| CSIRO Astronomy and Space Science Parkes Users Guide |
 |
|
|
This section describes some commonly encountered problems at the Parkes RadioTelescope and provides receipes for overcoming them. It is by no means an exhaustive list of all problems, but will hopefully provide a starting place. This section is constantly changing, check back regularly for updates. Also search the on-line fault reports and responses for problems and solutions. In the case of serious problems (those which endanger you or the telescope equipment), notify the Call–out person immediately.
If you strike problems using the DFBs, it is recommended you follow the checklist below to try and rectify the problem. Some of the more common error messages reported by TCS and/or the DFBs when configured are listed below:
The checklist is contained in fault report 6515 and can be viewed here (internal access only). A laminiated sheet should be located above PISCES in the lower control room.
Error messages associated with the Multibeam, Wide-band correlators may include:
It is usually best to stop observing and try recofiguring the correlator using the CONFIG button on the PKCOR GUI. If you need to restart the block control computer(s), they are ’standard’ diskless PCs with the familiar white reset button. Occasionally the EVG (Event Generator) cards in the BCCs get into an error state with a red LED showing, indicating loss of clock frame. It is usually necessary to power-cycle the BCCs, not just reboot them, to clear this red LED indication. (This can also occur with the EVG card in PKDESK). Any reset of a BCC requires a manual reprogramming of the associated correlator, by clicking on the CONFIG button of the correlator GUI.
The location of each block control computer is summarized below:
Occasionally the AT Clock display will crash and display only zeros. The system simply needs to be reset which can be done by pressing and holding the system reset button for 2 seconds to reset the display.
Rarely the Azimuth and/or Zenith drives will disable for no apparent reason. If this happens, first assess the state of the telescope then;
You may need to restart PKDESK in situations such as when the machine
you’re running the observing program (i.e. TCS) crashes and leaves an
open socket connection to PKDESK (The "Primary host" line on the
PKDESK screen shows the IP address of the host making the connection to
PKDESK). Currently the only way to clear this situation is the restart
PKDESK by typing quit.
The PKDESK computer is located in the second-level control room
in Rack 1. It has a black pull-out keyboard with monitor as shown in
Figure 7.1. You may have to check the display is set to
display #2 (DESK) and press the spacebar to activate the screen.
On the PKDESK console, simply type
quit and hit <Enter>. PKDESK will shutdown and restart in
10-15 seconds.
There are several clues that PKDESK has crashed. SHOWTEL may display
the error "PKDESK not responding." Go to the Computer
Room adjacent to the first-level control room and on the YOWIE console,
try typing NETCOM on the
YOWIE terminal. The "PKDESK not responding" message should no longer
be seen. However, if you see messages on SHOWTEL (or KPDESK
itself like:
You may need to restart PKDESK. Type quit then <Enter>
and wait 10 to 15 seconds while PKDESK restarts automatically.
SHOWTEL will complain "PKDESK NOT RESPONDING" while this happens,
but should recover quickly once PKDESK starts again.
If you need to reload the ME and/or SERVO code again, follow the
instructions on the laminated sheet (usually) located near the PKDESK
keyboard/monitor (or see ME or SERVO stops/crashes).
Figure 7.1: PKDESK location and display in Rack 1.
Recovering observing applications: TCS and PMCTRL will lose their connections to PKDESK (the antenna) if PKDESK is restarted. For TCS;
For PMCTRL, you might have to quit completely and restart.
Most likely the mouse is just on one of the other screens (there are three). Try moving the mouse in one direction until it appears in the furthest screen.
OPERFCC may report an error when selecting a new receiver, other than the multibeam. The error, which reads: “y2 Axis Following Error”, occurs when a new receiver is selected with the telescope is at or near stow. To fix, type the following in the “Command:” window on the OPERFCC GUI (may need to do twice):
take
stop
If the video camera for the focus cabin is on you should hear a click. Now re-select the receiver.
This may happen whilst observing using the Multibeam pulsar filterbank. The PDP11/73 Timing Computer (located in the upstairs control room, rack 1 supplies the 1MHz sampling clock and the 0.2pps (5-second) pulse for the Multibeam pulsar filterbank. An alternate source of the 5-second pulse is also provided from the PKDESK computer.
