[evlatests] Tests 12/20/05

[Ken Sowinski]

> Date: Wed, 21 Dec 2005 08:52:32 -0700
> From: Pat Van Buskirk <pvanbusk at aoc.nrao.edu>

> Failure: Ant 14 and 16 during 443POLC.  19.2Hz timing pulse not 
> available.  Antennas started file pointing in wrong direction. 
> Attempted aborting/re-starting script 3 times.  Also attempted putting 
> antennas in standby and back to DPM.  Attempted parking and re-pointing 
> as well.  Antennas would not show up in F display or go to correct 
> position.  Control screens were showing no AZ/EL errors, but antennas 
> were not pointing correctly, nor were they showing any errors in AZ/EL.

This is the first I had heard of this.  This behavior is consistent
with what Vivek reported yesterday noon, and Tom yesterday evening.

>   Finally parked antennas after 45 minutes.  Antennas were running fine 
> on previous file 443R570.  Possible script conversion error?

I doubt it.

> 12/20/05 -> FRM problem? Waiting on more information from the Operator.

?????

> 12/20/05 (evening): The same problem seemed to occur again.  The 
> antennas were pointing fine and following the loaded script but at some 
> point I started getting the 'sync_detect' error on the Alerts screen and 
> both antennas tried to go to the low elevation limit.  I tried putting 
> them in standby, then back into DPM and they continued to try to go to 
> the low el limit.  I was able to command them to park and they parked, 
> but would not follow the script.  (Tom Briscoe)
> 
> I believe Ken has more information.

I know no more than what is reported here.  I did talk to Tom
and suggested he try to park the antennas if he was unsuccessful
in getting them to do the right thing.

As already written I believe the L305 errors are a red herring and
irrelevant to the pointing problem.  It has been noted that this 
seems to have arrived with the new executor on Monday and that 
certainly seems to be the case.  Steve's experience this morning
suggests that no commands are makling it to antenna MIBs because he was
unable to switch bands until Rich restarted the executor.  All this
is consistent with someone having invoked that 'take' method on the
antennas in use.  Could that have happened? or is there something else
wrong?

Going back to an older version of the executor will require some
fiddling with the parameters DB which I do not understand.


Ken