cband calibration

aug,2002

     The cband  receiver covers 3.95 to 6.05 Ghz. It is a native linear receiver with the option to create circulars using a hybrid in the iflo cabinets.

Sections:

History
Recent system performance measurements
Daily monitoring of Tsys
Dewar temperatures
Calibration measurements
Miscellaneous
cal values

History:

Calibration measurements:

11feb13: calibration run after dewar reinstalled in the dome.
13jan13: calibration results: jun12-dec12. updated pointing offset.
02jul09: tsys vs freq for entire band.
may05-oct05: cband GAIN CURVES.
may05-oct05: System performance of data used for gain curves.
may03-mar04: cband GAIN CURVES.
may03-mar04: System performance of data used for gain curves.
Aug02-Dec02: cband GAIN CURVES.
aug02-dec02: System performance: Gain,Tsys,Sefd.. using calibration runs.
08aug02: horn moved down 1.08" to put phase center at focus.
sep01-aug02: cband GAIN CURVES.
sep01-aug02: System performance:Gain,Tsys,Sefd... using calibration runs.
22dec01: cband sefd,pointing errors (preview)
02aug01: gain/sefd after panels in the northern half of the dish adjusted
28jun01: gain/sefd on 3C48 after SE quad of dish adjusted
27jun01: gain/sefd with pitch,roll,focus correction after SE quad adjusted
   jan01: calibration after 1st surface adjustment
   sep00: calibration runs
 

Miscellaneous:

200210: Tsys PolB problem is dewar gain dropout at the crosshead frequency.
191212: Tsys polB jumps up on 12dec19
190501:  post maria gain fits.
19mar19: 11-19mar19 polB gain jumps. loose bias cable.
27jul16: installed new cal values measured back on 28apr16
15jun15: added page describing cband polbox from 2005
13apr14: p2865 4-5 ghz, average spectra and rms/mean
05jun13: 5-6 Ghz on the telescope after 5020  rfi turned off.
04jun13: 2nd trip to yiyi's tower. Source of 5020 rfi.
02feb13: polB instability at crosshead freq. Stage 2 drain current.
26sep12: sefd ratio PolA/PolB after new amp installed
cband instabilities 2012
09feb11: look at cband,iflo stability with the shutter closed.
19nov10: Tsys increases when horn gortex fabric cover comes loose.
27sep04: oscillations/jumping around of Tsys  in polA
29jul02: resonances in the cband omt 29jul0
07jun02: Polarized Tsys vs azimuth,za

190501: Post maria gain curves. jan18-apr19

    x102 calibration data 01jan18 thru 30apr19 was used to fit gain curves for the cband receiver. These curves reflect the degraded gain after hurricane maria.
 The gain curves were installed in the system 01may19. They are valid for the period starting 01dec2017 (after hurricane maria).

    Previous fits only used za terms. The current fits included azimuth dependence.

    gain= c0 + c1*(za-10) + c2*(za-10)^2 + c3*(za-10)^3  +
               c4*cosAz + c5*sinAz + c6*cos2az + c7*sin2az + c8*cos3az + c9*sin3az.


Curves were fit separately to the frequencies: 4500, 4860, 5000, and 5385 Mhz,. There were a total of 1987 points over the 4 frequencies (average of 500  pnts/freq).
For a point to be included in the fit:

System performance plots for fit data:

The first set of plots show the system performance plots for the data used for the fit (.ps) (.pdf)


Fit results:

    The plots show the fit results (.ps) (.pdf)
    The routine gainget() or corhgainget() will now return the cband gain for data after 01dec17 from these equations. The coefficients can be found in the ascii file  data/gain.datR9 (this is provided in the AO idl distribution for correlator routines). You can also find a copy of it at AO in gain.datR9.
   
     A description of the gain curves can be found at: gaincurves:
processing: x101/cb/apr19/doit.pro,dogainfit.pro

19mar19: polB gain jumps. loose bias cable.

plots:
the bias currents for 18feb19 thru  18mar19 (.pdf)
daily tsys measurements for cband 18feb19-18mar19 (.ps) (.pdf)
a3123  tsys vs freq for 12-15mar19 (.ps) (.pdf)

    Tests done tracking the moon at cband (18mar19) showed polB jumps of many db  in the cband gain.
The jumps were upstairs, since the upstairs power meter showed the power in polB increasing.
Testing the rf attenuator showed no problem. Felix looked at the cband bias and saw jumps in polB bias current for stage 1.

