2024 May 13

Plan is to reduce pulse width to suppress dome shake Fourier components, need to adjust ND filter accordingly.

images	Notes
13-16	Blocked each beam, one by one. Will have to re-do w/ re-alignment
25	pointing 89.1, 25.65, 0.1 s exposure at original pointing from nasymuth alignment
26	89.1, 25.9, .1s exposure continous beam
27	89.12, 25.9 1 ms pulse, 5s exposure (trying to center beams) <./pulses.py -n 1 -d 10 -f 100>
28	89.08, 25.9, same pulse/exposure
29	89.04, 25.95, same pulse/exp (found another beam!)
30	89.02, 26, same pulse/exp
31	89.02, 25.95 same pulse/exp
32	89.03, 25.9 same pulse/exp → moving back to side port to try to align a little better before transitioning
33	88.9, 25.8, .1s, continuous beam on <./pulses.py -n 1 -d 100 -f 300> key alignment takeaway: align at FOCUS of nasmyth port
34	same pointing, 5s exposure, 1ms pulse all four beams are in the frame! the PSFs don't seem ideal. we are going to see which beam is which, replicating tests 13-16
35-38	blocked beams, again from left to right (mirror closest to beam expander first, then left to right to far mirror. should be 4,1,2,3 in alignment image from above) the beams show up on image analysis, and are left to right on those images Elana wondering if they aren't on the focal plane at all and we are imaging the pupil - > we are going to do a focus sweep! Note that FOCUSZ is somewhere in the metadata If we are not in focus, it changes the plate scale - we are up the beam or down the beam path and the differential motion will not be as strong
39-138	Taking 100 images, 2s exposures 2s readouts, with 1ms pulses every 4 seconds <./pulses.py -n 300 -d 0.025 -f 0.25>
139-238	Taking 100 images, 2s exposures 2s readouts, with 1ms pulses every 2 seconds <./pulses.py -n 300 -d 0.05 -f 0.5>
239	Focus testing image: pointing 88.9, 25.8, 1ms pulse, 5s exposure The image seems extremely out of focus. Going to try to focus the telescope now.
240	Focus Sweep (all 1ms pulse, 5s exp) focusz = -0.025 (started from 0)
241	focusz = ~~-0.01~~ (= -0.035)
242	focusz = ~~+0.01~~ (= -0.025)
243	focusz = -0.01 # note these are relative offsets!!
244	focusz = +0.01
245	focusz = +0.04
246	focusz = +0.07 chris says make it more coarse
247	focusz = +0.12
248	focusz = +0.17
249	focusz = +0.08 (now testing around 0.07, which looks best so far)
250	focusz = +0.06 - chris brought up that we are not converging which might mean the mirror is not being given enough time to adjust
251	taking another at same focus offset to test this hypothesis - +0.06
252	focusz = +0.03 ; moving, waiting 10s for movement to conclude from here on
253	focusz = +0.00 ; moving, waiting 10s
254	focusz = -0.03
255	focusz = +0.07 (testing near initial point because it does seem better PSF in that direction)
256	focusz = +1.07
257	focusz = +1.07, CW, 0.1s exposure way saturated
258	focus = +2.07
259-263	biases because it's all very saturated
264	focus = +2.07, 5 OD now (up from 2.5), image is gone
265	focus = +1.07, 2.5 OD, pulse of 1 ms
266	focus = 2.07, 2.5 OD, 10 ms <./pulses.py -n 1 -d 10 -f 10>
267	focus = 1.57, 2.5 OD, 10 ms pulse
268-9	mirror refocus things
270	beginning to search for beams again. continuous exposure. pointing 88.9, 25.8
271	pointing 88.9, 25.9
272	pointing 88.9, 25.85
273-	biases
278	10 ms pulse, 5s exposure, pointing 88.9, 25.85
279-83	biases
284	1 ms pulse, 5s exposure, pointing 88.9, 25.82
285	1 ms pulse, 5s exposure, pointing 88.9, 25.87
286	1 ms pulse, 5s exposure, pointing 88.9, 25.89 - but did not seem to move... confusing. looking for fault in pointing – it did fault
287	same pointing as above, so that we can test for real
288	1 ms pulse, 5s exposure, pointing 88.9, 26 (testing a bigger pointing because telescope is faulting)
289	1 ms pulse, 5s exposure, pointing 88.9, 25.8
290	going back to 88.9, 25.85 to see if we spot it again - we do see it. trying to move down again
291	1 ms pulse, 5s exposure, pointing 88.9, 25.88 - matt and elana seeing tiny movement. trying a bigger step size.
292	1 ms pulse, 5s exposure, pointing 88.9, 25.95
293	same image as 292 with the lights off
294	starting to focus the telescope. we are going to offset ataos by +0.075, and then subtract 0.025 each subsequent time. we are still doing a 1 ms pulse with a 5s exposure need to enable correction before running
295	+.05 ataos fault when tried to run this. re-enabling need to take an image to see where ataos is now
	so it re-sets the ataos values after this fault. conclusion from this: we have no idea where the focus is. reaching out to someone more knowledgable. https://lsstc.slack.com/archives/C01K4M6R4AH/p1715647139373329 → need to reset the ATAOS offsets according to this channel there is a bigger issue and we have to reboot to change it.
296	88.9, 25.85 pointing, at ataos = 0 , 1ms pulse, 5 second exposure, initializing ataos sweep
297	z = 0.075, 1ms pulse
298	same thing - meghan and elana were talking and might have missed the first time this focus looks worse. moving on
299	+0.05 for z, same pulse duration and exposure
300	+.025
301	+0.0 (this should be a retake of 296)
302	-0.025
303	-0.05
304	+0.10
305-306	+0.15
307-316	ran 2s exposures w/ code 1ms pulses every 2s
317-416	same as above. PSFs look good - see notes below
417-1416	same as above, just more data
1417-1755	dome closed, same data as above
1756-2547	dome open, same data as above

