I'm stumped on this one.
I've constructed what is known as a monochromatic imaging spectrometer. It consists of a standard Czerny-Turner Monochromator, except that a lens in front of the entrance slit collimates the light from an object at its focal point and a lens at the exit slit refocuses the light onto a CCD at its focal point. The result is an image that consists only of light of a certain narrow wavelength range.
The monochromator is a McPherson 270, which looks similar to the one on
this page. The differences are that the McPherson monochromator has flat mirrors to turn the light so that the entrance and exit slits can be places on the sides, and that there are a few circular aperatures between the flat mirrors and the collimating mirrors.
Now, the problem... a beautiful monochromatic image appears on the CCD, but it is slightly stretched on the horizontal. The longer the wavelength, the more it is stretched. At zero order (when the grating simply reflects instead of selecting a wavelength) it isn't stretched. It obviously has something to do with the angle of the grating, which varies with wavelength. Perhaps something to do with differing pathlengths? Although I can't see how it would, I'm also kicking around the idea that it might have somethign to do with vignetting. I have tried a number of things, including using a different of the same type of monochromator, but haven't had any improvement.
The entrance lens has a focal length of 25 cm, the collimating mirrors have focal lengths of 35 cm (but the distances between them and the grating are actually only 34 cm) and the exit collimating lens has a focal length of 15 cm.
Any optics experts out there that might have an idea why this happens? I would really like to understand it and if possible correct it optically if possible instead of adjusting for it digitally afterwards. I'm stumped.
