JKT Duty Engineer Checks
For the JKT A&G box (JAG) with SITe2 and UltraDAS
First check the Operations Logbook for any restrictions
that may be still in force. Turn on the DOME LIGHTS from the console
and raise the control room blind.
Make a visual inspection of the observing floor to check that the telescope
is free to move. e.g. No ladders or steps are close to the
telescope or test instruments connected.
Check to see that the light tight BLACK BAG is fitted around the
bottom of the telescope. i.e. Enclosing the JAG and CCD
cryostat. If this is not correctly fitted or missing, it can affect
the balance of the telescope which may cause tracking errors.
Move the RISING FLOOR down to a safe level that will not endanger the
telescope when slewing.
- Push the LAMP TEST button on the console. Check that all
used indicators are working (e.g. the ALARMS).
Replace any defective lamps with new replacements (28v). There should
be a supply of these in the cupboard in the control room.
- Rotate the dome manually a small distance in both directions using
the DOME +/- buttons. If the weather permits, OPEN and CLOSE the dome
SHUTTER a small amount, but see important note below.
n.b. If these do NOT work from the console, check outside on
the observing floor. The switches on the dome supply wall box must be
pointing to REMOTE. They may have been switched to LOCAL.
The tests for opening the shutters only applies if they are working
correctly. If any mechanical problems exist then do NOT
do these checks. Provisional instructions on how to open/close the dome will
have been issued if this is the case.
- Switch ON the TELESCOPE POWER using the keyswitch on the top right hand
panel of the console.
Push the MIRROR OPEN button. Check the mirror petals open fully then push
the MIRROR CLOSE button.
LOGGING OUT FROM THE JKT COMPUTERS
The JKT observing system comprises of three computers. These currently being:
- jktics/jktdr (combined Data Reduction and Instrument Control System) Sparc
- jktdas1 (Data Acquistion System) Sparc
- jkttcs (Telescope Control System) Dec Alpha
It is best to bring the observing system down and re-start it. This not only generates
a new Observing Log for the day, but prevents problems due to possible memory leaks
which can build up with time. To shut down, proceed as follows:
Logging out from the Observing system (ICS) and uDAS Sparcs
- On the jktics/jktdr computer, enter in the pink (CI) window:
Do you want to wipe out the system Y/N? y
- In the orange (DAS) window, enter:
Do you want to wipe out the system Y/N? y
If there are any other windows seen, shut them down too. Here is what to type
for various prompts:
To shut down IRAF:
To exit the observing system:
jktobs@..... > exit
Graphics windows without a prompt can be exited as follows:
Move the mouse cursor over the "-" sign at the top left corner.
Click on it with the left mouse button.
Choose "close" or "exit".
Click the right mouse button over the background and choose the "exit"
option at the bottom of the menu. This will log you right out of windows.
Logging out from the TCS
The TCS, a DEC ALPHA computer on the ground floor, is run from the X terminal
in the console. In the TCS User window, enter:
Exit any other windows and click on the "cancel" button at the bottom right.
When the only remaining window is the console, click on "console" (at left)
and "reboot" (bottom of that menu) and "OK". This will reboot the X terminal.
STARTING UP THE TCS AND TELESCOPE CHECKS
When the X terminal has rebooted, start a session on jkttcs (lpas1)
Password: ********* (as shown on monitor)
- Start the TCS with:
The User and Infodisplay windows will appear.
- Push the Oil Pump RUN button. Wait until the OIL ALARM indicator
lamp has gone out.
Push the red CANCEL button to mute the audible alarm and push the green
RESET button to enable the telescope drive power amplifiers.
Check that the telescope is in ENGINEERING mode. The yellow section of the
ENG/COMP push button switch should be illuminated. This is the default
Zerosetting the telescope
Visually check that the telescope is at, or very near, zenith park. If
it is not at zenith park, move it as close as you can using the engineering
HA and DEC +/- buttons.
Put the "Rate" control knobs for HA and DEC to about half way.
n.b. There is NO NEED to set these controls to the green marks.
In fact, the telescope can be zeroset at full velocity. This was a 'left over'
from days when old filament lamp Baldwin encoders were used that loosed counts
at higher speeds.
Using the HA and DEC buttons, move away from zenith park in the -ve direction
for about 3 seconds. You can move HA and DEC simultaneously.
