• the source 3C273 (B1226+023) was tracked using the sbh (3-4GHz) receiver.
• 3 min on's were followed by 3 min off source
• The mock spectrometer sampled spectra at 2 millisecond. It used 8 10MHz bands.
• The on,off lasted from 12:51 to 12:58 ast
• The telescope position was aprox. az=349, za=16.7 deg.
  
• Tiedown 12 lost tension because of the high za and warm weather (about 91F).

The first plots shows the estimated pointing/focus error during the on,off (.pdf)

• Top frame:
  
• black pointing error from platform rotation.
• green: pointing error because platform is too low.
• Bottom frame:
• blue: height for correct focus
• green: actual height.
• The full width half max beam width is about 90 asecs.
  

The second set of plots shows the total power vs time and it's spectrum (.ps) (.pdf):

• The total power was computed from 10MHz about 3085 MHz.
• spectra were sampled at .0019968 secs.
• Page 1:
• Top: total power vs time
• black: 180 secs on source
• red: 180 secs off source.
• The strong oscillation dies out after about 100 seconds.
  
• Middle: spectra of total power (entire 180 seconds)
• smooth to 17 millihz.
• black: on source
• The power out to 8 Hz is scintillation from the sun
  
• red: off source
• The green dashed lines are blownup in the bottom frame.
• Bottom: blowup of oscillations
• oscillations occur at .239 Hz and .446 Hz.
• Page 2: break up  the 180  secs into 5 36 second junks to see when the oscillations died out
• Top: total power vs time
• plot each 36 sec block in a separate color.
• Bottom: over plot  total power spectra from the 5 36 second blocks.
• The  .239Hz signal damps out after the 3rd 36 second block (108 seconds)
• The .446 Hz signal decreases till the 4th 36 second block, and then  it starts to increase again.
• Both of these signals are not present in the off.

SUMMARY:

• The pointing procedure was:
• stop the telescope for 11 seconds to drift off source
• adjust the power
• move back on source (within 10 asecs),  wait 3 seconds to settle, then start taking data
• While on source we had oscillations of
• .239 Hz .. which died away after about 100 seconds
• .446 Hz.. which got weaker with time, but did not go away after 180 seconds.
• Off source did not show these oscillations.
• When over a half beam width away from the source, any small motion in the pointing causes a large change in gain (because you are on the edge of the beam shape.
• The oscillation frequency is probably a function of za (az you go up in za, the moment of inertia increases.