[evlatests] P band

[Ken Sowinski]

The P band receiver and T301 have been installed in antenna 14
and declared to be functional.  It looks much better than last
time we tried.  Sampler statistics and sync detector voltages
behave as expected when downconverter gain is changed, and 
reasonable levels are attained with abour 25 dB of attenuation 
in all IFs but IF A which seems to have less gain and require
less attenuation.

The antenna was setup at P band and fringes were immediately
obtained against VLA antennas, however the polarizations are
swapped wrt the rest of the VLA.  The AC data looked fine
against several VLA antennas, phases were as stable as expected
at P band.  The BD data has a uniform phase wind of about 
40 degrees in 10 seconds which is independent of baseline length.
There is a large spurious correlation in BD, but only on the
baseline to antenna 9 which, suspiciously, is the antenna nearest 
to antenna 14.  The AC data to antenna 9 seemed unaffected.

Observations were made with a first LO of both 12160 and 12968 MHz
and there were no readily discernible differences, but a more careful
examination should be made.

Walter took data from each of the samplers and applied a Fourier
transfrom.  There are differences in level between IFs but the gross
features are the same: about 40 MHz of noise that correpsonds to
the P band receiver, about 100 MHz noise at a sky frequence of
about 450 to 550 MHz of unknown origin and two large spurs at the
center of the sampler input band and at the 7/8 point.  Walter
said he would examine the spectra for interesting numerology.

We should spend some time tomorrow trying to optimize the power
levels.  It seems likely that the best diagnostic will be SD
voltage or correlation coefficient.  Sampler RMS will be misleading
because of all the other stuff in the 1 GHz bandpass the sampler sees.
This can be done even if the polarizations are not swapped before
tomorrow.