Summary
- Symptom
- When the cross disperser is moved, the XDANGL keyword
starts changing and never stops; it cycles continuously
within a certain range of values.
- Problem
- The cross disperser motor is confused.
- Solution
-
- From the window manager's pulldown menu, select the
HIRES Control Menu option, then release on the
item labeled Stop Cross-disperser Oscillation.
- Alternatively, type the following command on
makua:
stop_cd_drift
- If this does not halt the cross-disperser, contact your
Instrument Specialist.
- Symptom
- XHIRES reports a certain filter to be in, but exposures
look as if the wrong filter is in.
- Problem
- The filter wheel is lying and needs to be re-homed.
- Solution
- Re-home the stage by typing the following commands in a
manuka window:
m lfilcal=reset
m lfilcal=homed
- Symptom
- Observer clicked the button to turn on the ThAr lamps,
but exposure shows no lines.
- Problem
- Either:
- the lamps did not turn on, or
- the mirror did not enter the beam
- Solution
-
- Try turning the
ThAr lamps off and then on again. Check whether the
reported current changes from 0.0mA to about 25mA,
indicating power to the lights.
- Check whether XHIRES shows the mirror to be in the
beam. If not, insert it using XHIRES and try another
exposure.
- Symptom
- ThAr lamp image shows what looks like second-order light
- Problem
- This may not be a problem if the bands of light are
horizontal and only seen redward of about
7500 Å. In this case, you are seeing reflected
light resulting from the intense ThAr emission lines.
- Symptom
- Rotator will not initialise, motion seems wrong, or you are having trouble moving the unit into or out of the beam.
- Problem
- You are trying to initialise with the DCS gui. There have been
reports this is unreliable.
- Solution
- Use the rotator control panel obtained via XHIRES to initialize.
- Problem
- One or more of the watch_hirot daemons has died. Find out if
they are all present by typing ct at a lehoula prompt.
- Solution
- Login to lehoula as hireseng and manually kill any processes
that contain the string watch_hirot. Issue the command hires start watch_hirot
- Problem
- The telescope has been re-initialised or the tertiary has
been turned since the rotator was last initialised.
- Solution
- Re-initialise the rotator via the control panel from XHIRES.
- Symptom
- When you show one or more of the utility board voltage keywords you get a value much
different than you should.
- Problem
- The keyword which contains the mapping values for a given voltage may need to be re-set.
This could be a result of a problem with infoman or the file /kroot/data/music/info/data
- Solution
- Login to kanaha as esieng and check what the correct values are in the ESI
keyword library. If they differ from those in the HIRES library this is likely the problem.
Assuming this is the problem, here is how to fix each of the voltage keywords:
modify -s hires utbp30v = 0,35.71,2048,0,36
modify -s hires utbp15v = 0,89.60,2048,0,0
modify -s hires utbm15v = 0,89.60,2048,0,0
modify -s hires utbp5v = 0,278.42,2048,0,0
- Symptom
- Cross-disperser angle (displayed on the
xshow widget as XDANGL) continually
changes without ever stopping.
- Problem
- The motor controller for the cross-disperser is in a bad
mode and needs to be restarted.
- Solution
-
- From the background menu, select HIRES Control
Menu > Stop Cross-disperser
Oscillation
- Wait for the script to complete.
- Verify that oscillation has stopped.
- Set the cross-disperser to your desired angle.
- Note whether the move completes of whether the
oscillation returns. If so, repeat above procedure.
- If the oscillation continues after severel attempts
to run the script, contact your Support Astronomer.
- Symptom
- Attempts to open the cross-disperser cover or to change
the cross-disperser angle fail.
- Problem
- The cross dispersion motion control is set to manual.
- Solution
-
- Ask summit staff to verify that the cross disperser
is back in computer control mode.
- If trouble persists, have summit staff check cabling
to the cross disperser.
- Symptom
- Attempts to open the cross-disperser cover fail.
- Problem
- The cross-disperser cover is locked.
- Solution
- Ask summit staff to verify that the cross disperser cover
is unlocked.
- Symptom
- After an image is taken, it will not read out. Xpose
claims it "writing crate memory."
- Problem
- FIGDISP is being run on an 8-bit color display. It must
be run on a 24-bit display.
- Solution
- Quit FIGDISP and re-start it on a 24-bit display window.
- Symptom
- tklogger warns that certain FITS keywords are not being written
and/or it is noticed that the FITS files are smaller than they should be
- Problem(s)
- There may be problems with DCS.
- or
- Some low level processes may need to be restarted.
- Solution(s)
-
- Check if DCS communications are okay by trying to show
a DCS keyword.
- For example, in a nuu window type:
-
show -s dcs airmass
- If you get an error message, ask your OA if everything is fine
with DCS. If this is not the problem then you may need to restart some low
level processes such as watch_ccd and/or traffic. Look at your Hexpocon
gui or type ct at a nuu prompt. It will tell you which processes are running.
Rebooting nuu will restart these processes, but
this is a bigger hammer than needs to be used. Instead:
- Shutdown HIRES control from the openwin menu
- Login in as hirestst
- hexpo stop daemons
- Manually kill any processes that contain the string "hirot"
- hexpo start daemons
- hexpo start watch_hirot
- Startup HIRES and take a test image
- Note
- If you get the message "no match" after step 6, it may mean you did
not find and manually kill the hirot processes as instructed
in step 4 above.
- Symptom
- The computer is very slow to respond to input.
- Problem
- Obsolete processes are running on the computer, consuming
most of the CPU.
- Solution
-
- Go to a local window on the computer.
If you are sitting at hanauma, then open a
hanauma window (for example, use the
Login Windows submenu on the OpenWindows
pulldown menu to select the xgterm option).
- Find out whether another observer's processes are
running by typing the following command:
ps -aux | grep hires | more
This will list all processes owned by hires accounts.
If, for example, you are using the hires2
account and the output of the above command shows that
jobs are running under the hires1 account,
then you should stop these other processes as described
in the next step.
- Log in as user hires and issue the command
kill_em hiresN
to search for and destroy processes owned by the
hiresN account, (substitute the
appropriate number for N).
- Symptom
- Besides the lights not lit, other symptoms are to be determined.
- Problem
- The processes in question have died.
- Solution
-
- Reboot the crates.
|