End of Night Summary

Blocked beams so we had locations for previous data.
Re-aligned the telescope to the beams. Realized this was much easier if we aligned at the focal length of the nasmyth port.
Took 1ms pulse data (images 39-238)
Took images at many different focuses. Later realized that this is not really necessary with our system.
ATAOS fault and recovery
Re-aligned beams, did a smaller focus sweep
Took 1ms pulse data with dome open and dome closed (images 307-2547)

2024 May 14

Image Number	Notes
3-5	biases
6	pointing 88.9, 25.85; image continuous pulse, 0.1s exposure (beam finding)
7	pointing 88.9, 25.95, continuous pulse, 0.1s exposure (beam finding part 2 - meghan recorded 2 different heights last night)
8-11	blocking beams left to right as before
	biases
17-20	re-doing above test
	biases
27-30	same test, now w 5s exposures, 1ms pulses
31-34	same test, with 1ms pulses, at focusz=+.15
36-39	biases – beginning of nighttime
40	pointing 88.9, 25.95, continuous pulse, 0.1s exposure
41	pointing 88.9, 25.9 - trying to center beam → centered!
42	1ms pulse 5s exposure - came out super weirdly nan / nans show up when very saturated
43	took the same image - came out normal again
44-53	1 ms pulse, 2s exposure, 2s readout, 10 images. beginning of comparison of pulse lengths. OD 2.5 <$ python pulses.py -n 25 -d 0.05 -f 0.5>
54-63	10^-4 pulse, 2s exposure, 2s readout, 10 images. OD 2.5 counts are almost exactly an order of magnitude lower! beam shapes similar and sigma similar to 1 ms pulses <$ python pulses.py -n 40 -d 0.005 -f 0.5>
64-73, not including 72	10^-5 pulse, ZERO ND FILTERS, 2s exposure, 10 images it does seem like the beams change shape on this timescale. <$ python pulses.py -n 40 -d 0.0005 -f 0.5>
74-83	10^-4 pulse, 2s exposure, 2s readout, 10 images. r-band filter we were supposed to go up in magnitude from a ms
84-93	beam continuously on, 2s exposure, 2s readout, 10 images. r-band filter ghosty/angel streaks Chris: what makes it down and bounces back? <python pulses.py -n 1 -d 100 -f 120>
94-103	doing same thing to double check
104-113	filter z-band, beam continuously on 2s exposure we see streaks here too
114-123	filter now empty, y grating in we see super dim spots (max photons ~270-500) taking longer exposures to see what we see
124-128	y band with 30s exposures, continuous pulse we are weirded out by 124's max number of photons (47,000, where we were expecting ~3000); taking two more to have more data
129-130	see above
131-140	grating: quadnotch. 2s exposures, continuous pulse elana and chris think we are hitting the mask
141-150	g-band filter (we are right on the edge); 2s exposures, continuous pulse getting the white nan weird plots again on RubinTV elana checking USDF does seem pretty oversaturated in a way that taking 0.1s pulses would not help
Conclusion	Filters are not helpful
151-160	biases basically this is a persistence test- we oversaturated everything
	beginning large stacks of all pulse durations
161-201	10 ms pulse, 2s exposures, OD 2.5 we were going to take 100 but the images got super saturated going to try 10ms with g in <python pulses.py -n 25 -d 0.5 -f 0.5>
202-211 actually showed up as 202-213 skipped 208, 212	biases
214-223	g filter, 10ms pulse, 2s exposures, OD 2.5 seeing if this cuts some of the edge it did not. super saturated again
224-234	biases
	100 image stacks for smaller pulse lengths
235-334	0.1 ms pulse, 2s exposure, 2s readout, 100 images. OD 2.5 <$ python pulses.py -n 400 -d 0.005 -f 0.5>
335-374	0.01ms pulse, 2s exposure, 2s readout, 100 images. OD 2.5 FORGOT to take ND filter out <$ python pulses.py -n 400 -d 0.0005 -f 0.5>
375-474	0.01ms pulse, 2s exposure, 2s readout, 100 images. NO ND FILTER some blurriness/streakiness it seems not shocking - we saw the beams changing shape here before <$ python pulses.py -n 400 -d 0.0005 -f 0.5>

