Tsys= Tomt*alpha + (1-alpha)*Tsky + Tamp.Where Tsky includes the sky and scattered radiation, Tomt is the physical temperature of the OMT and Tamp is the amplifier temperature referenced to the input of the amplifier. alpha is the fractional loss of the sky signal caused by these resonances. Rewriting the equation:
Tsys= Tsky + Tamp + alpha*(Tomt-Tsky)Shows that the spectral density at the resonance will increase or decrease (relative to the adjacent frequencies) depending on whether or not (Tomt-Tsky) is positive or negative. If you are on cold sky then Tsky may be 6+ 14(scattered)=20 K. The OMT temperature is about 73 K so there should be a large positive bump. If you look at a 30 jansky source then Tsky= 20+30*9 = 290K is greater than Tomt so Tsys will have a negative going bump at the resonance frequency.
On 20sep01 we measured the cal values for sbw using
absorber (303K) and the sky (20K). The Omt physical temperature was
73K. The correlator was setup for 4x25 Mhz and a 5 second integration with
cal off followed by a 5 second integration with the cal on. We stepped
from 1800 Mhz to 3100 Mhz three complete times. This was repeated on the
absorber and on the blank sky.
The first set of plots shows the spectral density of the cal off spectra from 1800 to 3100 Mhz. The 3 separate measurements at each frequency have been overplotted. The vertical scale is normalized to the median spectral density over each frequency band (Tsys). A bandpass correction was made by averaging over the rfi free regions for sbc1, sb2, sb3, sb4. This was to correct for the 4 50Mhz analog filters in the correlator. The colors on the plot represent:
Figure 1 and 2 show the polarization A results. Figures 3 and 4 are polarization B (the receiver is native linear polarization).
black - spectral density on the absorber red - spectral density on the sky blue * - suspected resonances green vertical lines - the edges of each 25Mhz subcorrelator.
The second set of figures shows the individual 25 Mhz bandpasses that contain suspected resonances. Black is on absorber, red is on the sky. Solid lines are polA and dashed lines are polB.
There is a lot of rfi when on the sky (red) so some of the resonances may have been missed. The blue asterisks were placed wherever there was a positive bump on the sky and a negative going bump on the absorber. A list of the frequencies are:
1831.3, 2013.3, 2131.5, 2197.6 2259, 2580.8, 2662.7, 2711, 2746.7, 2752.6, 2817.5, 2869.6, 2931.7, 2988.8, 3054.3 Mhz.
Some of these may turn out to be ripples in the bandpass. To verify this, polarization data should be taken around these frequencies (c.heiles found that polarization measurements near the resonances in lbw gave polarized results).
The following frequencies have bumps on the sky but no dips on the absorber:
2832 2915 2995 Mhz.