In the TCS User window, enter:
USER> zeroset ha target
USER> zeroset dec target
Using the HA and DEC buttons, move back towards zenith park in the +ve
direction until the TCS beeps and tells you the zeroset is successful. Again,
you can move both axes simultaneously.
Testing the telescope in Computer Mode
Put the telescope into computer mode by turning the COMP keyswitch. The green
section of the ENG/COMP push button switch should illuminate showing that
Computer Mode has been selected.
Check the Local Sidereal Time display. The telescope can be safely moved in RA +/- 3
hours away from the LST as displayed.
Enter a value of RA either +/- 1 hour away from the displayed LST.
In Dec, a good position for checking the telescope is at the celestial equator.
This will position the telescope tube at right angles to the polar axis.
Format for entering telescope co-ordinates is :
USER> source name hh mm ss dd mm ss equinox
name No more than 20 characters
hh mm ss RA hrs mins secs
dd mm ss DEC +/- degs mins secs
equinox Usually J2000
Example: LST is at 15:30:00
USER> source TEST 16 30 00 00 00 00 J2000
Check that the telescope gets into "Tracking" mode and the dome moves.
n.b. It is normal for the dome angle to be different from the telescope
azimuth. This is because the telescope tube is not in the centre of the
Put the telescope back to zenith park.
USER> park zen
Send the dome to 235 degrees.
USER> dome 235
and check that the arrow on the dome rail lines up closely with the mark
on the dome. These marks are almost opposite the control desk and can
be seen through the window.
n.b. If there is a significant difference in the position between
the marks, it is highly probable that the dome encoder sprocket has jumped the
chain! This can cause the dome to vignette the telescope and would need to
be fixed urgently. To rectify this, proceed as follows:
Move the dome manually using the +/- dome buttons to line up the marks.
Climb onto the balcony above the control room and pull back the spring loaded
arm (with the encoder and gearbox attached) to dis-engage the sprocket.
Rotate the sprocket manually until the dome position on the infodisplay reads
as close as is possible to 235 degrees.
Without turning the sprocket, re-engage it back into the chain.
Return the telescope back into engineering mode by pressing the ENG/COMP
button. Enter "ack" on the keyboard to stop the TCS alarm.
Turn off the telescope power. This will also switch off the oil pump. Push the
CANCEL button to mute the alarm. Close the roller blind.
STARTING UP THE uDAS AND ICS COMPUTERS
n.b. Always bring up uDAS before starting the ICS.
When shutting down, do this in the reverse order.
Log on at the console of the JKT Observing System Sparc :
Password: ****** (as shown on monitor)
Two xterm windows appear: one with a pink background, logged in to the
System computer and one with an orange background logged in (via
rlogin) to the DAS computer (located in the control room annex).
In the DAS orange window, enter :
Check the whiteboard for the current observing system option numbers. These are
7 Instrument = JAG
6 Detector = SITe2
Enter these numbers (or what is shown on the whiteboard) for the observing system
requirements. A message will then appear asking if you want to start the observing system.
In the same window enter :
and check the messages for any errors.
Eventually, you should get the prompt :
If there were no errors, the DAS is now ready.
In the pink window, type :
Enter the menu option numbers that were used when starting up the DAS.
These must be the same.
In the same window, enter :
The Observing System will now be loaded. During this process several windows will
be generated. The important ones being:
A TALKER window.
A LOG of observations.
A TCS INFODISPLAY (Check that the UTC and LST time is updating).
A MIMIC for the JAG
The UltraDAS mimic window.
Check that you have the prompt: SYS>
Check that you can send commands to the TCS from the ICS.
Check you can send commands to the TCS. e.g.
SYS> tcs page
The Infodisplay on the Xterm should change. There are three pages
available which can be toggled using the PAGE command.
The Observing System is now ready to rock n'roll !
CCD DATA ACQUISTION AND BIAS CHECKOUT
Taking a BIAS frame
Make sure the dome lights are out and the control room blind is down.
On the ICS Sparc in the pink window, enter :
The uDAS mimic window will show the operations and progress as the
CCD is readout, processed and written to disc.
Displaying the image
To avoid screen congestion, open up a new screen (using the CDE control bar
buttons). This keeps the image data reduction packages seperate from the
Observing System windows.
Start up IRAF and XIMTOOL using the procedure below :
- Move the mouse over the background and press the right mouse button.
- Keeping it pressed, select Programs and then xgterm.
An XGterm window will appear.
