[evlatests] Time error fixed.

[Rob Long]

I could find absolutely nothing wrong with either the hardware or the 
master timing.  I believe that we would have to see a difference in the 
master 1PPS from the GPS and the timing outputs from the L350 (meaning 
the L350 was not really sync'd properly), confused NTP time in the MIBs, 
or incorrect timing to the computing system/correlator to produce this 
kind of problem.

As far as I know, the MIBs believed they had correct time and the 
largest 'offset' I could find in the master timing was slightly less 
than 300 ns.  After re-syncing, this measured difference between GPS and 
our timing was unchanged.  Since it seems that re-syncing/resetting 
solved the problem, I am still pondering what the real problem was!

Rob

Vivek Dhawan wrote:
> The EVLA knickers have been untwisted. I heard that a re-synch of the L350
> was done, followed by all the antennas. Ken (+ Hichem or Rob) may wish to
> elaborate.
> 
> Here is a timeline, based on VLA online log entries (IAT/UT). I have verified
> the first and last comments, the comments in between are by deduction.
> 
> UT date      startTime  project
> 
> 2009-02-11	07:14	POINT    ! Known to be good.
> 2009-02-11	10:18	AB1321   ! Probably good.
> 
> 2009-02-11	15:43	MAINT    ! Likely start of problem
> 
> 2009-02-12	05:42	AB1321   ! First science affected.
>                                   ! everything between affected.
> 2009-02-17	12:50	AW761    ! Last science affected.
> 
> 2009-02-17	23:50   SOFTW    ! Problem fixed
> 
> I tested corrections I had derived earlier, by applying them to other data.
> This did not go well. I suppose we can go ahead and contact the affected
> observers, but it may take me a day or 2 to ensure I have a workable fix to
> pass on to them.
> 
> Vivek.
> _______________________________________________
> evlatests mailing list
> evlatests at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests