[evlatests] Tests, Aug 30 2005

[Ken Sowinski]

                        EVLA Tests
                      August 30, 2005

Since last time the L305 in 14 was replaced and surgery was performed
on the deformatters for 16A and 16D to make the strong birdies go
away.  Eventually 14 came up and all the DTS modules worked when
prodded enough.  The transfer switches in both UX convertes were 
hardwired so that both outputs are from the direct path as expected
by the software.

The changes to the pointing model derived from the X band pointing
run last Saturday definitely improved 16; K band was made to work
and delays and focus were optimized.  We probably have problems
because the collimations that were best for 3c84 while setting were
not good for 3c273 rising a few hours later.  We will attempt a 
C band pointing run tonight to see if the curious X band results
are a properly of X band or of the antenna.  The crossed hand
products at K band were not alarmingly large, but not too small
either: between five and ten per cent.

No attempt was made to look for birdies after the DTS modules were
repaired.  We should do that on Thursday.  

I tried tooking at ANTSOL gains with the T_sys correction turned off
to get an idea of relative G/T between VLA and EVLA antennas.  More
work is needed to get conclusive results but the initial result is
not encouraging.  I think the best way to do this is in line mode at
25 or 12.5 MHz so that the amplitude can be estimated from the line
channels and be free of delay errors and not be washed out by the
phase behavior across the bandpass.  This is something we can try to
do more carefully on Tuesday.  If we do, pointing offsets should be 
measured at the same time at all bands.

It was noted that the means of the sampler values were not very close 
to zero for antenna 14.  The range of mean values over the 4 IFs was 
from -2 to -4.  In principle this is unimportant because it is introduced
withing the phase switching loop and so ought to be removed.  Nonetheless
we should probably not intentionally introduce DC offsets if it is easy 
to adjust them out.