Intro:

• Lbw, linear polarization.
  
• start thru 24oct10:1230-1470 filter with 1280 hipass filter
  
• 25oct10 onward: 1320 hipass filter
• 20oct10: switched from 750 wide IF to 750 narrow band If.
• 08apr13: 1370 hipass with 1230-1470 filter.
  
• Interim correlator using:
  
• 9 level sampling, 1 second integerations 240 or 300 sec on ,offs.
  
• sbc 1: 2024 channels, 25 Mhz. polA,B each sbc
• blanking is used when the 1350 radar lies in the observed band.
• They are turning off the tiedowns (to prevent motion between the on and the off)
• Dates:
• 20oct10: switched 750 wideband to 750 narrow band IF
  
• 25oct10: switched 1230-1470 and 1280 hipass filter to 1320 hipass filter
• 01Feb11: 1320 hipass polB connector was found to be damaged and it was repaired.
  

A2335 2011

18may11: 1367 rfi present  (top)

• page 1: 600 second average spectra.
• black-pola, red:polB
• green dotted lines: freq for puntasalinas chan 18 long range pulse
• blue dotted lines: freq for puntasalinas chan 18 short range pulse
• Page 2: Baseline removed from the plots on page 1.
• Page 3: rms/mean by channel for the 600 1 second spectra.

• punta salinas chan 18 was broadcasting during this observation.
• Looking at the hilltop monitor data:
• They were not broadcasting at all frequencies
• occasionally the hilltop monitor saw power at the channel 18 freq.
• This channel must have been left partially on...
• You can see the different bandwidths for the long and narrow range pulses. This is more obvious in the rms by channel plots.
• To see what frequencies belong to punta salinas:
• http://www.naic.edu/~phil/rfi/rdr/puntaSalinas.html
• Look at the 2nd table : punta salinas freq channels.
• At each ipp, two frequncies are sent (col 2 and col 4 of the table).
• The radar can then jump between frequencies on an ipp to ipp basis.
• When lband observing is being done, punta salinas has agreed to stay in modeA of the table.
• For those using idl at AO i've written an idl routine to return the puntas salinas frequencies:
• @phil
• @rfiinit
• rfipuntasalinasfreq,rdfinfo
• explain,rfipuntasalinasfreq   .. will give doc within the idl session
• software --> idl routines related rfi   also has the documentation.
  

30jan11: polB baseline instabilities  (top)

• Page 1 thru 3
•  30jan11
• rfi 1360,1370 (last 2 on/off pairs where blanking was off).
  
• Pages 4-6
  
• 31jan11
• rfi:
  
• gpsL3 1381
• 1360,1370 for last 2 pairs where radar blanking was turned off.
  
• Pages 7-9
• 01feb11
• rfi: 1360,1370, and gpsL3 at 1381.
  
• For each day the 3 pages are the 12.5 Mhz band and then the two 25 Mhz bands.
• Black is polB, red is polB.
• Offsets have been added for display. The source names are show at the right of each spectra.
• The red polB bandpass has much more ripple than the polA especially in the 25 Mhz bands.
• radar blanking enabled:
• 30jan11: all on/off pairs except last 2.
• 31jan11: all on/off pairs except last 2.
• 01feb11: radara blanking off
  

• The first page overplots the total power samples for polA and polB
• The 2nd page overplots the on and off source for polA or polb.
• The total power samples within an on,off pair  have been normalized to the median value for each  off (Tsys).
  
• The yaxis is total power [Tsys).
  
• total power vs time 30jan11 (.ps) (.pdf)
• total power vs time 31jan11 (.ps) (.pdf)
• total power vs time 01feb11 (.ps) (.pdf)

• Going through previous days data:
• the ripple was not present 28jan11 observing 1 to 6am (friday)
  
• the ripple was present on 29jan11 observing 1 to 6am
• On 01feb10 the polB connector to the 1320 hipass filter was found to be damaged. It was then repaired.
• The damage probably occurred during the maintenance day of 29jan11.
• Other Notes prior to finding the damaged 1320 hipass filter connector:
  
• The ripple is mainly in polB (the receiver is in linear pol mode).
• The ripple is seen in the 4 interim correlator bands so it is probably before the 2nd if2 mixers.
• The ripple appears whether radar blanking is on or off.
• When radar blanking is on, there are still dips every 12 secs in the images. This is probably from the aerostat radar compressing the front end.
• The polB bandpass shape has a lot more structure than the polA bandpass shape.
  