Repeat the above procedure and select ximtool from the Programs
- Start IRAF by entering: cl in the XGterm window.
In the IRAF window, move into the directory where the images are stored.
The directory format is: yyyymmdd e.g.
cl> cd /obsdata/jktb/20020124
n.b. The data path to the image files can be found at the bottom of UDAS mimic.
There are 2 obsdata discs, currently : jkta, jktb. The data will be stored on the disc that has the most space available. A new "date" directory will be automatically created at midday.
When the bias frame has read out, the computer will beep. Check in the OBSERVING LOG
window that the image has been received and that the headers have been written. The
images are stored as "rnnnnnn.fit" files, each having a sequential run number.
In the IRAF window, type:
cl> display r123456 1 fill+
Where the run number corresponds to the last image taken. ximtool will
now display the image.
IMPORTANT : You must add the  extension (which denotes the window
selected) after the run number.
n.b. This applies to all IRAF commands that you may use.
n.b. The WINDOW syntax for uDAS is:
SYS> window 1 "[xtart : xfinish, ystart : yfinish]"
The fill+ option will make ximtool display the ENTIRE ccd image.
Examining the image
In the IRAF window, enter:
Move the mouse into the area of the displayed frame and left click.
Move the mouse around the image area, typing: m. This will produce
statistics for a box containing 25 pixels centred on the mouse cursor.
Check that the noise(e) levels correspond to the CCD data held in:
If there are white spots, move the mouse over them and type: s. This
will produce a surface plot. Check that the white dots are single pixels.
To exit imexam, leave the mouse cursor over the image area and type:
TESTING THE JAG
n.b. If there are any problems with the JAG, refer to the document:
Taming the JAG which should solve most problems.
Move the ccd FILTER WHEEL. (positions are 1 to 6) e.g.
SYS> filter 3
You will see the filter on the JAG mimic turn blue and the position change.
n.b. You may also see the autoguider filters and the telescope focus change if
a filter of a different thickness is selected.
- Test the ccd SHUTTER.
SYS> shutter open
Check that the JAG mimic shows the shutter is OPEN. The shutter status will also
change colour in the uDAS mimic window. Close the shutter with:
SYS> shutter shut
Move the AUTOGUIDER probe ready for the autoguider tests.
You should see the probe X / Y positions on the mimic turn blue and the
numbers change. When the probe is at the demanded position, the numbers
should be green. There will also be a message on the mimic saying that
the probe is close to the BLIND SPOT.
Move the TV acquistion MIRROR into the beam and open the TV SHUTTER.
SYS> tv on
You should see the TV mirror and shutter turn blue. The TV flat will
move IN to the light path and the TV shutter will OPEN. They will turn
green again if the operation was successful.
CHECKING OUT THE AUTOGUIDER
The JKT autoguider uses an image dissecting PMT mounted on an X/Y slide
occupying one half of the JAG. The tube can be easily damaged by over
illumination and for these reasons, precautions are taken to switch OFF
the main dome lights when the autoguider is activated.
Power on the autoguider (the red "ON/OFF" button.). This should also
test the over illumination safety interlock and turn off the dome lights.
Important Make sure this happens and that
there are no other lights on in the dome.
Switch on the autoguider TV monitor (small right hand screen).
In the TCS user window, enter:
A small window will appear within the main autoguider display.
You should see very little or no counts in this window.
Move the autoguider probe from the ICS:
The probe will now move out of the `blind spot` and pixels at different
light levels should be building up in the small autoguider window.
SYS> autoxy 3000 23000
In the TCS user window, enter:
The autoguider screen should say "ACQUIRING" and then "PULLING IN".
Check that the autoguider probe X and Y values are changing on the mimic.
n.b. You won't be able to check the guiding without a real star.
To finish testing, enter:
Switch off the autoguider and monitor. The dome lights should come
CHECKING OUT THE INTEGRATING TV SYSTEM
The JKT uses an intensified Westinghouse TV camera in conjunction
with a GRINNELL SYSTEM integrating frame store for image acquisition.
Testing out the TV camera in direct viewing mode.
Make sure the TV flat is IN and the TV shutter OPEN.
Switch on the Grinnell TV display and the direct viewing mode TV monitor
(small left hand screen).
On the TV remote control box. Make sure the GAIN control is backed
off (fully anti-clockwise).
Push the ON button on the TV remote control box. This will also TURN
OFF the dome lights.
