View  Edit  Attributes  History  Attach  Print  Search

PPTAMan

The PPTA Observing Manual

TROUBLE SHOOTING

Startup Procedure

Receiver control:

  • Go upstairs and check that on Rack #8 the Remote Switch Control Panel shows the chosen receiver is "on". The 20cm MB is switch #8 and it should be in the down position (ignore the green LED). The 1050cm receiver is switch #1 an it should be in the up position with the green LED on.
  • On the PKMC program click "show" on the Multibeam and 10-50 receivers. Turn the required receiver's LNA's on (and the other's off). Check that the Cal controls are on active.
  • Check the CAL CONTROL UNIT (CCU) upstairs that the CCU is set to "remote".
  • Then from the PKMC GUI "show" the CCU display.

PDFB Setup:

Start the correlator control program(s) on pkccc3 (PDFB3), pkccc4 (PDFB4):

  • Open an xterm in a bourbon window (the top row of monitors)
  • ssh to pkccc3 (pkccc4) as corr
  • at the prompt, type;
      		corkill (to kill any processes still running in the background)
                pdfb3

Next start the SPD program:

  • Open another xterm in another bourbon window
  • ssh to pkccc3 as corr
  • at the prompt, type;
                spd 
  • then from within the spd program type;
        	mode b 
        	sel pp11 
  • other useful commands are:
        	a   to view the passband
        	chan to view a range of channels, eg. chan 10 1000

APSR Setup

CASPSR Setup

TCS Setup

Up to three separate instances of TCS can be run in parallel - a main TCS and two alt TCSs. Each must be run on a different processor, normally pisces (main), pavo and sagitta for the alt TCSs. We are now using the current version of TCS which automatically sets up the required receiver and backend configuration.

Main TCS:

  • log into pisces as pulsar
  • type
                atcsver

This should give "current". If not,

                atcsver current
  • then start TCS by typing;
         	tcs 
  • select the pulsar and expert mode options
  • select the relevant recall state
  • connect to the relevant PDFB (normally PDFB3). Make sure that Fold mode is selected.
  • APSR is run from PDFB3 sub-window - switch on
  • Switch on CASPSR
  • select the relevant schedule in the P456 directory (under Own), select View and enter the required start and end lines.

Alternate TCS:

To setup an alternate TCS for simultaneous observation with another, primary TCS.

In an xterm on pavo type (NOTE: It must be on a different machine to the primary TCS);

        	start_alt

then;

        	tcs alt

Then continue as normal, being sure to select a schedule for the alternate TCS, for example P456_MB_PDFB4A.sch (note the "A" for alternate). Note, it is important to have the "dummy" telescope systems started by "start_alt" running; otherwise file header parameters will be incorrect. Note that "start_alt" kills existing processes before restarting them.

Second Alternate TCS
To setup a second alternate TCS, open an xterm on a machine that is NOT running TCS (either primary or secondary), e.g. sagitta, and type;

       		start_alt 3

then;

                tcs alt3

Then proceed as normal for an alternate TCS as above.

Setup for normal P456 observations:

  • Run the primary TCS on pisces running PDFB3, APSR and CASPSR.
  • Alt TCS on pavo running PDFB4.

For 20cm (MB) obs:

  • On a pisces or pavo window:
               lorun ~/losetup/mb.cmd
  • Schedule files: P456_MB_DFB3.sch, P456_MB_DFB4A (now located in a separate P456 directory)

For 1050cm obs:

  • On a pisces or pavo window:
               lorun ~/losetup/3100+732.cmd
  • Schedule files: P456_50cm_DFB3.sch, P456_10cm_DFB4A.sch (now located in a separate P456 directory)

Hydra files are similarly named, as hydra_P456_XXcm_DFBX.sch.

  • Note: Before exiting TCS, switch off CASPSR.

Pulsar ephemerides

The pulsar folding is controlled by ephemerides which are accessed through psrcat -all -e. There is a link in /psr1/runtime/psrcat for obs99.db to /psr1/cvshome/pulsar/soft_atnf/psrcat/database/ppta.db. Apart from obscat.db, ppta.db is the last .db file accessed and so overrides all other .db files. The database file ppta.db is under cvs control from Epping.

If a new pulsar is added, as well as editing ppta.db at Epping and cvs updating the Parkes copy, the following must be done on a Sun machine (e.g. pisces):

                   cd ~/tcs/catalogues
                   makePSRCAT

This produces the file PSRCAT in the local directory; this file is read by TCS to give the pulsar position etc. The .eph and .pred files actually used for the folding are stored in ~/tcs. They are updated each time an observation is started.

You must also update the catalogue used by APSR:

                   ssh apsrcat

You must then enter a passphrase to complete the update.

Observation monitoring

It is important to monitor the observations as things can go wrong.

  • When starting a new observation, especially after a receiver change, use spd to check the receiver bandpass (a) and cal signal (sel pp11). Plots similar to those below should be seen.




