[evlatests] tests, Dec 9 2005

[Ken Sowinski]

                        EVLA Tests, December 9 2005

Much time was used in continuing to recover from all the hardare
work of the last week.  DTS modules in 14 were all reset and delays
had to be found again.  Eventually things settled down and the 
most obvious problem is that something is wrong with 16A.  Its 
response is much less than expected at all bands.  Changing levels
in the downconverter does not improve things; Kerry says the DTS
module and deformatter looks OK.  Is it the downconverter?

The suspected aliasing at C band was explored by retuning the first
LO so that the part of the receiver bandpass which is being aliased
is masked by the 8-12 GHz filter at the dowwnconverter input.  This
did ineed make C band look much cleaner and apparantly increased G/T
by a few tens of per cent.  The amount of improvement varied with IF.
The synthesizer setting used was 13184 MHz instead of the usual
14076 MHz.  Do we wnat to modify the tuning algorithm for EVLA antennas
to place the desired sky frequency near the 8 GHz edge rather than the 
middle of the IF?
Here are some IAT times:
1952 begin observing with AC shifhted, but not BD
2002 begin observing with both AC and BD shifted
2026 return to standard synthesizer settings
2030 finished

Funny phase behavior was noted.  whenever the executor was started
the phases wandered around by radians for about a minute before 
settling down.  This was seen in 14 x 16 as well as on baselines
with either EVLA antenna against a VLa antenna.  I can think of no
simple explanation.

The code which applies the K term was tested and did not obviously break
anything.  I will commit it shortly and we will see how the antennas
behave at the next baseline run.

Q band now half works at antenna 16.  Walter found the beam yesterday
and the pointing correction from that allowed it to work today.  Delays
and focus and pointing are now about right.  However RCP  does not fringe.
Steve noticed today that the SD voltage is lower than expected for RCP.
Walter had no trouble seeing Venus in TP yesterday in RCP.  More work is
needed.