Daytime Checkout Before Every Run
No mechanism moves and no lamps required
- Check conking. Log into hqcronserver as
user esi. Verify that ESI numbered accounts
to be used during the next run have been conk-ed by
using conk-check esi to check the
conk date. Manually initiate conking on
unconked accounts by logging into
hqcronserver as esi and running the
following command:
autoconk -f -m m -M M
where m is the first account number to conk and
M is the last.
- Check rebooting. Log into esiserver
as user esieng . Issue the uptime
command to check how long the computer has been up. If more
than 7 days, request a reboot from sysadmin. Repeat this
check for esiserver2.
- Check disk space. Log into esiserver
as user esieng.
- Verify that ESI data disks have been cleaned off by running
datadirs all and checking the output. If many old
directories are present, work with software group to verify that the
data have been savier-ed before deleting files.
- Run the df command to check that the disks have a
reasonable amount of space available. If any of the
/sdata
disks have less than 20% usage, then there is sufficient space.
- Check disk speed. Log into esiserver
as esieng. Check the speed of the disks using the
command
~esieng/scripts/run_test_disk_io_speed
Check the logfile in
~esieng/log/test_disk_io_speed.log and see
whether current I/O speed is in line with previous values.
If not, then suspect a problem with the ESI RAID array and
seek assistance from sysadmins; this test is completely
independent of the ESI software and simply tests the time
required to write data to disk.
- Check image acquisition sequence speed (7
minutes). Bring up the ESI software from
esiserver as esieng. Check the speed of
the image acquisition sequence by running the script
~esieng/scripts/run_test_goi_time
For each data disk, this will acquire 5 spectral-mode bias
images and compute the time to write to disk. Check the
resulting output in ~esieng/log/test_goi_time.log,
which looks like this:
# date disk n erase wcrate wdisk total
# ---- ---- - ----- ------ ----- -----
2010-Jun-06 17:19 /sdata700 5 11.44 38.96 4.16 60.66
2010-Jun-06 17:24 /sdata701 5 11.45 39.79 3.04 59.63
2010-Jun-06 17:29 /sdata702 5 11.44 39.28 4.24 60.81
2010-Jun-06 17:34 /sdata703 5 11.46 39.84 3.74 60.48
2010-Jun-06 17:39 /sdata704 5 11.44 39.81 4.31 61.03
2010-Jun-06 17:44 /sdata706 5 11.44 39.81 3.77 60.43
2010-Jun-06 17:49 /sdata707 5 11.43 39.54 3.75 60.31
Compare the time required to complete various phases to what
it took in the past. If problems are seen, check software
and hardware:
- if the script does not return from first exposure, use
ct on kanaha to check that
two lickserv processes are running on kanaha as esieng. Restart
if necessary. You may also need to
reboot the CCD crate.
- unusually long erase time may indicate a problem with
the CCD controller or CCD keyword library
- unusually long wcrate time may indicate a
problem with the CCD controller
- unusually long wdisk time could be a
problem with the gathering of keywords or the
write_image daemons
- abnormally long total time could indicate
a problem with lickserv
Mechanism moves required, no lamps
- Check safety. Phone the summit and verify that
ESI is available for moving motors, including opening the
ESI hatch
- Start software. Log into
esiserver as esieng and launch the
instrument software.
- Home motors.
NOTE: This will open and close the ESI
hatch. Ensure summit personnel near ESI are aware.
- Click the Start Here button on the ESI
dashboard window to access the Setup/Options
panel.
- Click the INITIALIZE INSTRUMENT button to
get the INITIALIZE INSTRUMENT panel.
- Click the button reading Yes, I Really Want To
Initialize ESI! to start homing motors.
- Wait while homing completes (about three minutes).
- Verify that all stages home properly. If a stage
again fails to home, you may need to enlist the help of
summit personnel to fix it.
