A 430 Mhz test feed was installed on carriage 1 for john mathew's to look into the 430 interferometer he wants to install.  4 cross scans on 3C286 were done on 12sep04 while the feed was 5 feet below the paraxial surface (the bottom of the carriage house was used as the reference).On 22 sep04 the feed was lowered about 3 feet and more tests were done. 4 cross scans were done on 3C286 using the test feed and 4 cross scans were done using the 430 Mhz linefeed (for comparison). The setup was;

• The signal from the receiver was passed through a 1 Mhz butter worth IF filter and then detected with a .2 second time constant.
• The data was sampled at 10 Hz.
• The cross consisted of a 1 deg strip in azimuth and a 1 deg strip in za (centered on the source). Each strip took 1 minute. The data was sampled synchronously with the motion at 10 Hz.
• Great circle az,za offsets were installed to point this feed using the ch model. The offsets were .5 degrees in azimuth and -1.6 degrees in za (these value were added to the computed carriage house position).

• Fig 1. The data and fits for the test feed. The top are the azimuth strips and the bottom plot has the strips in zenith angle. The units are Tsys.
• Fig 2. The data and fits for the line feed. The top plots are the azimuth strips while the bottom plots are the za strips. LRR1 was used. It has a warm receiver with a Tsys of about 90K.
• Fig 3. The top two plots are the az and za pointing error (vs az and za) for the test feed. The bottom two plots are the same thing for the line feed.

Some comments on the results:

• The srcDeflection increased by 50% after moving the feed down by 3 feet.
• The Beamwidth decreased by 10%
• The ratio of test feed to linefeed is not as bad as the 12sep04 values since this time i measured the results of the line feed. This comparison includes the system temperatures (since the source deflections are in units of Tsys). The test feed is a warm receiver while the line feed measurement used long range receiver 1 (which is supposed to be cooled ).
• The results for the line feed show that the retensioning of the support cables on the line feed did not change the pointing a lot.
• The measured SEFD for the line feed is not very good. At 15 degrees this feed should have a gain of close to 12 K/Jy. The receiver is not cooled and has a Tsys of about 90K. This would give an SEFD of 7.5 rather than 12.5.

How good is the test feed.

• If uniform illumination then fwhm=1.02*lambda/d --> d=395 ft (120 m)
• If close to cos^2 illumination fwhm=1.29*lambda/d --> d=509 ft (155 m) . this used g(r)=(1-2*lambda*D/r)^2)^1.1 for the taper.

• za=asin(509*.5/870)=17 deg.
• the halfa angle of the horn is then (2*za)=34 deg out to taper = 0.
• The report "cosine squared point feeds for ao - numerical results"  (jj condon 19??) defines the power pattern of the feed  as P(B,th)=NCos^2(B*th) where B is the feed beam width parameter. For 430 Mhz B optimum is 2.24 and z=2.92 (where z is the horn position below the paraxial surface). From our fwhm measurement and setting cos^2()=0 (B*th=90) we get B*34=90 or B=2.64. For this  value of B, the optimum value of z is  between 2 and 3 meters below the paraxial surface. These parameters give a theoretical effective area of 10^4 square meters.

 home_~phil