Regarding point 3, I propose 3 different experiments. I have indicated their priority in the text with ``minimal'', ``medium'' and ``larger''Total times assume following overheads => ccd reading overhead = 45 sec, dot-roach fiber movements = 30 min, cobra move = 60 sec. In case that dot-roach time can be reduced, some of the times could be significantly reduced. Also, I assume that for every experiment, we retake dot-roach. As such I hope these estimates are upper limits.

Total time needed for ``minimal'' : 145 minutes (2h 25 min)
Total time needed for ``medium'' (minimal included): 405 minutes (6h 45min)
Total time needed for ``larger'' (medium included): 16 hours

##### study misalignment
# daytime activity, but closed dome

# Order of activity
# select fibers that are avaliable in dcb and do not overlap
# These are:
# 2, 63, 192, 255, 339, 401, 464, 525, 587, 650
# 12, 32, 60, 161, 210, 223, 259, 289, 341, 360, 400, 418, 449, 497, 518, 545, 593, 621, 643

expose arc neon=140 duplicate=1 (one side of defocus)
expose arc neon=4 duplicate=2
expose arc neon=140 duplicate=1 (other side of defocus)

all together for one position of cobras = 140+140+4*2+45*4+60 = 440 seconds = 7.5 minutes

# Priorities
# minimal - at least 10 steps along phi or theta (145 minutes - assume 30 minutes for dot roach)
# medium - at least 10 steps along both phi and theta (210 minutes)
# larger - 10 steps in theta at different phi (e.g., at 3 different fixed values of phi) (360 minutes)

##### illumination of all fibers
# daytime activity, but closed dome

# Order of activity
# Select every 10th fiber - estimate 30 min

expose arc neon=140 duplicate=1 (one side of defocus)
expose arc neon=4 duplicate=2
expose arc neon=140 duplicate=1 (other side of defocus)

Priority - medium
3 runs to get good sampling = 195 minutes
Priority - larger
10 runs to get everything = 360 minutes

 ####### Take data that mimics dummy cable B data
# daytime activity, but closed dome

# Order of activity
# select fibers 2, 63, 192, 255, 339, 401, 464, 525, 587, 650
# run the command below
# then select fibers 12, 32, 60, 161, 210, 223, 259, 289, 341, 360, 400, 418, 449, 497, 518, 545, 593, 621, 643
# run commands

expose arc neon=340 duplicate=3 (one side of defocus)
expose arc neon=4 duplicate=6
expose arc neon=340 duplicate=3

expose arc argon=850 duplicate=3 (one side of defocus)
expose arc argon=10 duplicate=6
expose arc argon=850 duplicate=3 (other side of defocus)

Priority - larger
8300 second for one set of fibers + 30 min to select fibers = 2.8 hours
5.6 hours for both ``sets of fibers''

Early on, I think we should explore and understand what we can do during the day, under various conditions. The main things we can adjust are the room lights, the Cass cover, the mirror covers, the flat-field screen, and the choice of quartz lamps. The easy configurations are:

• With the dome dark and the cass cover open, we can do nearly all FPS work.
• With the dome dark, the mirror covers open, the FF screen in place, we can do much SPS work. To use the HSC lamps, we would need staff support, as I understand it.

Beyond those, ee should explore a couple of ways we might be able to run some procedures during the day. One background question is whether there are room light configurations which are acceptable for some staff work which also allow some useful PFS work. Just talking about "room lights on" vs. "dark dome" is, I believe, a bit simple.

The biggest questions are:

• can we do anything with the MCS with any room lights on? Specifically, is there is some room light configuration such that we can run convergence? Without this, it is hard to imagine doing anything with the SPS and cobras.
• can we do anything with the SPS with room lights on and the mirror covers closed? Specifically, whether the dot obscuration is such that we can run the dot-roach procedure. Alternately, can we do anything with the SPS with the room lights off and the mirror covers closed?

To explore this, one afternoon with a helpful observer?

Thoughts, @rhl?

Being a bit more specific about the dotroach datasets we want :

1. dotroach
   1. Take at least one dataset where we don't stop any cobras: eg we pass through the dot measuring fluxes with sm1, this will allow us to find the best/robust criteria for stopping the motion.
   2. Then we can play, we are thinking about a command :
   iic dotroaches hide=(all|odd|even|none|dcb|filename)
   we can try those different configuration and see how good we do

I checked the timing from last time :
sps expose took 30s in total (with exptime=10), reduction(20s), with some overhead around ~55s per step.
we can envision 20-30 steps so, 20-25 minutes per dataset.

Hassan and I talked about this ticket and how many hours are expected for day-time work. And the below is what we talked:

• I'm planning to split the allocated nights into daytime and nighttime work as follows, following the experiences from the past runs.  
  • In the past engineering runs, we started the observation at ~5pm and ended at ~6am (i.e. 13 hours for work, and 30min for preparation and cleanup each). 
  • The sunset and sunrise is ~6:50pm and ~5:45am, respectively in the middle of May. We can start opening the dome and observation ~+/- 45 min after/before that time.
  • So, roughly, 5pm-7:30pm in the evening and 5am-6am in the morning will be allocated for daytime work (or twilight work) which requires telescope movement and/or darkness, namely, dot-crossing with different EL, and isolated fibers. Note that we should do command flow test too in particular in the evening, so not all the time can be used for DRP.
• Since MCS installed on Cs this Friday (29 April), we can check the impact of the room light on the background. (I personally don't think it much since MCS is covered..) If the impact of the room light is small, we can carry out convergence test, dot crossing without moving telescope, before 5pm and after 6am as we discussed with Julien before. (Maybe ~11am to ~4pm on weekdays at maximum, when day crew and others usually works at the summit.)
  • On the first night, we can try dot-roach with the room light (and bright dome lamp) is on. If it works, we could test dot-roach during daytime, if it is allowed to keep the bright light. For example, you can prepare a given position and standby for data acquisition during twilight..
  • (We should discuss how many hours are needed for convergence test, but this is out of scope of this ticket)
  • The 2nd and 3rd day are weekend, so we should use some twilight/night time for cobra activity if we cannot keep cell cover open.
• I have a question to cloomis : Could you please tell the background why you need to test dot-crossing at different InR? Is it related coordinate transformation? (Dot rotates with PFI focal plane, so I'm confused..)

Added current working PU data acquisition activities PU_MayEngRun.md

Consider this closed for an past engineering run.