The Lowspec spectrometer

[Thalestris24]

Just now, Paul Gerlach said:

You know what? Turn that ring upside down. Print it 180 degrees rotated so that the side you you are going to thread in is on top.

I have tried that before. I'll have another look tomorrow.

[drjolo]

Today some stars appeared at my night sky, so I attached spectroscope to the 10" Meade ACF f/10 telescope and made first light. I was able to obtain first spectra of bright stars, however I am a little bit concerned about the shape of the stars image at the reflective slit. I have slit mounted with reflective surface pointing to the camera. The star (Betelgeuse) image looks like this:

       

These 6 images shows star at different focus positions. Telescope is collimated well, because I checked it before I put spectroscope into the focuser. The resulting spectra is extended vertically (along the slit) and also slanted in different directions at different wavelength:

  

Left image is blue part of the spectrum, right image is red part. This spectrum was done at the focuser position that corresponds to the third image from left in the upper row of focusing images. 

I wonder if I should reverse the reflecting slit, so its reflective surface points to the telescope.

[Merlin66]

OK.

A bright star like Betelgeuse will show multiple images in the guider.  The incoming light is reflected from the front surface of the glass slit plate (about 4%), then reflected from the chrome slit surface, ( about 90%) (1.5mm behind the front surface), then internally reflected from the rear of the front surface, reflected by the chrome etc etc. Giving fainter "companion stars". This is normal with a glass reflective slit plate. Don't reverse it!.

 

The guide image shows astigmatism - the first images are distorted vertically.....the last images distorted horizontally. I'd put this down to mis-alignment of the guide optics. Not originating in your scope.....

The exposure/ settings of the guide image could possibly be reduced to minimise the bloat of the star image....

Looking at the spectral image - I'd say that the telescope focus was not exactly on the slit gap, hence the double line - I think that's your secondary mirror out of focus.  Refocus the telescope to get a narrow consistent band, then re-check your guide image, move the guide camera to get a tight image - the slit gap should also show a  clear and tight line.

The slant should be reasonably consistent across the spectrum - not a major issue, can be corrected in processing. Could be that the slit gap is not 100% parallel to the grating grooves.

Lucas, all in all, congratulations on your first light spectral image.

Onwards and Upwards.

 

 

 

[drjolo]

Ken, thank you for your suggestions. I will review guider path and check guiding mirror and lens once more. I have 25um artificial star that should help to figure this problem out.

Despite such poor guiding image PHD did a good job and RMS tracking error was about 0.6" total in both axes. And here is my first star spectrum, calibrated with Relco starter and processed in BASSProject software. 

I am very happy about it  Resolution is quite good, both Na doublet and Mg triplet were separated. 

[Thalestris24]

I managed to get the Ovio lock ring reprinted - and screwed in! Woo! So, hopefully, turning the slit selector will work properly now. I nearly threw the thing out the window trying to get the slits lined up with the corresponding size marker! Gave up in the end, and just settled for the slit lining up with the holder opening. I've not actually tested it again yet. I could just about make out a couple of stars last night so got the 800mm/qhy5l-ii and 400mm/178m focused on some stars. I successfully plate solved with the 178 using SharpCap. Currently, the AVX PA must be way off and it's nowhere near balanced at the moment. However, I couldn't get the 178 image properly aligned with the qhy5l-ii image. It's really not an easy thing to do... But today I managed to get a much better alignment (than I previously had) on an electricity pylon. Everything's seems easier in the day. It's supposed to be clear later tonight (not that I believe the forecast) so hopefully I'll be able to try again to get the two images aligned. The 800mm scope image is about 20.6' x 15.5' - quite a small area of sky to align the 400mm scope to. I set a roi on the Travel Scope 70/178m to be 1024 x 768 and that pretty well matches the image size of the qhy5l-ii on the apo, so used that roi for fine adjustment. I should have done that before, really... Anyway, if it is clear (but also too cold for me...), I might even try to take some images, but at less than an arc sec per pixel they are likely to come out rather fuzzy especially as my guiding is only +/-4". The Travel Scope 70/178 at 3072 x 2048 should maybe appear better even though the scope optics are poor (the scope only cost me £39 off Amazon 4 1/2 years ago - bargain! ).

Louise

[drjolo]

If you align spectroscope slit along E-W direction, then tracking and guiding errors will move star along slit. You will not loose light then when star moves off the slit, only the spectrum will be little bit higher.

