As mentioned in our last post on "more fun with grisms", our next incremental improvement would be to try to improve the optics using darkroom enlarger lenses (versus standard Pentax SLR lenses). Our logic for trying this is that while the old M42 style lenses are very good general purpose lenses, enlarger lenses are specialised in a number of ways that are probably favourable to spectroscopy. First of all, they are made to project a flat image onto another flat image (ie a film negative onto photographic paper). Essentially in a spectroscope we are trying to project an image of a slit onto a CCD. Second, very high resolution is needed; ideally the silver grain in the film negative needs to be resolved. In many normal SLR lenses, a "soft look" is often considered desirable. Third, enlarger lenses must control all types of chromatic aberrations over a very broad range of wavelengths. In fact, printer paper is quite sensitive to blue and not very sensitive to red (hence the red darkroom lights) so in theory, they should be designed to work effectively towards 400nm.
In fact, these special properties of enlarging lenses have in the last few years made them very popular as "macro lenses". A lot of the nice insect and plant macro photos you see are made using enlarger lenses. We've had some experience using enlarger lenses for macrophotography so we had the required extension tubes, bellows and some lenses around. Here is our previous "grism module" upgraded with a Rodenstock Rodagon 80mm f4 enlarger lens:
- rodagon
- rodagon.jpg (63.19 KiB) Viewed 1993 times
The Rodagons are some of the best enlarger lenses ever made. They typically have 6 elements and even the non-Apo models are quite Apo by astronomy standards. Their lenses are very highly curved so it's unusual to see an enlarger lens (or a true marco lens) with an aperture larger than 25mm. Here is out test setup, this time using a quite high-end monochrome camera:
- grism setup
- grism-enlarger.jpg (117.39 KiB) Viewed 1993 times
The camera lens is a Nikon El-Nikkor 63mm f2.8. This is also one of the classic enlarger lenses that are really popular these days for macro picture taking. They are so popular for macro that they are now almost impossible to find. It turns out the focal length and aperture (22.5mm) is just right for the size of grism we are using (600 l/mm and 15mm diameter from Paton Hawksley). The 80mm collimator is optimised for about an f6 telescope but in this case we are not using any objective telescope in order to just test the spectroscope optics. Here is a solar spectrum taken of direct sunlight on a 10 micron slit using the Atik 460EX:
- spectrum
- spectra_013.jpg (130.89 KiB) Viewed 1993 times
As you can see, when oriented diagonally, we can just squeeze in the whole visible spectrum. But what I find extraordinary is that over this quite broad spectrum (and over a quite large CCD chip) the spectral lines are all about equally sharp! Usually, using normal camera optics, if you focus on the green, the red and blue will be not in focus at all. In the following picture I've split the spectrum in two pieces and done some Photoshop:
- spectrum split
- spectra_013_disp.jpg (156.81 KiB) Viewed 1993 times
Again, what stands out to me, compared to other experimental setups I've tried, is that the whole spectrum is sharp and in focus. Believe me, that is usually not the case. In the next image, I've cut out some of the spectral lines and presented the pixels 1:1 (the previous images were reduced by about a factor of two for presentation):
- mosaic
- spectra_013_mos_disp.jpg (107.36 KiB) Viewed 1993 times
It looks to me that the resolution is about constant across the whole spectrum from below Ca-K to above H-alpha. The region near 390nm is noisier (since there are fewer photons) but the resolution does not look very different than at longer wavelengths. The previous photographic lenses we used never gave as good resolution below 400nm as above 500nm.
The theoretical resolving power of this setup is 2,100 and the resolution is 2.6 angstroms (which looks to be achieved). The end goal of building this device was for night-time astronomical observation. But I think the setup could also be refined to make a useful spectroheliograph. It is for this reason I think the good performance of the enlarging lenses near Ca-K (and the uniformity across the whole solar spectrum) is interesting.