475-574	1 ms pulse, 2s exposure, 2s readout, 100 images. OD 2.5 <$ python pulses.py -n 400 -d 0.05 -f 0.5>
	DOME 2% OPEN HERE
575	image to make sure we are still aligned
576-675	slit open, 0.1ms pulse, 2s exposure, 2s readout, 100 images. OD 2.5 <$ python pulses.py -n 400 -d 0.005 -f 0.5>
676-775	slit open, 0.01 ms pulse, 2s exposure, 2s readout, 100 images. OD filter in for first image or two but then meghan flipped <$ python pulses.py -n 400 -d 0.005 -f 0.5>
776-875	1 ms pulse (ND filter back in) <$ python pulses.py -n 400 -d 0.05 -f 0.5>
	DOME 9% OPEN
876-925	1 ms pulse (ND filter back in) preliminarily, psfs as expected <$ python pulses.py -n 400 -d 0.05 -f 0.5>
926-975	.1 ms pulse (ND filter still in) this also looks as expected from pulling up a few of the image analyses <$ python pulses.py -n 400 -d 0.005 -f 0.5>
976-1025	.01 ms pulse (ND filter OUT) <$ python pulses.py -n 400 -d 0.0005 -f 0.5>
	ND Filter Testing - dome still open
1026-1125	10 ms pulses, OD 3.5 <python pulses.py -n 400 -d 0.5 -f 0.5>
1126-1225	100 ms pulses, OD 4.5 thinking that although its the same number of pixels we are getting large FWHM so pixels are being deposited on a much broader area this is why peak flux is so much lower try with dome closed at some point <python pulses.py -n 400 -d 5 -f 0.5>
1227-1230	beam blocks as before (4,1,2,3 in labeled piezo image below) OD 2.5


	data collection and physical next steps: ~~dome open - same plots~~ ~~ND filter testing~~ ~~block beams again~~ NEEDS TO HAPPEN TODAY IN THE DOME lock down flip mount set screw screw OD 1.5 to end of beam expander flip mount - different OD more images with new OD filters to confirm they work dome rotation
	remote next steps: define sine conventions for piezos - remotely plot flux rotator angle check angle on square plots

End of Night Summary

Filter testing
ND filter and pulse length testing
Beam blocks for tonight data
Pulse duration testing with changed ND filters for different amounts of dome open (0%, 2%, 9%)

2024 May 15

Note: first test is pointing tests, to make sure we understand what happens when the telescope moves in az and el.

Image Number

Notes

0 - 231

1 ms pulses, OD 2.5

el: 25.9, az: 88.9

232 - 266

1 ms pulses, OD 2.5

continuation of previous pulses

el: 25.9, az: 88.93

267 - 314

1 ms pulses, OD 2.5

continuation of previous pulses

el: 25.9, az: 88.87 - half off the image sometime

315 - 382

1 ms pulses, OD 2.5

continuation of previous pulses

el: 25.93, az: 88.9

383 - 1403

1 ms pulses, OD 2.5

continuation of previous pulses

el: 25.87, az: 88.9

1404 - 2852

dome open somewhere around 2440 - check on efds coronagraph

100 us pulses, OD 2.5

check saturation towards the end - ATCamera faulted

<$ python pulses.py -n 400 -d 0.005 -f 0.5>

el: 25.9, az: 88.9

biases from fault before in case thing was still integrating with shutter open - ATCamera faulted again

2853-2862

~~1 ms pulses, OD 2.5 in, 2s exposure, confirming that new ND mounting gives us the same results as before~~

failed but nothing faulted the first and second times Elana tried to run this. messaging Yijung.