10cm bandpass

10cm cal signal

20cm bandpass

20cm cal signal

50cm bandpass

50cm cal signal
  • On the correlator control GUIs click on the "samplers" button and ensure that the RMS values are around 10.0 (between 9.0 and 12.0 is OK).
  • Look for blue warning messages on the PDFB control GUIs. If there are more than than one or two isolated warnings, stop the observation, reconfigure and try again. If this fails, follow the trouble-shooting guide below.
  • Look for red messages on TCS - something is seriously wrong!
  • Ensure that the System Status on TCS is "RUNNING SCHED FILE"
  • Monitor operation of the PDFBs and APSR using http://pulseatparkes.atnf.csiro.au/dev/.
  • Check that APSR starts Preparing after startup and then Recording.
  • Especially for low-DM pulsars, check that the S/N after Tobs/8 is roughly as usual (see typical profiles after Tobs/8). If the pulsar is in a scintillation null and the S/N is significantly less than usual (and not growing), kill the observation and come back later (after at least a couple of hours).

PKMC:

The programme, PKMC is used to switch the receiver LNA's on and off.

PKMC should be running on bourbon, if not then start it up by;

  • Opening an xterm in any bourbon window and type;
	                pkmc
  • Then click on the SHOW button for the appropriate receiver (See Fig. 1a).

Figure 1a Click on image to enlarge


When the receiver monitor and control GUI pops up (See Fig. 1b), turn on the LNA's by clicking on the four LNA buttons.

Figure 1b Click on image to enlarge

NOTE: Before observing with the 1050cm receiver, be sure the Switches settings indicate the following:

	10cm Noise Cal	Active
	10cm LNA A	On
	10cm LNA B 	On
	Vac Gauge	Off
	50cm Noise Cal	Active
	50cm LNA A	On
	50cm LNA B 	On

Monitor Control for 10cm Observing ONLY

  • When observing with the 1050cm receiver, be sure to turn the monitoring off, by clicking on the "monitor" button in the popup GUI. Turn it back on when not observing with it. NOTE: Red indicates that it is off and green indicates that it is on.

CAL Control

  • From the PKMC GUI, click on the "show" button in the Cal Control Unit row. A popup window will appear that will allow you to control the CAL provided the CCU upstairs is set to "remote" first.

50cm RFI Mitigation

PDFB3 has the facility to mitigate narrow-band RFI from a specific source. This is mainly used at 50cm - with the new 50cm band (700 - 764 MHz) there are strong narrow-band RFI signals from the Peak Hill analogue TV repeater. (This is expected to be decommissioned in 2011.) A reference antenna pointing at Peak Hill has been set up on the tower near the 18m antenna. Signals from this come into the control room on a cable connected to RFI REF (socket 8) on the FEEDTHROUGH panel near the middle of Rack 2. Required connections are as follows:

RFI REF -> Mixer hanging between racks 3 & 4 through a 3db pad to R

LO from HP-2 synthesizer (Rk 8) -> IF Distribution 5 (Rk 8 -> Rk 2) -> Mixer L

Mixer I -> 6db pad -> BPF 96/T52 -> +10db+10db+10db+10db amplifiers (Rk 3) -> Variable atten (Rk 3) -> -11db coupler -> splitter => Spare Input Panel FC3/SP2 and FC4/SP2.

To set the HP-2 synthesizer (to 828 MHz, +14dbm), run losetup/50cm_rfi.cmd. NB: This must be done AFTER running 3100+732.cmd.

The spectrum analyser can be connected to the coupler, set to centre freq 96 MHz, BW 100 MHz, Resolution bw 300 kHz, Video resolution 3kHz. The central part of the bandpass should be around -56dbm and prominent RFI should be seen at 69.6 (758.4), 85.2 (742.8), 90.8 (737.2), 106.2 (721.8) and 111.8 (716.2) MHz.

On a Solaris machine (pisces, pavo, etc):

                  cd smsetup
                  smrun 50cm_rfi.cmd

To activate the RFI mitigation system on PDFB3, switch RFI on. Normally Timescale = 0.001 and Loop gain = 0.1. These can be set from the schedule line.

Trouble Shooting

Add all trouble shooting tips here.

Trouble Shooting with the PDFB's

  • If pdfbn fails to properly start the PDFB gui (blank lower windows), kill the GUI (^c, then exit glish) and type:
               corkill            then restart pdfbn.
  • If the PDFB exhibits problems during observing, then try clicking on the "config" button first.
  • If the configuration fails, then in the popup config-window that appears, type ENTER then "prog" to configure it again.
  • If the PDFB complains of a XFER Error, then try the above two procedures. If this still fails, then:
    • change to a different configuration, reconfigure, then go back to the troublesome one and try again.
    • quit from the GUI and reboot the problem machine (as indicated in the error message). On an xterm logged into pkccc3 or pkccc4 as corr, log into the problem machine. For example, if the problem is on l-bcc06 type:
                ssh root@l-bcc06	then the password
                reboot
  • after about 2 min hit ENTER to check that the reboot is complete (it will return you to the pkccc prompt)
  • type:
                bcckill
                corkill
  • It this fails, then power cycle the relevant PDFB. The power switch is at the rear of the unit. Turn it off, wait for 10 sec and then turn it on again. Then reboot both l-bcc06 and l-bcc11 and do bcckill and corkill as above.
  • As a last resort reboot or power cycle the pkccc machine.

Trouble Shooting with APSR

  • If APSR doesn't work then stop the observation, disconnect APSR from TCS, type "disable apsr" <enter> in the command box on tcs, hit "Restart everything" on the APSR webpage, wait until it's done, type "enable apsr" in tcs, reconnect APSR and then restart the observation