The Timing Computer derives its timing via a clock bus connection to the station clock. When it is working correctly you will see a green LED flashing at 11Hz on the events distribution panel immediately above the computer. (You should also be able to see the 5-second pulse on output "EVD8" if you look carefully).
If this isn’t running the Multibeam pulsar filterbank may not work correctly and the Timing Computer needs to be started.
The console for this computer is displayed on the flat-screen display near the MCP which also shows terminals for SERVO, ME, WINDS and SHOWTEL. The Timing Computer window has a grey background. To restart the program;
ctrl c at least twice in succession (three times
is also good!) and wait a few seconds. You should get the "."
(period) prompt after a few seconds. If so skip the next step.
WPP <ENTER>
3 <ENTER>
|
to select option "3" (5-sec pulse + 11Hz pulse). You should see the 11Hz flashing green LED start again after a few seconds.
Common problems encountered with the Pulsar data aquistion system are described in the pulsar observing documentation.
For convenience we have reproduced the text here.
Occasionally, PUMA crashes for no obvious reason, bringing PMCTRL down with it. If this occurs in mid-observation (as it usually does), there are two possible ways to recover.
You may one day come across a bus error, or otherwise find that PUMA does not respond (e.g., you can’t log on.) The CPU may have halted. The most straightforward way of recovering is to press the reset button on the front of PUMA (that’s the button on the upper left of the case, not upper right, which is the power button.) Note that for upwards of a minute it will appear as if nothing is happening on PUMA’s console (a vt320 across from the computer.) Don’t despair, eventually all should be well.
If PUMA has been rebooted, and maybe on other occasions, there can be a DMA timeout in the digitiser interface to PUMA. This is indicated by a sequence counter error followed by a PUMA INTERNAL ABORT and a general crashing sound. A fix which usually works is to power cycle the digitiser. The power switch is a black switch at the back of the rack, mounted on the left wall of the rack adjacent to the digitiser. Another possible cause for this error is a broken connection between the digitiser and PUMA - check the data cable and plugs.
After a mains failure, there are sometimes problems with the ME (re)locking to the dish. Also, sometimes there is a rather large discrepancy (>150 arcsec) between the ME and DISH and and SHOWTEL displays the message ’ME BAD’. This last situation might be a fault with the Observatories 1PPS (1 Pulse Per Second) signal which is required by (amongst other things) the ME computer (this is particularly bad and will require technical staff to fix).
For SERVO, problems will show themselves if SHOWTEL displays messages like “SERVO problem?”, or “SERVO dead”. If the above fits your symptoms, then do the following:
load me and hit <ENTER>
OR load servo and hit <ENTER> - whichever process
(ME/SERVO) SHOWTEL is complaining about.
go on the PKDESK computer to load the program code.
ctrl p at the ME (or SERVO) console to restart the program.
If after restarting, the ME/SERVO screen is not updating correctly (a crash state is indicated by an ’’ prompt) then you should:
173000G (ME) or 1000G (SERVO) at the ’’ prompt (uppercase ’G’ is required.)
ctrl p.
ctrl p w.
If any of these proceedures fail to restart the ME or SERVO, contact the technical staff or the Call–out person.
If after depressing the green reset button the safety timer fails to reset try pressing an a different green reset button (upstairs control room, etc.). If this fails turn off the infra-red detector. The detector is located on the ceiling at the center column in the downstairs control room. It is isolated with a toggle switch located on the underside of the enclosure to which it is attached. Try to reset the timer.
If the timer still won’t reset try unplugging the two reset buttons on the control desk. The outlets for these buttons are under the desk. Try to reset the timer.
If all this fails call the Call–out person and Matt McFarland or Andrew Hunt.
First try typing a to display the amplitudes. If this doesn’t
work try exiting out of SPD and restarting. After the correlator has
been reconfigured the SPD display may have trouble redisplaying the
data. A restart can fix this. If that doesn’t solve the problem,
then it is symptomatic of a more serious problem. Try stopping the
observation and reprograming the correlator.
This document was generated by Stacy Mader on October 13, 2011 using texi2html 1.82.