[evlatests] August 18 tests

[Ken Sowinski]

                              EVLA Tests
                            August 18, 2005

On Tuesday the L302s refused to lock.  Today the network was not
working.  Eventually cycling power on the switch at the top of
the 10.88 network fixed it.  This was happening yesterday as well
but was only diagnosed and fixed today.

By the time the network had returned Jerry had left the VLA and
once we had visibility to the systems again all was not as well
as might be.  Most of the problems probably involve DTS or deformatter
modules.

Deformatters for 14B, 14C and 16D do not respond to telnet sessions.
This remained true even after they were rebooted.  Perhaps they have
slot ID problems, but I did not look for them in DHCP IP space.  
I tried turning on samplers for all the IFs in antenna 16, but that
seemed to make no difference.  

Antenna 14 was pretty hopeless.  Only IF C looked reasonable at the
T5 in both TP and SD, but since the deformatter was not accessible
we couldn't set the filter parameters.  None of the other IFs seemed
to work.  The synthesizers locked and reasonable power levels were seen 
at the downconverter output detectors.

Antenna 16 half worked.  IFs A and B delivered a reasonable signal to
the T5.  With adjust ments both TP and SD seems OK and there was 
correlated signal in L, C and X bands.  There was even a weak correlated
signal at K band in IF A, but not IF B.  There is no reason to think that
pointing is good enough for K band yet.

IF C has no TP at the T5, but IF C has a nominal 5.5 V which is enough 
to level the ALC.  The SD voltages for C and D are variable, but identical 
which makes me suspect a wiring problem somewhere.

If the antenna is not disturbed much Friday it might still be worthwhile
to try a pointing run during Rick's time even if only one polarization
is working.  It would be safest to do this at C band.

I think we need a cooperative effort with the hardware people to try to
get all four IFs in shape in at least one of these antennas.  This can be
done even without time on the array.