View  Edit  Attributes  History  Attach  Print  Search

ObsProblems

Trouble-shooting

The Parkes general trouble-shooting guide can be found at:

http://www.parkes.atnf.csiro.au/observing/documentation/user_guide/pks_ug_7.html#TroubleShooting

Please jump to the following sections:

CASPSR stops recording data
CASPSR is offline
DFB problems
PSRMON problems
CASPSR digitiser saturated

CASPSR stops recording data

To fix CASPSR in event of a failure, do the following:

  • Stop the schedule on TCS.
  • Disconnect CASPSR on TCS Primary.
  • Bring up the CASPSR web page http://caspsr-srv0.atnf.csiro.au/caspsr
  • Click on the 'Controls' button.
  • After the new window pops up, hit 'Hard Stop'
  • Wait until it says it is safe to close the window.
  • Once all the lights are red, it should be stopped. Restart the instrument by hitting 'Start'.
  • Wait until everything goes green.
  • Reconnect TCS Primary to CASPSR.
  • Restart observing.

Please see the caspsr user guide for detailed information

CASPSR is offline

You may experience an 'timeout' error in TCS. This may occur for example, if the previous observing session was using BPSR, and CASPSR was stopped.

  • Stop the observation
  • Uncheck CASPSR on TCS.
  • Ensure BPSR is stopped. Open the BPSR Monitor, then 'Controls' and 'Stop'.
  • On the CASPSR Monitor, click 'Controls' and 'Start'
  • Wait until everything goes green.
  • Connect CASPSR on TCS. In the TCS Command entry box (bottom-right corner of TCS gui), type:

disable caspsr (and press return)
enable caspsr (and press return)

  • Restart observing.

DFB problems

[Taken from the troubleshooting section of the Correlator Guide]

If you have issues getting DFB4 going, or getting multiple XFER timeout errors, follow the procedures below:

1. Try reconfiguring the DFB (click CONFIG button on DFB GUI); Try this at least 2 times before moving to #2.

2. If #1 is not sufficient, quit and restart DFB GUI (pdfb4/sdfb4 according to the observing mode) Try this at least 2 times before moving to #3;

3. If #2 is not sufficient, use the bcckill and corkill procedures:

  • quit DFB GUI
  • type bcckill
  • type corkill
  • restart DFB GUI

Try this at least 2 times before moving to #4

4. If #3 is not sufficient, reboot the two l-bcc06 (DFB3 only) and l-bcc11 boards:

  • quit DFB GUI
  • from two pkccc3/pkccc4 terminals (user corr), login us root
    ssh root@l-bcc06 (passwd required; NOTE: DFB3 only)
    ssh root@l-bcc11 (passwd required)
  • on each of them, type "reboot"
  • wait 2 mins
  • type bcckill
  • type corkill
  • start DFB GUI

Try this 2 times before moving to #5

5. (No longer an option for remote observing!)
If #4 has been unsuccessful, power cycle the DFB (very last resort!):

  • quit DFB GUI
  • power-cycle DFB
  • wait 2 minutes
  • reboot the l-bcc boards; see #4
  • type bcckill
  • type corkill
  • start DFB GUI

PSRMON problems

PSRMON suddenly stops working.

  • ensure that you're on a computer on the CSIRO network
  • go to psrmon webpage
  • click on "Controls" on the top left
  • a drop-down box will appear. Click on "All: restart"
  • wait a few seconds

CASPSR digitiser saturated

During a good observation, CASPSR's digitiser histogram should look something like this, preferably with the two pols - red and green lines - overlapping, and the counts clustered around state 0, within the range +/- 20 being ideal:

In certain circumstances such as continuous or large bursts of RFI, you may notice that the CASPSR histogram however looking something like this:

<picture of saturated digitiser>

This is an indication that that the digitiser has become saturated, clipping occurs and the system is driven non-linear as a result, and therefore detrimental to observation quality.

Under these conditions it may be necessary to adjust the attenuator levels in loGui.

Please follow the steps below - if you are unsure about what to do, make a note in the Dashboard log and contact Matthew K. or Ryan:

  • Ensure that the current observation has been stopped
  • Locate the loGui in the VNC window
  • Click 'refresh' to update any changes made on other instances (marked with the magenta oval on the image below)
  • Look for the C30 attenuators on the right-hand (FC3/4) side, marked as red ovals in the image below. For both H-OH and 40cm observations, the 'lorun' script sets the C30 attenuators to 13 dB.
  • Adjust the C30 attenuators up or down accordingly
  • Click 'refresh' again
  • Add a new entry to the Dashboard log describing the changes you have made
  • Start the next observation and check the cal and bandpass