cbandHi cal values measured 28jun12

Links to SECTIONS:
Why the cals were remeasured.
Measuring the high correlated cal using blank sky and absorber
Measuring the other cals on blank sky relative to the high correlated cal.
Diagnostics

Why the cals were remeasured.

• The cals had not been measured for 8 years.
• The stokes calibration was showing an relative error in the polA,polB cal values for up toe 45% for some frequencies.

• below 7400MHz the resonances are mainly in polB (the pol that goes off at an angle in the omt?)
• above 7500 MHz both pols show resonances.
• The many resonances make it difficult to measure the cal values accurately using sky and absorber.
  

Measuring the high correlated cal using sky and absorber: (top)

• absorber measurement: morning of 28jun12
• on Sky measurements:
  
• 29jun12  - 16:00 to 17:30 (horcal then other cals).
  
• Lots of clouds during the sky measurements. This may affect Tsys.
  

    The temperatures used in the computation were:
 

Tabsorber	304.05  (87.6F avg temp sensor)
Tsky	8K
Treceiver	from ant test range feb04
Tscattered	15 K

    The recording and processing steps were:

• the band from 5900-8100  MHz was covered in 3 sections:
  
• Set 1: 5900 to 6800 MHz
• Set 2: 6750 to 7650
  
• Set 3: 7600 to 8100 (only 4 mock boxes used). 
• The mock boxes used 172 MHz bw. sets 1,2 used 7 boxes, set 3 used 4 boxes.
  
• The 25 Hz hardware winking cal was used.
  
• Spectra were sampled at 2 milliseconds
• Each 20 millisecond calOn,Caloff block was averaged (throwing out spectra before after each cal transition)
• The median in each channel (over the 120 seconds) was used as the average value for calDeflection/Tsys
  
• The  5950 to 8100    MHz data  was resampled to a fixed spacing of (172/1024=.168KHz). This was needed since the  spectra were overlapped.
  
• A robust fit (35th order harmonic, 1st order polynomial) was done to the calDeflection/Tsys  ratio (18000+ points). For the sky fits, 6 points about each excluded point were also removed.

    The results of the reduction are:

• Fits to the Average CalDeflection/Tsys  and HcorCal in Kelvins(.ps)  (.pdf):  This shows the average Tcal/Tsys data with the fits over plotted in red.
• Page 1: Tcal/Tsys plots
  
• The top two plots are on the absorber (polA,polB)
• The bottom two plots are on the sky.
• The units are Tsys (about 30K for sky and 300 K for absorber).
  
• The fitRms is computed for the fraction of the spectra used in fitting. The rms and fraction of spectrum used are printed on each plot.
  
• The radiometer equation should give: rms=sqrt(2)./sqrt(172bw/1024chan*120secs)=.0003
• The abs plots are within a factor of 3, the sky plots are a factor of 30 larger.
• This is probably because the fits don't do a good job of fitting the many resonances in the receiver.
  
• Page 2:The  high correlated cal in Kelvins.
  
• The first two plots show the cal fits in kelvins measured from the Sky, absorber, and the sky, absorber ratio (Y factor). The top plot is polA, the middle plot is polB. The dashed line is the receiver temperature used for calSky.
• The Tsys from the sky alone is a little hight. This may be that the Treceiver curve may be a little off.
  
• The bottom plot is the cal In kelvins from the Y factor.
•  The * are spaced every 25 MHz. PolA is black and polB is red.
  

Measuring the other cals using sky and the high correlated cal (top)

• hcorcal(on,off)
• hcal(on,off),hxcal(on,off),h90cal(on,off)
• hcorcal(on,off)
• lcorcal(on,off),lcal(on,off),lxcal(on,off),l90cal(on,off)
• hcorcal(on,off)

• Each 10 second set was processed similar to the hcorcal above.
• The ratio (calDifX/caloffX was computed (X is each cal type) for each 10 second set and then a robust fit was done (same as hcorcal above).
  
