---

Zpectrometer observing notes

Basic pattern for subreflector-nodded observing

1. Set up for observing
2. Make an instrumental phase calibration if one is needed
3. Select sources, slew to source, and point and focus near the target source.  If possible, make the first target a calibrator and sky pair at a typical elevation.
4. Check correlator power levels
5. Obsererve a target and nearby sky position
6. With a maximum time span of one hour, do another point and focus
7. Loop to step 5
8. When finished, go through the simple shutdown procedure.

Set up for observing

• Ask the telescope operator to configure for Ka-band observations.  This means:
  
• Ka-band receiver at the telescope focus.
• Active surface switched on.
• Your login will be "in the gateway" for telescope and system control.
  
• Open telnet connections to the Zpectrometer's correlator and power supply (reference details)
  
• First open 5 terminal screens in a window and, following some sensible geometrical order, open telnet sessions in each screen with
• telnet zpectrometer0.gbt.nrao.edu           (this one provides switching signals to the total power detector)
  
• telnet zpectrometer1.gbt.nrao.edu           (this one provides the master timing signals)
  
• telnet zpectrometer2.gbt.nrao.edu
  
• telnet zpectrometer3.gbt.nrao.edu
  
• telnet zpectrometer4.gbt.nrao.edu           (this one is the power supply, also other monitor points)
• A  zpec>  prompt should appear for each connection.  A brief command reminder will appear if you hit a simple <return>.
  
• In another window, start CLEO by typing cleo & at the command prompt.  Use the tabs to launch the following CLEO screens:
• Launch -> Status       (brings up the status window to monitor observations)
  
• In another window, start Astrid by typing astrid & at the command prompt. 
  
• Choose to be online with control of telescope
• In the Run tab, choose
• Project: AGBT06C_033
• Session number: keep the default
  
• Observer: Andrew Harris
• Operator: as appropriate
• In the Edit tab, choose
• Project: AGBT06C_033

Select the source pairs for fast nodding

• In Astrid select the Observation Management 1 -> Edit tabs and open zpec07-obs-fastnod-duo
• Add sources to the source lists as needed, in either J2000 or B1950 coordinates.  The sources are in pairs; if the secondary source is sky, preceed its name with SKY (e.g. W51, SKYW51), and give it a position about 5 arcmin away (a Dec offset is easiest).  Following the pattern in the file, comment out all other sources and add the source pair to the list.
• Adjust integration time as needed, following the notes in the file.
• Save the file.

Slew to source and point nearby

• Prepare or check the pointing script
  
• In the Astrid editor, open  zpec07-point
• Add the target source to the source list if needed, in either J2000 or B1950 coordinate sections.
  
• If the source is not known to be strong, Astrid will choose an appropriate source from its database near the target if you use the AutoPeakFocus() command in the script.  To force it to use a specific source, for example a planet, comment out AutoPeakFocus() and uncomment the Focus(...) command.
  
• Save the file.
  
• Run the file to slew to the source and point, either nearby or on the science source.  Commands at the beginning of the file disconnect the Zpectrometer from the IF and connect the DCR in its place.  Mode (under Tools -> Options -> Data Processing) should be Raw, X/Left.
  
• During auto-point
• Verify that the pointing is running well and the offsets are reasonable by looking at the data and fits by selecting Astrid's Data Display 1 -> Pointing tabs. 
  
• Write down the pointing offsets for futrue reference; it may be useful to know that we have drifted by half a beam since the last pointing
  
• During auto-focus
• Verify that the focus offsets are reasonable by looking at the data and fits by selecting Astrid's Data Display 1 -> Focus tabs. 
  
• It may be necessary to select the "Total Power" mode to get good fits; Tools -> Options -> Data Processing, then pick the Total Power button.  If this is the case, the telescope operator may have to enter the fit results manually.
• Write down the gain change and focus offset for future reference if only to reassure ourselves later that the gain hasn't changed significantly.

Observe the source

• Configure for Zpectrometer observations by highlighting
• Highlight zpec07-config-totpwr  in Astrid's Observation Management 1 ->Run tab
• Hit the "submit" radio button on the screen once
• If the weather or elevation have changed substantially since the last observations, check the correlator power levels.
• Update the step attenuator values in the zpec07-config-totpwr script as appropriate: atten0, atten1, atten2, atten3  for zpectrometer0, 1, 2, 3
  
• highlight zpec07-config-totpwr  in Astrid's Observation Management 1 ->Run tab
• Hit the "submit" radio button on the screen once
• Verify that the source pair is correct in the current  zpec07-config-fastnod-duo script.
• Start the integration:
  
• Highlight the zpec07-config-fastnod-duo script in the Run window
  
• hit the "submit" button one or more times.
• Note: Point and focus at least once per hour!  Pointing moves the nodding secondary to ensure that its actuators are properly lubricated.

Shutdown

• Highlight and run the zpec07-shutdown script in Astrid
• The telescope operator can now remove you from the gateway
  
• Go to zpectrometer0...3 and type l 60 at the prompts
• Go to zpectrometer0...4 and type quit at each prompt to break the telnet connections
  
• Close all screens and log out
  

Check the correlator power levels

• Configure for Zpectrometer observations by choosing
• zpec07-config-totpwr  in Astrid's Run tab
• hit the "submit" radio button on the screen once
• Set the step attenuator for zpectrometer1, which also generates the master clocking signals.  Connect with telnet zpectrometer1.gbt.nrao.edu.  At the command prompt, type (a full list of the status word bits in the microcontroller is here)
  
• b 1                      (sets this correlator as master)
• l 3                      (set the step attenuator to its default value)  
  
• poke 0.4 1    (sets quadrature phase switching)
• poke 0.3 0    (turns of the noise diode switching, just in case it was left on)
• m 100 128      (get means and variance for all lags, 100 iterations)
• Look at the third column of data; most values here should be from about 20-40. Ignore the values in the fifth column.
  
