Intro

Processing the data:

• a band pass correction was computed using data from 12.0822 to 12.093 (before the ripples started0
• All spectra were divided by the band pass correction (to remove the filter shape)
• The dynamic spectra were then made.
• The abs(fft(spc - mean(spcAllspc))) was taken for each .25 second spectra. this gives the ACF.(actually it should be the cosine xform:).
  

• dynamic spectra
• The vertical axis is freq, the horizontal axis is time (.25 sec resolution).
  
• The ripples start around 12.09 hours and they last until 12.865
• The ripples are in polA and in polB.
• There is no large rfi that starts when the ripples appear. It looks like a standing wave problem.
  
• the phase of the ripples in changing with time.
  
• acf
  
• The vertical axis is lag delay
• the horizontal axis is time.
• When the ripples start there is a correlation around lag .4 usecs. It lasts until the ripples turn off at 12.86.

The plots show the total power and average acf's (.ps) (.pdf)

• Page 1: total power and spectra
• top: the total power vs time
• the total power between 30 and 80 MHz is computed for each .25 second spectra.
• the total power jumps up at 12.095 hrs and comes back down at 12.86 hrs (while the ripples are present).
  
• When it jumps up, there is a large increase for 1 .25 second spectra
• middle: a blowup of the total power when it jumps.
• the large jump lasts for 1 .25 second sample
• bottom: spectra around the jump at 12.095
• The black spectra is .25 secs before the jump
• the red spectra is at the jump
• There is a large oscillation with a spacing of about 2.25 MHz.
  
• the green spectra is after the jump
• Page 2 average spectra and acf during the ripples
• Top: compare spectra during the ripples and before the ripples.
• PolA is black, polB is red
• the solid line is the average during the ripples
• the dashed line is prior to the ripples.
• the large spectra at the jump was not included in the average.
  
• The spectra during the ripples is  a little higher than outside the ripples. There is no large rfi spike.
• Bottom: the average acf.
• the acf for each .25 second spectra was computed and then averaged.
• There is a spike at .405 usecs in both polA and  polB. (this corresponds to the horizontal line in the acf images above).
  
• For n =1 this would be 60.8 meters.
• if n=.7 (a cable) it would be about 42 m. (140 ft).

SUMMARY:

• Ripples appeared across the entire spectral band pass and lasted for about 1 hour.
• The ripples had a spacing of about 2.469 MHz or .405 usecs in lag delay
• The ripple had a very large oscillation for the first .25 sec spectra.
  
• It looks like a bad match problem, but i  don't think there is a cable of this length in the system.
• It is seen in both polarizations. This argues against it being in a single cable (since the 2 pols have separate cables).
  

processing: x101/220127/callisto_ripples.pro

callisto page