[Paul Gerlach]

4 hours ago, Thalestris24 said:

I managed to get the Ovio lock ring reprinted - and screwed in! Woo! So, hopefully, turning the slit selector will work properly now. I nearly threw the thing out the window trying to get the slits lined up with the corresponding size marker! Gave up in the end, and just settled for the slit lining up with the holder opening. I've not actually tested it again yet. I could just about make out a couple of stars last night so got the 800mm/qhy5l-ii and 400mm/178m focused on some stars. I successfully plate solved with the 178 using SharpCap. Currently, the AVX PA must be way off and it's nowhere near balanced at the moment. However, I couldn't get the 178 image properly aligned with the qhy5l-ii image. It's really not an easy thing to do... But today I managed to get a much better alignment (than I previously had) on an electricity pylon. Everything's seems easier in the day. It's supposed to be clear later tonight (not that I believe the forecast) so hopefully I'll be able to try again to get the two images aligned. The 800mm scope image is about 20.6' x 15.5' - quite a small area of sky to align the 400mm scope to. I set a roi on the Travel Scope 70/178m to be 1024 x 768 and that pretty well matches the image size of the qhy5l-ii on the apo, so used that roi for fine adjustment. I should have done that before, really... Anyway, if it is clear (but also too cold for me...), I might even try to take some images, but at less than an arc sec per pixel they are likely to come out rather fuzzy especially as my guiding is only +/-4". The Travel Scope 70/178 at 3072 x 2048 should maybe appear better even though the scope optics are poor (the scope only cost me £39 off Amazon 4 1/2 years ago - bargain! ).

Louise

Good work Louise! Perseverance pays-off in the end.
Recently bought a second 3D printer. A more mainstream one (Sidewinder X1) to get a more realistic picture off what these machines produce. Kinda strange to realize that I can start a print farm with five of these for the money I spend on one Felix Pro 3....

Paul

[Thalestris24]

8 minutes ago, Paul Gerlach said:

Good work Louise! Perseverance pays-off in the end.
Recently bought a second 3D printer. A more mainstream one (Sidewinder X1) to get a more realistic picture off what these machines produce. Kinda strange to realize that I can start a print farm with five of these for the money I spend on one Felix Pro 3....

Paul

Thanks! I don't know much about 3D printers - I only bought the Ender 3 Pro to print the Lowspec! Maybe I'll find some other things to print/make - who knows.

Louise

 

[Thalestris24]

28 minutes ago, drjolo said:

If you align spectroscope slit along E-W direction, then tracking and guiding errors will move star along slit. You will not loose light then when star moves off the slit, only the spectrum will be little bit higher.

Yeah, I'll try it when it comes to attaching the Lowspec to the scope.

Louise

[Thalestris24]

Have just about managed to plate solve and check that the 178 and qhy fov's are pretty much lined up. Transparency has been poor so not easy to get enough stars to do the plate solve, and exposures have needed to be rather long. I'm staying up and running and I'll try again later if it's clearer. Wish I could pick up more stars from here... The Coma Berenices cluster is quite bright and will be about later so hope to try on that.

Louise

[Thalestris24]

So I did manage to get some images of some stars in the Coma Berenices cluster last night, which was good. But then I had a moment and thought "Oh, wait a minute..." - am I thinking about this wrong?? I've set up with the qhy5l-ii as an imaging camera in the 115/800mm scope. But in the Lowspec I'm using the qhy5l-ii with a 31mm lens focused on the plane of the radial slit and looking at the image projected onto the surface by the 800mm scope. These two situations are not the same, are they? Or are they? I should know but my old brain gets stuck in the groove sometimes! I'd previously thought I would only see a 20' x 15.5' view of the sky via the guide cam but that can't be right. Surely it will see the size of the image effectively projected onto the surface of the Ovio. I don't know exactly how you calculate the effective fov of the guide cam and lens - it must be more akin to macrophotography with a small sensor? There is probably a calculator somewhere which will give me a better idea of actual fields of view and dimensions of images...

Anyway, fwiw, here are a couple of images of the cluster comparing the 178 and the qhy5l-ii:

178 1280 x 1024:

qhy5l-II:

16 Com 178 1280 x 1024:

 

I should probably attach the Lowspec to the scope and see what I can really see in practice! At least I've proved I can closely align images with a small fov!

Louise

[Bajastro]

Hi everyone,

I'm new in this forum.

