white pupil back end for L200 or LHires III lspectrographs?

[Hamish Barker]

I was thinking about the possibility to increase the visible range of wavelengths and increase resolution for my L200 without changing the grating. I have access to a C14, which with its long focal length of 3800mm or so, gives quite a big star image circa 50 microns at 3 arcsecond seeing, and therefore needs a corresponding large slit width for efficiency. But a big slit reduces resolution.

Has anyone ever configured these littrow type instruments with a secondary collimator and additional camera lens in a white pupil configuration?

The secondary collimator would be positioned after the primary focal plane. It would be imaging the incident collimated beam which lands on the grating, reimaging it onto the entrance pupil of the final camera lens.

For example my L200 spectrograph has a 200mm lens in littrow configuration with 600lpmm grating. I am proposing a 100mm focal length secondary collimator, positioned 100mm after the normal focal plane. This would make it about 350mm from the grating, and would recollimate the diverging monochromatic beams, but also redirect them back to a so called white pupil 140mm after the secondary collimator. Then a 50mm camera lens could be positioned at this white pupil.

It is two additional optics (secondary collimator and final camera lens), but what would be the gain? The slit would be re-imaged at a reduction of 50mm/100mm, closer to the resolution of the ccd. Because it is cheaper to make a larger secondary collimator than to get a huge size ccd, the un-vignetted spectral range available in one image can be larger. The flexibility to change the camera lens means potential to adapt the instrument to different focal length telescopes.

The downsides would presumably be more reflection losses, more aberrations (which could lose more resolution than the potential gain from the reimaging size reduction of the slit) and a longer, heavier instrument, potentially more vulnerable to mechanical flexure (and consequent calibration inaccuracy).

I have the 100mm and 50mm lenses (100mm is an old slide projector f2.8 lens, the 50 is f1.8 pentax lens), and everything can be on-axis so is mechanically simple to construct with T2 (42mm thread) tubes. I will try it out and post any results.

Has anyone tried this before?

[Hamish Barker]

I redrew it on paper and this doesn't look useful. The dispersion of the beams means that by the time they reach the secondary collimator at 350mm from the grating, for 3800-7000A range, 11.2degrees of dispersion gives 68mm plus the 100mm/f11= 78mm secondary collimated beam diameter if sized for no vignetting. i.e. f1.28 very expensive and likely to have lots of aberrations.

So although the advantage of reduced slit image size would still be there, the difficult specification required of the secondary collimator seem to outweigh it.

I guess white pupil schemes are only useful for echelle spectrographs.

I originally had the idea because looking through a 32mm eyepiece I could see almost the entire visual spectrum on the L200 and figured that if I could see it, a camera could too. Looking more closely, it seems that most of the range is somewhat vignetted.

[Olivier GARDE]

I guess white pupil schemes are only useful for echelle spectrographs.

Here's an example of a white pupil design with this spectrograph and a resolution R=30000.
https://www.shelyak.com/le-woppshel-un- ... n/?lang=en

This spectrograph use 2 greats optics (FSQ106 Takahashi) in a symmetrical way.

To use it with a telescope bigger than 400mm in diameter.

[Hamish Barker]

is the purpose of the second collimator to re-image the incoming collimated light which strikes the echelle grating, reimaging it onto the camera entrance pupil, or onto the cross-dispersing prisms? (i.e. is the white pupil being used to minimise the cross disperser size, or the camera f ratio? What is the path length from the exit from the second refractor until the prisms and to the camera entrance pupil? It would be great to see the optical / mathematical analysis of the whoppshel light path. Perhaps you did publish a paper on it already?

[Ken Harrison]

Hamish,
An interesting concept....
With the 200 mm fl collimator to be honest the “sweet spot” at focus is around 12mm.
Outside this range the spectrum gradually goes out of focus with a loss of resolution.
You can see this when a DSLR is used......

[Hamish Barker]

yeah I guess so. I read the manual and noted the 12mm sweet spot limit. Maybe the benefit of white pupil is anyway a thing for bigger instruments where the required detector size is bigger than available detectors. I wonder if correction elements downstream could expand the in-focus range. Do you know if the sweet spot is limited by chromatic aberration limit or by other off-axis aberrations (coma/astigmatism/spherical?)?

[Robin Leadbeater]

Hi Hamish,

Do you know if the sweet spot is limited by chromatic aberration limit or by other off-axis aberrations (coma/astigmatism/spherical?)?

Christian's ray traces and practical measurements for the LHIRES suggest the main aberration is astigmatism, producing an elongated image. This is not a problem for resolution at the centre of the field but this becomes slanted at the edges of the field, reducing resolution
http://www.astrosurf.com/buil/lhires_eval/optique.htm

Cheers
Robin

[Robin Leadbeater]

Christian suggests there that this slant might be corrected for in processing but I dont know anyone who has tried this (Note this is different from the normal slant correction due to the slit not being orthogonal to the dispersion direction. This effect is local to the star spectrum and varies across the field)

Cheers
Robin