• If the third column values are not in range, you will need to adjust the internal step attenuator away from its nominal good-weather value of 3 dB.  To do this, change the attenuation to a vaule you would like to try with the l (lower case ell) command; that is, to set the attenuator to 5 dB, type l 5
• Repeat the   m 100 128   command and iteratively adjust the step attenuator to get the third column values to about 20-40.
• Note the value for later use -- you'll need to edit the zpec07-config-totpwr file with this value.
  
• Finish by typing quit at the command prompt.
  
• Set the step attenuation values for zpectrometer0, 2, and 3 by telnetting in to each in turn.  At the command prompt, type
  
• sync                   (synchronizes the correlator to the master)
• l 3    (or 4)         (set the step attenuator to its default value of 3 dB for zpectrometer0 and 4 dB for zpectrometer2 and zpectrometer3)  
  
• m 100 128      (get means and variance for all lags, 100 iterations)
• Look at the third column of data; most values here should be from about 20-40. Ignore the values in the fifth column.
  
• If the third column values are not in range, you will need to adjust the internal step attenuators away from their nominal good-weather values of 3,  4, and 4 dB for the correlators.  To do this, change the attenuation to a vaule you would like to try with the l (lower case ell) command; that is, to set the attenuator to 5 dB, type l 5
• Repeat the   m 100 128   command and iteratively adjust the step attenuator to get the third column values to about 20-40.
• Note the value for later use -- you'll need to edit the zpec07-config-totpwr file with these values. 
  
• Finish by typing quit at the command prompt.
  
• Run for each correlator in turn.  The values should all track approximately: if one goes up by 1 dB, the rest should too, more or less.
• Note the telescope elevation; also the TotPwr value with a q (query) command at the power supply micro prompt (zpectrometer4)

Make a phase calibration (cwsigs)

• Configure for Zpectrometer observations:
  
• highlight zpec07-config-totpwr  in Astrid's Run tab
• hit the "submit" radio button on the screen once
• Change the Astrid session to the next number (unless you are starting with a phase cal) -- the idea is to get the phase cal data in its own session.  In subsequent notes, this session number is denoted SS.
  
• Configure for phase cal:
  
• highlight zpec07-config-cwsigs  in Astrid's Run tab
• hit the "submit" radio button on the screen once
• Set up cal tone LO
• At the CLEO launcher,  Launch -> I.F./L.O. -> L.O. 1
• On the L.O. 1 screen, Managers -> LO1B -> Standby      (LO1A should be on)
• Trigger the manager to start the cal
  
• At the CLEO launcher, Launch -> Utilities/Tools -> Device Explorer
• Scroll down to the Zpectrometer entry in the Manager column (first column), and double-click Zpectrometer manager
• Then click Managers -> Start
  
• Look at the LO1 monitor screen.  LOA should be at 14616666667, and LOB should be incrementing fairly rapidly.  It will pause for 5-10 seconds at the beginning and occasionally therafter for a zero measurement. In start:stop:increment notation, frequencies in MHz, the default calibration sequence is 35670:40000:8, 32170:36500:8, 28670:33000:8, 25170:29500:8.
  
• When the cal is done (LOB gets to 1474900..0)
  
• Open a terminal window and go to the data directory /home/gbtdata/AGBT06C_033_SS/Zpectrometer  to see that the cwsigs has finished properly (SS is the session number).   There should be two files with date-stamp names, one datestamp.fits, the other datestamp.parm. 
  
• An ls -l should show that the fits file has size 7744320 bytes.  If the file is smaller, the cwsigs did not complete properly -- a backend may not have responded, and was disabled, for instance.   Check the backend status with waspset verb=true | grep beenab.
• Open the zpec07-config-totpwr  in the Astrid editor.  Near the top, change the names of the cwsigsproj and cwsigsfile variables to the directory and file names of the phase cal fits file.  Save the modified file and run it to check for errors.  If the the file as named is missing, a cwsigsLoc error will be generated.
• In Astrid's Run tab, increment the session number by one to store subsequent astronomical data in a new session.
  
• Go to the L.O. 1 screen and select Managers -> LO1B -> On
• Note the telescope elevation; also the TotPwr value with a q (query) command at the power supply micro prompt (zpectrometer4)

Use total power and receiver gain monitors

• Total power monitor
• Query zpectrometer4 with command q , then record the TotPwr offset value (typically -0.209 V) when the receiver is not connected to the Zpectrometer electronics
• Query zpectrometer4 with command q , then record the value when the receiver is connected to the Zpectrometer electronics.  A typical high-elevation value on a good sky is  0.9 V (or 1.1 V counting the offset).
• Receiver gain monitor
• Set up the zpectrometer1 micro by entering the commands
• b 1                       (set as master)
  
• poke 0.4 1     (set quadrature phase switching)
  
• poke 0.3 1    (noise diode modulation on)
• Set up the synchronous detector's phase reference with the zpectrometer0 micro
• sync        (synchronize internal clocks to master)
  
• poke 0.3 1     (generate noise diode demodulation signal)
  
• Query zpectrometer4 with command q , then record the SwPwr signal.  This should be about -1.06 V.
  
• Turn off noise diode modulation: at the zpectrometer1 micro prompt, enter poke 0.3 0    (noise diode modulation off).  Turning off the noise diode is an important step, as it will run during observations otherwise!  The other settings are not critical.
  
• Query zpectrometer4 with command q , then record the SwPwr offset signal.  This should be very close to zero.

Zpectrometer backend microcontrollers