• Looking at the total power vs time, there does not appear to be any more jumps in polB then there is in polA.
• ToDo:
• I ran for 80 minutes with the dome sitting still
• 4 sbc, polA,B 25 Mhz each. 1320 hipass filter in. No radar blanking.
• Didn't see polb jump or have ripples any worse than polA.
• simulated on/offs by taking 300 secs as on, next 300 secs as off. ended up with 8 on/off pairs.
  
• see if things change when we use a different rf filter.
• switch polA,B out of the dewar and see if the problem switches to polA.

Nov10: switching to 750 narrow band IF caused power adjustment problems. (top)

• Prior to using the 750 nb IF, they would have cima auto adjust the power levels at the start of each source.
• When the 750 nb IF was used, they had to disable this:
• The power adjust sequence is:
• adjust the upstairs IF1 power levels
  
• adjust the downstairs IF2 power levels
• adjust the interrim correlator power levels.
• the program does this sequentially and does not go backwards to readjust any levels.
• the 60 MHz bw of the 750 nb IF was narrow enough that the above sequence did  work automatically.
• The user had to manually add 5 db in the power adjust menu.
• Once the power was set, you didn't want the auto adjust sequence at the first loop of the on/off pattern to try and readjust it. So this was disabled.
• Everything worked fine until they moved to  a new source with different frequencies.
• changing the frequencies entails a complete setup of the if/lo.
  
• This includes resetting the power levels to the default values.
  
• These default values were lower than what was needed.
• Since they had no more power adjusts, the system ran with these lower power levels.
• So the observing would start and run ok  until a source with a different set of frequencies was selected.
• The power levels would get reset to the default values and the observing would continue with power levels too low.

• black is polA, red is polB.
• The plots show the total power for sbc 1,2. These are the narrow band spectra.
• The y axis is the measured power/The Optimum power .
  
• It should be around 1.
  
• The attenuator steps are +/- 1 db so we can typically set it to +/- 10 or 20 %.
• You can see where it drops down to .1 to .5 when a new source/set of freq are setup.
• The positive going narrow spikes are the calon,caloff at the end of each on,off pair.

Checking Tsys vs correlator power levels.  (top)

• Top: Tsys versus the correlator power level (1=optimum power).
• black is polA, red is polB
• * are sbc1, sb2: 12.5 MHz bandwidth
• + are sbc3 with 25 MHz bandwidth
• There is not much correlation of Tsys and mean power level.
  
• Middle: Tsys vs Za.
• Bottom: Tsys Vs time.
• Each point is a calon,off tsys measurement.
• The dashed green lines are the start of each datafile (day). The gaps are not real.
• You can see a large jump (-7 kelvins) on 31oct10. This is after the dewar was warmed up and then cooled back down (it had been making a ratcheting noise).
  

• switching to the 750 nb IF caused a power level problem because they could no longer auto adjust the power levels at the start of each source.
• On 06dec10 the auto adjust  power in cima was fixed to work with the 750 narrow band filter. A2335 should now select the adjust power "at the start of each loop".
• Computing the tsys vs correlator power level showed little change in Tsys with the decreasing power level. So they data is probably ok.
  
• There was a large jump (-7K) in polA on 31oct10 after the lbw warmup/cooldown. We need to figure out if this was real or just the cal values changing.

Daily plots of oct10 data: (top)

• The total power for each on,off is normalized to the mean off power.
  
• The total power vs time was plotted with polA,polB overplotted..
• This was done for the 3 bands
  
• The On source,off source fore each pol were overplotted.
• band 1 polA,polB and band 3 polA,polB are plotted.
• 01oct10 (.ps) (.pdf)
  
• 02oct10 (.ps) (.pdf)
  
• 03oct10 (.ps) (.pdf)
  
• 04oct10 (.ps) (.pdf)
  
• 06oct10 (.ps) (.pdf)
  
• 07oct10 (.ps) (.pdf)
• 13oct10 (.ps) (.pdf)
  

notes on the data

• PolA and polB do not track together for some on,offs of 02,03,06 oct.
• This may be coming from the polarized power vs az that comes from the dish.
• 07oct10: variations in  power within on or off.
• 11754 source:
  
• 02oct10,06oct10: on power = off power for polA,polB
• used 240 seconds on,off
  
• 01oct10,07oct10: on power < off power for all 3 bands. and polA,polB
• used 300 secs on,off
• So the off positions are different for the two sets.
  
• The 300 second off position must have a continuum source of 1% of tsys. With sefd=2.5 this is about 25 mJy.
  
