[evlatests] X-Band performance last night

[Rick Perley]

    I got 3 hours time to check the performance at all bands, using the 
usual modes.  Here I report on X-band only. 

    Overall -- very good quality.  But there are notable exceptions.   
By antenna:

    13  All looks very good, except there is a very small scale (2%) 
ramp in gain on a 20 second timescale on IFs B and D.  The ramps slope 
in opposite directions on these two IFs:  For 'B', the amplitudes ramp 
up and clunk down.  For 'D', it's the other way around.  The 'clunks' 
are not synchronous. 

    14   No data!  The antenna was clearly operating, as the Tsys values 
are normal, and I have data from this antenna on P-band.   The reported 
Tsys values for IFs B and D are far too high (~70K). 

    16   Still out for rewiring.

    17  All looks very good.  A single 0.4 second amplitude dropout was 
seen on IFs A and C (simultaneously).  The amplitude loss is only ~10% 
(so that, say, in a 10-second average, it would never be seen).  The 
reported Tsys is far too high (about 85K on all IFs), but antenna 
sensitivity is normal, indicating something wrong in the Tcal entries. 

    18  Stellar behavior, as usual.  But not perfect -- the A and C 
sides had two short-term (< 0.41 second) amplitude dropouts.

    21  This antenna has serious issues still, particularly on IFs A and 
C. 
          - The bandpasses for A and C are not normal, showing a 'hump' 
of ~4 MHz width in the middle of the 12 MHz passband I used.
          - The delays are fine now.  (21 C is a little off still -- 
about 5 nsec, which should be subtracted from the current value). 
          - The phases on all four IFs undergo a sinusoidal variation, 
of amplitude 30 degrees, and period about 90 seconds.  All four vary 
with the same phase (i.e., they rise and fall together).  Other than 
this variation, the phases behave normally. 
          - The reported Tsys is too high (65 K), but vary in a normal 
fashion -- evidence that the switched power and total powers are 
operating as intended. 
          - IFs A and C have essentially zero sensitivity -- a G/T a 
factor of 10 too low.  IFs B and D are about 50% low. 
          - The fast-scale amplitudes for IFs A and C have two states:  
The low-amplitude state lasts 1.25 seconds, the high amplitude state a 
single 0.416 record.  So far as I can tell, the period is precisely 
1.667 seconds.  The 'low' state is a factor  about two below the 'high' 
state -- no where  close to explain the loss of sensitivity (a factor of 
10) noted above.   Note also that observers using any integration time 
at or over 1.67 seconds won't see this effect, other than horrendously 
low sensitivity. 
          - The gains for IFs B and D look completely normal.

    23  Perfect behavior!  ABsolutely nothing to complain about.  Our 
new poster-antenna. 

    24  IF 'C' has 5 different amplitude states, apparently randomly 
distributed.  The different in amplitude between highest and lowest is 
about a factor of 2 (or four in power).  The highest amplitude state 
(which is only rarely seen) is about the right level.  The bandpass for 
this IF is not normal:  a sinusoidal oscillation with period of about 6 
MHz.  This is seen in bandpass phase also, so it has the appearance of a 
standing wave.  All other IFs are normal.  IF 'A' had six low-amplitude 
fast dropouts (they probably are present on C as well, but are buried in 
the multi-states described above).  B and D are perfect. 

    26  IFs A and C have many fast dropouts, of amplitude 10%, and which 
are perfectly synchronized on these two IFs.  Over a 4-minute period, 
these dropouts occured once every 10 seconds.  Outside this 4-minute 
window, the dropouts were much rarer.  The effect is completely absent 
on B and D.    Otherwise, the antenna operated completely normally.  (In 
particular, the simulataneous phase clunks are gone -- proof that their 
origin is in the RT phase, which has been turned off).