Note: if you encounter trouble, you can try to execute the
initialization from the command line using this syntax:
/opt/tcl803/bin/tcl /kroot/bin/ktrun /kroot/data/esi/Tests/ESIinit
- Check CCD shutter. In an xterm on
esiserver, issue the command
cycle_shutter -100 to verify that the shutter
will open. You will initially see errors reported, but the
final lines should read:
setting cshutter = open (wait)
setting cshutter = close (wait)
- Check CCD bias.
- Check triple wheel renishaw encoder voltages (3 hours).
Mechanism moves and lamps required
- Check safety. Phone the summit and verify that
ESI is available for moving motors, including opening the hatch.
- Start software. Log into
esiserver as esieng and launch the
instrument software.
- Home motors.
NOTE: This will open and close the ESI
hatch. Ensure summit personnel near ESI are aware.
- Click the Start Here button on the ESI
dashboard window to access the Setup/Options
panel.
- Click the INITIALIZE INSTRUMENT button to
get the INITIALIZE INSTRUMENT panel.
- Click the button reading Yes, I Really Want To
Initialize ESI! to start homing motors.
- Wait while homing completes (about three minutes).
- Verify that all stages home properly. If a stage
again fails to home, you may need to enlist the help of
summit personnel to fix it.
Note: if you encounter trouble, you can try to execute the
initialization from the command line using this syntax:
/opt/tcl803/bin/tcl /kroot/bin/ktrun /kroot/data/esi/Tests/ESIinit
- Check internal lamps.
- As esieng@esiserver, execute the
script test_lamps script, which will:
- Configure for low-dispersion spectroscopy with
LowD_1.0 slit
- Configure CCD for full-frame readout
- Acquire exposures on each lamp
- Verify that you get light from each lamp by checking
for a spectrum on the corresponding image.
- Typical peak counts:
- Ar: 50,000 DN
- Cu: 1,600 DN
- Ne: 21,000 DN
- Qz: 65,000 DN (saturated in parts)
- Check focus settings (2 hours for full, 40 minutes for
quick).
- Launch IDL from any HQ machine to acquire and analyze data:
- Open an xterm window at HQ by typing
kixrsh remopshost
- In the HQ xterm, launch IDL via
~esieng/idl/esieng/do_esi_focusloops
-OR-
~esieng/idl/esieng/do_esi_focusloops_quick
- inspect resulting plots in /home/esieng/data/focus
- compare results to previous fits as sanity check
via:
compare 2013nov07 2013sep25
substituting dates as appropriate
- Often there will be issues in the fit for 2-3 of the
modes that require separate analysis.
- If you perform separate fits on any of the data and obtain
alternate best focus values, edit them in to the
{mode}.{date}.dat
files in the /home/esieng/data/focus directory.
- Perform separate analysis of low-D images:
- In the data directory as esieng on esiserver, create a text file listing
the 9 focus images for the lowD setting via:
obslog | fgrep lowD | awk '{print $1 ".fits"}' > lowD.lst
- Open an xgterm window on
kealia and launch IRAF.
- Load the keck and lris
packages.
- Change to the data directory.
- Execute the specfocus program with the
following parameters:
specfocus.images = "@lowD.lst"
specfocus.focus = "collfocr"
specfocus.corwidth = 20
specfocus.level = 0.5
specfocus.shifts = yes
specfocus.dispaxis = 1
specfocus.nspectra = 1
specfocus.ndisp = 1
specfocus.slit1 = 990
specfocus.slit2 = 1010
specfocus.logfile = "logfile"
Note that slit1 and slit2 are
chosen to be at the point where the spectral lines are
parallel to image columns (near row 1000).
- Inspect the resulting plot and confirm that it
looks good (like this).
