20 September 2010 Night 4 (last night) of LBC Checkout
From the data we took on 17 and 18 Sept UT, I confirmed that there was a tilt in the focal plane of LBC-Blue: mostly left to right, such that the best focus on chip 3 was about 50 microns greater (M1 higher) than best focus on chip 1. Doug went up to the camera and tried to correct for this.
Opened a little past twilight - missed evening flats - because of the work on Blue.
We are focusing/collimating at elevation 46 deg. Rick is testing IMCS and we are waiting for temperatures to equilibrate. Seeing is horrible right now - pupils have no central hole. We hit a Y limit on DX. Temperature is 10.8 C, about 1.2 C cooler than when we implemented the model. But the limit was due to a large value in the PSF GUI active optics row - may not be a problem with collimation model, but instead crazy offsets.
Some bad pupil images: /newdata/lbcr.20100920.023817.fits and lbcr.20100920.024005.fits are Red pupil images where the algorithm locked onto noise, although there are many brighter pupils (if we can call these smudges pupils...). FPIA converged in 13 iterations and estimates seeing ~1.5 arcsec. We let it continue, despite poor seeing.
On to a higher field (Collinder399) to see if seeing is good or stable enough to allow us to take superfoc data. We are focusing/collimating and Red is taking a while. InformationTechnologyhere was coma in red (visible on the first in-foc image after the slew), and so far FPIA is not very effectively removing this. See images lbcr.20100920.030953.fits, lbcr.20100920.031146.fits and lbcr.20100920.031345.fits for examples. Blue on the other hand has already converged in only 2 tries. Last correction was huge: -5411 -5395 -1114 -1487 5090 -28 -32 y 0.00 3.58 /newdata/lbcr.20100920.031345.fits r-SLOAN
We will clear active optics on DX again (clear forces & active optics).
FPIA after clear AO & forces on Red: Pupils start out squashed - sort of like they have Z5. Rotator angle is 27 deg, the tip of the "A" points to 11 o'clock and elongation is from 7 o'clock to 1 o'clock. FPIA is giving large -Z5 and making pupils rounder. lbcr.20100920.032632.fits is an image where there are several poorly fit pupils.
1 sec exposure on this field. Collimated after 7 iterations, >~ 10 minutes. The images do not look good and there is a focus gradient across blue, but seeing is horrible - about 7px FWHM.
focus/collim again on this field. Now taking medCoarse (0.06 mm instead of 0.04 mm focus step) superfoc data at PA=0 --- seeing is still very bad, but RH is rising. 035237 ... 033059...
focus/collim again and take PA=180 medCoarse sequence. Analysis of both of these sequences indicates that the Blue FP is now flatter. Despite the poor seeing, fits were good and there is little variation (< 10 microns) in best-focus across the array.
Trying to focus/collim to take another 2 sequences. Stopped with Shift-S after blue was reported to have converged but red had not. Taking usual finer grid (0.04mm vs 0.06mm step in focus) superfoc now. 042751... for PA=0 and then PA=180. First look at these results are not so conclusive. Much better fits earlier. Field is setting and humidity is rising.
Converged after 6 iterations in poor seeing - but we are closing now as RH=94%.
10:12 Still closed. A note about adjusting CRPIX1,CRPIX2 in LBC headers.
Determine the New rotator center on Chip 2 and the delta_x, delta_y offsets from the Old rotator center.
delta_x = NewRCx - OldRCx
delta_y = NewRCy - OldRCy
NewCRPIX1, NewCRPIX2 can be computed from the Old values and delta_x, delta_y:
For chips 1-3:
NewCRPIX1 = CRPIX1 + delta_x
NewCRPIX2 = CRPIX2 + delta_y
For chip 4 (because it is rotated 90 degrees from chips 1,2,3):
NewCRPIX1 = CRPIX1 + delta_y
NewCRPIX2 = CRPIX2 - delta_x
When you enter these values into the LBC header templates, you should add +1 to all (chips1 through 4) CRPIX2 values.
10:30: Starting a series of 25 biases. Dome lights off. Two blue biases, e.g,105207 show horizontal step pattern (over all 4 chips). and initial red ones show some pickup noise, but later ones do not. NOTE at 110300 Steve is moving azimuth.... biases do not seem to change.
05:17: Turned off all but HouseKeeping. Running modified allseeing script in Sandbox, called skyseeing to determine IQ data and Sky Background count rates from 20100919 images.
Closing at 21:51Still to do:
Remeasure rotator center. Red was within 2 pixels of old value, but Blue was off by +8 in X (1032 -> 1040). Perhaps this was due to the tilt, which is now removed.
Run flat field scripts to check relative counts in all filters.
These are /home/LBTO/Calib_OBs/NewSKYFLAT
V4a cycles through 4 LBCR filters (V,R,I,r) with V-BESSEL in LBCB. exptime ratio is B:R 1:1
V4b cycles through 4 LBCR filters (i,z,Y,F) with V-BESSEL in LBCB. exptime ratio is B:R 1:1 except with F where it is 1:5.
RV6 cycles through all 6 LBCB filters with V-BESSEL in LBCR. exptime ratio is B:R 1:1 except 2:1 with U and S filters.
One exposure only, no dithers. Coordinates are set to go to dec=27, and about 15 min before LST at sunrise on 20100920.
If seeing is good and stable,
LBCB FP flatness: run finer focus step (0.04) superfoc sequence. But the medCoarse superfocus fits were good and confirmed focal plane flatness.
measure filter focus offsets.
Repeat a science type observation: North America Nebula is at 21h 44d and accessible during first half in September. Little Dumbell Nebula is available second half.