Cokin #40 Diffraction grating

[bojan]

Someone may be interested in this:

Cokin #40 diffraction filter (clear aperture 55mm, 240 lines/mm).
Way cheaper than Star Analyser !!!
http://cgi.ebay.com.au/ws/eBayISAPI.dll?ViewItem&ssPageName=STRK:MEWAX:IT&item=130315277127

The efficiency for the first order spectra is not great though, but enough (i guess) for the first stage of this project.
http://www.astrosurf.org/buil/us/spe1/spectro22.htm


I have just obtained this filter for $25 (together with postage, from another vendor on ebay), and it is bran new :-)

[Merlin66]

Certainly worth trialing. The published results show low efficiency but used as an objective prism, may still be OK. Look forward to seeing some results.
BTW Paton-Hawksley do a "standard" range of transmission gratings for about the same price.....

[robin_astro]

The cokin "filter" is very poor for spectroscopy (It is designed to give a little rainbow effect on photographs and as so most of the light goes stright through (only 5% into a useful spectrum) The "Jeulin" grating immediately below on Christian Buil's webs page is 8 times more efficient and is in fact the one manufactured by Paton Hawksley that the Star Analyser was developed from. As you say they can be bought cheaply from PH
http://www.patonhawksley.co.uk/nuffield.html
(model TE218b)
note the other l/mm gratings on this page are not particularly efficient unfortunately)
As you might expect, they are not as efficient or consistent in quality as the Star Analyser but they are much better than the Cokin filter and is what I used to start with.

Cheers
Robin
www.threehillsobservatory.co.uk (http://www.threehillsobservatory.co.uk)

[bojan]

Thank you for this info, Robin, I will certainly ask them for a quote :-)
Well..
If we are talking low cost, then even better option may be Edmund Scientific, here:
http://www.edmundoptics.com/onlinecatalog/displayproduct.cfm?productID=1490
For the same price you can get a pack of 15 gratings.. 12700 l/mm (possibly too much for general spectra, but great for line profile details.. )

[robin_astro]

Yes I looked at those a while back (they are quoted in lines per inch so 12700 is 500 l/mm) I asked them to give me some figures for efficiency like they do with their other gratings. The declined, so I assumed they were not particularly efficient. It would be interesting to know what they are like though.

For a simple grating in an uncollimated converging beam setup there is not really any advantage going to more than 100-200 lines/mm as the optical aberations of the configuration swamp any potential increase in resolution. The disadvantage with using a transmission grating in a collimated setup is the in line configuration means the spectrograph is rather long and rapidly become unweildly compared with a folded reflection grating design. Also the efficiency of transmission gratings falls off rapidly as you increase the number of lines (proportionally less gap for the light to pass through) This is why most spectrographs use reflection gratings. Now if only there was a source of cheap but efficient reflection gratings..:shrug:

Cheers
Robin

[bojan]

Robin, you are quite right about dispersion..
Last night I obtained my fist spectrum :D (used in front of 135mm lens)
I wanted it to be of Alpha Cen.. But, in a hurry I forgot to orient the grating, so I ended up with spectrum of Beta Centauri :P (outside the frame).
The spectrum was spread over almost quarter of the frame of my 400D, and 45 sec @ F4 exposure was about right for such a bright star (but it will be nowhere near enough for anything much fainter).
Also, it is contaminated with stars on its path.

Anyway, it was a good exercise :-)

The spectrum plot (Iris) is not calibrated, and axis marks are in pixels (blue to the right). However, some absorption lines are visible, especially in the blue region (where my un-moded 400D is more sensitive, btw) .
Next time it will be better..
Last night I ordered TE218b to try with brighter spectrum and lower spreading..

