"Second Light" Engineering Run for ShaneAO

First night saw light overhead cirrus and reasonably normal seeing; about 1 arc second. The goals of the first night were to finish a few NGS mode engineering tasks and start on the LGS ones, then to continue on with deeper exposures of the science targets we started on the first light run. In particular we were able to take another exposure of the open cluster M92, using an 11'th magnitude natural guide star. The guide star appeared dimmer than last time, probably due to the cloudiness. This gave us a chance to tune up on dim star AO performance:

AO with a dim guide star. This was after adjusting settings for low signal-to-noise. 

The second night was beset by clouds and bad seeing, and we were barely able to get even the brightest star to have any signal on the wavefront sensor. The third night was heavy fog and rain and we didn't even open the dome.

Tonight, Friday looks hopeful for clouds clearing by evening. The laser is turned on for the next three nights and the LGS mode AO controller is in place, so we hope to get our first LGS guided images by the end of the run (through Sunday night). Several LGS mode operations and systems need to be tested for the first time on sky, so we hope to get a jump on that starting tonight.


Engineering tasks completed:

• Connected and verified operation of the high-voltage driver for the laser uplink fast steering mirror
• Calibrated 22 more field-steering positions for the wavefront sensor, adding to the set of 4 cardinal points we got on the first light run. We're going to fit a model to these in order to enable arbitrary position offsetting of the natural guidestar.
• We unveiled the GUI for image sharpening. Image sharpening is a calibration process that peaks the PSF on the science image by adjusting the wavefront sensor's definition of "flat wavefront." This cancels the internal aberrations in the science camera. We tested the theory that the image-shapening offsets need to have a scale factor of d/r0 to account for the difference of starlight and internal source size as seen by the wavefront sensor, and this works. We verified with improved on-sky PSFs over no offsets and over not applying the scale factor.

Results of image sharpening

• Validated the dim guidestar AO tuning parameters and closed down to signal-to-noise = 2 on the wavefront sensor. This is an about 11'th magnitude star. Noise is completely dominated by "sky background" noise. We need to investigate now whether some of this "sky background" is in fact caused by glowing lights from electronics boxes and other powered items on the AO bench! If this accounts for say a factor of 2 in background it could make a really big difference in terms of sky coverage in NGS mode - almost a factor of 10 improvement in sky coverage per additional magnitude deeper on a statistical star count basis.