Stability of bump in center of bandpass
A bump was seen
in the center of the bandpass on 20sep07. The bump was not present when
we bypassed the mixer chassis. The bump was an increase in gain (not
extra noise). A test was done on 20sep07 to check the stability of this
bump. The setup was:
The plots show the results of
the averaging (.ps) (.pdf):
- Red noise generator feeding new mixer chassis.
- Box 103,104 sp0,1 were used in full stokes mode with 8192
- Hires spectra with decimate by 8 and 1 second integrations were
taken for 3600 seconds.
- Average the data to get 12 5 minute averages.
- Normalize the 2nd thru the 12 spectra by the first 5 minute
- Compute the rms across frequency channels to see if the rms
increased as we go farther away from the original 5 minute
- Page 1 top is first 5 minute integration used for normalization
- Page 1 bottom: The 11 normalized spectra for box103sp0pola. The
bottom spectra is 5 minutes after the normalization spectra. The top
spectra is 55 minutes after the normalization spectra.
- Page 2: The rms/mean for each of the 11 normalize spectra. The
horizontal axis is the minutes since the normalized spectra was taken.
All spectra (box103,104,sp0,1, polA,B) are included. The dashed line is
the expected rms/mean using the radiometer equation.
- The bump from the mixer chassis in the center of the spectra is
stable over time periods of 1 hour using 5 minute integrations.