Lab alignment procedure: Difference between revisions

From Monnier Group Research Wiki
Jump to navigationJump to search
← Older edit
No edit summary
No edit summary
 
(9 intermediate revisions by 3 users not shown)
Line 1: Line 1:
'''1. make sure the input collimated beam is parallel to table:'''
'''1. make sure the input collimated beam is parallel to the table:'''
1a. put an iris in front of the beam to shrink the laser beam size to ~1mm<br />
1b. iteratively place the same target at two places along the beam, the two places have to be far apart for high precision
and the connection of the two places has to be parallel to the table<br />
1c. adjust the tiptilt of the mirror  to make sure the beam hit the same position of the target at the two places. The last
mirror is for fine adjustment for target at farther position, and the second to last mirror is for coarse adjustment for the target
at closer position.<br />


2. make sure the home-made front metal panel is perpendicular to the beam.
It is relatively easy to mount the camera parallel to the optical table, so I think it is useful to align the beam parallel to the table too. I also find it is useful to mount the input beam source on a linear stage so that when you align the beam parallel to the table, you can
easily move the beam around to shoot it into the camera.
 
Depending on the detailed setup, the alignment procedure could be different. Usually it is useful to have a mirror or the source
that can be steered in tiptilt. Then a common alignment procedure will be to use the tiptilt control to align the same target close to and far away from the beam source. The procedure takes several iteration steps.
 
'''2. make sure the home-made front metal panel is perpendicular to the beam.'''


2a. assemble the metal pieces, no lens or camera.<br />  
2a. assemble the metal pieces, no lens or camera.<br />  
2b. open the iris to be about 5 mm<br />  
2b. keep the beam size as small as possible (<~ 1mm) <br />  
2c. use the edge of a rectangular card to block half of the beam, and let the other half pass through<br />
2c. put in the hole-made cylindrical target into the S mount on front metal panel, and move the beam source to be aligned to
2d. put a flat mirror tightly against the front metal panel (I used one hand to hold and push the mirror) to reflect the passed-through
the target. This step can be rough.
half beam<br />  
2d. install the lens tube with flat mirror (edmund, #45598, #63951) to the S mount on the front metal panel <br />  
2e. loose the four screws that tight the metal baseplate to the linear xyz stage. Rotate the whole metal piece so that the reflected half beam
2e. loose the four screws that tight the metal baseplate to the linear xyz stage (newport ultra 562). Rotate the whole metal piece so that the reflected beam matches the incoming beam <br />
matches the blocked half beam horizontally (meaning they become a circle, but it is not true if they are not vertically aligned)<br />  
 
2f. If the reflected beam and blocked beam don't match vertically, we can put some shims between the triangles panels and the front metal panel.
'''3. align the beam with the camera'''
But I didn't find it necessary this time.<br />  
 
3a. attach the camera to the front metal panel<br />
3b. install the first reimaging lens, you should be able to see a spot on the detector. Then loose
the screws on the camera, try to move the camera to put the spot at the center
of the final readout area. the readout size should start with full frame so that you can see where the spot is on the detector. Note: we are in crop mode, so if you see
light in a subarray readout, this doesn't mean the light is really in the subarray area. That's why you want to start out with full frame. <br />
3c. Then take out the lens, you will be see a pupil on the detector. try to move the linear xyz stage of the camera to move the center of the pupil to match
where the spot was. <br />  
3d. put on the lens again to check if the spot has moved. If it has moved a little, iterate the above steps to get the spot match the center of the pupil.
Here is my understanding of how this step works.
[[File: WFS_lens_misalign_handdrawing.JPG]]
 
'''5. align the reimaging lens pair with the camera.'''


3. align the center of the beam to the center of the threaded hole of the front metal panel (rough)
5a. Then put on the second lens, you will see the pupil again on the detector. The center of the pupil should
be close to  the spot (< 5 pixels). If not, there must be some misalignment in the previous steps.<br />


3a. put a second iris close to the camera (theoretically you can first iris, but it is too far away to reach), and beam size small ~1mm. And fully
'''6. align the lenslet array'''
open the first iris.<br />
3b. put the cylindrical target into the hole of the front metal panel. Note, the target doesn't tightly fit into the hole, it can be shaken, that's why
this step is rough.<br />
3c. move the second iris to put the laser beam at the center of the target.<br />


4. align the beam with the camera
6a. install the lenslet array, and you should be able to see many spots from the lenslets.<br />
6b. un-tighten the screws holds the rotation stage to move the rotation stage so that the spots-of-interest are not on the edge of the detector readout array. Usually I will keep at least 3 pixels away from the edge.<br />