In the meantime, I will continue work on my version of Littrow.. This exercise is a good boost for it:thumbsup:

[robin_astro]

Yes, putting the grating in front of a camera lens works nicely as the light from the star is collimated. This way you can get a better resolution compared with putting the grating behind the telescope, but only for bright targets as you say. I wrote the method up for the AAVSO Epsilon Aurigae eclipse outreach project www.citizensky.org (http://www.citizensky.org) You can see the writeup on my website here.
http://www.threehillsobservatory.co.uk/astro/spectroscopy_11.htm
At the other end of the resolution scale here is the same object using my LHIRES III Littrow
http://www.threehillsobservatory.co.uk/astro/spectra_40.htm
Interestingly although the LHIRES is now sold commercially, all the drawings and design information are still on the web
http://www.astrosurf.com/thizy/lhires3/index-en.html

Good luck with your spectrograph building
Robin

[Terry B]

Good job.
:thumbsup:

[bojan]

Me again, this time with spectrum of Antares.
I still have to master calibration and that sort of stuff, but one thing at a time..
This spectrum was obtained by stacking 15 frames, 30 sec, ISO 1600, and using a screen to block unwanted star images..
The Cokin filter was placed in front of the lens.
Then I aimed the camera at Antares, and placed it in centre. After that I moved the camera 7.5° to the west, and placed the black screen 50mm in dia, 15cm away from the lens.. blocking everything in the direct field of view.. however Antares spectrum was not blocked by this arrangement, and was centred nicely in the frame.
The result is below.. much, much better :thumbsup:

[Merlin66]

I assume you used the 135mm lens as per previous?
It would be interesting to see how the camera responce comes out. My Canon lenses seem to have a "hump" in the red....
Can you upload a .fits file ex-IRIS to work through Vspec. ( Red to rhs please ;))

[bojan]

Yes, it was the same lens (Tair-11A, 135mm).
You want Fits saved from Iris, right? Or plot (*.dat)? Or separated channels?
The camera is standard 400D, not modified.

[Merlin66]

The .fits file from IRIS should be OK to use in Vspec. Upload what you have available, and we can see if we can assist with calibration etc.
I use a 350D for the same sort of imaging.
BTW I've just mounted a 7 deg 110mm diameter objective prism in a cell to fit the front of the ED80; weather permitting I'll trial it soon.

[robin_astro]

Ken,

Remember the instrument response is (roughly) the camera response * the grating efficiency curve * the response of the optics * atmospheric extinction. The first two terms normally dominate but unless you know the Cokin grating response you cannot get back to the camera + optics response.

Robin

[bojan]

OK, here you go... spectrum banner in fits format from IRIS.

I have not tried to use it in Vspecs yet..
I know that fits from DSS does not work.

[Merlin66]

The fits file needs to be converted to a 1D format in IRIS -Digital Photo "48 to 16bit" before import to Vspec.
I've just done the conversion...OK it goes into Vspec, but there's lots of noise...
I'll have a go at calibrating and post what I can.

[Merlin66]

Bojan,
Well I managed to get something.
Assuming your 200lpm at a distance of 135mm, the dispersion should be around 2.6 A/pixel.....
If I apply a x5 Gaussian smoothing filter to your spectrum, use the M1III comparison, (purple curve) and assume that the absorption lines out in the red are O2 and Atmosphere, I can get a result at 3A/ pixel....
I've also included a spectrum from the ETX website for reference.

[bojan]

Thanks for this, mate, it looks very promising now that I know it actually shows spectrum and not just coloured noise :-)

[Karls48]

I have been watching this tread with interest, as I would like to try spectroscopy without shelling out much of cash. As luck has it there are new Cokin filters currently available on eBay for$10. So I have bought one – but I don’t know if it is right one. It got written on it B.40 Cosmos France.

[bojan]

That's the one :thumbsup:
Good luck :-)

[Karls48]

Thanks Bojan. If anyone is interested, those Cokin gratings are still available from eBay for $A10 including postage. http://myworld.ebay.com.au/onlyfoolsandhorses51/

[Merlin66]

Karl,
Thanks for the link!
An ideal way of starting basic spectroscopy with minimal outlay.
Objective prisms (or in this case gratings) mounted on a CCD camera, with a telelens 100mm fl or so will give you results ( as per Bojan's excellent images)
Don't expect to discover new planets or measure red shift of 12 mag galaxies; the brighter stars will show their absorption lines and a bit of practise with Vspec will so get you "hooked" to try something more demanding.....