Important Check this happens and that there are no other
lights on in the dome.
When the red DELAY led goes off, the yellow RESET led should go off also.
n.b. If this doesn't happen, check that the GAIN control is
turned back past a `click stop` when rotated fully in the anti-clockwise
- Push the DIRECT button on the Grinnell control panel.
Slowly increase the GAIN until you see scintillating speckles (ion events) on
the direct viewing mode TV monitor. Toggling the H and V scan direction change
switches and seeing the `target' moving around the screen is a good test that
the TV camera is healthy.
Important Always return the TV scan switches to their `marked`
Checking the GRINNELL system
This enables an image to be built up within a frame store memory over a series
of integrations. Using the GRINNELL, the TV system can resolve much fainter
objects than when using the DIRECT viewing mode.
Back off the GAIN control and push the INT (Integration) button.
n.b. If the TV RESET button has been pushed, the `black level` will
default to a value of 2048
Set up the Grinnell TV Integration values as follows:
- Set the INTR 2n FRAMES switch to 2
- The BLACK LEVEL to 512
- The CONTRAST to 6
n.b. The BLACK LEVEL and CONTRAST switches have continuous
rotation. Check in the small window in the lower right hand corner of the
Grinnell monitor to see the values change.
Watching the direct TV monitor, increase the GAIN slowly. You should see
a pulsating patch of light.
Watching the Grinnell monitor, increase the CONTRAST until you see a
build up of light somewhat `pixelly` in appearance.
Decrease the GAIN until the light patch disappears.
Push the AVERAGE button.
Slowly increase the GAIN until you see the build up of light again. The
light patch will now have a smoother profile.
Turn the GAIN down, push the OFF switch on the TV remote controller and
switch off the TV monitors. The dome lights should come back on.
On the ICS, in the pink window enter:
SYS> tv off
to move the TV flat out of the light path and to close the TV shutter.
Filling the Cryostat
The CCD should be filled early afternoon. It's best to do this AFTER
running the telescope checks as the cryostat can empty out when the
The liquid nitrogen dewar must be earthed to prevent static discharge.
Clip the dewar earthing cable onto the grounding point. A copper earth strap
is provided on the mirror cell.
FULLY insert the filler tube into the cryostat.
On the dewar, CLOSE the VENT TO AIR valve (4).
OPEN the DECANT valve (1).
OPEN the PRESSURE BUILD valve (3).
Check the pressure does not exceed 20psi for a armoured steel
hose. 10psi if the hose is silicon.
If the dewar is low on liquid nitrogen, the BYPASS valve (5) can be opened.
Use precaution when doing this as the pressure can build up very quickly
if the dewar is close to full.
Wait until liquid nitrogen is SEEN to be coming out of the cryostat.
CLOSE the DECANT and the PRESSURE BUILD valves.
OPEN the VENT TO AIR valve. If the flexible hose is silicon, wait
for it to de-frost before removing the filler tube.
Always remember to leave the PRESSURE RELEASE valve on the dewar
OPEN. If this is forgotten, pressure can build up over time and burst the
relief valve on the dewar. Especially if the dewar is close to full.
Before leaving the JKT
Replace/zip up the black bag around the JAG.
Lower the rising floor so that the telescope can't hit it.
Back in the control room. Turn OFF the dome lights and ...
Don't forget to tick off the checks as completed and to sign off in the logbook
Rev 1.0 First release: Nov 10th 1998 Sheila Crosby
Rev 1.1 Revised and extended: Nov 26th 1998 by John Mills
Rev 1.2 Revised after s/w upgrades: Dec 4th 1998 Sheila Crosby
Rev 1.3 Revised IRAF startup and image file access in the Data Acquisition
section: Mar 29th 1999 John Mills
Rev 1.4 Cryostat filling procedure corrected: Apr 21st 1999 John Mills
Rev 1.5 LN2 dewar earthing added. SITe2 noise value added : Oct 5th 2000 John Mills
Rev 1.6 Changes made for UDAS operation: Jan 9th 2002 John Mills
Rev 1.7 Removed machine names. Generic names only now: July 16th 2002 John Mills
Rev 1.8 Wording changed from 'Shutting down' to 'Logging out from' the JKT
Observing System : July 16th 2002 John Mills
THIS FILE: http://www.ing.iac.es/~eng/ops/jkt/jkt_checks.html