The Lowspec spectrometer

[Merlin66]

Oleg,

The small flip mirror designed into the Spectra-L200 directed a reference beam from the external lamp source (connected to a 12V supply) to the slit gap.

The lamp has a diffuser and there's a 6mm hole in the housing to let the light through.

Opposite this hole is a 10 x 10mm mirror mounted on a support shaft. The mirror holder has two small magnet inserts. When the mirror is swung out of the beam it's held against the inside of the housing by one of the magnets.

When it's flipped down the centre of the mirror sits immediately over the centre of the slit gap at 45 degrees to the reference beam, and again held in place by a magnet.

This provides a reference beam wide enough, centred on the middle of the slit height, to illuminate the whole height of the slit.

These are some images of the original prototype.

[Valerio]

Hi

I finally finished my lowspec assembly and tested it. First light seems promising. I tried with a normal lamp in front of the opening where the telescope should be.

This is the first order spectrum with a 600 lines grating.

After that I searched for the central maximum and I found it to the right (micrometer completely pushed in)

but when I try to use a longer exposure time the result is this

As you can see besides the slit I see around it a strange image, that seems the border of the hole in the plastic of the piece holding the slit disk

this is the piece photographed in normal light

this is the piece photographed with a light that reflects on the inside of the hole, that shows the orange part around the slit

Is this normal? I understand that I was using a very bright light and probably with a star I cannot see this border, but I wanted to know if this happens to you too...

Thank you!

Valerio

 

[Paul Gerlach]

Valerio,

Are you sure you've placed the slit disc in the correct orientation? Meaning that the mirror surface is facing the inside of the spectroscope. If not, it could be that the bright light is being reflected between the two surfaces of the slit disc.

[Valerio]

the disk I bought is perfectly the same in both sides. Mirror on boh surfaces

[Merlin66]

Valerio,

On first glance that's how it appears, but it is only chrome coated on one surface. This surface is where the slit gaps are etched through the coating. This should face as Paul says towards the inside of the spectroscope.

If you look carefully you'll see the difference.

Ken

 

[Merlin66]

Valerio,

I wouldn't worry too much about the "halo" - I have seen similar when high gain and very bright light sources are used. Will not impact on the usual target objects.

(You should also flip your spectrum to bring the blue edge over to the left hand side. This is how all spectra are subsequently measured and calibrated. I've slant corrected and calibrated your spectrum using BASS Project. resolution, R=319)

[Valerio]

Thanks Ken, very helpful. I have not yet installed BASS Project and will do soon. I must say that the image I posted was a screen grab with cel phone so probably I will have a better resolution with the camera image (ASI294MC PRO) :-))
I will also flip my grating to have the spectrum with blue on left as you say. I will check again the slit disk but since I had difficulties in assembling it in the plastic holder I don't know if I can remove it now --- hahaha --- without breaking it or bending it, but if you say that this probably is not a problem I will try this way.

I will post soon another test with full resolution after these changes

Question: is it better to use the 1st order spectrum (brighter) or a 2nd order (darker but more spread out?). It depends on the source?

Thanks

Valerio

 

[Merlin66]

Valerio,

The spectrograph is designed to use the blazed 1st order spectrum.

(The angle of rotation for the 2nd order is twice as great....just use the bright 1st order.)

If you want a higher resolution use a grating with more l/mm.

 

[Valerio]

5 hours ago, Merlin66 said:

Valerio,

The spectrograph is designed to use the blazed 1st order spectrum.

(The angle of rotation for the 2nd order is twice as great....just use the bright 1st order.)

If you want a higher resolution use a grating with more l/mm.

 

Thank you Ken, always very kind and precise!

[Valerio]

Hi,

this is a second test with the inverted grating and full resolution capture.  I used a RELCO SC480 lamp

What do you think about collimation and focus?

I still have to start using BASS Project and of course try to connect the telescope (Celestron C6 f/10) with some real tests

Thank you for any feedback

 

Valerio

 

[Merlin66]

Valerio,

Pretty good result!

Using your spectral image, correcting for slant and calibrating it against the RELCO lamp in BASS Project

The dispersion is 0.76 A/pixel

The resolution

R= 949 @ 4764 A (5.02 A FWHM)

R= 821 @ 6962 A (8.4 A FWHM)

The Ne line at 5852 A is very wide (This is usually bright enough to over expose?) and the focus is better towards the blue.

