:14 05/12/2006  Change: 149 after 11395 seconds, dF = 
9.933E-12 Hz
58EB at 15:38:46 05/12/2006  Change:  80 after  8192 seconds, dF = 
5.333E-12 Hz
5925 at 19:40:44 05/12/2006  Change:  58 after 14518 seconds, dF = 
3.867E-12 Hz
5917 at 21:57:16 05/12/2006  Change: -14 after  8192 seconds, dF = 
-9.333E-13 Hz
58C9 at 01:57:17 06/12/2006  Change: -78 after 14401 seconds, dF = 
-5.200E-12 Hz
5872 at 04:13:49 06/12/2006  Change: -87 after  8192 seconds, dF = 
-5.800E-12 Hz
57DD at 07:48:47 06/12/2006  Change:-149 after 12898 seconds, dF = 
-9.933E-12 Hz
5801 at 10:05:19 06/12/2006  Change:  36 after  8192 seconds, dF = 
2.400E-12 Hz
5854 at 12:21:51 06/12/2006  Change:  83 after  8192 seconds, dF = 
5.533E-12 Hz
58F7 at 16:18:57 06/12/2006  Change: 163 after 14226 seconds, dF = 
1.087E-11 Hz




More information about the time-nuts mailing list