Design Pupil-Imaging Mode for SCExAO Path

[Jashcraf]

Ideally just a lens that can be dropped in to get an image of the pupil on the new detectors, but journal the design here.

[Jashcraf]

SCExAO as of 2019

There's one accessible pupil, so the collimating optic needs to come after that. The two lenses following are also being replaced with a telephoto lens configuration to slow the beam down to take advantage of the larger detector that is coming in, so the design may need to be modified to help with that.

Consider reaching out to Miles about the design?

First Thoughts

Just looking at the design, there should be a solution to throw a negative singlet between the two lenses or after the second lens to collimate the beam with a Gallilean telescope.

[Jashcraf]

Galilean Solution Between two Lenses

Start with a paraxial lens w/ negative focal length in between, variables are:

• Airspace in between
• Paraxial lens focal length

Optimization parameters are

• Pupil Position
• Wavefront Error

The optimizer can't come up with a solution that meets these requirements for a perfect singet. Two lenses also doesn't work terribly well. What about after the two lenses?

The pupil is after a focusing space, which is suboptimal for polarimetry where you really need a collimated space. The wavefront error is also quite bad because the pupil isn't in focus, so this isn't an ideal result.

Revised Thoughts

A Keplerian Telescope is necessary to re-orient the chief ray. Ask Miles for design updates and maybe you can collab on something that zooms from telephoto to afocal?

[Jashcraf]

Miles' Design

The COTS lens that's in the latest design has a nominal 150mm focal length but it 180mm from the pupil, so we don't have a lot of room to play with.

At first glance these ray angles look larger than the ones in the 2019 design file, maybe Julien's model didn't have all of the fields? Otherwise the achromat looks good, go Miles.

Diffraction limited spot size for an F/20 beam in the visible is ~12.2um, so this design is limited by lateral color at the edges of the field, but maybe that's fine for this application - especially if you use narrow-band filters and can steer with an upstream DM/FSM.

I guess the question is how much we care about the full field of view, or if we can just characterize the on-axis pupil. This would give us the aberrations from the Primary mirror

Questions to Ask

• Do we need to measure the telescope pupil for all fields of view?
  • If not, this is as simple as popping in a negative lens
• What are the free parameters of the design?
  • Space from pupil to detector
  • Walloston prism size
  • Filter Wheel size
• What's the budget for lenses & polarization optics? What do we already have?

Remember that you need to fit a lens, polarizer, and quarter waveplate in here. And the quarter waveplate needs to be able to rotate.

[Jashcraf]

So Mile's design won't be implemented this year. This means that we have to go with the current design, which implements a telephoto lens after the pupil stop (see Issue #6). For a simple gallilean configuration, I was curious as to if we can just translate the negative lens to get there. So I looked to the Gaussian reduction equation for thin lenses

If you solve this equation for t, you get

Which for the lenses above, equals a thickness of 165.2 mm. If this isn't too far from their current separation and we have the space, this is a very easy implementation. If you put the entrance pupil at the front focal point of the positive lens, the pupil is virtual. I wonder if placing it inside makes the pupil real?

I wrote some code that simulates the ABCD matrix of a generally telephoto lens in an afocal configuration and propagated a chief ray through it. The entrance pupil needs to be at a distance of ~1.2m from the telescope in order for the ray angle to be negative after the negative lens and produce a real exit pupil. So we need a way of relaying the pupil.

Reminder: Check the total distance we have between the pupil mask and the camera.

[Jashcraf]

It looks like the lens in the file Julien sent me were wrong, use the ones from Barnaby's Paper

updated design will be put in scexao-designs folder, but it needs some work. The two lenses in their current configuration are not presently focusing.