Littrow Spectrograph?

[Merlin66]

The SimSpec Spreadsheet will help you.....

IMHO I'd go for a 30 x 30 mm grating.

The grating will probably have to rotate to access the full spectrum, you'll need a holder and some means of fine adjustment.

Ken

 

[DaveS]

Thanks, I only just remembered seeing 30 mm gratings in the product matrix.

Unfortunately I don't have office or other spreadsheet program on any of my computers. Just going to check Edmund for grating holders, as I misdoubt my engineering (non)ability.

[Merlin66]

I'm pretty sure there's a "freeware" spreadsheet program which reads Excel files...

https://articles.bplans.com/4-free-alternatives-to-microsoft-excel/

For the Spectra-L200 which uses 30 x 30 mm gratings, I used Optometrics (Now Dynasil) https://www.dynasil.com/company/optometrics/

I made interchangeable holders to allow different gratings to be used - 300/ 600/ 1200/ 1800 l/mm.

 

Edited November 15, 2019 by Merlin66

[Thalestris24]

2 hours ago, andrew s said:

It's the focal ratio that counts and the collimator .ens should match the focal ratio of the feed telescope. The camera lens needs to be faster to catch the divergent Ray's of the spectrum. The SIMSPEC design software gives the details.

Regards Andrew 

 

Sorry Andrew, didn't mean to be short before - was a bit distracted at the time. I think Simspec recommends slightly larger lenses but they are difficult to fit in physically. My collimator lens fl should match the slit distance ok. I've read (well, skimmed ) Philp Massey's paper on astronomical spectroscopy . The Lowspec design has given the collimator lens a diameter of 24.5mm and fl of 125mm - that gives it a focal ratio of 125/24.5 = ~f 4.9. However, the original collimator lens aperture apparently gave blurry images and the lens holder was changed to mask the lens down to ~12.75mm giving a fr of ~9.8 (f10 in round numbers, and I understand that worked better. I'm probably going to use my f6.96 apo so I'll probably have to change the mask to give a bigger diameter of about 18mm. It's all a learning curve and I'll have to see what works best with my setup. I'll be pleased if I can get it working at all!

Regards

Louise

[Merlin66]

If you use the spectrograph on an f7 scope, the beam after the slit will emerge at f7 towards the collimator. This acts as an "aperture stop" - if the beam f ratio is > collimator f ratio then the illuminated part of the collimator will be smaller without adding a mask.

I use a 30/200 (nominally f7) collimator which works (no masking) well on f8  to f12 scopes. 

 

[Thalestris24]

4 minutes ago, Merlin66 said:

If you use the spectrograph on an f7 scope, the beam after the slit will emerge at f7 towards the collimator. This acts as an "aperture stop" - if the beam f ratio is > collimator f ratio then the illuminated part of the collimator will be smaller without adding a mask.

I use a 30/200 (nominally f7) collimator which works (no masking) well on f8  to f12 scopes. 

 

I'm constrained by the Lowspec design and physical layout. I think Eric @ejp1684 found the aperture stop improved the image. Perhaps it was more to do with focus. When I get the lenses from surplus shed I'll hopefully have something to play around with. I've tried to fit a larger camera lens but there's no room really. Of course, I'll post my findings. I'll obviously do some basic tests without a scope first.

Louise

[DaveS]

Fortunately I'm not constrained by a pre existing design, so can design it around the optical elements I already have, or can readily source.

At the moment I'm looking at a cage arrangement using lengths of SS studding as a skeleton to hang the optics on. The SX flanges for their filter wheel look to be a good possibility for connecting the spectrometer to the telescope, and the imaging camera to the spectrometer. The guide camera may well end up within the cage.

[robin_astro]

The faster the optics the better they need to be.  That is why stopping down the lowspec (or using a higher f ratio telescope) gives better results with the simple achromatic doublets used there.  Dont forget to allow for the fact that the grating is at an angle when choosing the size to avoid vignetting (eg at high resolution where the angle get large, the LHIRES III uses a 25X50mm grating.

Compact designs with lightweight components ( ie high dispersion grating, small diameter short focal length lenses, the ALPY is an extreme example ) are easier to make mechanically rigid than using long focal lengths. Even a few microns movement is significant and can degrade the spectrum resolution in long exposures.

Robin

[DaveS]

Thanks Robin, some points to ponder. I just weighed the Ektanon and it comes in at 393g with its flange, a bit heavy I think. I do have a few achromatic doublets available, will do some calculations based on an f/7 telescope, remembering this is a littrow design, not the Lowspec, so my collimating lens needs to be bigger than the bare f/7 figure might suggest.

