[evlatests] EVLA tests, Sept 8 2005

[Jim Jackson]

I am having Jerry Langevin go to the site tomorrow (Friday) morning to do a 
complete power cycle on the LO/IF racks and DTS modules in both 
antennas.  It sounds like the DTS modules in Antenna 14 were started today 
when the L305 in the antenna was out of sync.  In a discussion I just had 
with Kerry Shores, neither of us is convinced that the current set of DTS 
modules would follow suit when the L305 is resynced so we have no idea what 
those modules may be doing.  Antenna 16's fiber was also worked on earlier 
in the week which could have upset the timing in that antenna as well.  The 
only way I know to be sure is to shut everything off, then turn on the 
LO/IF rack,  sync the L305, then finally power up the DTS modules.  I think 
the new FPGA and MIB firmware that is being developed for the DTS module 
will take care of this in the future.

Steve Durand and I are also looking at adding additional surge protection 
to the 48VDC supply to the Vertex room.  Hopefully that will help prevent 
problems like occurred on Wednesday evening.

Jim



At 04:47 PM 9/8/2005, Ken Sowinski wrote:
>                         EVLA Tests, September 8 2005
>
>It was such a confusing day it is hard to say what was done or
>learned.  Communications with 14 were lost yesterday and today
>it was found that the inverter which powers the network swich had
>died during a thunderstorm.  Once it came back 14 still did not
>work for a while because tne MIBs in all four DTS modules were not
>responding.  Antenna 16 was sort of working during this time.
>The 16B donwconverter was reset and Rick's sesnitivity tests was
>run during lunch.
>
>After lunch an attempt was made to get as much as possible to work.
>Two more nonresponding downconverter MIBs were found: 16A and 14B.
>After a little more struggle the cause of the weak correlation was
>found to be faulty phase switching.  Just as I was about to investigate
>different time offsets to the L302 16 began to work again.  Without
>any intervention by me.  If this was not caused by people working in
>the control building it is the most important thing learned today.
>
>What happens is that at random times antennas 14 and 16 cease to
>correlate well against VLA antennas but continue to correlate
>reasonably well against each other.  The most straightforward
>explanation is that the L302s, all of them!, lose the correct idea
>of time at the same time, and, if we are patient enough come back
>again when they feel like it.  I am hoping that someone will confess
>to causing this by disturbing something in the control building.
>
>There is a pointing run scheduled for tonight which will be done at
>X band and include EVLA antennas.  George's polarization tests following
>that will also use EVLA antennas.
>_______________________________________________
>evlatests mailing list
>evlatests at listmgr.cv.nrao.edu
>http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests