This is an old revision of the document!
FS clock errors
If the FS is complaining frequency of a formatter to FS time difference > 0.5s, it may be necessary to recalibrate the drift rate
The procedure is outlined in /usr2/fs/misc/fstime.txt The guts of the procedure is:
the “station” procedure library. It should go as early as possible
in the procedure.
- Make sure there is a “sy=run setcl adapt &” in the “midob”
procedure in "station".
- Set-up the “time.ctl” file with the following steps. Please note
that this procedure should be repeated whenever you change FS computers or upgrade kernels. Another case where you should redo this step is if the crystal has either because of aging or other reasons, like replacement, changed significantly in rate.
non-comment line:
0.000 0.010 rate
B. Start the FS
C. Execute the SNAP command: "sy=run setcl offset &"
D. Leave the computer running for about one hour. You can start and stop the FS, login and logout as many times as you wish. Just don't reboot the computer.
E. Execute the SNAP command: "sy=run setcl rate &"
F. Execute the SNAP command: "sy=run setcl save &"
G. At a shell prompt execute:
cd /usr2/control mv time.new time.ctl
As long as there were no communication or permission errors, your "time.ctl" file should be updated with a reasonable rate value.
***
We can perform this after the observation, and it should be fixed. Until we fix it properly by getting a permanent machine to replace the dead hobart.
T
On Tue, 2010-08-10 at 09:45 +1000, Brett Reid wrote:
> Hi Guys,
> We are running a different pcfs computer.
> On the new system we are getting error such as below.
>
> 2010.221.23:15:39.44?ERROR sc -13 setcl: formatter to FS time
> difference 0.5 seconds or greater
> 2010.221.23:17:44.13?ERROR sc -13 setcl: formatter to FS time
> difference 0.5 seconds or greater
> 2010.221.23:19:45.09?ERROR sc -13 setcl: formatter to FS time
> difference 0.5 seconds or greater
>
> Running sy=run setcl offset produces the following:
>
> 2010.221.23:19:45.09?ERROR sc -13 setcl: formatter to FS time
> difference 0.5 seconds or greater
> 2010.221.23:20:11.20;sy=run setcl offset
> 2010.221.23:20:11.26#setcl#time/1757666966,0,2010,221,23,20,11.09,-4.259,0.789,-11
> 2010.221.23:20:11.26#setcl#model/old,1263819362,-2043,1757382975,-0.778,1.543,rate,0,unknown,11,0
> 2010.221.23:20:11.15#setcl#model/new,1263819362,-2057,1757666966,-0.778,1.543,rate,0
>
>
> However the problem is not permanently fixed.
>
> We have seen this before. I just cannot remember how to fix it
> permanently. Any ideas greatly appreciated.
>
> We are running R1433 currently.
>
> Regards,
> Brett
>
>
– ############################################## Timothy Hoban VLBI Developer http://www-ra.phys.utas.edu.au/auscope School of Maths and Physics University of Tasmania ############################################## Guilt should be a momentary pang to spur one into corrective action. Anything longer is wasted, and worse, selfish. - Mark Belletini