- From an HQ window, edit the corresponding focus
file via
emacs ~esieng/data/focus/lowD.YYYYmmmDD.dat
to update the focus settings with the value derived
from the specfocus analysis. Update the
# best focus value in the example below.
# mode = lowD
# date = Wed Nov 16 14:24:12 2016
# number of pinholes = 5
# best size = 5.53403
# best focus = 975607.
0 /s/sdata707/esieng/2016nov16/e161116_0021.fits -5000.00
7.66837
1 /s/sdata707/esieng/2016nov16/e161116_0022.fits 15000.0
7.21367
2 /s/sdata707/esieng/2016nov16/e161116_0023.fits 35000.0
7.43573
3 /s/sdata707/esieng/2016nov16/e161116_0024.fits 55000.0
6.92101
4 /s/sdata707/esieng/2016nov16/e161116_0025.fits 75000.0
7.59082
5 /s/sdata707/esieng/2016nov16/e161116_0026.fits 95000.0
7.35732
6 /s/sdata707/esieng/2016nov16/e161116_0027.fits 115000.
7.09436
7 /s/sdata707/esieng/2016nov16/e161116_0028.fits 135000.
6.82815
8 /s/sdata707/esieng/2016nov16/e161116_0029.fits 155000.
7.01923
- Update focus setup files:
Daytime checkout Occasionally
Mechanism moves and lamps required
- Check CCD parameters. [TBD]
Don't perform these tests yet; scripts to be
written to automate. Below is a list copied from DEIMOS.
- Acquire 3 bias exposures (spectral mode)
- Acquire 3 internal flats using 1200G at 7800A, slitless
- Analyze to derive gain, readnoise, bias levels on each CCD
- Compare to baseline values
- Also analyze data to derive bad pixel mask
- Check guider.
- Bring up guider software:
- Call summit support personnel and request
permission to use K2 MAGIQ software for 30 minutes.
- Log into k2obs via
ssh -X k2obstcs@k2obs
- Launch TCSGUI via tcsgui < /dev/null &
- Click on TCS >
MAGIQ > Current to launch the
MAGIQ-K2_Operations menu.
- Click on MAGIQ-K2_Operations and
seelct Start Cam Diagnostic.
- Wait for MAGIQ user interface
to appear.
- Configure instrument for bias frames:
- Launch a ESI dashboard GUI from the
OpenWindows menu or instr_menu widget via
ESI Control Menu > Subcomponents... >
Restart Dashboard GUI
- Turn off internal lamps by clicking the OFF button in the lamps
area.
- Click on the TV Guider
button to launch the TV panel.
- On the TV panel, click on the filter selector
button and select the ND_1 filter, then
click on the Go button to
change the filter.
- Acquire guider bias frames using MAGIQ:
- Select the ESI guider
- Set binning to 1
- Set exposure time to 300ms (shortest available)
- Turn image update on
- Save 3 camera images by setting Save
next N images to 3 and clicking the
Save next N images button.
- Configure instrument for guider flats:
- On the TV panel, click on the filter selector
button and select the R filter, then
click on the Go button to
change the filter.
- On the main dashboard, turn on the Qz
lamp.
- Acquire guider flats using MAGIQ:
- Save 3 camera images by setting Save
next N images to 3 and clicking the
Save next N images button.
- Analyze images for gain and readnoise
- Analyze flux level to check camera sensitivity
- Focus guider.
Use this procedure to acquire and analyze focus sequences
for each of the ESI TV filters. To check the full
complement of filters requires 1-2 hours from start to
finish, but the R-band filter can be checked in a few
minutes.
Approximate TV focus offsets (from R-band)
| TVFILNAM |
Focus Offset |
TVFOCRAW (2021-10-07) |
|---|
| R | 0 | -2000 |
| B | 0 | -2000 |
| V | 0 | -2000 |
| I | -43 | -2043 |
| Clear | +680 | -1320 |
| ND_1 | +260 | -1740 |
Nighttime Checkout
These procedures are run at night to calibrate the instrument.
- Check throughput (procedure for DEIMOS, not ESI)
- Use setup of 1200G@7800A, slitless
- Put standard star at SLIT P.O. The preferred
standard star is BD+28deg1142 from the flux standards list.
- Acquire 5s exposure
- Extract and compare to Ricardo's data
Note: would useful to have 3 or 4 chosen standards spaced in
RA such that at least one is always accessible
|
