MOSS (Multi-beam Optical Seeing Sensor)

Dec 7 2023. 

Thinking about direct measurement of seeing right down the light path, no substitute for using the actual LATISS CCD and sending light down the optical axis. This measures exactly what we want. 

Figure from CWS MNRAS paper on strobed dome imaging: 

We could, in principle, use CBP for this on main telescope. So let's make one for AuxTel first. 

To get 1 arcsec diffraction limit we need about a 4 inch aperture, at 400nm. 

If system has 1m focal length, 1 arcsec is 5 micron diameter. How about this:

This is an f/7.8 RC system. 10" full aperture which with obscuration should give us about 4 inches for wedges. Assumes isoplanatic patch is at least 4 inches in extent, so we'll see about that. 
Ten micron pinhole is one arcsec. Back focal length is 24.2 cm though. We've faced this issue before and done OK. Manual: 53809_1_EN_Omegon_Ritchey_Chretien_Full_Tube_IM_REV_C.pdf
Obscuration is 110 mm for 254mm aperture so annulus of illumination is (254-110)/2= 72mm or about 3 inches. Close enough. So we need 3-inch dia optical wedges that deflect the beam by up to a few arcmin

We could fit 6 wedges of 3 inch dia into this annulus: 

What deflection do we want? For Auxtel the FOV is about 6 arcmin. From center to side it's 2000*0.1 arcsec = 200 arcsec = 3.33 arcmin. So a deflection of 2 arcmin seems decent. If we're 4m away from primary, spot is displaced by only 2mm. 
Problem with this config is beams are not very separated, spatially. We need a precision periscope on some to get more displacement. 

SEMESTER PLAN:

• end of Jan - conceptual design review complete
  • Tasks for Meghan
    • make a 4-5 slide slide deck where I make pictures describing the system (include part numbers, make list)
    • software architecture plan for controlling the mounts (can do during the lead time for hardware)
      • read control manual for whatever mount setup we order
      • If the adjustable hardware impeding timeline we abandon
    • Place beamsplitter order and put below deadlines on calendar

Checkpoints:

• Prototype one single collimated beam that we can steer around
• Optical source and control the light on/off (engineering milestones)
• Then, we scale this up to six steerable beams
• Then test these six beams on a telescope that we have HERE to test input (tight-beam demonstration)
• check in with telescope mounting after prof. stubbs trip (phone app that captures inside of telescope dome/Mario's Solidworks model)

Beamsplitter options, CWS Jan 26 2024

Alignment: 

CWS Jan 27 2024

If we use bounced-back beam from front face of beamsplitter cube, or else retro-reflected return beam, what displacement corresponds to what angle? 
Assume 50 cm distance from cube face to detector. If we want to align to 5 arcmin (size of a Rubin detector) then the angle is 5*60*5e-6 rad = 1.5 mrad. 
That is a spot displacement of 1.5E-3*50cm = 7.5 mm. So a detector 10mm on a side would be good for this. Thor labs 126MU camera is 14mm x 10mm. 

If we send the beam down to a corner cube and back again, that allows for in situ alignment, with a flip-into-beam beamsplitter. 

camera:

Camera software:

Flippers:

3/8 inch diameter micrometer mounting is compatible with Thor Labs Piezo PIAK10 actuators

This Newport stage is too. We'd need two in order to steer each beam. 

The other alternative is Thor labs small tip-tilt stage

After considerable looking around for an off-the-shelf way to control 6 different tip-tilt platforms, arrived at this from Newport:

Nope- inadequate force from actuators. Didn't order this. 

Acronym options

Multi-beam Optical Seeing Sensor- MOSS
Precise Emitter of Aligned Rays - PEAR
Parallel Aligned Ray Projector- PARP
Parallel Photon Projector- P-cubed
Laser Light Aligned Multibeam Apparatus-  LLAMA 

Feb 3 2024 CWS. Adding another beam expander, adapter hell: 

laser diode to beam expander mount, parts ordered Feb 5 2024: 

MOSS II