Three-port Image Motion Monitor

Overview

By placing a mask with apertures over the input pupil of a telescope, we constrain the light that hits the focal plane to only the cylinders of light that pass through the holes. Seeing makes the angle-of-arrival of those wavefront differ, even if the light comes from a point source at infinity. 

In order to separate the images that are made on the focal plane. optical wedges are placed in the apertures. That introduces a static angular offset that separates the images in the focal plane. For typical glasses, an optical wedge with an angle theta will deflect the beam by roughly  theta/2. We want that deflection angle to be big enough to separate the PSFs, i.e. more than tens of arcseconds, but considerably smaller than the field of view of the camera, so that all images land on the focal plane. That means in practice we want a deflection of a few minutes of arc. 

We could buy custom wedges that do this, but that's expensive and takes a long time. An alternative is to use a pair of off-the-shelf wedges and configure them as a Risley prism pair. IF the two are rotated so the deflections add, it's maximal. If they are rotated to cancel, the beam is undeflected. Rotation in between lets us dial in any desired beam deflection angle. 

Thor labs makes two inch optical wedges, but they have larger wedges angles than we want. But by using them in pairs we can tweak it to obtain any desired deflection up to some maximum. 

So the efficient way to do what we want is to use off the shelf Thor Labs SM2 (two inch) optical components. That gives each aperture enough area to get decent photon collection, and we don't need to buy expensive optics. 

Making optical masks with SM2 threads is therefore sensible. I had the shop make mask disks with three SM2-threaded holes. The idea is then to put an appropriately rotated pair of wedges in each of them (actually only two need wedges but that's a detail), and mount the mask on an optical bench that also holds one of our 150mm aperture telescopes. 

For measuring atmospheric seeing, we would put this on a telescope mount. 

For measuring horizontal seeing over, say, 1 km path length we'd focus the telescope on an LED or laser diode source far away. 

For measuring nearby seeing, we'd put a point source at the focus of an identical optic (as a collimator) and point them at each other. The collimator makes plane waves and the masked receiver telescope would be focused at infinity. 

Requisite Sensitivity

Astronomical seeing is typically 0.5-1.5 arcseconds. 

Implementation

If we use one of the Omegon Ritchey-Chretien Pro RC 154/1370 OTA telescopes, we need an adapter to a camera. This device converts from 2-inch eyepiece tube to T-threads:
T adapter from telescope

That will bolt onto a T-thread to Canon EOS bayonet adapter, for example, allowing DSLR camera to be mounted. 

Design

1) Transmitter

Assume we make a source of collimated light from a matching OTA. Typical laser diode has few-mrad divergence and 2 mm sized beam. If beam is expanded to 150mm then angular divergence is down by 150/2 = 75X

That means emerging full-aperture collimated beam has divergence of 3 arcsec X (divergence/1 mrad). 

So a pulsed, collimated laser diode isn't a bad source to use, except for speckle issues. 

Note we will be diffraction limited by DIMM aperture. 2 inch diameter at 650 nm gives theta=lambda/D = 2.6 arcsec FWHM. 

If we use a telescope eyepiece plus telescope as a beam expander, the angular magnification is the ratio of telescope to eyepiece focal lengths. For a 10mm eyepiece and 1370mm OTA focal length (Omegon RC telescope) then magnification is 137X. 

But we also need to match the f-number of the telescope. Typical collimated laser diode circular beam size is 3.5mm. OTA f-number is f/D=1370/154=f/8.9. So that means we need an 'eyepiece' with focal length of 8.9*3=26.6 mm. 

One good option that would do this with no chromatic aberration is (yet again) an off-axis parabolic mirror. Thor labs makes one with 1 inch focal length- MDP-119-01, one inch diameter. The one inch version has a 90-degree cage mount.

Fluxes- if laser diode output is a few mW, imagine factor of two loss from input to output (two reflections, etc), so conservatively assume 1mW of output power. If that impinges on receiver input mask with 10% mask-to-detector throughput, we get 100uW of incident optical power, divided into three separate PSFs. At optical wavelengths 4W of optical power is 10^19 photons/sec. So we'll get around (30E-6/4)*1e19 photons/sec/PSF = 7.5e13 photons/sec/PSF. A 1 millisec strobed source would give 7.5e10 photons per PSF. 

2) Receiver

Three-hole mask in front of Omegon OTA, with 2-inches wedges in the beam for 2 (or all 3?) of them. Use Canon monochromatic camera at its focus, with no lens. iPad remote data collection from that device.