How to setup the ARGOS system to measure LUCI pointing offsets
The principle is to measure the LUCI pointing shift during a full rotator rotation while in closed loop on the pyramid in conditions as close as possible to the on-sky ones.
Since we are interested in the position of the PSF in the mask focal plane, a simple measurement of the PSF coordinates on LUCI detector is not enough as it will be affected by LUCI internal flexures: to account and correct for such flexures we perform differential measurements between the position of the PSF on LUCI (in pixels on the detector) and the position of the image of the N30 field spot mask that has several aligment holes.
The alignment holes are visible on LUCI detector if some dome light is on. In particular the Chamber/Visitor gallery lights seem to be the proper ones, when LUCI is configured with the N3.75 camera and FeII
The AO system is setup for loop closed using WhiteLightSource
only, without LGS sources. The system is configured using the ArgosArbitrator
forcing a reconstructor that corresponds to the FLAO bin 4 one.
A critical aspect is to find the proper illumination levels that allow to measure a non saturated spot on LUCI: experimenting with the WhiteLightSource
levels is typically needed with the caveat the below a critical threshold the shutter is closed completely and illumination drops to zero, crashing the AO loop. During our experiments, levels of 60 on SX and 82 on DX were found to be OK for both the AO loop and LUCI illumination.
Once the entire system is setup (loop closed and LUCI illumination such that star position and mask position can be properly measured) the rotator can be moved. To avoid breaking the loop we must leave enough time to the hexapod to offload the ASM tilt: for that purpose, the rotator speed must be reduced. A value of 1deg/s (instead of nominal 5deg/s) is ok.
Dave Thompson passed a script to Iskren that allows to take LUCI images and rotate the rotator in steps of 10deg.
- 08 May 2017