You're on your way! I'd certainly set it on the scope for further trials.....

Onwards and Upwards.

 

Edited September 26, 2020 by Merlin66

[Merlin66]

Valerio,

I'm a little confused....

The SimSpec spreadsheet I have for the 3D print, based on 125mm Collimator/ 80mm camera lens and a 600 l/mm grating coupled with your ASI 294 (4144 pixel x 4.63 micron pixel) should give a dispersion of 0.95 A/pixel and a resolution (30 micron slit gap) of R=983.

This doesn't seem to match well with the results.

On 15/09/2020 at 20:23, Valerio said:

I will also flip my grating to have the spectrum with blue on left as you say.

The way to address the spectral blue-red direction would be to either rotate the camera by 180 deg or just flip the spectral image prior to processing.

I think by flipping the grating you've ended up imaging on the -1st order rather than the +1st order. The difference being the anamorphic factor which would change from 0.91 to 1.1 this impacts on the dispersion.

The dispersion would drop to 0.87 which is closer to your results.

(See "Astronomical Spectroscopy for Amateurs", p150-152)

I would reverse the grating to the original positioning and confirm the slit gap used.

 

 

SimSpec V4.3_Paul3D.xls

[Valerio]

Ken,

thanks for your great analysis. I flipped the grating and tried again with Relco lamp and a blue day sky. The slit is 30um. I tried to keep the central part in focus. What's the better way to obtain a full spectrum in focus? Should I take more images rotating the grating with the micrometer and then compose them together?

Are the dark horizontal lines a problem?

 

Valerio

 

[Merlin66]

Valerio,

Your latest RELCO shows improved resolution R= 1219.

The size of your CMOS sensor (19.1mm) is showing the field curvature of the imaging lens. It's always difficult to get a full width spectrum all in focus. (The "sweet spot" for my Littrow Spectra-L200 is only around 12.5mm.) You can cut and merge sections of the spectrum (in BASS project) with slightly different grating settings to display a larger spectral coverage in focus.

Normally, in actual use, you would be focused (excuse the pun!) on a particular spectral feature/ wavelength for your target object i.e. Ha for Be stars. It's not necessary then to have large spectral coverage.

The horizontal lines are "Transversalium" lines (!!) and are caused by dust/ dirt on the slit gap. You could use a blower to clean the slit plate....it's of no major consequence.

Looking at the solar spectrum, obviously has the same focus issues... If I assume you took the solar spectrum at the same micrometer setting as the RELCO then there appears to be a shift of 7.2 A between the two.

The close-up crop around the Na lines shows the good registration after correction.

Something's moving in the assembly? It's important that the reference matches the target, this is how you will calibrate your future stellar spectra.

I'm also having difficulty in reconciling your dispersion with the camera pixel size (4.63 micron), the grating (600 l/mm) and the imaging lens focal length I thought was standard in the LowSpec design (85mm fl)

To get a dispersion of 0.75 A/Pixel it infers an imaging lens closer to 100mm fl. Can you confirm the imaging lens.

 

All in all you're doing very well! It shows your LowSpec is capable of producing a quality spectrum. Just comes down to some final "tuning".

I would certainly be setting it up with your scope to get a "first light" stellar spectrum. Just remember the sequence - take a reference spectrum immediately prior to the target.

Good work!

 

 

[Merlin66]

I was just wondering, following Valerio's journey into LowSpec spectroscopy, how the others are doing.

There's obviously a few (many?) of the 3D LowSpec instruments "in progress"

Any updates from the members????

[skybadger]

Hi Merlin, all.

Mine is now assembled, on the scope and going through the commissioning process o f learning to use. 

I have a lodestar as guider and asi174 as camera. 

I have spectra of a CFL lamp and daytime fraunhofer spectrum, and learning to guide on the slit using phd2.

Interesting thing is focusing. It looks in focus in the guider but has a distinct split spectrum in the camera. That means the star isn't focused in the slit. Focus until the spectrum has the minimum vertical dispersion. With no central split and refocus the guide camera...

Using Altair as my trial star, had some problems with intensity but wasn't using first order. Have identified the micrometer setting for that now and taken some spectra using the simple line profile tool in maxim. 

So making progress, was about to have a go at del sag last night  but the batteries went flat.. next time. 

 

 

 

[Merlin66]

Sounds promising!

