Pupil Imaging in CD15

While in pupil imaging mode, we scanned the position of the echelle parabaloid mirror to try to get a handle on where to move the pupil lenses to achieve better focus.

Conclusions: We see almost no difference in sharpness on the target wires over the mirror travel range for either lens.

 It is strange to see that one target wire almost 3x sharper than the pair of wires perpendicular to it.
Scaling to the pupil diameters, the sharp wire seems equally well-focused for ZnSe and KRS5. The blurry wires are imaged much more sharply by the ZnSe lens.

Here’s what we did:

The pupil target looked as follows for CD15:

Diameter and wire separations are to scale, (unlike the schematic drawings in post : http://codgermmck.blogspot.com/2012/12/cold-stop-target.html).

The resulting images at 1T on echelle look like:

ZnSe pupil image

KRS5 pupil image

Remarks:

The striations are due to the echelle grooves- they are perpendicular to the medium resolution dispersion direction.

The single wire in the echelle-echelon direction is clearly visible. Why are

the two parallel wires not?

The physical diameter of the pupil is 38mm. In pixels, the diameter is 168 pixels and 186 pixels, for ZnSe and KRS5 images, respectively. The separation between the center and clearly visible wire is  43 and 50 pixels, actually 10mm. 

The scale is therefore, on average: 

               ZnSe: 4.35 pix/mm

               KRS5: 4.95 pix/mm

In order to line things up along rows and columns, I rotated the images by 24.5 deg. 

Then I measured the Gaussian widths – to remove the profile of the pupil, I divided out a Gaussian function that approximated the profile of the pupil. 

Rotated pupil image with summed profiles in vertical and horizontal directions.

After the pupil profile was divided out, I fit a second Gaussian function to the wire features (just the central one for the blurry pair of wires.) The FWHM of that fit is shown below.
Then I scaled the FWHM by the pixel/mm scales in the second plot. The sharply imaged wire has the same width for both ZnSe and KRS5 lenses, but the blurry wires are much more sharply imaged by the ZnSe lens.

So what does it mean when the pupil image is astigmatic?
Also, it is not surprising that a motion of 14 T on the echelle parabaloid (x 2 = 9mm) would do so little to the focus of the lens. The lens has a f=38.1mm, and if the effective "object distance" was around 1000mm, then a change of 9mm would only change the image distance by a few hundredths of a mm.