I also built Low Spec spectrograph designed by @Paul Gerlach in second version. I use 1800 l/mm holographic diffraction grating:

I received first profiles of spectral lines few bright stars in higher resolution, examples:

But I have similar problem as @drjolo . Image of star in my guider module is very low quality.

See below:

I know that f/4 Newtonian telescope isn't optimal for Low Spec. Soon I will buy Barlow 2x.

Marius

Edited January 19, 2020 by Bajastro

[Thalestris24]

27 minutes ago, Bajastro said:

Hi everyone,

I'm new in this forum.

I also built Low Spec spetrograph designed by @Paul Gerlach in second version. I use 1800 l/mm holographic diffraction grating:

I received first profiles of spectral lines few bright stars in higher resolution, examples:

But I have similar problem as @drjolo . Image of star in my guider module is very low quality.

See below:

I know that f/4 Newtonian telescope isn't optimal for Low Spec. Soon I will buy Barlow 2x.

Marius

Hi Marius

Welcome to the fold! Your Lowspec looks very neat! What guide scope are you using? Your guiding doesn't look too bad - much better than I'll be able to achieve. Looks like you just need to collimate? But I know nothing! I've not done a star spectrum yet but am getting there...

Louise

[Thalestris24]

3 hours ago, Thalestris24 said:

So I did manage to get some images of some stars in the Coma Berenices cluster last night, which was good. But then I had a moment and thought "Oh, wait a minute..." - am I thinking about this wrong?? I've set up with the qhy5l-ii as an imaging camera in the 115/800mm scope. But in the Lowspec I'm using the qhy5l-ii with a 31mm lens focused on the plane of the radial slit and looking at the image projected onto the surface by the 800mm scope. These two situations are not the same, are they? Or are they? I should know but my old brain gets stuck in the groove sometimes! I'd previously thought I would only see a 20' x 15.5' view of the sky via the guide cam but that can't be right. Surely it will see the size of the image effectively projected onto the surface of the Ovio. I don't know exactly how you calculate the effective fov of the guide cam and lens - it must be more akin to macrophotography with a small sensor? There is probably a calculator somewhere which will give me a better idea of actual fields of view and dimensions of images...

I should probably attach the Lowspec to the scope and see what I can really see in practice! At least I've proved I can closely align images with a small fov!

Louise

Well, I just attached the Lowspec to the imaging scope and got it more or less focused. I think it still is quite a small fov. I took some pics of a pylon:

Using qhy5l-ii:

nb - slit across the middle! I will align it along RA (or maybe DEC) at some point.

 

178 on Travelscope:

Look pretty similar in scale to the Coma Berenices comparison. Of course, substituting the Lowspec for the qhy5l-ii means the two images are no longer aligned as before. So I'll have to do that all over again...

It doesn't look like there will be any clear sky tonight

Louise

Edited January 19, 2020 by Thalestris24

[Thalestris24]

11 minutes ago, Thalestris24 said:

Well, I just attached the Lowspec to the imaging scope and got it more or less focused. I think it still is quite a small fov. I took some pics of a pylon:

Look pretty similar in scale to the Coma Berenices comparison. Of course, substituting the Lowspec for the qhy5l-ii means the two images are no longer aligned as before. So I'll have to do that all over again...

It doesn't look like there will be any clear sky tonight

Louise

Just did a quick align:

Close enough, I hope! I suppose I'll have to keep doing the alignment...

Louise

[Bajastro]

@Thalestris24 I use built-in Low Spec guide module with achromatic lens with 30 mm focal length and ASI120MM Mini.
I don't need any external guide scope.

[Thalestris24]

1 minute ago, Bajastro said:

@Thalestris24 I use built-in Low Spec guide module with achromatic lens with 30 mm focal length and ASI120MM Mini.
I don't need any external guide scope.

Similar to my qhy5l-iim then - same sensor. How to do you get a particular star on the slit? I'm using an e-finder - a 70/400mm scope with a 178m camera plus plate solving. Maybe you have a much better mount and alignment than I have?

Louise

[Bajastro]

After the calibration of axis RA in my mount I calibrate GoTO usually with using one star. Next I center first star with visually observation in finderscope 8x. Important is moving on the same side of sky where was calibrated star for GoTo system. Becasue after rotating mount on the opposite side of the sky usually scope isn't pointed on a particular star.

I'm using HEQ5 mount with Rowan modification kit.

[Thalestris24]

7 minutes ago, Bajastro said:

After the calibration of axis RA in my mount I calibrate GoTO usually with using one star. Next I center first star with visually observation in finderscope 8x. Important is moving on the same side of sky where was calibrated star for GoTo system. Becasue after rotating mount on the opposite side of the sky usually scope isn't pointed on a particular star.