    The plot shows the bias currents for 18feb19 thru  18mar19 (.pdf)

    The 2nd plot shows the daily tsys measurements for cband 18feb19-18mar19 (.ps) (.pdf)


    Project a3123 was using cband to do position switching and drift scans on different objects.
I used the a3123 cal on off data from 12-15mar19 to see how their observations were affected.

The plots show a3123  tsys vs freq for 12-15mar19 (.ps) (.pdf):

Resolution:

processing: x101/190319/cbpolBjumps.pro

13apr14:p2685 spectra and rms's using 4-5 ghz   (top)

    p2685 took full stokes data with the mock spectrometer covering 3985 to 5055  using  the cband receiver.

the setup was:

processing the data:

The first plots show the average spectra for the entire 46 minutes (.ps) (.pdf)

dynamic spectra using the power and the rms/mean

    Dynamic spectra were made using the average power of each file as well as the rms/mean for each file

processing: x101/140413/p2865_joanna.pro, plotbms.pro
 

11feb13: calibration run after cband reinstalled in dome.

    The cband receiver was reinstalled in the dome on 11feb13. A calibration run was done that evening (the receiver was cold).
The setup was:
The first set of plots show the calibration results for 11feb13 (.ps) (.pdf):
The 2nd set of plots compares the source B0600+219 before and after the installation (.ps) (.pdf)

Summary:

processing: x101/130211/x102cb.pro

02feb13: cband instabilities in polB.

    On 02feb13 chris and tapasi reported an instability in cband. They saw the band going up and down at about .83 seconds. 1/.833 is 1.2 hz which is the crosshead frequency.
    I looked at the x111 data taken 21:00 01feb13 as well as the dewar bias monitoring:

x111 cband data 01feb13 (.ps) (.pdf):

cband bias monitoring (drain current)  jan-feb13 (.ps) (.pdf):

Summary:

processing: x111/130202/cbandripple.pro



26sep12: SEFD ratio polA,polB after new amp installed in polA

    A new lna was installed in cband polA on 25sep12. After installation, Tsys for polA (at 5GHz) was running about 40 K while Tsys for polB was about 30K.
This difference could be caused by:

To check which of these has occurred, calibration scans were done on 26sep12:
The plots show the results of the calibration measurements (.ps) (.pdf):

SUMMARY:

processing: x101/120926/x102cb.pro




19Nov10: Tsys increases when horn gortex fabric cover comes loose.

    Tsys for the cband receiver increased from 30oct10 (polA 26K) to 13nov10 (polA 38K). It then decreased back to 26 K on 19nov10. After inspecting the feed it was found that the gortex fabric horn cover had come loose. Water drops were found inside the horn. The change in tsys was probably a function of how much water precipitated on the the inside of the Horn. This would be a function of the horn temperature and the dewpoint.
    The top of the reciever is inside the gregorian dome room which is air conditioned. The bottom part of the receiver  and the horn are under the rotary floor. This part is not air conditioned.

    The plots show the Tsys and gain variations for this period (.ps) (.pdf):
Summary:
processing: x101/101119/cbandTsys.pro

May05 to oct05  fit GAIN CURVES to calib data.  (top)

    On 25may05 new amplifiers were installed in the cband dewar. Calibration data was then taken from 25may05 through 31oct05 (see the cband gain data ). Gain curves were then fit to the data.  The gain curves were installed in the system 22nov05.
    The functional form used was:
    gain= c0 + c1*(za-10) + c2*(za-10)^2 + c3*(za-10)^3

Curves were fit separately to the frequencies: 4100,4400,4500,4700,4860,5000,5400,5690, and 5900 Mhz. There were a total of 1774 points over the 9 frequencies (average of 200 pnts/freq). There was no attempt to fit to the azimuth dependence because of a lack of azimuth coverage.
    The plots show the gain data (black) and the fits (red) (.ps) (.pdf) for the various frequencies:     The routine gainget() or corhgainget() will now return the cband gain for data after 25may05 from these equations. The coefficients can be found in the ascii file  data/gain.datR9 (this is provided in the AO idl distribution for correlator routines). You can also find a copy of it at AO in /pkg/rsi/local/libao/phil/data/gain.datR9.
processing: x101/cb/may05/dogainfit.pro

