Troubleshooting guide

We typically use alignment star boxes that are 24 – 30 pixels wide. It is very important that this
parameter be set to within a few pixels of the actual values.
2. rot_4m
This is listed as the 4-m rotation offset, and it should directly correspond to the instrument position
angle. We have experimentally determined that it works for position angle of 90 degrees:
If ROT_PA = 90: set rot_4m = -90.
Running xbox causes a graphics terminal to pop up. It steps through the boxes, in the order listed in the box
file, and presents a cross-cut profile for the x and y directions, along with a red graphics cursor. It will show the
estimated location of the box along the bottom of each cut, along with a horizontal threshold level through the
cut, and a vertical line indicating the estimated location of the peak of the alignment target; the first time through
there likely will be no targets in the boxes, and the line will be centered on a noise peak.
You can adjust the position of this vertical line, say if xbox detects the wrong peak, once the alignment targets
are in their boxes by moving the cursor to the correct location and hitting the f key. Hit q after looking at the
fits for every box, and the final graphical presentation of the estimated adjustments. Afterwards, xbox will print
out a lengthy set of values used for computing the offsets. The final bit of this indicates how much to move the
telescope and rotator. Here is example output from xbox for a plate with ROT_PA = 90 (and rot_4m = -90):
====================================================
*** MOVE TELESCOPE/ROTATOR by the following offsets:
Offset PA by -0.06 (0.01) degree
Offsets: -4.38" (1.09) 30.60" (0.21)
====================================================
The above translational offsets are printed in the order (δ RA δ Dec); also shown are estimated errors.
Translational offsets smaller than 0.5 pixel = 0.16" may be ignored; rotational offsets down to 0.01 degree
probably should be done.
Here is a step-by-step list of how to align the MOS plate:
1. Set instrument PA before acquiring target.
This must be done first, as the telescope must be at zenith in order to move the rotator.
For safety reasons, have the telescope operator make the rotation from the Cassegrain Cage.
Run config.mos.dither.kp4m.pl, and set the parameter ROT_PA to the new value. Now the
WCS compass arrows will display correctly in ds9.
2. Verify that the xbox parameter rot_4m is correctly set for the PA.
3. Have telescope operator verify pointing on the North Port TV Camera.
Use an SAO star or a Fixed Bright Star near to target field.
There should be a grease pencil mark on the TV monitor indicating where to zero the telescope
pointing.
Take a quick image (2-5 seconds) on FLAMINGOS, and make certain the SAO star is well centered
on the array. If is not, center the star with relative.offset.kpno.pl, then take another image.
Once it is well centered, Z the telescope (cf. § III Nightly Startup Tasks: B. Startup on the sky).
If the target field is too faint, or does not have any stars, consider adjusting focus on this pointing
check field (cf. 5 & 6, below).
4. Load the mask design image and region file into buffer 3 of ds9.
FLAMINGOS@4-m, Ver. 2.34, 2006 Apr 04 Page 25 of 44