The Adaptive Optics Bonnette

Operation of the AOB continues to be straightforward and trouble free. As with other complicated systems, much of the ease of operation on the sky can be attributed to intense run preparation by the technical staff.

The planned purchase of upgraded APD's has been put on hold pending comparative tests between old and new technology APD's (a sample new APD is on loan from the JNLT). Initial estimates of potential gains in sensitivity appear to be overly optimistic. A single APD in the Wavefront Sensor failed and was replaced by a spare.

Optical throughput efficiency calculations of the wavefront sensor subsystem predict a factor of 2.6 higher sensitivity than is measured on the sky, and in the laboratory. Attempts to locate the source of this loss have been unsuccessful to date.

Tests of the 2x focal enlarger for the AOB using FOCAM and STIS were very successful. The system provides critical sampling of the psf below 1.0 µm. This system is now an available option for observers.

The focal enlarger was also modified slightly to accommodate our Fabry Perot etalons. The system was successfully tested on the AOB with a comparison spectral source earlier in the semester. We hope to test the combination on the sky in January.

Three new beamsplitters have been ordered for use in the visible with OASIS. The beamsplitters will have bandpasses for the V, R, and I bands These should improve throughput to the WFS by about 50% over the current '50/50' beamsplitter and will increase throughput in the science path to about 90% in the bandpass.

The diameter of the AOB's small artificial star was reduced from 8 microns to 3.0 microns for work at wavelengths in the 0.5 to 1.0 um range.

The user interface is now working quite well; the number of system glitches has been reduced to at most one or two a night. Still, substantial improvements to the software are underway to improve overall reliability. The MOSAIC tool used in combination with MONICA has been particularly successful.

Overall ease of use for visiting scientist - one of the initial design goals - has been validated by two observing runs by non-specialist observers. Guide star acquisition and loop closure have proven to be trouble free and straightforward. AOB specialists are no longer required on site once initial observer familiarization has taken place.

A large amount of background work by the technical staff has been aimed at making setup operationally "friendly" through hands-on training and updating of documentation.


Much time was spent preparing and supporting the 27 night run of FTS. The system was operated in several different modes (InSb, InGaAs, BEAR-imaging, and seismology-mode). Time spent earlier in the year to eliminate 60 Hz interference with the InGaAs diodes clearly paid off. These InGaAs diodes provide a 4x to 6x signal gain over the InSb detectors at 1.2 µm and are operationally quite easy to use since they require neither a dewar nor cryogenics.

Flexure measurements of the FTS in the BEAR configuration were made to determine if the modifications made to the mounting mechanics of the REDEYE camera had reduced image motion. Unfortunately the stiffer camera mount did not reduce motion in the image plane. The flexure data suggests that the major source of image motion, which reduces BEAR's effective FOV by about 10% for long scans, is probably in the optics between the output of the FTS and the REDEYE camera. The upgraded mechanics does however greatly ease the mounting and alignment of Redeye.

The newly re-coated IR1 beamsplitters were installed and worked well. Increasing problems with reference laser throughput for the IR2 beamsplitters caused several servo unlocks during the July / August observing run. Replacement beamsplitters are currently being fabricated (re-coating was not an option). The new beamsplitters should be ready in time for the next FTS runs. The reduced reference laser signal with the IR2 beamsplitters made switching between the beamsplitters difficult and required modifications to the reference signal circuitry.

A new "seismology mode", consisting of small scans (typically 16 steps) around a selected carriage position has been added..


Development of OASIS continues in Lyon. The CFHT Electronics and Optics Group will be heavily involved over the next 8 months to one year with the acceptance and delivery of OASIS instrument. Recently, scheduling issues, as they impact the CFHT staff and requests for observing time at the CFH telescope, have been discussed and agreed to by the development team from Observatoire de Lyon in France. Review of the final mechanical, optical and electronics design drawings has just started at CFHT. Over the next three months, members of the CFHT technical and scientific staff will generate an acceptance test plan, which will be used to accept OASIS as a "`Guest Instrument".


The cabling and connectors between the octagon and the control electronics rack were completely rebuilt to help with overall system reliability. As a means of reducing power consumption at the cassegrain environment controls for both the Fabry-Perot etalon and the cassegrain spectral lamps have been moved from CAMAC to the MOS / OSIS controller.

MOS / OSIS controller software has been upgraded to eliminate several problems and to provide better test capabilities. In particular, problems with system hangups during OSIS fast guiding should now be solved.

A pair of high resolution IR grisms for OSIS have been received and are undergoing tests. We have also received a replacement for the low resolution H grism which was originally made out of specification.

New light baffles have been fabricated for MOS and OSIS. These should be installed by the end of October. These baffles will hopefully solve light leak problems at the edges of the Prontor shutters.

The inside of the OSIS camera barrel has been painted a flat black in an attempt to reduce the amount of scattered light seen in OSIS.

MOS / OSIS setup has represented a heavy load to the technical staff. A great deal of this load is associated with interchanges of grisms between cassettes. A first step in freezing each of the grisms to a specific cassette was started this semester with a contract to DAO for the manufacture of 11 grism cells .

Finally, bi-prism focus capabilities for OSIS and the MOS ARGUS mode has been provided by a pair of new bi-prisms.

Coude f/4 (GECKO)

A larger 100 mm diameter iris shutter is being added at the detector environment to eliminate vignetting of the 4096 x 200 thinned UBC ccd. A custom driver mechanism for this shutter is being designed to keep power dissipation at the shutter low.

As with MOS / OSIS, the spectrograph's DUCK control software has been updated to improve system reliability.

Work on the fiber feed system is still on hold.


The mask stage motor controls have been significantly updated with a controller which permits rapid mask motion between slits and simultaneous motion of both axes. The new controller also supports cut-file generation straight from AutoCad files.

System safety has been significantly enhanced with the completion of a wrap-around laser shield.

Finally, work has started on a laser auto-focus mechanism to help improve cut precision and speeds.

Cassegrain Spectral Calibration (gumball)

The existing spectral calibration system available at cassegrain (gumball), will not meet the needs of OASIS when it arrives in 1997, and the existing control system is quite difficult to maintain. A project to "upgrade" the existing gumball to meet the needs of OASIS and be more maintainable is slowly getting started. A rough conceptual optical design to modify the existing lamp holder and a rough design of the control system have been generated. The detailed design phase for both the optics and control system are expected to occur in the last quarter of 1996 (time allowing). The time line for this project is being driven by the arrival date of OASIS.
Copyright © 1997, Canada-France-Hawaii Telescope Corp. All rights reserved.