• For each of the "other cals" (hcal,hxcal,h90cal,lcal,lcorcal,lxcal,l90cal)" the ratio (calDifX/calOffX)/(HcorCal) was computed. The hcorcal value used was interpolated (in time) from the 3 sets of hcorcal measured. Each cal type measurement was re sampled to a fixed frequency grid covering 5950 to 8100 MHz.
  
• The hcorCalFit in kelvins from the hcorcal measurements (see above) was then multiplied int the other cal fit to give each caltype  in Kelvins.
  

     
  The results of the reduction are:

• OtherCals/Hcorcal and the fits (.ps)  (.pdf):
• These plots show the ratio of the otherCal/hcorcal data and the fits (in yellow) to this ratio.
  
• Page 1: high cals
• page 2: low cals
• top frame polA,bottom from polB
  
• Over plotting all of the cals (in deg K) (.ps) (.pdf). 
  
• Top frame has the high  cals, the bottom framse has the low cals
• The solid lines show polA, the dashed lines show polB.
• The colors show the different cal types.
• the + on the black plot on the upper frame show the 25 MHz samples used for the cal lookup table.

Diagnostics: (top)

• Over plotting the new and old cal values (.ps) (.pdf)..
  
• The solid lines are the new cals. The dash lines are the old cal values.
  
• Top HiCalsPolA: black Diode1 -> polA, Red diode2->polA.
• 2nd HiCalsPolB: black Diode1->polB, Red diode2->polB
• 3rd LoCalsPolA: black Diode1 -> polA, Red diode2->polA.
• 4th LoCalsPolB: black Diode1->polB, Red diode2->polB
• The polA cal values differ between the old and the new, the polB cal values are closer.
• The "Old" cal measurements did not have very good frequency resolution, so they do not show much frequency structure.
• Comparing Tsys measurements using 28jun12 and feb04  cals  (.ps) (.pdf):
• The plots show the tsys measurements jan12 -> 17sep12. This data is measured at 7000 MHz.
  
• Tsys is plotted using the old cal values (polA, polB) and then replotted using the new cal values (polA, polB)
• A horizontal dashed lines shows 30K
  
• Top Frame: hcal (diode1 -> polA, diode2->polB)
• 2nd frame: hcorcal (diode 1->polA, diode1 ->polB)
• The new cals have TsysA,B the same value (about 32K). TsysPolB from the old cals is close to the new TsysValues. Tsys PolA was about 5 kelvin lower than TsysPolBOld and TsysPolA,B new cals
  
• Bottom frame (TsysPolA - TsysPolB)
• Black is Tsys difference  for  the old cals, red is tsys for the new cals.
• The 5K difference TsysA,B for the old cal values is evident.
• The stokes calibration cal diode relative error (deltaG) was probably coming from this difference.
  

• Stokes Calibration check:
  
• The stokes calibration processing computes deltaG which is the error in the the cal Difference (CalPolA - CalPolB)/(mean(calPolA+calPolB))
• The old cal values had vary large deltaG values  (> 50%).
• The plots show the change in the cal values and how the stokes deltaG error improved (,.ps) (.pdf)
• Top frame: plot hcorcal values:
• solid line: new cals
• dashed line : old cals
• black polA , Red polB
• Bottom frame: plot (calOld - calNew)/((calOld+calNew)*.5)
  
• This shows how much the cal values changed between old and new
• block is polA, red is polB
• There is  a large change in the polA values
• * these points are the stokes deltaG error with the old cals
• The old deltaG errors follow the change in the polA calvalues. This shows that the old polA calvalue were wrong.
  
• + the deltaG errors with the new cals

Summary

• The new cal values were measured on 28 and 29jun12.
• the new cal values have more frequency resolution.
• The old cal values were measured with a tuneable filter, and square law detector on the test range.
• The higher freq resolution shows the resonances in the omt. 
• There was  a large change in the polA cal values. The polB cal values did not change too much.
• Tsys for polA and polB is now the same (at 7000 MHz). Before it differed by 5 degK
  
• The cal values were installed on 19jun12.
  
• the cal values were back dated to be valid from 01jan09.
  
• The stokes deltaG error was improved for data taken feb09.
  

home_~phil