May05 thru oct05 : System performance data used to compute gain curves.  (top)

     Heiles calibration scans done from 25may05  (after the new amps were installed) thru 31oct05 were used to measure the  system performance. This data was then used to compute the gain curves for data taken after 25may05.
    The first set of plots show the system performance with all frequencies over plotted (.ps) (.pdf). The sources are identified by symbol and the frequencies by color. The second set of plots has the system performance plotted separately for each frequency (.ps) (.pdf).  The figures are:
  • Fig 1 4100 Mhz Gain,Tsys,sefd,bmwid. 105 points
  • Fig 2 4100 Mhz  coma,sidelobes, beam efficiencies
  • Fig 3 4400 Mhz Gain,Tsys,sefd,bmwid. 106 points
  • Fig 4 4400 Mhz  coma,sidelobes, beam efficiencies
  • Fig 5 4500 Mhz Gain,Tsys,sefd,bmwid. 244 points
  • Fig 6 4500 Mhz  coma,sidelobes, beam efficiencies
  • Fig 7 4700 Mhz Gain,Tsys,sefd,bmwid. 106 points
  • Fig 8 4700 Mhz  coma,sidelobes, beam efficiencies
  • Fig 1 4860 Mhz Gain,Tsys,sefd,bmwid. 242 points
  • Fig 2 4860 Mhz  coma,sidelobes, beam efficiencies
  • Fig 1 5000 Mhz Gain,Tsys,sefd,bmwid. 447 points. The source B1645+174 looks like it has a bad cal value. The gain is high, the Tsys is high, but the sefd was normal.
  • Fig 2 5000 Mhz  coma,sidelobes, beam efficiencies
  • Fig 1 5400 Mhz Gain,Tsys,sefd,bmwid. 341 points
  • Fig 2 5400 Mhz  coma,sidelobes, beam efficiencies
  • Fig 1 5690 Mhz Gain,Tsys,sefd,bmwid. 98 points
  • Fig 2 5690 Mhz  coma,sidelobes, beam efficiencies
  • Fig 1 5900 Mhz Gain,Tsys,sefd,bmwid. 98 points
  • Fig 2 5900 Mhz  coma,sidelobes, beam efficiencies
    • processing: x101/cb/may05/doit.pro

    27sep04 oscillations/jumping of tsys in polA.  (top)

        The system temperature for polA has been jumping around. On 24 sep04 data was taken with the radar interface (ri).  40 Mhz around 4850 Mhz was detected with a 20 usecond time constant and sampled at 10 useconds. 90 seconds of data was taken. The plots show the variations in pola.     The 1.2 hz power is probably coming from the refrigerator for the dewar. Somehow it is modulating the gain at this period. This is probably the culprit for the jumps in Tsys in polA.
    note: When the dewar was brought down to the lab, one of the cables in the dewar had lost its insulation and was intermitantly  shorting out. It may have been that the vibration of the refrigerator caused the cable to short.
    processing: x101/040924/doit.pro

    May03 to Mar04  fit GAIN CURVES to calib data.  (top)

    link to gain curve plot

        Gain curves were fit to the cband gain data from 05may03 through 29feb04.  On 05may03 new  cals were installed and the cal cable was tightened. This data is described below.The plots show the gain data (black) and the fits (red) for 4500, 4860, 5000, 5400,5690, and 5900 Mhz. These gain equations were installed on 08mar04.

    The only puzzling thing is that the gainfits for 4500Mhz and 4860 Mhz are almost identical..

        The routine gainget() or corhgainget() will now return the cband gain for data after 05may03 from these equations. The coefficients can be found in the ascii file  data/gain.datR9 (this is provided in the AO idl distribution for correlator routines). You can also find a copy of it at AO in /home/phil/idl/data/gain.datR9.

    processing: x101/cb/may03/dogainfit.pro

    may03 thru feb04 : System performance data used to compute gain curves.  (top)

