Summary of the State of Bent Gregorian Focus -- April 2008
The following is a summary of the state of performance of Bent Gregorian F/15 operation
at the end of the first two weeks of commissioning activities during 11-26 April 2008.
Pointing and Tracking
With about 20 pointing stars used to make a pointing model which is 6 arcsec rms across
the sky, pointing and tracking are quite good. The pointing and tracking are already in
the range needed for LUCIFER.
The main limitation on the pointing is caused by temperature gradients in the telescope
structure. When the temperature changes by more than 5 degC per day, we have problems
acquiring the star in the 30 arcsec FOV. On typical nights with 2-3 degC temperature changes,
we reliably place the star within the 30 arcsec FOV.
The pointing model should be redone in June after we have completed all the alignment and
collimation refinements with closed-loop active optics.
The by-eye collimation model model with temperature corrections is working decently well.
On stable temperature nights, we can hold 1-2 arcsec images for hours without active optics
corrections. The collimation model certainly can be improved once we have closed-loop active optics.
The main problem is that the collimation model doesn't account for the telescope structure
being out of temperature equilibrium. Temperature gradients appear to be the limiting factor
for open-loop collimation.
So far, we have only adjusted the primary mirror collimation to correct the on-axis images.
Work to measure the field aberrations to position the secondary is anticipated for May.
Left Bent Gregorian Instrument Rotator
The left bent Gregorian rotator has been performing reliably. It's tracking error is in the range of
5 arcsec rms.
The only limitation for observing is the limited range of rotator angle between -1 and 359 degrees.
PCS (Pointing Control System) Sub-system
After some struggle to get PCS to make the guiding transformations correctly, the PCS is pointing, tracking and guiding reliably at bent Gregorian.
Problems to be addressed for next month:
- Understand and implement PCS angles to adjust instrument position angle in coordination with adjusting the guiding transformations.
- Resolve the residual timing issues between PCS and MCSPU.
IRTC (InfraRed Test Camera)
The IRTC has been performing quite reliably. The rapid frame rate and real-time video has been useful to diagnose a number of telescope issues. We had to reboot the Windows PC a couple of times, but
no problems seen beyond that.
Needed for next month:
- the complete documentation.
AGw (Acquisition, Guiding and Wavefront) Unit #3
The AGw unit #3 has been operating reliably. We have been using it for guiding and wavefront
sensing, but we've not yet gone through the detailed on-sky checks of the AGw parameters.
The on-sky commissioning work for the AGw is planned for May 11-15.
GCS (Guiding Control System) Sub-system and Guiding
The good news is that the newly deployed GCS subsystem is able to find guide stars and reliably guide on
them for periods up to an hour. The bad news is that it stops frequently during the night
and has to be restarted.
Things needed for next month:
- Fewer crashes (unexpected deaths) of GCSL
- Fix bug where wavefront images show up in guider images and vice versa.
- Understand why GCS sometimes loses the guidestar after one guider frame.
- Ability to offset the guide star onto the WFS hot spot. We've written a python script to do this with command line commands, but it is still tedious.
- GCS GUI
- Fix bug related to acquiring stars in the outer part of the field.
- GCS wavefront data collection and processing
All the machinery for collecting and processing intra/extra-focal pupils with EF is in place.
We have not closed the active optics loop using any EF results.
This work was postponed to make time for testing GCS guiding on-sky.
Initial results using the S-H WFS through the engineering commands have been quite promising.
We have analyzed these spots through Doug's IDL software. We have worked out the rotational transforms between the rotator and the M1 pupils. We don't understand some strange factor
of two scalings between one Zernike term and the next. We dialed the telescope into decent
collimation one time using Doug's S-H analyis, but we have not yet closed the active optics
loop by having the software automatically send the Zernike corrections to PSF. We have
demonstrated that we can send polynomials for either M1 or M2 through DirtI
PSF (Point Spread Function) Sub-system
PSF has been working reliably for managing the collimation of multiple optical elements.
Needed for next month:
- We need to implement pointing-free coma corrections so that
active optics corrections don't push the star out of the 5 arcsec guide box.
The downside of this change could be limiting the collimation range of LBC on nights that are
out of temperature equilibrium.
DirtI (Command Line Interface to IIF and IRTC)
The command line interface has been an invaluable tool for controlling the IRTC and interacting with
TCS. After some initial problems with timing on the socket communications, it has been quite
Has some minor problems with timing on the GetSequence
command. Otherwise the communication
with IRTC and TCS are both reliable.
Needed for next month:
- consolidated man pages and user manual
- improved GetSequence timing
OSS (Optical Support System) and M2 Hexapod
OSS has been running satisfactorily to operate the secondary mirror. It has had some problems with
getting confused and need to have the sub-system restarted.
The hexapod has been operating smoothly without any obvious jumps in the image.
Need for next month:
- IIF commands are returning success on commands that move the secondary before the mirror finishes moving. We've patched around this in our observing scripts, but this needs to be corrected for efficient data taking.
- user manual for the OSS subsystem.
PMC (Primary Mirror Control) Sub-system and SX Primary Mirror
The SX primary mirror cell has operated reliably during the commissioning.
The two problems that need to be addressed for next month are:
- calibration of the replacement LVDT in Hardpoint 0
- resolution of the cell problem that causes 1-2 arcsec of pointing jitter each time the mirror is repositioned.
Image Quality and Vibration
We've seen images as good as 0.4 arsec FWHM in H even without the benefit of closed loop
active optics. So things are good from the perspective of the initial alignment and collimation.
The main problem with image quality is that the fans in the treehouses are apparently exciting
a resonant vibration in the swing arms and causing image vibration of 0.5 to1 arcsec.
This is the dominant error in the image quality when the seeing is better than 1 arcsec.
Rigid Secondary Mirror
We found the bent F/15 focus remarkably within millimeters of where it was expected to be.
And we do not see any excess spherical aberration. Thus, we can say that the rigid secondary has
the correct asphere and focal length.
There are no problems that we can attribute to the secondary mirror.
No remote motions of the tertiary mirror have been made since Andrew made the initial optical
alignment. JMH isn't in touch with the details of the remaining M3 electromechanical issues.
- 27 Apr 2008