WPO Littrow Spectrograph- first light

[Merlin66]

Finally got the pick off mirror in the right position to allow imaging with the ToUcam. With a 100mm collimator/ camera and a narrow slit ( more on that in a moment!) managed to get my first images of the solar spectrum to check resolution etc.

The results exceeded all my expectations!!!! I'm actually getting 0.43 Angstoms/ pixel resolution!!!!!! I had to download the detailed specta from Kitt Peak to find a comparison. The Na lines are wide apart and you can see the Fe line between them. All this without even connecting it to the telescope, just bright sunlight into the slit.

The attached photo is a stacked 16 images, no corrections with the Kitt Peak comparison spectra just below. The Mg bands etc.

My effort with razor blades for the slit was so-so, it needs to be quite far into the 1 1/4 adaptor so the size of the "body" is limited to about 29mm diam. Need to find a better solution before the next trials.

[janoskiss]

Nice spectral lines Merlin. Can you explain a bit more what we're looking at? Along the horizontal is obviously wavelength as apparent from the colour change from left-to-right. But what is the quantity that is changing along the vertical?

[RAJAH235]

Steve, they are the spectral lines of the various elements stated.
By taking readings at diff times, they can tell if it's red or blue shifted too. I think that's how it's done. :D L.
ps. all the same elements on earth. eg; sodium/mercury vapour lamps etc
They set up a control lights spectrum on one side/edge of the plate & the object/stars spectrum on the other & compare them.

[Merlin66]

The horizontal lines are the dust and edge of the razor slit, the vertical lines are due to the various elements in the sun.

[ballaratdragons]

Steve, The full spectrum looks like a rainbow and the colours relate to elements. Wherever the vertical lines cross the spectrum tells you what chemicals/elements are in the star/object.

Here is a site with a fairly simple explanation. http://outreach.atnf.csiro.au/education/senior/astrophysics/spectrographs.html

[janoskiss]

Then you could stack the data along the vertical direction as well to get better signal to noise.

[janoskiss]

Here it is stacked along the vertical, i.e., just blurred completely in the vertical direction, plus a final level adjustment to improve contrast. (hope you don't mind me playing around with your pic...) You could get more detail by stacking with realignment.

[ballaratdragons]

That cleaned it up Steve! Tricky fellow aren't you!!!!

[Stu]

Yeh the vertical lines are from the razor blade. I can get you a mechanical variable slit mechanisim that you can pull apart and fit to get a cleaner edge, I have to look for it though.

I always throw these things out just before someone needs them...

No actually I have an actual slit wheel with a set of 8 or so laser cut slits in it at different widths. That's much easier to use. I'll put it in the "for Ken" box.

[ballaratdragons]

Gee thanks Stu, but what do I want it for? Merlin is the one doing the Spectral Imaging :lol:

One of the (now deceased) members at out Ballarat Society had a Spectrum Analysis EP and it was magic. Bung it in and see what each star is made of! If I remember, he paid a massive amount of dollars for it 20 years ago.

The patterns were pretty too :P

[Merlin66]

Janoskiss, that was a tremendous improvement, many thanks! I'm using PSP7 for other stuff ( like painting Spitfire models for FS!) but never saw the blurr "vertical" will have a look in a minute.

Any help you can give me Stu, would be appreciated VERY much!!!

[janoskiss]

I use the Gimp mostly, but PS should be the same. There is no "blur vertical" as such, I just go to Filters->Blur->Gaussian Blur. Set "Vertical" to something very large, say 1000 (this effectively averages all pixels values along the vertical) and "Horizontal" to 0 (no blurring along horizontal). Motion blur is another good one, and you can adjust the blur angle if the spectral lines are not quite vertical.

Actually I just tried it and motion blur gives sharper lines in the Gimp. Needs to be applied repeatedly (max length is 256 pixels) until all detail along the vertical is completely blurred out. See attached pic.

Edit: replaced pic. First one wasn't cropped properly before blurring and some of the comparison inset got mixed in.

[Kieken]

Hey, that's cool :). Can you use it also to examin the spectral lines of stars and DSOs?

[Merlin66]

The design should be capable of getting usable spectra from 8-10 mag stars and DSO's. Maurice Gavin has been using it to get spectra of Supernova!!

http://www.astroman.fsnet.co.uk/newspec.htm

When I get the slit sorted out and improve the camera I anticipate trying it out on stars, double stars etc. Keep you posted.

[Robert_T]

Wow Merlin this is fantastic stuff, real astro science. Can't wait to see your results on stars. This thread is also a microcosm of the IIS cooperative spirit - already we have Steve providing a means to clean up the spectra and Stu digging out of his kit bag for a better "razors edge". Great stuff guys :thumbsup:

[Merlin66]

After the initial trials I ended up re-doing the pick off mirror support ( the most fiddly bit of building this design... believe me!) I then found that the grating had to be inclined by about 1.5 degree to "throw" the return beam onto position (ToUcam/ 6mm Mono eyepiece). Now done and everything back to working order. Have re-done a couple of slits ( If Stu can me help it would be great!!!!!) one very narrow ( 10-15 micron?) and one wider ( 0.5/ 0.7mm) for stellar work.
Won't get the chance to try it before the weekend, but keep you posted.

[Stu]

I'll try and drop of the stuff before I go to perth on friday.

Hey I asked around about the grazing angle! The grazing angle is the angle at which the grating is the most efficient, ie. reflects the most light. At the moment you are using a grating that has a blazing angle at 240~280nm I think (in the UV range anyway). So if you find the you don't get enough light in the red, then you can change the grating to one that has the same number of lines per millimeter (resolution) but has a higher blazing angle. The exact type that I happened to have accidentially not thrown out this week. :)

Unfortunately I don't know what to do about the resolution problem you are having. I found the old dual sided 600/1800 lines/mm grating I was looking for a month ago but they are to big to fit in your spectroscope and they are quite badly scratched. :(I think one of them was even drawn on by someone. :confuse3: :scared:

[Merlin66]

Hope I can catch you before you head off. Looking forward!!!

[Stu]

Slight correction it's not "grazing" angle it's blaze angle. :o

[[1ponders]]

Merlin, I've been following this discussion with real interest. Unfortunately I'm a bit lost as to how your spectroscope works/is put together. Would you be able to include a bit of a diagram or a link to a webpage that shows what it is you're doing and a bit of an explanation on how it works.

[Merlin66]

The link in the earlier post takes you to M. Gavin's page. I'm using his WPO design which is a very basic spectroscope. It uses a reflection grating (30mm sq, 600/1200 lines/mm), a standard tele lens ( I use a 135mm f2.8) and a small mirror to "pick-off" light from the telescope. I'm using a ToUcam as a camera ( hopefully to upgrade SOOOON to a MX7c)
In the telescope adaptor I have a narrow slit at the focus of the telescope and also at the focus of the tele lens this gives a "cleaner" spectra.
Depending on the grating and lens you get differing resolutions which allows more detail to be seen in the spectra.
I use IRIS to manipulate the raw image and spectraview to prepare a 2D graph which allows better analysis.
With a 10" telescope should be able to get usable spectra of 9 mag stars etc etc.
I'll post a photo of my set-up at the weekend.

[Merlin66]

The attached photos show the set up. The connector to the telescope with the same pick off mirror, the 135 telelens and the grating which I mounted in a plastic project box. The grating box connects to the lens by using step-up filter rings.