Telescope gain during and after surface adjustment (Sband)

14jan01 (last mod 16feb01 for focus gain)

    The main reflector surface was measured in oct/nov00. It was found to be way out of adjustment:  rms near 1.2 cm and points near the southern rim up to 5 cm.  The measurements consisted of only 3000 targets situated along the support cables (there were 2 small patches where targets were place on every panel).

    Adjusting the surface started in dec00 and by 08jan01 the eastern half of the reflector had been adjusted and by 13jan01 the adjustements had been completed  (they did the 2nd half of the dish in 1 week???). This adjustment was to take out the large scale variations in the surface. Another adjusting cycle will be needed with all 38000 panels targeted to do the fine scale  panel adjustments.

    On 04jan01 through 08jan01 a number of calibrator sources were tracked using the sband narrow receiver at 2380 Mhz and the standard calibration pattern (4 cross scans at 45 degrees to one another). This pattern took about 5 minutes to do.  Each pattern was then fit for the main beam with coma and sidelines. The 04thru08jan01 figures show the results of these measurements.
    On 13jan01 another set of measurements was done with the dish completed. The 13jan01 figures show the results of these measurements.
 

  1. Figure 1 shows the point source gain in K/Jy versus azimuth angle.  The solid lines are sources rising in the west while the dotted lines are setting in the east. The dotted dotted vertical lines show  transit for northern and southern sources. The fluxes come from chris salters flux.cat catalog. From the plots,you can see that some of the fluxes are  a bit off. The lower figure shows a different set of sources tracked during the sep00 calibration run (B2249+185, and B0134+329(3C48) were the only sources in common).
  1. Figure 2 shows the same point source gain in K/Jy versus zenith angle with the bottom plot coming from the sep00 run.
  2. Figure 3 plots the gain for B2249+185 and B0134+329(3C48) which were common for both dates.  The gain in K/Jy is shown at the top while the ratio gainJan01/gainSep00 is on the bottom. The gain increase varies from 15 to 28 % for the 18 degree dec source and 40% for the 32 deg dec source (rise only).  Comparing 3C48 on 13jan01 and 08jan01 the setting values differ by a few tenths of Jy'.s
  3. Figure 4 plots the great circle azimuth pointing error versus azimuth and zenith angle.
  4. Figure 5 plots the zenith angle pointing error versus azimuth and zenith angle.
  5. Figure 6 shows the mainbeam and total beam efficiencies in the top two plots. The former is the integral over the main beam while the latter includes the main beam and the first sideline.

  6.    The ratio of the beam efficiency to the point source gain is independent of the source flux. These are shown in the bottom two plots (beam efficiency ratio and total beam efficiency ratio).
       The beam efficiency ratio above 15 degrees za increases since the beam spills over and the main beam widens. This causes the point source gain to fall. The lost power remains within the main beam so the beam efficiency falls less rapidly.
       The  total beam efficiency ratio has more scatter since the side lobes are a function of azimuth and zenith angle.
  7. Figure 7 corrects the point source gain for pitch, roll, and focus errors using a model that has been constructed.
The figures show that:
  •     The adjustment is working (they're turning the turnbuckles in the right direction!). It is a bit surprising that the increase for B2248+185 is greater in the west where the adjustment has not yet been done (the truth is the reflector adjustements were probably a lot farther along than "half way")
  • You need to take some of the gain values with a grain of salt (or shot of rum..) . In particular B1749+06 in sep00 and probably B0127+233,  B1602+014, and B001+154 look a bit high so maybe the flux is off a little. B0134+329 (3C48) should be the most reliable.
  • The adjustment is changing the pointing. Errors of 20+ arcsecs will affect the gain at higher frequencies. We probably need to redo the focus curves too. The rms pointing error for all the sources in 11.4 arcseconds. Some of the source position maybe a bit off since they came from the calibrator catalog and not the pointing catalog (although in general the source errors are pretty systematic).
  • The totalBeamEfficiency/(point source gain) is higher for rising sources than for setting for za < 15. You don't see this as much for the mainBeamRatio. Larger  side lobes for rising sources are probably causing this.
  • After correcting for the pitch, roll, and focus errors,  the gain for setting sources (where the dish has been adjusted) is consistently better than the rising side.  After the adjustements were completed, the only full track shows 3C48 stronger on set.
  • processing: x101/010104/doit.pro x101/010113/doit.pro for a description
    plotting  : x101/010104/plotit.pro x101/010113/plotit.pro
     home_~phil