I'm using HEQ5 mount with Rowan modification kit.

Ok - so that gets a particular star close to the slit?

[drjolo]

2 hours ago, Bajastro said:

But I have similar problem as @drjolo . Image of star in my guider module is very low quality.

I have checked yesterday my guiding lense - it was tilted about 1mm (!!) in the holder. The guiding lense holder has very small holding ring (0.2mm), and setting lens position precisely is not an easy task. I redesigned it a little, so holding ring width is now about 1mm. The lens aperture is then little bit lower, so the image may be little bit darker, but also better quality. This lens focal ratio is f/2.4, so it is extreme especially for simple doublet lens. Fraction degree tilt may degrade image significantly.

With this new holder I adjusted and checked all using 25um artificial star. Results are promising, of course reflections are still there, but the star image quality was good, and once placed at the slit it almost disappeared - so most of the light was focused well. Now I need to wait for another clear night for real tests.

[Bajastro]

Thanks @drjolo, today I also checked my lens in guiding module, it was... tilted. Unfortunately, I don't have artificial star and I have to wait for clear sky and test it again.

I think that focal ratio 2.4 for achromatic lens is too low. I will try to reduce this aperture up to f/3 - f/3.75, external aberrations should be much lower. Guiding lens also was stressed (squeezed).

Edited January 19, 2020 by Bajastro

[Paul Gerlach]

Great work @drjolo ! Unfortunately I still don't have my own setup (for a couple of years now 🙄) so I can't test these things myself.
By that 'small holding ring'you mean that little protruding edge (see picture red circle)?
The inner diameter of the guide lens holder is 12.75 mm and the lens itself has a diameter of 12.7 mm. This tolerance is a bit of guess work. I can set my printer at 12.7mm and it will fit but every 3D printer is different. The lens has to be pushed in until it hits this small protruding edge. The flange on the holder should help in placing the lens flat against the edge of the opening in the main body. Is the holder a tight fit in your case or is there still some space for the lens to move around? If this little edge turns out to be to small for some printers, maybe I should enlarge it a bit. If I double the edge the opening will be 12 mm which means it could accept a f/6 light cone.

Paul

[Thalestris24]

5 minutes ago, Paul Gerlach said:

Great work @drjolo ! Unfortunately I still don't have my own setup (for a couple of years now 🙄) so I can't test these things myself.
By that 'small holding ring'you mean that little protruding edge (see picture red circle)?
The inner diameter of the guide lens holder is 12.75 mm and the lens itself has a diameter of 12.7 mm. This tolerance is a bit of guess work. I can set my printer at 12.7mm and it will fit but every 3D printer is different. The lens has to be pushed in until it hits this small protruding edge. The flange on the holder should help in placing the lens flat against the edge of the opening in the main body. Is the holder a tight fit in your case or is there still some space for the lens to move around? If this little edge turns out to be to small for some printers, maybe I should enlarge it a bit. If I double the edge the opening will be 12 mm which means it could accept a f/6 light cone.

Paul

FWIW, my lens was 13mm diameter so I just made a new lens holder

Louise

[Bajastro]

I think that smaller diameter of aperture will be better and cut peripheral aberrations this achromatic doublet (about f/2.4), for example:

This value gives me focal ratio about f/3.5 for guiding lens with diameter 12.5 mm and focal length 30 mm. I will test it soon.

Edited January 19, 2020 by Bajastro

[Merlin66]

Louise,

Think of the reflective slit plate as a on- axis guider or a star diagonal. All it does is reflect the FOV at the slit position.

The transfer lens - depending on the spacing between the slit plate and the final guide image then acts as a reducer - giving a slightly larger FOV in the camera. As far as I can see the reduction is x0.72.

The actual guide FOV will depend on the area of the slit plate visible to the guider, the reduction factor and the guide camera chip size.

In my case the slit plate has a front cover, with an aperture of 8 x 6 mm, centred on the slit gap. This effectively limits the actual area of the sky visible ( the camera FOV is much larger, but obviously blocked by the cover plate!)

In my C11 @ f10 I get 9.8 x 7.4 arc min coverage.

I can easily put a target star close to centre in the 60/220 eFinder and find the same field in the spectrograph guider. If you use PHD2 you can set up the virtual slit and nudge the target into/onto the slit gap.

Much easier than you think.