300 l/mm Objective grating trials

[Merlin66]

To improve (?) the results from the transmission grating - a la Uwe's tremendous efforts, I set up a 50mm Sq 300 l/mm Thorlabs blazed grating in a Cokin filter holder mounted on a Canon 85mm prime lens and a 1000D (modded) camera.
I used a 60mm finder as an acquisition/ guide scope with the camera off-set by 7.5 deg. Both mounted on a Vixen dovetail on the HEQ5 mount.

The FOV of the 85mm lens is 15 x 10 deg (orientated with the grating dispersion (2.2A/ pixel) along the Dec axis.) This allows the target star to be positioned on the right hand side of the image and the full spectrum recorded within the frame.
A 1 or 2 min exposure gives good results on the brighter stars.

Initial results: Hmmmm - honestly not as exciting as I hoped. Yes, you get good spectra of the field stars and any stars with H alpha emission (Del Sco) show up very well, the final resolution is limited by seeing. The major issue I found was the number of background stars "interfering" with the spectra - it's a lot of work analysing and cleaning the result.

One good "spin off" found, was the the image could still be plate solved (by Astrometry.net) notwithstanding the obvious spectra!

I'll continue the trials a bit longer to establish limiting magnitude v's spectral resolution......
May be do some comparison spectra with the Baader 207 l/mm (only 28mm aperture)

[Terry B]

Interesting.
When I used the little SA100 I also found that the background stars interfered. The simple fix was to mask half of the field and place the target near the edge of the mask so that the spectrum goes into the masked area with no background stars.
Not perfect but still good enough to detect emissions.


Terry

[Merlin66]

Terry,
I was thinking the same...
We show a similar mask idea in "Astronomical Spectroscopy for Amateurs" p182.

[Rob_K]

Good stuff Ken! A Thorlabs 300 l/mm 50mm sq grating is not much more expensive than the SA100 grating and might be the way to go when my SA100 finally succumbs to years of use and mis-treatment! Less vignetting when used over a camera lens than the SA100 (25mm diameter) for a start. Thanks Terry for the tip about masking too (also to Ken, must have missed it in the book!).

Cheers -

[Merlin66]

Another option may be to mount the Baader 207 grating off centre to the camera lens, rather than putting it in the "normal" central position???

[Merlin66]

Guys,
Covering 50% of the grating didn't work - wishful thinking!
When you think about it, the camera lens is still producing an image of the star field whether is 50% covered or not...

The "temporary" solution is to slightly tilt the grating/ camera to move the spectral image from the worst of the interfering background stars.

The aperture stop mentioned (Astronomical Spectroscopy for Amateurs, p 182) works with the grating "in the converging beam" and isolates the target star "image" (at the grating) from the background.....

I will however try the Baader grating for comparison.

[Merlin66]

Interesting observation...
The 85mm prime according to Astrometry.net gives a 15 x 10 deg FOV with the Canon 1000D.
While studying the FOV around Antares I notice there are spectra recorded from stars ( eps Sco and Mu1 and 2 Sco) obviously well outside this 15 x 10 deg field.
This infers that the coverage of the 85mm prime lens is close to 34 deg!!!!!

Just need a bigger sensor chip!

The attached screen dumps from CdC show the field of view and coverage.

[robin_astro]

Hi Ken

Try mounting the mask some distance in front of the lens so it can form an (out of focus) image. If you use two vertical masks you can effectively form an out of focus wide slit so the spectrum of the star is projected against a dark background and the spectrum formed by the sky to the left of the target is also removed. (arrange it so the star is in the unvignetted region and the star spectrum falls in the fully vignetted region)

Do not be tempted to use a hole aperture. (use a slit instead) A hole aperture makes it impossible to effectively subtract the sky background which becomes a low resolution spectrum of the sky when a mask is used.

You can see the effect of using a slit mask (in this case placed at a focal plane) here. (Venus in daylight)

http://www.threehillsobservatory.co.uk/astro/spectroscopy_4.htm

Cheers
Robin

[robin_astro]

Hi Ken,

The FOV should be as calculated for any given wavelength. It is just shifted to the right, relative to the zero order image, by the diffraction grating .

Cheers
Robin

[Merlin66]

Robin,
Yes. That’s normal to include the zero order and the diffraction spectrum.
In this case the spectrum shown in the image were generated by a zero well outside the FOV of the camera.

[robin_astro]

Hi Ken,

It does not mean though that you would neccessarily be able to see the zero order if the sensor was larger as it may be outside the unvignetted field of the lens.

Cheers
Robin

[Merlin66]

Yeah, true...
The object was to try to obtain a collection of "wide field" spectra and reasonable resolution in the single exposure.

The fast lens (f1.8) picks up the sky background quickly, it will be interesting to compare the results when using say f4 (just less than the Baader aperture)

[robin_astro]

The optimum grating dispersion v lens focal length is an interesting one. There are a few pluses and minuses to consider.

http://www.threehillsobservatory.co.uk/astro/spectroscopy_17.htm

Cheers
Robin

[Merlin66]

Robin,
Thanks for the reminder...

The TransSpec spreadsheet, see attached provides a useful guide. I've added the Thorlabs 300 grating for completeness.

[Merlin66]

Just a follow-up....
Being disappointed in the 300 l/mm results, I tried some exposures with the 200 l/mm (28mm aperture) mounted on the 85mm Prime and Modded Canon 1000D.

Similar results, the 30sec started to show excessive background and the 60 sec even worse.
Good result was the faint IR stars recorded in the image (some 12 IR stars > 10mag around Musca/ Crux.

I'll try a few things...
1. Change to the "grating in the converging beam" on the ED80 to allow lo-res spectra recording of nominated objects.
2. Put the 20 deg objective prism back on an ED80 for comparison trials.
3. Do some 30 to 60 sec exposures with just the 85mm Prime to see the possible worth as a "Nova search" set-up.
Onwards and Upwards.