Operations of the Keck AO
Download
Report
Transcript Operations of the Keck AO
Keck AO
Operations
D. Le Mignant
for the Keck AO team
Topics
AO refresh
AO operation at a glance
About the operation tools
Troubleshooting session
AO = Adaptive Optics
Control loop for canceling out the optical
aberrations introduced mostly by the
atmosphere (WFS, WFC, TT, DM,etc).
Tracking devices (ROT) and slewing
devices (FSMs, FCS, FSS)
Telescope guiding and focusing.
Is this a transparent process?
ROT
Pupil re-imaging
Dichroic
TT
DM
FSMs
WFS
Software
Architecture
obs
eng.
screen
AO
supervisory
control
Telescope
DCS
epics channels
Optics Bench Devices
pro
files
IDL
wfc
eng.
screen
AOA camera
Wavefront Controller
Epics kwd server
slk
Java
User Interface
autom.
units
cshow
AO at a glance:
what it could be…
Run
nighttime script
Slew to reference star
Acquire star on WFS
Run auto AO setting script
Go to next object
Acquire star on WFS
Etc..
Run Halt AO script
What is it :
1.
2.
3.
4.
5.
6.
Nighttime script
Slew, acquire reference
star
Check FSM POs and
run auto AO setting
Coordinate rotator slew
Watch for problems
etc
1.
2.
3.
4.
5.
Need to be simplified
Ok, watch PO and/or use
pixel location
Reliable, yet not
optimized for IF and
survey mode
Automation needed
Auto-recovery and
automation needed
What has been our strategy so
far..
K2AO/K1AO need to be modified for dvlpt.
Yet, we routinely operate them for science.
After many years, OPS software still evolving..
New feedback from OAs
Changes being implemented
AO handover will provide a strict baseline for
NGS AO operation
Science version vs dvlpt version
Reliable, high performance and easy-to-use
IDL
JAVA
JAVA
IDL
AO
supervisory
control
obs
eng.
screen
Telescope
DCS
epics channels
Optics Bench Devices
pro
files
IDL
wfc
eng.
screen
AOA camera
Wavefront Controller
Epics kwd server
slk
Java
User Interface
autom.
units
cshow
IDL widgets and scripts
AO Acquisition widget:
Watchdog tasks
Rotator, FSMs fault
WFC crash
Set AO!
Scripts:
Auto AO setting
Auto-recovery scripts
Set FSMs
Set AO parms
Record WFS backgrd
Iterate once and exit
AO/DCS coms
FSMs, Rot fault
Update Cog
Java and other tools
1.
2.
3.
Java screens
Some value displayed are
- EPICS keyword server
- automation units kwds
WFS config. / rot position call
automation units
IDL
JAVA
JAVA
IDL
The most common problems
(random order)
WFC crash
FSMs Fault
Rotator Fault
DCS/AO communication
AO loops not closed
GUIs not responsive
Poor AO correction
Telescope nod with AO
NIRC2:
sends nod request to telescope
waits for AO (wait4tt, wait4dm, wait4fcs)
DCS/AO handshake :
DCS
AO:
request to AO : pause
DCS :
moves telescope
request to AO : resume
AO:
AO
In parallel to FSMs moves, WFS
focus repositioned
NIRC2:
Resumes
Open AO loops
(aoloop=open/restore)
acknowledges
converts nod in FSM X, Y
moves
Send FSM moves
Close AO loops
(aoloop=restore)
acknowledges
Nodding using FSMs
FSMs : 2 mirrors / 4 actuators + encoders
Moves image on the WFS
Maintains DM to lenslet pupil registration
--> Any coordinated moves requires driving all 4
actuators
Moves take <10sec
Tight space
See hardware
AO/DCS communication
SC
TT offloading
Focus offloading
FSM nod
AO loops (aoloop kwd)
DCS
Telescope guiding
Secondary focus adjustment
Restart AO/DCS communications
Open all loops (TT, DM, TT and focus offloading)
Set aoloop kwd to open
Reset automation units to record WFS
background
Zero-out the DT control offset (used for DAR
and KAT offload)
Reload the last-used centroid origin file for the
WFC
Run DCS script to set AO TT offload as inputs
for telescope guiding
Close TT and focus offloading
Does not close TT and DM loops
Telescope
Pointing
TTO
Secondary
Mirror
Piston
AO Loops
WFO
DCS
TTM
Supervisory
Controller
TT Loop
Wavefront
Controller
WFS
DM Loop
DM
WFC crash (see web page)
WFC running at 672Hz
TT controller and DM controller, Fault likely
comes from DM controller
Fault detection not obvious
Bad image quality!
WFS intensity not updating, wyko frozen
AO loops not responsive
Once Fault detected, one needs to reboot WFC
No Calibrations needed BUT need to run the
nighttime script again and auto AO setting script
We are working on
Better fault detection
Full auto-recovery script
Rotator operation
Rotator is a tracking device with two
modes (Position Angle, Vertical Angle)
Position is fixed by the astronomer
Rotator slew are cumbersome:
AO loops open
Star on-axis (within few arcsec)
After rotator slew
Close loops
Wait for new reconstructor
Rotator fault
OBS/DCS communication used to drive
the rotator
Fault modes:
Invisible: rotator does not report any fault
Visible: FACSUM reports FAULT
But acquisition does not work
Mechanical? OBS/DCS coms?
See web page for troubleshooting