Intro

    After making the oct22 movie of the pulsar observations dynamic spectra, a  dispersive signal was seen on 02oct22  in polA while looking at B1641-45. It was not seen in polB. The  the dispersive signal started around 15:03 ast.

---

The data:

dynamic spectra for B1641-45 for 02oct22  (.jpeg)

• there is 1800 seconds of data
• the avg been averaged to 1 second
• top : polA
• bottom: polB
• You can see the signal varying in freq with time.

Blowup of 1st 400 seconds of  dynamicspectr (.gif)

• This is taken from the first fits file of the observation.
  
• Top: polA
• The signal starts around 2330MHz (in the image)
• It moves to about 2240 MHz after 100 seconds
• this is .9MHz/sec a 2250MHz.. Too large for any ism dispersion.
• Bottom: polB
• The signal is not seen  in polB.
• So it is probably instrumental.

Plots of spectra for 1st 100 seconds (.ps) (.pdf)

• Page 1: over plot spectra for the first 100 seconds.
• The spectra were band pass corrected using spectra from 280 to 400 seconds (when the signal was not present).
• The spectra have been offset for display purposes.
• the first spectra is at the bottom.
  
• Top: polA
• You can see a large wide  bump at the bottom spectra around 2310 MHz.
  
• As you move up (later in time) the bump moves to lower frequency. This is the dispersive looking signal
  
• bottom: polB
• The bump is not seen in polB.
• Page 2: Plot the total power vs time
• The total power was computed throwing out any rfi.
• Black is polA, red is polB
• Top: 1st 400 seconds
• Bottom: blowup 1st 100 seconds.
• During the 100 seconds the total power is changing by about 2% in polA
• Page 3:  temperature and rainfall for 221002
• the temperature and rainfall for 02oct22 was measured at the ao weather station
• this is on top of the lidar building, probably 500meters from the 12meter
  
• The red line shows where the data with the drifting pulse was taken
• the temperature is changing rapidly during this time period.
  

---

Discussion

• The drift in freq is  to large for dispersion from the ism.
• The drifting signal is not seen in polB
• Even if an external signal was 100% circularly polarized, our isolation between pols would probably have a leakage into the other pol.
• So the signal is probably internally generated in the polA receiver.
  
  

When Tsys vs freq  was plotted for the 90 second pulsar cal files taken during oct22 (more info) 

• we saw ripples in the band pass that moved in freq with time.
  
• It was stronger in polB than polA.
  
• These plots were each  averaged over 90 seconds of data (unlike the spectra on this page that were averaged to 1 second.
  
• Plotting vs hour of day showed that the the motion was a function of the hour of day (morning to afternoon). It is probably temperature dependent. It largest during the middle of the day (when it was warmest).
• Looking at the temperature plot during the 02oct22 drift, the temperature was changing very rapidly.

We could blame this problem on the sband polA receiver. But:

total power changes vs freq with xband across 1 Ghz.

• xband observations showed a variation of power with frequency
• total power was computed for each of the 7 172MHz bands covering 8100 to 9200 MHz
• Plotting the  total power vs time for each of the 7 bands would occasionally show a variation in power that increased with the freq of the172MHz band.
• reviewing the data showed the variation with frequency (the hours is how long we took data).
  
• 210430 - 14 hours. seen in polB
• 210504 - 19 hours. seen in polB
• 210521 - 19 hours. seen in polB
• 210601 - 23 hours disconnected lna output and drove the system with cal diode (with amp)
•  no freq variation was seen in polA or polB
  
• 210603 - reconnected polA, polB to lna outputs
• 210608 - 7 hours. may have seen it in polA
• 210701 - 17 hours . freq difference now seen in polA
• 210919 - 12 hours . seen in polA
• 211110 - 26 hours - seen in polA
• For some reason the tp variability with freq  move from polB to PolA after the lna was reconnected to the polA,B cables on 02jun21
• If sband and xband both see a change of power with freq when the temp changes, then the problem might be in a common element (say the polA fiberOptic transmitter).

---

SUMMARY

• A dispersed signal was seen i the dynamic spectra of polA on 02oct22 when observing B1641-45.
  
• It lasted for about 100 seconds and moved about 90 Mhz. This dispersion could not be generated by the ism.
• It was not seen in polB, so it is probably an instrumental affect.
• When plotting the oct22 pulsar tsys vs freq we saw  ripple motion in the bandpass (but mainly in polB)
• This motion was a function of the time of day/temperature.
• The 02oct22 motion occurred when the outside temperature was changing rapidly
• The xband receiver has shown a total power variation with time when looking at separate 172Mhz bands.
• prior to jun21 it was only in polB
• after jun21 it was only seen in polA
• When we drove the xband chain with the cal  diode (rather than the horn,lna) no variation in either freq was seen.
  
• These measurements compared 172Mhz  avged data points covering 1Ghz.
• If the sband, xband variation have the same cause, then the problem may be a common element (say the fiber optic xmiter) rather than polA of the sband lna/horn.
• The freq variation did not appear when xband  was  driven by the cal diode, so this would argue against the fiber optic xmiter.
• If i really wanted to pursue this i could go back and make dyanmic spectra of the xband data when the power was changing rapidly with frequency.
  

processing: x101/221002/pulsarsig.pro