< python pulses.py -n 40 -d 0.05 -f 0.5>

2863-2872

some sort of rubin ingest problem

biases

2873-2882

1 ms pulses, OD 2.5 in, 2s exposure, confirming that new ND mounting gives us the same results as before

psfs went wonky, we are unsure why
- flux counts are a little too low because of this
- update: we think putting on the ND filters moved the outer lens and changed collimation. we have corrected w/collimation
will test collimation

2883-2892

0.1ms pulses, OD 1.5 in, same test as previous

2886

moved be05 collimation one tick down - 100 microsecond pulse

this made it worse. we are at end of range so switching to collimate other beam expander after moving this back

2887

tai215259

moved be20 collimation one down, same pulse

2888

two up "

2889

one more up "

2890

back down 1 "

2891

1/2 down "

this is the best collimation improvement, seemed to make a BIG difference

2892

1 up- still.5 offset

worse than other 1/2 offset so set to 2891

2893

back to position of 2891 - looks good!

2894-2897

beam blocks across (same procedure as before, left to right)

didn't see all four images immediately, but then they came through. rubintv is slow tonight

2898-3825

stack of <2000 images, 0.1ms pulse, 2s exposures, OD 3.5

NOTE

telescope pointed at slit is 52 degrees

8pm

preparations to go on-sky.

- beginning to correct on-sky pointing

3826

ran correct_pointing script, which failed

could be it didn't find a star

3827

another attempt at pointing correction

3828

pointing: 60 el, 52 az; taking an on-sky image to see what we get. g-filter in. exposure time 5s

MIRROR COVERS WERE CLOSED - opening now

3829

30s exposure 60 el, 52 az

expect streaks
MIRROR COVERS WERE CLOSED - opening now

3830

60 el, 52 az; g-filter in. exposure time 30s

if successful, we turn on tracking, turn on another image, and follow a star up/down
got a dim streaked star!! turning on tracking
focus?

3831

trying track_target/find_target, then have to take_image_latiss

60 el, 52 az; mag_limit = 12, radius = 1 deg, ignore ATDOME, ATDOMETrajectory

did not find a catalogue object so we will see what we get
took an image, but it took an image of the inside of the dome (azimuth moved super, so we got no target)
pointing out the slit and seeing what we get
pointing in alt/az and having it track
eric says we dont have a script set up
- do track target and do a wide range

3832

tried decreasing the mag_limit in the script to 8 and doing the same thing as above

we got stars!! HD 113127

3833-3842

we are trying to run focus sweep - latiss wep align

images labeled focus offset or curvature wavefront sensing

got some donuts
script finished but failed to find a focus

3843

taking an image at what curvature wavefront sensing thinks at the best point

3844-3852

FOCUS SWEEP - 60 degrees (compromised with no filter)

ATAOS fault when we originally set the focus - things are not syncing though

image number	focus-z	FWHM (arcsecs)
3844	-0.075	3 2.9
3845	-0.05	~~3.6~~ 2.6
3846	-0.05	2.5
3847	-0.025	2.6
3848	-0.01	2.8
3849	0
3850	+0.01
3851	+0.025
3852	+0.05
stopped at 0.05	~~+0.075~~

3853

compromised - no filter

az: 52, el: 80, taking image, set z to zero

fwhm 2.98
we think we have a pointing offset and it puts the star it finds in the same spot each time, field is not same

3854

compromised - no filter

az: 52, el: 40

fwhm 3.35

3855

compromised - no filter

az: 52, el: 20

3856-3905

stopped tracking, pointing to MOSS (25.9, 88.9)

turned MOSS on 100 microsecond pulses: <python pulses.py -n 6000 -d 0.005 -f 0.5>

taking stack of 50 images

points are a little out of the field - we want to repoint for the next round.
chris thinks we need more zenith angles to do what we are having - every 10 degrees from 80 to 20

3906

beginning of sweep every 10 degrees

doing a find_target every time

need to make stars fainter for this to work
out of focus - getting astigmatism

az: 52, el: 80

3907

az: 52, el: 70

star saturated

3908

az: 52, el: 60

3909

az: 52, el: 50

saturated again

3910

az: 52, el: 40

3911

az: 52, el: 30

3912

az: 52, el: 20

3913-3962

pointing to MOSS (25.93, 88.9) and with beams more centered (go elana!)