Wondering if I *can* get away with an achro doublet at f/7 without having to refocus, I suspect not.

[Merlin66]

I use a Edmund Optics achromat in the Spectra-L200 (similar to the LHiresIII) it's a 30/200.

The attached curve shows the required focus adjustment required over the visible spectrum. I use a Borg helical focuser.

 

[DaveS]

Thanks.

I may have to use the Ektanon, assuming it has better correction. I'm hoping to avoid having to refocus.

I've got some figures to push around, but ATM am in constant pain from what feels like sciatica so finding it difficult to concentrate on the maths.

[Thalestris24]

14 hours ago, Merlin66 said:

I use a Edmund Optics achromat in the Spectra-L200 (similar to the LHiresIII) it's a 30/200.

The attached curve shows the required focus adjustment required over the visible spectrum. I use a Borg helical focuser.

 

Hi Ken

I take it that curve applies to the camera - grating view? Do you manage to get away with not having to change focus within the visible spectrum? Or do you take exposures in sections?

Louise

[DaveS]

I did a few "sums" and it looks like a 300 l/mm grating in the first order will put the range from 400 to 700 mm on the ASI1600 sensor with the Ektanon. However I think this would be a waste of a fine lens, and a waste of sensor area. Am now looking at options for an ASI183 with its narrower, higher pixel count sensor with a theoretical resolution of 2,800.

I may reserve the Ektanon for something a bit wilder later, thinking Eschelle. Thorlabs do some 25 x 50 mm eschelles that look doable.

Of course, beyond even the Eschelle spectrograph we get into the real head-banging territory of the laser referenced interferometric spectrometer and FFT analysis. No, I'm not going there, least I come back in a canvas jacket with very long sleeves.

[Merlin66]

Louise,

Yes, it impacts on the spectral image. I use a hi res, narrow bandwidth, R=10000 Spectra-L200 and have to re-focus between the various target wavelength.

It's  very seldom that I add sections of the profile together...if you want a full spectrum spread, then a 300 l/mm grating or similar is used and set to the "average" focus.

I have seen examples where the camera focal plane is tilted across the length of the spectral image, but when compared to the curve above, it's still a compromise.

 

 

[Thalestris24]

5 minutes ago, Merlin66 said:

Louise,

Yes, it impacts on the spectral image. I use a hi res, narrow bandwidth, R=10000 Spectra-L200 and have to re-focus between the various target wavelength.

It's  very seldom that I add sections of the profile together...if you want a full spectrum spread, then a 300 l/mm grating or similar is used and set to the "average" focus.

I have seen examples where the camera focal plane is tilted across the length of the spectral image, but when compared to the curve above, it's still a compromise.

 

 

Hi Ken

Did you actually design the L200? I see your name appears on the sample spectra  

I wonder what you think of Christian Buil et al's Uvex3? I was thinking of having a go at building that originally - until Robin pointed out the lack of guiding in the basic design. However, knowing what I know now, I could incorporate the Lowspec reflective slit guiding  and modify the case to use the Uvex3 layout. But I doubt I'll be trying that any time soon! The Lowspec is coming along. If I get the lenses this week I could be able to do initial testing

Louise

 

 

[Merlin66]

Louise,

Yes, I designed and built the Spectra-L200.

I've also designed a FC 120 (folded Classical)  and an MG80 (80mm collimator/ 300 l/mm/ Zuiko 50 mm ) using standard components. Details on the forum.

I don't have access to 3D printing so I tend to skip these designs....I haven't looked at Christian's Uvex3 design.

Paul G. and I communicated on his early design concepts.

Ken

 

[DaveS]

On 16/11/2019 at 21:57, Merlin66 said:

I use a Edmund Optics achromat in the Spectra-L200 (similar to the LHiresIII) it's a 30/200.

The attached curve shows the required focus adjustment required over the visible spectrum. I use a Borg helical focuser.

 

This very clearly shows the "blue bloat" that is the bane of the imager's life, especially with less than perfect correction. since I wish to keep the design mechanically simple without having to either refocus or move the grating I will look again at off axis paraboloids.

And since this thread has wandered away from littrow designs into the lospec spectrograph I will leave it too.

Edited November 17, 2019 by DaveS

[DaveS]

Off to pursue my designs by myself.

[Merlin66]

Dave,

if you need help.....we're always here.