         Heiles calibration scans done from 05may03 (after the cals installed) thru 29feb04 were used to measure the  system performance. This data was then used to compute the gain curves used after 05may03.
        The first set of plots show the system performance with all frequencies overplotted. The sources are identified by symbol and the frequencies by color.
  • Fig 1 shows the distributions on the dish of the measurements.
  • Fig 2 has the Gain in Kelvins/Jansky. This relies on the cals and the source flux. Ths  next plot is Tsys vs za in Kelvins followed by the SEFD  (System Equivalent Flux density) in Janskies / Tsys. At the bottom is the average beam width in arc seconds.
  • Fig 3 plots the coma parameter, first sidelobe height below the peak, the main beam efficiency, followed by the main beam + 1st sidelobe beam efficiency.
  • Fig 3 has the pointing errors in az,za, and total (added in quadrature) for the data.
    • The second set of plots has the data plotted separately for each frequency. The colors and symbols are used to differentiate the sources. The figures are:
  • Fig 1 4500 Mhz Gain,Tsys . 275 points
  • Fig 2 4500 Mhz sidelobes,beam efficiencies
  • Fig 3 4860 Mhz Gain,Tsys 275 points.
  • Fig 4 4860 Mhz sidelobes,beam efficiencies.
  • Fig 5 5000 Mhz Gain,Tsys 567 points.
  • Fig 6 5000 Mhz sidelobes,beam efficiencies
  • Fig 7 5400 Mhz Gain,Tsys 566 points.
  • Fig 8 5400 Mhz sidelobes,beam efficiencies
  • Fig 9 5690 Mhz Gain,Tsys 292 points.
  • Fig 10 5690 Mhz sidelobes,beam efficiencies
  • Fig 11 5900 Mhz Gain,Tsys 292 points.
  • Fig 12 5900 Mhz sidelobes,beam efficiencies
    •     Use the second set of plots to identify the sources.
    processing: x101/cb/may03/doit.pro

    Aug02 to Dec02 fit GAIN CURVES to calib data.  (top)

    link to gain curve plot

        Gain curves were fit to the cband gain data from 28aug02 (after horn focus change) to 30nov02. This data is described below.The plots show the gain data (black) and the fits (red) for 4500, 4860, 5000, and 5400 Mhz.

        The routine gainget() or corhgainget() will now return the cband gain for data after 28aug02 from these equations. The coefficients can be found in the ascii file  data/gain.datR9 (this is provided in the AO idl distribution for correlator routines). You can also find a copy of it at AO in /home/phil/idl/data/gain.datR9.
    processing: x101/cb/dec02/dogainfit.pro

    aug02-dec02: System performance:Gain,Tsys,Sefd... using calibration runs.  (top)

         Heiles calibration scans done from 28aug02 (after the horn focus) were used to show the system performance after the horn focus.
        The first set of plots show the system performance with all frequencies overplotted. The sources are identified by symbol and the frequencies by color.
  • Fig 1 shows the distributions on the dish of the measurements.
  • Fig 2 has the Gain in Kelvins/Jansky. This relies on the cals and the source flux. Ths  next plot is Tsys vs za in Kelvins followed by the SEFD  (System Equivalent Flux density) in Janskies / Tsys. At the bottom is the average beam width in arc seconds.
  • Fig 3 plots the coma parameter, first sidelobe height below the peak, the main beam efficiency, followed by the main beam + 1st sidelobe beam efficiency.
  • Fig 3 has the pointing errors in az,za, and total (added in quadrature) for the data.
    • The second set of plots has the data plotted separately for each frequency. The colors and symbols are used to differentiate the sources. The figures are:
  • Fig 1 4500 Mhz Gain,Tsys 241points.
  • Fig 2 4500 Mhz sidelobes,beam efficiencies
  • Fig 3 4860 Mhz Gain,Tsys 245 points.
  • Fig 4 4860 Mhz sidelobes,beam efficiencies.
  • Fig 5 5000 Mhz Gain,Tsys 279 points.
  • Fig 6 5000 Mhz sidelobes,beam efficiencies
  • Fig 7 5400 Mhz Gain,Tsys 273 points.
  • Fig 8 5400 Mhz sidelobes,beam efficiencies
  • Fig 9 5690 Mhz Gain,Tsys 35 points.
  • Fig 10 5690 Mhz sidelobes,beam efficiencies
  • Fig 11 5900 Mhz Gain,Tsys 35 points.
  • Fig 12 5900 Mhz sidelobes,beam efficiencies
    •     Use the second set of plots to identify the sources.
    processing: x101/cb/dec02/doit.pro

    28aug02 horn moved down 1.08" to put phase center at focus.  (top)

    link to new plot: system performance before and after the horn move.