• nvss shows no continuum source in the off position. The 408 haslam survey doesn't show much either (although its not sensitive enough..)
  

Total power plots by month: (top)

• The total power for each on,off is normalized to the mean off power.
  
• The total power vs time is plotted with polA,polB overplotted..
• The green vertical lines show the start of an on position.
• The source name is printed up against this line.
  
• The blue vertical lines show the start of the corresponding off position.
• 
  
• This is done for the 3 bands
  
• The On source,off source for each pol are overplotted.
• band 1 polA,polB and band 3 polA,polB are plotted.
• processing: usr/a2335/doit_tp_mon11.pro
  

Jun11:

• 01jun11 (.ps) (.pdf)
• 02jun11 (.ps) (.pdf)
• 03jun11 (.ps) (.pdf)

May11:

• 16may11 (.ps) (.pdf)
• 18may11 (.ps) (.pdf)
• 19may11 (.ps) (.pdf)
• 26may11 (.ps) (.pdf)
• 27may11 (.ps) (.pdf)
• 28may11 (.ps) (.pdf)
• 30may11 (.ps) (.pdf)
• 31may11 (.ps) (.pdf)

Apr11:

• 04mar11 (.ps) (.pdf)
  
• 26apr11 (.ps) (.pdf)
  

Mar11:

• 10mar11 (.ps) (.pdf)
  
• 14mar11 (.ps) (.pdf)
  
• 18mar11 (.ps) (.pdf)
  
• 19mar11 (.ps) (.pdf)
  
• 20mar11 (.ps) (.pdf)
  
• 21mar11 (.ps) (.pdf)
• 23mar11 (.ps) (.pdf)
• 24mar11 (.ps) (.pdf)

A2749: 2013

08apr13: polB drifts at start of run. (top)

• avg on, off source, compute on/off -1 then convert to Jy.

• (on/off -1)  plots (.ps) (.pdf)
  
• Page 1: rms in Jy for each on/off -1.
• channels 850-11200 where excluded from the rms computation.
• black is polA, red is polB.
• The green line is the expected rms:
• sqrt(2)/(chanWidth*time*bwFactor)
• bwFactor:
• hanning smoothing: (1.5*1.5) = 2.25
• sinx/x channels (rather than rectangle): 1.2
• The rms is computed on the unsmoothed on/off -1
• Page 2-n: individual on/off -1 3x3 per page
• This data has been smoothed by 5 channels.
  
• The vertical scale is in Jy.
• The source name  is placed at the top of each plot.
• Time goes left to right, and then down a row.
• The vertical scaling is -.05 to .02 Jy for the 1st 4 plots, then -.02 to .02 Jy
  

• Page 1: plot sequentially in time
• black polA, red polB
• green lines show start of onsrc, blue lines show start of off src
• I only plotted the first 3 sbc: (12.5 MHz pola, 12.5 MHz polB, 25 MHz polAB
• The red polB has a large slope that differs from polA for the first 3 on/off patterns
• Page 2:
  
• black On src, red Off src
• frame 1 polA 12.5 MHz,
  
• frame 2 polB  12.5 MHz
• frame 3 polA 25 MHz
• frame 4 polB 25 MHz
• The polB drift was present for the first 5 patterns
  
• the drift is about 2% of Tsys.

Summary:

• Pol B  drifted by about 2% (gain?) for the first 5 patterns (about 50 minutes) of the evening (which started around 21:00)
• PolA looked stable
• The rms on the bandpasses were reasonably closed to the expected rms's
  
• They used the 1370 hipass filter along with the 1230 to 1470 filter
• We should probably do a wiggle test on the cables for these filters.

A2749 monitoring  (top)

• col 2: total power vs time:
• Page 1:
  
• black polA, red polB, green lines show the start of onsrc, blue lines show start of off src
• I only plotted the first 3 sbc: (12.5 MHz pola, 12.5 MHz polB, 25 MHz polAB
• Page 2:
• black On src, red Off src
• frame 1 polA 12.5 MHz, frame 2 polB  12.5 MHz,frame 3 polA 25 MHz,frame 4 polB 25 MHz
• Col 3: rms and on/off-1 3x3
• Page 1: rms for each on/off -1 . exclude channels 850-1200
• Page 2-N: on/off -1 9 per page. the vertical scale is Jy.
  
• Col 4: on/off   5x5
• Set of .png files  showing all the on/offs (25 per page). the srcname and scan number is at the top. The vertical scale is Jy.