On 12jun01 blank sky (ra 8:40:40, dec: 20:00:00) was tracked rise to set. This was AST 14:30 to 17:20 AST. An sband radar run had just completed. During the track, 10 sec cal on,offs were done using the correlator and a 12.5 Mhz bandwidth centered at 1410 Mhz. The lbw hybrid was out so the native linear polarization of the feed was used. The attenuator values were adjusted at the beginning of the run and then they were not touched. Any changes in cal levels would then come from the electronic gain or the cal itself.
The conversion factor from correlator counts to kelvins is:
(Kelvins/Cal ) / (correlator counts/cal)This value can change if the gain of the system after the cal changes, the size of the cal changes, or both. If the gain of the system after the cal changes then this ratio will change but the system temperature will not jump (since the gain variation changes the cal and tsys equally). If it is the cal that changes, then the ratio and the system temperature will jump. The plots show the results:
Images were made of the 407 spectra versus time. The 4 images are: calon A, caloff A, calonB, and caloff B. The processing divided each spectra by the average bandpass and then divided each channel by the frequency average of all channels. The last step removed the total power time variation ( za dependence and the jump at 16:00). You can see standing waves toward the top of the plots and also some interference. I think this interference is out of band interference being moved into the band via the digital filters (this needs to be looked at further).
Top fig. This shows the temperature in the rotary floor room for 12jun01. The sband experiment started around 10 am and stopped around 14:00. The area between the dotted lines is when the data was taken. Mid fig. The black and red lines are the cal scale factor from the cal on/off measurement. The green line is the (temperature - 35 F). The cycle is from the air conditioner. The blue line is the dome zenith angle (scale to fit on the plot). It went from 19 deg to 1.6 degrees. The scale factor is not cycling with the temperature . There is a jump in polB around 16:00. Bot fig. The bottom figure plots Tsys versus time computed using the cal scale factor and the cal off band passes. There is no jump in polB around 16:00.
The cal scale factor and Tsys do not cycle with the measured temperature (at least at the 1%/degF that was measured with the cband system). The jump in polB at 16:00 was in the gain after the cal since the system temperature was stable.
The null result is not as easy to interpret as the positive cband measurement. Not seeing a correlation means that it doesn't exist, or that the ambient room temperature cycling was not being seen by the amplifiers. It is a bit suspicious that the cycling occurred close to 80 F rather than the 75 F air conditioner cycle that occurs at night.