[evlatests] Azimuth-Dependent Noise Power Variations on Antenna 4, at K, Ka, and Q bands

[Rick Perley]

    Bob Hayward and I journeyed to antenna 4 last Thursday (19 June) to 
antenna sensitivity tests at K, Ka, and Q bands.  These include tips, 
which are taken at an azimuth of -20 degrees.  The antenna is located on 
pad W9.
    After the completion of our last tip, at 23560 MHz, with the antenna 
pointed to the zenith, we rotated it to the 'stow azimuth' at Az ~ 80.   
We were monitoring the receiver total power in the T304 1-2 GHz TP 
monitor, and we noted very distinct changes in power at Az ~ 26 and 50 
degrees.  The changes were by about 1K , or 2% of total system power.  
We reversed the rotation, and noted the same power bumps at the same 
azimuths -- these are not due to any receiver gain changes!  We then 
drove from an azimuth of -20 through to azimuth of 180 -- and noted the 
same pair of noise bumps (but stronger now -- 2K-- than the initial 
passes), plus a very wide 'bump' extending over 40 degrees of azimuth, 
centered near Az ~ 140. 

    We then changed frequency to 33160 MHz (Ka-band), and drove from Az 
= 180 through to -20.  Three distinct 'bumps' were seen, at azimuths of 
147, 100 and a double-bump near 59 degrees.  This latter pair were very 
narrow in width -- less than 10 degrees.  All are of about 2 K in amplitude.

    We changed frequency to 40072 MHz, in Q-band, and rotated from -20 
through +180.  Again, power bumps of about 1K amplitude were seen, at 
azimuths of 70 and 140, with numerous others -- both 'up' and 'down', 
along the way. 

    These were all seen in the 1 GHz-wide TP monitor in the T304.  When 
we had completed these scans, we retrieved the analog 'Milhouse' data, 
which was running with a 100 MHz-wide filter.  The same patterns were 
seen as in the T304 data -- strongly indicating that the source is 
wide-band noise.  But where from? 

    We note that in our previous testing at K and Q bands (in 2005), no 
such variations were seen (we did the tests in an identical manner) -- 
but were were not in the D-configuration at that time. 

    Any ideas?  Any suggestions for useful further tests?  These can be 
done remotely, as there is no need for the analog Milhouse system to 
record the data. 

    I have plots in my office, for those wishing to see the scans.