Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Why isn't there a 'spectro-telescope'?

equipment observing optics
  • Please log in to reply
13 replies to this topic

#1 JoeVanGeaux

JoeVanGeaux

    Explorer 1

  • -----
  • topic starter
  • Posts: 94
  • Joined: 01 Jan 2010
  • Loc: SETX

Posted 23 February 2020 - 06:06 PM

Pardon my taking liberties with the terminology for equipment relating to observing/recording astronomical spectra.  From the multitude of ray-traced diagrams and other postings of scope designs I've seen, all attempt to avoid chromatic aberration... and understandably so!  However, some optical designs seem to naturally create a really well-defined spectrum - which is the point of the tools for spectral observation, is it not??

With such designs in mind, whether done intentionally or the result of just poorly performing designs, can't these designs be used as is or be lightly 'tweaked' for use in spectroscopy?  What I mean is that some designs just seem to need neither slit nor diffraction grating (reflective or transmission) to generate a clear spectrum (some vastly clearer than others, of course). Or, is it I just don't know how to interpret those diagrams and designs that show all the pretty colors all neatly separated and lined up - ripe for the picking by eye or camera?

I am hoping for some insight so it can help me understand more detail and theory involved in the design of some spectrographs out there.  I've seen many web pages that go through the calculations for what follows once a "pinpoint" of starlight is obtained but it seems to first combine all that light that was once (maybe) already well on its way to producing a full blown spectrum.

 

It just seems like a lot of "re-engineering" is being made to undo what is naturally occurring in some basics optics - only for it to be redone by, yet, more optics.  Just look at some of those designs in the "Post your Optical Design!" thread for some examples.  What am I missing, here??

Joe



#2 sg6

sg6

    Cosmos

  • *****
  • Posts: 7,658
  • Joined: 14 Feb 2010
  • Loc: Norfolk, UK.

Posted 23 February 2020 - 06:27 PM

For spectroscophy measurement you need a precise measurement of angle and wavelength, and in a way any old spectrum does nothing useful.

 

Also if you look at OIII and Hg they are very close to each other. A sort of normal scope will not allow you to differentiate visually whereas a spectroscope will have them as 2 distinct lines.

 

You can get odd results from a spectroscope - someone at university found he was blind to one bit of red in an experiment. We had 4 red lines and he could see 3. Assumption was the end one was out of his range but it was actually the one before - as in lines 1,2,3,4 and he could see 1,2,4.

 

One club has an SCT and a camera on it and they can display the spectrum via the camera. Unsure of the details but have seen the camera image as in EAA and underneath the objects spectrum.

 

Nice as I suppose an extra but not adaquate for spectral measurements. Thinking radial velocity measurement of double stars.



#3 descott12

descott12

    Vendor - Solar Live View

  • *****
  • Vendors
  • Posts: 1,961
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 23 February 2020 - 08:12 PM

Check out the Spectroscopy sub-forum.

Many of us use the very simple but powerful Star Analyzer.

 

https://www.cloudyni...l-astrophysics/

 

Great results with a very simple device.


Edited by descott12, 23 February 2020 - 08:49 PM.

  • jtsenghas likes this

#4 MitchAlsup

MitchAlsup

    Skylab

  • -----
  • Posts: 4,306
  • Joined: 31 Aug 2009

Posted 23 February 2020 - 08:29 PM

A spectroscope uses the main optics (i.e., the telescope) as a light concentration device.

 

The slit is used to choose which part of the original image enters into the spectroscope.

 

The spectroscope splits the part of the image coming through the slit perpendicularly to the slit creating a spectrum for each unit-of-length across the slit.

 

The way prisms and diffraction gratings work, the light being bent into spectral contents has to be (for all intents and purposes) collimated. This means that after the slit, there is an inverse telescope that collimates the image through-the-slit. Once collimated, the light is refracted through a prism or diffracted off of a grating. Once the spectrum has been bent into colors, it is focused again at which time a sensor can capture the spectrum.

 

So, a spectrograph has 3 telescopes in it; 1 big one, and 2 little ones. The resolution of the spectrograph is almost entirely dependent on how large the beam is when it encounters the prism or grating, but can easily be degraded by the optics. How consider the problem of vibration, the spectroscope must be continuously fed the portion of the image "under the slit", so the telescope has to track (from an amateur point of view) exceedingly well, and the spectroscope has to admit that light continuously, while being carried on the telescope, with absolutely no vibrations, and if you want to get high quality data, the spectroscope must remain at a single temperature during the whole observation run.