100 microsecond pulses

50 more images

we will try to focus again next

3963

pointing az: 52, el: 60, g band filter

Elana re-enabled corrections and we will compare - from eric message
on previous sweeps did not do a filter, so hopefully this will be ok
g filter is in for this one
the astigmatism could have been from lack of filter
Elana: we could write a block to do this sweep down the sky in the future - not worth doing on the fly though

3964

az: 52, el: 80

this one is more astygmatic
chris:
- particular arc that goes right over obs.
- if we are tracking that star, right when it gets to the zenith, the azimuth must instantly spin around and go back to the other side
- required rotation rate singular at this point
- tracking problem in azimuth at this point?

3965

az: 52, el: 70

3966

az: 52, el: 50

3967

az: 52, el: 40

3968

az: 52, el: 30

3969

az: 52, el: 20

3970-4019

pointing to MOSS (25.93, 88.9) and with beams more centered (go elana!)

100 microsecond pulses

50 more images

4020

az: 52, el: 70

4021

az: 52, el: 60

4022

az: 52, el: 50

4023

az: 52, el: 40

4024

az: 52, el: 30

4025

az: 52, el: 20

4026-4075

pointing to MOSS (25.93, 88.9)

100 microsecond pulses

50 more images

4076

az: 52, el: 70

4077

az: 52, el: 60

4078

az: 52, el: 50

4079

az: 52, el: 40

4080

az: 52, el: 30

4081

az: 52, el: 20

4082-4574

taking a bunch (2000) of MOSS images. Elana and Yijung are going to open the vents to see if seeing improves. will timestamp when that happens and image number

10:37pm - opened vent gate, turned on fan

4157 image number

4575

az: 52, el: 70

note: mount motion image degradation

4576

az: 52, el: 60

note: mount motion image degradation

4577

az: 52, el: 50

4578

az: 52, el: 40

note: mount motion image degradation

4579

az: 52, el: 30

4580

az: 52, el: 20

4581-4680

MOSS 100 images

4681

az: 52, el: 70

note: mount motion image degradation

4682

az: 52, el: 60

note: mount motion image degradation

4683

az: 52, el: 50

4684

az: 52, el: 40

4685

az: 52, el: 30

4686

az: 52, el: 20

4687-4714

MOSS run but it was too far off the frame of the image. moving to new pointing

4715-4814

repointed to 25.96, 88.87, 100 images

4815

az: 52, el: 70

note: mount motion image degradation

4816

az: 52, el: 60

note: mount motion image degradation

4817

az: 52, el: 50

note: mount motion image degradation

4818

az: 52, el: 40

4819

az: 52, el: 30

4820

az: 52, el: 20

4821-4859

MOSS at repointed angle same everything else

need to re-repoint - beams are off again

4861-4872

25.99, 88.87

still needs more repointing

wind seems strong outside according to Yijung
actual wind speed check ~13-15m/s

4873-4884

26.02, 88.87

4885-4905

26.05, 88.87

4906-4917

26.08, 88.87

4918-5076

26.06, 88.87

Yijung going to close the vents

5077

az: 52, el: 70

note: mount motion image degradation

5078

az: 52, el: 60

note: mount motion image degradation

5079

vents closed

az: 52, el: 50

5080

az: 52, el: 40

5081

az: 52, el: 30

5082

az: 52, el: 20

5083-5097

26.06, 88.87

tons of mount degradation on first image

5098-5134

26.09, 88.87

beams traveled completely across the field of view
we are shutting the slit. we believe wind is causing this

5135-5145

MOSS with dome closed. initially nothing on images

5146-5156

returning to 25.9,88.9

- new images: MOSS in frame

moving up again to 25.93, 88.9 (pre-wind coordinates)

5157-5169

25.93, 88.9

still close to the edge, but beam position is much more stable

5170-5196

25.96, 88.9

seems centered, but still moving around a little bit

5197-5297

25.99, 88.9

last stack of 100 images

leave dome away from scaffolding!!

End of Night Summary

ATCamera faulted during the day. Seems they exceeded the number of images they can have on the stack. We should put this in the night log.
Mounted and tested new ND filter setup. Adjusted collimation of system accordingly, and re-blocked beams.
Took stack of ~2000 images, stopped when went in dome for on-sky preparations.
Prepared to go on-sky with a super bright star; somewhat cloudy.
We ended up going on-sky, and were able to take runs of comparisons up and down the sky in zenith angle interspersed with stacks of MOSS images!
Took comparison data till around 1:15 am when it got too windy for the MOSS beams and common mode motion exceeded the limit where we could image them consistently.
Moved the dome - worked! MOSS seems secure up there.

stubbslab

MOSS Observing Log

2024 May 13

End of Night Summary

2024 May 14

End of Night Summary

2024 May 15

End of Night Summary