4a. attach the camera to the front metal panel<br />
'''7. align the collimated lens'''
4b. switch the light source from laser to white light, and turn on the camera, set exposure to be 0.01s, and keep the second iris small.<br />
4c. the readout size should start with full frame so that you can see where the spot is on the detector. Note: we are in crop mode, so if you see
light in a subarray readout, this doesn't mean the light is really in the subarray area. That's why you want to start out with full frame. Then loose
the screws on the camera, and move the camera to put  the light roughly at the right area (the left up corner is (0,0), ideally we want the center of the light
to  be at (45,45)). When the light is roughly at the center of the 90X90 array, tight the screws.<br />


5. align the reimaging lens pair with the camera.
7a. The device (optosigma, # 111-1310) that holds the collimated lens is not very  secure. It is good enough for lab test, but probably not enough for on sky. So use glue to attach the lens to the optosigma device. <br />
7b. The device has a long arm, and doesn't have the adjustment of fine rotation. The question is how do I know if the collimated lens is parallel to the lenslet array. The method I used is to perfect vertical reference like a ruler, and a piece of very thin paper like lens paper. Then push the ruler against the device and use the paper to feel the friction on the top and bottom part of the touched places. If you feel strong friction on both part, then the lens is placed vertically and parallel to the lenslet array.


5a. there are three images needs to be aligned: the pupil image without any lens in the front;
[[File: alignment_of_collimated_lens1_2013Oct20.jpg]]
the image with one lense in the front; the pupil image with two lens in the front. If they are not overlapped,
I think it is because the optical axis of the lens is not aligned with that of the pupil, as shown in the attached figure.<br />


5b. use the actuators on the linear xyz stage to move the images on the detector. The two pupil images
[[File: alignment_of_collimated_lens2_2013Oct20.jpg]]
move in the opposite way if you turn the actuators in the same direction. This confirms my understand of the misalignment.
So it is possible to align these two images. You just need to iterate a couple of times.<br />
5c. In this case when the two pupil images are align, the image of the focus seems pretty well aligned automatically. If it is not aligned, I haven't
thought of how to fix it.<br />

Latest revision as of 19:13, 20 October 2013

1. make sure the input collimated beam is parallel to the table:

It is relatively easy to mount the camera parallel to the optical table, so I think it is useful to align the beam parallel to the table too. I also find it is useful to mount the input beam source on a linear stage so that when you align the beam parallel to the table, you can easily move the beam around to shoot it into the camera.

Depending on the detailed setup, the alignment procedure could be different. Usually it is useful to have a mirror or the source that can be steered in tiptilt. Then a common alignment procedure will be to use the tiptilt control to align the same target close to and far away from the beam source. The procedure takes several iteration steps.

2. make sure the home-made front metal panel is perpendicular to the beam.

2a. assemble the metal pieces, no lens or camera.
2b. keep the beam size as small as possible (<~ 1mm)
2c. put in the hole-made cylindrical target into the S mount on front metal panel, and move the beam source to be aligned to the target. This step can be rough. 2d. install the lens tube with flat mirror (edmund, #45598, #63951) to the S mount on the front metal panel
2e. loose the four screws that tight the metal baseplate to the linear xyz stage (newport ultra 562). Rotate the whole metal piece so that the reflected beam matches the incoming beam

3. align the beam with the camera

3a. attach the camera to the front metal panel
3b. install the first reimaging lens, you should be able to see a spot on the detector. Then loose the screws on the camera, try to move the camera to put the spot at the center of the final readout area. the readout size should start with full frame so that you can see where the spot is on the detector. Note: we are in crop mode, so if you see light in a subarray readout, this doesn't mean the light is really in the subarray area. That's why you want to start out with full frame.
3c. Then take out the lens, you will be see a pupil on the detector. try to move the linear xyz stage of the camera to move the center of the pupil to match where the spot was.
3d. put on the lens again to check if the spot has moved. If it has moved a little, iterate the above steps to get the spot match the center of the pupil. Here is my understanding of how this step works.

5. align the reimaging lens pair with the camera.

5a. Then put on the second lens, you will see the pupil again on the detector. The center of the pupil should be close to the spot (< 5 pixels). If not, there must be some misalignment in the previous steps.

6. align the lenslet array

6a. install the lenslet array, and you should be able to see many spots from the lenslets.
6b. un-tighten the screws holds the rotation stage to move the rotation stage so that the spots-of-interest are not on the edge of the detector readout array. Usually I will keep at least 3 pixels away from the edge.

7. align the collimated lens

7a. The device (optosigma, # 111-1310) that holds the collimated lens is not very secure. It is good enough for lab test, but probably not enough for on sky. So use glue to attach the lens to the optosigma device.
7b. The device has a long arm, and doesn't have the adjustment of fine rotation. The question is how do I know if the collimated lens is parallel to the lenslet array. The method I used is to perfect vertical reference like a ruler, and a piece of very thin paper like lens paper. Then push the ruler against the device and use the paper to feel the friction on the top and bottom part of the touched places. If you feel strong friction on both part, then the lens is placed vertically and parallel to the lenslet array.

Retrieved from "https://monnier.astro.lsa.umich.edu/research/index.php?title=Lab_alignment_procedure&oldid=3191"