 

A spectrograph that spreads the light from a star to 500-Airy disks wide has a resolution of 500 and one can hardly make out anything other than major spectral contributors. The individual locations along this spectrum are (more than) 500 times dimmer than the star, too.

 

So, basically anything you do with a telescope will take 500 times as long if you hook up a spectroscope. Then imagine using a high resolution spectrograph with a resolution of 50,000 good enough to easily separate the 2 OIII lines. Now consider that the large telescope like Keck might use an entire hour taking one spectrum, or an entire week when trying to take low-medium resolution spectra of "stuff" a 26M (such as galaxies in a HDF)

 

Spectroscopes are what most professional time on big telescopes are about.

There are NGs devoted to amateur scale spectroscopes (but I can't seem to find it right now).

There are semi-commercial spectroscopes one can buy.

 

Spectroscopy is significantly harder that astrophoto. But if it floats your boat: go for it.


  • JoeVanGeaux likes this

#5 JoeVanGeaux

JoeVanGeaux

    Explorer 1

  • -----
  • topic starter
  • Posts: 94
  • Joined: 01 Jan 2010
  • Loc: SETX

Posted 23 February 2020 - 09:32 PM

Thanks for the replies. 

 

Yes, Scott, I have been following the spectroscopy threads (and numerous other webpages and sites) and know about the Star Analyzer and much of  the other common equipment at the tail end of some scopes. 

I think, though, from a practical point of view your statement, Mitch, sums up the "why?" to my question pretty well by reminding me that the 'target' object has to be isolated from the rest of the field (and with precision tracking).

Thanks!  Now I can quit obsessing over that detail that should not have escaped me!

 

Joe



#6 BGRE

BGRE

    Mercury-Atlas

  • *****
  • Posts: 2,859
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 23 February 2020 - 10:08 PM

There's also the issue of the shape of the focal surface when the CA is significant.


  • JoeVanGeaux likes this

#7 JoeVanGeaux

JoeVanGeaux

    Explorer 1

  • -----
  • topic starter
  • Posts: 94
  • Joined: 01 Jan 2010
  • Loc: SETX

Posted 23 February 2020 - 10:39 PM

There's also the issue of the shape of the focal surface when the CA is significant.

Yeah, thanks for bringing that up, but I was having less problem reconciling that phenomenon. 



#8 BGRE

BGRE

    Mercury-Atlas

  • *****
  • Posts: 2,859
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 23 February 2020 - 10:50 PM

UV spectrographs using a singlet camera lens often used to tilt the glass photographic plate to compensate for the varying focal length with wavelength.

 

If one uses an objective with significant CA with an objective prism maintaining the entire spectrum in focus for each star in the field may be somewhat problematic when the objective has significant CA. 


  • JoeVanGeaux likes this

#9 Spectral Joe

Spectral Joe

    Messenger

  • -----
  • Posts: 422
  • Joined: 28 Feb 2011
  • Loc: Livermore CA

Posted 24 February 2020 - 01:09 AM


 

A spectrograph that spreads the light from a star to 500-Airy disks wide has a resolution of 500 and one can hardly make out anything other than major spectral contributors. The individual locations along this spectrum are (more than) 500 times dimmer than the star, too.

 

A spectrograph with a resolution of 500 can resolve two spectral lines that are separated by 1/500th of the mean wavelength of the line pair. Dispersion is how much the light is spread out, measured in wavelength units per length unit, i.e. Angstroms per millimeter.



#10 Oberon

Oberon

    Gemini

  • -----
  • Posts: 3,248
  • Joined: 24 Feb 2013
  • Loc: Hunter Valley NSW Australia

Posted 24 February 2020 - 04:36 AM

Google terms like “full aperture slitless spectroscopy” will find you stuff like this...

”There are two basic photographic options at the UKST:

direct imaging of the sky in different wavebands from ultraviolet to infrared, selected by the appropriate choice of photographic emulsion and filter;
slitless low dispersion spectroscopy through thin, full-aperture objective prisms.”

 

I’ve seen the photographic plates taken by the 1.8m UKST (United Kingdom Schmidt Telescope) with the full aperture prisms, and they show an entire field of stars, galaxies and quasars as small spectral streaks...like a banded stripe of light approx 10-20mm long iirc. I’ve also removed the prism from the telescope (ie, removed it from the front aperture, prism diameter was 1.8m or 6’) as part of routine operations...must be at least 30 years ago now.

Both HST and the JWST utilise slitless spectroscopy (albeit with a grism, not a full aperture wedge), below is an image from the HST (the effect is similar to what appears on the Schmidt plate)...