Re focusing: A two stage process. I use the reference lamp to illuminate the slit plate then focus the guide camera on the slit gap, to get the tightest image of the gap. This should ensure the guide camera is OK to use with a target star.

Then focus the target star on the slit using the telescope focuser, it should be in pretty tight focus in the guide and present a reasonably narrow spectral band in the imaging camera. If this is not happening then the slit to collimator spacing is probably incorrect.

 

[Valerio]

9 hours ago, Merlin66 said:

Ken,

I am happy my spectrum are improving! :-)))

Now I am studying the BASS project and I hope to be able to analyse my own spectra soon.
It's possible that something moved in the assembly with these two spectra because I didn't completely close the screws of the cover. I will test with stars and close everything perfectly next time!

Camera and grating data are ok but the imaging lens is 100 mm. probably this is the error in your calculations?

When sky gets clear I will try with a star. I hope I will not have issues with focus and guiding...

Thank you again for your great help

Valerio

 

[Merlin66]

Valerio,

Thanks for the update!!

I’m glad the imaging lens was 100 mm fl. It makes 100% sense and matches all the SImSpec results.

You’re doing very well, I’m sure you will make great progress.

We're here to help, when you need us.

Edited September 28, 2020 by Merlin66

[skybadger]

13 hours ago, Merlin66 said:

Sounds promising!

Re focusing: A two stage process. I use the reference lamp to illuminate the slit plate then focus the guide camera on the slit gap, to get the tightest image of the gap. This should ensure the guide camera is OK to use with a target star.

Then focus the target star on the slit using the telescope focuser, it should be in pretty tight focus in the guide and present a reasonably narrow spectral band in the imaging camera. If this is not happening then the slit to collimator spacing is probably incorrect.

 

That's pretty much how I started, but what was clear was that the star image is dependent on the exposure, adjusting the output by focusing the scope  enabled me to refocus the guider. 

It's all in focus now though. 

 

[Gregj888]

I'm not as far along as Valerio  but did get the first spectra out of my LowSpec last night.  I could not get my camera lens to focus on the DSLR at infinity so purchasing a another lens, 108mm and hopefully that will do improve things.  This is with just a crude bench focus, pegged the camera lens and focused best I could on a piece of paper with line covering the slit holder slot.  Did not attempt to focus on the spectra.  Did not try a Relco but a neon and another I am playing with.

More later...

[Merlin66]

Greg,

Thanks for the images, they show promise.

6 hours ago, Gregj888 said:

I could not get my camera lens to focus on the DSLR at infinity

Not sure I understand this comment..

Are you using the DSLR body mounted on the LowSpec, if so the design back focus should allow the DSLR to get a focused spectrum.

You could double check that the collimating lens is positioned correctly and focused on the slit gap. If this is misplaced then you'd see a difference in the imaging focus position.

(Your lamp images shows some tight lines which says you're not to far away)

 

[Merlin66]

Greg,

I applied slant correction to you neon image, and it calibrates well.

You said you were using a DSLR? Did you crop and compress the image you posted? It seems to be only 1024 pixel wide, which is much less than a DSLR. Also to match the dispersion of 1.76 A/pixel the pixel size would have to be close to 10.5 micron, based on a 600l/mm grating.

The resolution is R= 1003 at 6143A

The lamp spectra is similar width, but doesn't seem to match the CWL wavelength settings of the neon. I was unable to match the emission lines shown. Which lamp did you image??

 

 

Edited September 30, 2020 by Merlin66

[Valerio]

I finally had an almost clear sky and i tried to get my first "real" spectrum. I used my Celestron Nexstar evolution C6 f/10 with Lowspec version 3.

The only visible bright body tonight was Mars (with clouds everywhere) and after a few trial and errors to focus everything I think I got a pretty good result. I'm not sure which is the best spectrum in terms of brightness (exposure time). I tried 4 versions, all with 30um slit:

Probably the first one is overexposed... which one is usually the best to use in Bass project to have the best analysis?

After that I captured the RELCO calibration lamp, with the same exact positions in the spectrograph, except I had to detach it from the telescope because the lamp was to faint to capture though the telescope lens.

Best

Valerio

 

[Merlin66]

Valerio,

Good to see your "first light" results!

In BASS project it's easy to load all the results and compare the quality of each profile, just hide the "bad ones" and work with the "good".

I'll have a look at your results and post my findings.