        The cband horn was lowered 1.08" toward the tertiary to place the phase center at the focus (this is explained on the lband focus change page). The source J0738+177 was tracked on 26aug02 before the horn was moved and then again on 28aug02 after the horn positioning. The plots show the results of the focus change.

        Above 8 degrees za the gain improved. Below 6 degrees it degraded. The za pointing error for the source setting is up to 25 asecs. The change in pointing because of the move was up to 10 asecs for rising azimuth.

        The focus curves done by moving the  platform (bottom plot. cband is bright red color) show that the cband horn was in focus (platform height=1256.35 feet) at za=5 degrees. Below za=5 the platform focused higher (the ray path is too short) while for za > 5 the platform was focusing below 1256.35 (the ray path is too long). Making the ray path shorter by 1 " should have hurt the gain below za=5 and helped it above za=5.
    The model of the focus height using AO9 as the reference agrees with the platform focus curves and the data we took here. The model of the focus height with the reflector at the origin (about 2" difference in focus) does not.

    sep01 to 27aug02 fit GAIN CURVES to calib data.  (top)

    link to gain curve plot

        Gain curves were fit to the cband gain data from 01sep01 (after dish adjusted) to 27aug02 (before horn focus changed). This data is described below.The plots show the gain data (black) and the fits (red) for 4500, 4860, 5000, and 5400 Mhz.

        The routine gainget() or corhgainget() will now return the cband gain for data after 01sep01. The coefficients can be found in the ascii file  data/gain.datR9 (this is provided in the AO idl distribution for correlator routines). You can also find a copy of it at AO in /home/phil/idl/data/gain.datR9.
    processing: x101/cb/aug02/dogainfit.pro

    sep01-aug02: System performance:Gain,Tsys,Sefd... using calibration runs.  (top)

        All of the heiles calibration scans sept01 to aug02 for the cband receiver were used to plot the system performance. This includes dedicated calibration runs and individual calibration scans done by observers. Scans with pointing errors greater than 30 Asecs, extended sources, tsys greater than 46K, or uncertain flux at 5 Ghz were not used. This data was taken under differing weather conditions. After the above selection, there were 1697 points (all frequencies) and 30 different sources.
        The first set of plots show the system performance with all frequencies overplotted. The sources are identified by symbol and the frequencies by color.
  • Fig 1 has the Gain in Kelvins/Jansky. This relies on the cals and the source flux. Ths  next plot is Tsys vs za in Kelvins followed by the SEFD  (System Equivalent Flux density) in Janskies / Tsys. At the bottom is the average beam width in arc seconds.
  • Fig 2 plots the coma parameter, first sidelobe height below the peak, the main beam efficiency, followed by the main beam + 1st sidelobe beam efficiency.
  • Fig 3 has the pointing errors in az,za, and total (added in quadrature) for the data.
    • The second set of plots has the data plotted separately for each frequency. The colors and symbols are used to differentiate the sources. The figures are:
  • Fig 1 4500 Mhz Gain,Tsys 355 points.
  • Fig 2 4500 Mhz sidelobes,beam efficiencies
  • Fig 3 4860 Mhz Gain,Tsys 353 points.
  • Fig 4 4860 Mhz sidelobes,beam efficiencies.
  • Fig 5 5000 Mhz Gain,Tsys 427 points.
  • Fig 6 5000 Mhz sidelobes,beam efficiencies
  • Fig 7 5400 Mhz Gain,Tsys 420 points.
  • Fig 8 5400 Mhz sidelobes,beam efficiencies
  • Fig 9 5600 Mhz Gain,Tsys 71 points.
  • Fig 10 5600 Mhz sidelobes,beam efficiencies
  • Fig 11 5900 Mhz Gain,Tsys 71 points.
  • Fig 12 5900 Mhz sidelobes,beam efficiencies
    •     Use the second set of plots to identify the sources.
    processing: x101/cb/aug02/doit.pro

    22dec01: sefd, pointing errors.  (top)

        A calibration run was done on 22dec01 using the heiles scans. The plots show the sefd and pointing errors. A more detailed analysis of this data is included in the sep01 to aug02 system performance. 
     home_~phil