EL-glx-on-WFC3-grism-PMcC-10.jpg

 

...which I found on this website which has loads of stuff about novel spectroscopic techniques (and curiously features a photo of the AAT’s 2DF taken by yours truly...). lol.gif


Edited by Oberon, 24 February 2020 - 05:23 AM.


#11 Oberon

Oberon

    Gemini

  • -----
  • Posts: 3,248
  • Joined: 24 Feb 2013
  • Loc: Hunter Valley NSW Australia

Posted 24 February 2020 - 05:20 AM

Pardon my taking liberties with the terminology for equipment relating to observing/recording astronomical spectra.  From the multitude of ray-traced diagrams and other postings of scope designs I've seen, all attempt to avoid chromatic aberration... and understandably so!  However, some optical designs seem to naturally create a really well-defined spectrum - which is the point of the tools for spectral observation, is it not??

With such designs in mind, whether done intentionally or the result of just poorly performing designs, can't these designs be used as is or be lightly 'tweaked' for use in spectroscopy?  What I mean is that some designs just seem to need neither slit nor diffraction grating (reflective or transmission) to generate a clear spectrum (some vastly clearer than others, of course). Or, is it I just don't know how to interpret those diagrams and designs that show all the pretty colors all neatly separated and lined up - ripe for the picking by eye or camera?

I am hoping for some insight so it can help me understand more detail and theory involved in the design of some spectrographs out there.  I've seen many web pages that go through the calculations for what follows once a "pinpoint" of starlight is obtained but it seems to first combine all that light that was once (maybe) already well on its way to producing a full blown spectrum.

 

It just seems like a lot of "re-engineering" is being made to undo what is naturally occurring in some basics optics - only for it to be redone by, yet, more optics.  Just look at some of those designs in the "Post your Optical Design!" thread for some examples.  What am I missing, here??

Joe

So, in short, its an excellent question, and yes there is, and yes it works. Professionals have exploited the technique for well in excess of a century. Just as aperture prisms are used to correct atmospheric dispersion, so they may in effect be reversed to create a low dispersion spectra. And low dispersion spectra is not just a toy, it is essential to science; for example, at the UKST (a survey telescope) the low dispersion spectra obtained via the slitless aperture prism technique were used to identify quasars (QSO - quasi-steller object) for follow up higher resolution studies with the 3.9m AAT (and later the HST).


Edited by Oberon, 24 February 2020 - 06:28 AM.

  • gregj888 and JoeVanGeaux like this

#12 Oberon

Oberon

    Gemini

  • -----
  • Posts: 3,248
  • Joined: 24 Feb 2013
  • Loc: Hunter Valley NSW Australia

Posted 24 February 2020 - 05:50 AM

An example from history...(ever heard of the HD (Henry Draper) catalogue?).

 

E399040C-1754-4888-ADF8-42A7309D624C.jpeg

 

https://faculty.virg...11/lec8-f03.pdf

 

Also...

 

”Draper Catalogue of Stellar Spectra

In 1882 Edward Charles Pickering, the Director of the Harvard College Observatory began a programme of astronomical spectroscopy using objective prisms. This type of setup enabled up to 200 stellar spectra to be captured on a single photographic plate. This work was continued under the auspices of the Henry Draper Memorial, a fund set up by the widow of Henry Draper to honour his work in the field of astronomical spectroscopy.

In 1890 the Draper Catalogue of Stellar Spectra is published by Edward Charles Pickering, which contains the photographic spectra of 10,351 stars, nearly all of them north of 25° south declination. In this work the spectrum of each star was classified according to a scheme developed by Williamina Fleming (1857-1911). In her system known as the ‘Draper Classification’, the letters A to Q (omitting J) were used to classify stellar spectra.”

 

http://www.catcherso...strophotography


Edited by Oberon, 24 February 2020 - 06:29 AM.

  • gregj888 likes this

#13 MitchAlsup

MitchAlsup

    Skylab

  • -----
  • Posts: 4,306
  • Joined: 31 Aug 2009

Posted 25 February 2020 - 09:35 PM

Here is the place I could not find a couple days ago.

 

http://www.astrosurf...ctrographs.html



#14 JoeVanGeaux

JoeVanGeaux

    Explorer 1

  • -----
  • topic starter
  • Posts: 94
  • Joined: 01 Jan 2010
  • Loc: SETX

Posted 25 February 2020 - 11:28 PM

Here is the place I could not find a couple days ago.

 

http://www.astrosurf...ctrographs.html

Yes, I agree!  Its an excellent site.  I have been visiting that site very frequenlty over the last few weeks.

 

Joe




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: equipment, observing, optics



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics