15 replies to this topic

[AllanDystrup]

Classic Stellar Observation

 

    
     Stars are often mostly observed by amateur astronomers in aggregates, -- from constellations, associations and asterisms  to globular and open clusters, and down to multiple and double systems.  And of course as ”stepping stones” and ”faint fuzzy” galaxies, when star hopping to DSO.

    

     But stellar physics and evolution are also interesting topics ”per se”, and with simple amateur spectroscopes it is possible -- even with small telesciopes and from suburban environments -- to take the ”fingerprints” of the brightest stars and study their astronomical features such as chemical composition, temperature, movements and more.

    

   This early morning I am out under the night sky at 01:30 local time (UT+2) in nordic (Danish) nautical twilight (SQM 16.3, NELM 4.2^m), and I can just see the main outline of the summer constellations (Cygnus, Lyra, Aquila), but not much more. DSO and even low power rich field observing is pretty much out of the question tonight.

    

     Never the less, I have my classic Vixen FL-80S/640mm refractor out this early morning, together with a classic Zeiss ocular spectroscope :

 

*click*

    

The Zeiss star spectroscope screws right onto the top of a standard CZJ eyepiece. It consists of a Amici dispersion prism followed by a cylindrical lens for widening the spectrum. The prism dispersion is non-uniform, so that the short (violet end) wavelengths are more widely spread out.  This device works best at good seeing and at a magnification of 5-10 x Dcm (ie for my 8cm scope : 40-80x. The magnification of the CZJ O-16 is @ 68x (using a 1.7xGPC in front of my diagonal) , which gives a small but nice bright spectrum of Vega. At -0.6^m +0.03^m Vega is the second brightest star in the N. hemisphere (after blazing Arcturus at: -0.0^m -0.05 ^m) and with my 3” Vixen refractor I should be able to use the EP spectroscope on stars down to 2-3^m. (With a 8” scope it should be useable down to ~5^m). --
^note: [Thanks to cincosauces for providing the correct stellar magnitudes for Vega & Arcturus!]

    
 

    
     I start out by viewing Vega with the CZJ eyepiece spectroscope; You have to defocus a little inwards, to broaden the spectrum enough to be able to see the lines. The spectrum of Vega is dominated by two absorption lines: one strong  where green fades into blue (between 480-490nm), plus a weaker absorption line in the far end of the blue (between 430-440nm). The ultra far violet and red parts of the spectrum (below ~420nm and above ~670nm) can not be seen in the CZJ okular-spectroscope :
 

*click*

 

     The two absorption lines I am observing in Vega match the hydrogen lines : H-beta (486nm) and H-gamma (434nm) of the Balmer Series, ie. photons absorbed in the stellar atmosphere by Hydrogen atom electrons being ionized (”kicked”) from the 2. quantum state (”shell”) to the 4. (beta) rsp. 5. (gamma) state :

    

*click*

 

     The prominent Balmer series Hydrogen absorption lines are characteristic of massive (>1.3x M_solar) and hot (core: T_C>18M K) blue/white stars of spectral type Secchi-I (corresponding to type A-F of the Yerkes/MKK system). These stars fuse hydrogen protons to helium nuclei through the CNO-cycle (carbon-> nitrogen->oxygen), which  results in a convection zone around the core, that distributes the 'ash' from the fusion below an overlying atmosphere in radiative equilibrium. The photosphere shows strong H- absorption lines, as seen in for example: B: Rigel, Bellatrix  -- A: Vega, Deneb, Altair – F: Polaris Procyon.

 

*click*

    

 

     I now change to my 100li/mm transmission difraction grating spectroscope, for a closer analysis of the Vega spectrum.

 

(to be continued...)

Allan

 

 

Edited by AllanDystrup, 03 July 2017 - 10:44 AM.

[DHEB]

Very interesting, thank you for posting. I understand well the problem with nordic twilight.

I am a bit surprised about your quoted magnitude value for Vega. You write -0.6 (I assume in the visible, right?) but this is not what we read in the VSX database, which gives values -0.02 - +0.07:

https://www.aavso.or...l.top&oid=18630

In fact, Vega has long been considered the reference star and given magnitude 0.0 by definition, a practice that is now left aside though.

https://en.wikipedia...ega?wprov=sfla1

May be you just swapped the values of Vega and Arcturus? ☺

Edited by cincosauces, 03 July 2017 - 09:11 AM.

[NEOhio]

Allan, neat post, I had asked recently in another thread about whether there are any "eyepiece spectroscopes" where you view the spectrum directly the same way you look through an eyepiece (or "ocular", but sadly it is not yet Friday), rather than using a camera/computer setup. Sounds like this "classic Zeiss ocular spectroscope" is exactly that device. 

 

Any idea if such a device is being manufactured/sold today?

[AllanDystrup]

@ cincosauces:

Right you are -- I've corrected the magnitudes to those of Cartes du Ciel: Sky Charts (V.3.10)
Thank you for the corrections!!

 

@ NEOhio:

The Zeiss EP spectroscope is not produced anymore, but may be found for sale 2.hand (it is a collector's item though and thus has become very expensive).

A better solution, I think, is to buy a blazed grating, like the Star Analyzer 100. I'll describe that in detail in my next post, where I used it on my small ASI 120MC camera, -- but it can also be used on an eyepiece, together with a cylinder-lens to spread out the spectrum. I'll describe that too, when I've had a chance to test it.

 

Thanks and cheers, --

Allan

Edited by AllanDystrup, 03 July 2017 - 10:55 AM.

[NEOhio]

Would love to see your description of the visual approach. Looked up the Star Analyzer 100 at this site, and it had the interesting comment: "What’s this 'Star Spectroscope' that we refer to above?  The Star Spectroscope is an older 200 line/mm grating designed primarily for visual use." So apparently that had been used for visual before, but is discontinued. The site seems to suggest the Star Analyzer 100 would work for visual. Big difficulty it would seem to me would be adding a reticle showing the wavelength marks, without which it would be difficult to read visually. I suppose if you know which lines you should get, and can see the line colors, you can figure out what you are seeing without a wavelength labeling reticle. 

[AllanDystrup]

Vega Spectrum -- continued

 

    

     I now switch to my Star Analyzer 100 li/mm diffraction grating. This is a low-res transmission grating mounted in a std. 1.25” D threaded filter cell.

    

    

     For best resolution I now mount the SA100 grating on my small camera (instead of on an eyepiece). The distance from the grating to the sensor determines the length of the spectrum, and thus the resolution. I want to capture both the 0. order star and 1. order spectrum on my sensor chip (which is a 1/3” CMOS on my ZWO ASI 120MC camera). The useable distance for my chip is 15-75mm, and the resolution in Angstrom/pixel should then be in the range 12-15 A/pix.  As my telescope is a f/8 APO, the focusing of the spectrum turns out to be relatively easy (which may not be the case for Achromats with significant CA and/or for long-FL scopes without a reducer). The seeing this early morning however is below medium, and this will degrade the resolution, but there’s not much I can do about it (apart from setting a short exposure time while still keeping the image of the spectrum reasonably bright).

    

    

    

Allan

[AllanDystrup]

Would love to see your description of the visual approach. Looked up the Star Analyzer 100 at this site, and it had the interesting comment: "What’s this 'Star Spectroscope' that we refer to above?  The Star Spectroscope is an older 200 line/mm grating designed primarily for visual use." So apparently that had been used for visual before, but is discontinued. The site seems to suggest the Star Analyzer 100 would work for visual. Big difficulty it would seem to me would be adding a reticle showing the wavelength marks, without which it would be difficult to read visually. I suppose if you know which lines you should get, and can see the line colors, you can figure out what you are seeing without a wavelength labeling reticle. 

 

     Yes, the SA100 should work for visual observation. I plan to test it mounted on my 12.5mm micro guide orthoscopic eyepiece, that has an illuminated reticle, which I may be able to roughly collimate (using an A0V star like Vega) to match the dispersion in Angstrom from the zero order star.

 

     But you are right that translating spectral absorption lines to wavelengths visually -- for simple prism as well as grating spectroscopes -- is not easy; The Zeiss EP prism spectroscope can probably only be used visually to qualitatively determine the Secchi type spectrum of the brightest stars (>3^m for my 80mm refractor), -- but that should also be an interesting project :

 

 

     If I'm able to collimate the grating spectrum to the micro guide reticle, -- well that would open up for a more quantitative mapping of the spectrum lines to wavelengths. We'll see.

 

     I also have a classic Unitron spectroscope with a Pellin Broca prism for visual eyepiece observation. I'll try to get that up&running on a 12.5mm Unitron 18mm K eyepiece.

 

Ahh well, lots of ideas, but no starry starry nights over here. -- Clouds and rain, more rain and clouds...

CS!! -- Allan

Edited by AllanDystrup, 05 July 2017 - 10:42 AM.

[AllanDystrup]

SA100 grating spectroscope -- Visual

 

 

     I’m out this early July morning, in so-so conditions: 02:30 UT+2, still nordic nautical twilight with a NELM ~3.9^m (SQM 15.7). The seeing is stable, well above medium (8-9/10), while the transparency is below medium due to a veil of thin, high cirrus (3/7). All in all I’m facing a Bortle White (inner) city sky this morning, but still the main outline of the summer triangle constellations is visible, so I proceed with my small stellar spectroscope project.

 

     My main goal this morning is to try out the SA100 grating spectroscope for visual use (http://www.patonhawksley.co.uk/staranalyser.html).  I first screw  the SA100 1.25” filter cell into my ATC Nagler 8mm EP, and slide this into my Zeiss 2x barlow. With a 1.7x GPC in front of my diagonal, this gives me a total magnification of 187x. I center the 80mm f/8 refractor on Vega, and get a 0. star image plus a bright 1.order spectrum in the field of view. The SA100 grating  disperses the spectrum evenly, with a large yellow-orange strip in the center, and equally sized smaller blue and red zones to the left rsp. right (contrary to a prism spectroscope, where -- as previously described for the CZJ EP gizmo – the short wave / violet end is more widely spread out ).  A slight defocusing of the image broadens the spectrum a bit, and with some effort I’m able to get a glimpse of the H-beta line of the Balmer series; The H-gamma line, that was easily seen in the CZJ prism EP spectroscope, is hard in the SA100.

 

*click*

 

     I now proceed with my visual test of the SA100 grating by mounting it in the barrel of my Celestron Micro Guide  12.5mm Ortho eyepiece (174x magnification, w/ 2x barlow + 1.7x GPC). Again the image is bright and sharp, though the resolution for visual observation is only ”borderline” with respect to identification of spectral lines. It is however possible to align the 0. order image at zero of the laser-etched reticle linear scale, with the 1. order blazed spectrum along the scale in such a way, that calibrating the SA100 image to rough wavelength values in Angstrom should be possible.

 

*click*
(iPhone4 snapshots...)
 

  

     My conclusion for this mornings experiments however is, that for visual spectrum observation with the SA100 grating (and preferably also with a prism spectroscope like the CZJ), a higher resolution is a necessity, and this will require an aperture of at least 6”-8”. So for further observations with my small 3” refractor, I plan to play visually a little more with my CZJ EP spectroscope on the brightest stars (say down to 2^m), while for more ”serious” spectrum studies, I’ll use the SA100 grating on my simple mono camera, and have software like Rspec help me display and analyze the captured spectra.

 

 

Allan

Edited by AllanDystrup, 10 July 2017 - 08:18 AM.

[AllanDystrup]

SA100 grating spectroscope – Monochrome Camera

   

   

     It’s an hour past midnight local time DST (UT+2) this mid July (2017-07-13), and we're already into astronomical dawn with a NELM of 4.7^m (SQM 17.3). There’s a 19dy (83%) last quater moon hanging low at ~11° Alt. towards the SE in Aquarius, with a faint halo around the globe due to thin, high haze in the atmosphere (transparancy  3-4/7). The seeing however is well above medium (7/10) and the main outline of the summer triangle constellations are traceable naked eye, while I’m out observing naked arms (in shirt sleeves) this early summer morning at a fresh 11°C temperature.

   

     I have mounted my small monochrome PGR CM3 camera on my Vixen FL-80S/640mm refractor, with the SA100 li/mm transmission diffraction grating in a std. 1.25”D filter cell threaded into the nosepiece. The CM3 an inexpensive machine vision camera with a 1/3” Sony ICX445 CCD, and with a 1.7x GPC in front of my diagonal, the sensor can just accommodate the 0. order + blazed 1.order spectrum on the chip. I now aim the small refractor at Vega.

   

     Visually (on the pc screen) the 1.order spectrum shows three evident absorption lines, corresponding to the predominant Hydrogen lines of the Balmer Series typical for Secchi type-I stars (~Yerkes/MKK Type late B, A and F) : H-beta, H-gamma, H-delta (listed from far red towards near blue).

 

     The three absorption lines are also clearly seen on the raw light intensity curve as displayed in RSpec, together with some minor dips, that require a closer analysis. The exposure time for Vega is ~1s with the gain at minimum (~250); This setting results in the 0. order star image being saturated on the CCD, whereas the 1. order spectrum is only 53% saturated so I should be able to increase the gain somewhat for the spectrum recordings to get a brighter image.

 

*click*

    

     For a closer analysis of the Vega spectrum, I now calibrate the spectrum dispersion in pixels to the corresponding dispersion in light wavelengths as measured in Angstrom : Setting the 0. order star image to zero and the prominent H-gamma absorption dip to 4340 A, I now get the following Vega-spectrum, with light intensity calibrated to wavelengths (6.5 A/pix).

 

*click*
  

*click*
    

Allan

Edited by AllanDystrup, 14 July 2017 - 11:40 AM.

[desertlens]

Allan, Thanks for this... a lot of practical information. It's been a long time since I bookmarked a CN thread. I'm using a Rainbow Optics blaze grating and cylindrical lens for purely visual observing. My interests are more aesthetic than metric. I get the best results with the grating ahead of the diagonal.

[robin_astro]

Hi Allan,

 

A very nice result. The Star Analyser works very well with small short focal length refractors like this provided they are well corrected. 

Just one small point. The instrument response includes not just the response of the camera sensor but also the response of the optics (in particular the grating response which is as significant as that of the camera) and the atmospheric extinction which depends on the altitude of the target in the sky.    

 

Cheers

Robin

Edited by robin_astro, 15 July 2017 - 01:58 PM.

[AllanDystrup]

Hi Robin,
Yes, that makes perfect sense, what you state about the instrument response.
Thank you for the clarification! -- I'm learning as I go with this project

 

____________________________________________________________________________

 

SA100 grating – Mono Camera – DENEB

    

    
     It’s midnight local DST (UT+2), and we’re now transiting from nautical twilight to astronomical dusk, -- as dark as it gets in mid July up here at 56° N latitude just north of Copenhagen, Denmark.  There’s no moon, and it’s calm with a fresh temperature of 10°C. Both transparency and seeing are hovering around medium (3-4/7 and 5-6/10 respectively), and the NELM in my suburban backyard tonight is ~4.8^m (SQM 17.5).

 

     I have my small Vixen 80mm refractor out, with a SA100 grating in the nose of my PGR CM3 mono camera, and I now aim my scope at Deneb (α Cygni), which is conviently high up (~65° Alt) towards the SE. As I focus on Deneb, it becomes evident that the seeing is rather wavering, which is bound to introduce some noise in my observation. I try different settings of camera exposure (½-1s) and gain (400-250), until I get a reasonably bright 0. order image and 1. order spectrum on the CCD (and the PC screen).

 

     Here’s the Deneb 1. order spectrum from the SA100 grating, as it shows up on the PC screen, calibrated in Angstrom to a dispersion of 7.6A/pixl;  I have not corrected the luminosity curve for instrument response, as the seeing this morning is not good, and the quality of the data is correspondingly wavering. Maybe I’ll try that later – or preferably redo the spectrum with my 4” Vixen refractor in better seeing. Instead I’ve overlaid the spectrum of the type A2 Ia supergiant Deneb with the type A2 V main sequence/dwarf spectrum of Vega – for easy comparison of these two early type A stars :

     

*click*

 

*click*

 

Allan

Edited by AllanDystrup, 16 July 2017 - 09:21 AM.

[robin_astro]

Hi Allan,

 

Another nice result. Picking up the Balmer lines in Deneb is quite difficult with the Star Analyser because they are so narrow (around 2-3A compared with 20-30A for Vega).  Significantly narrower than the resolution of the Star Analyser which is why they appear so much weaker. Also Deneb shows H some alpha emission which tends to cancel out the absorption at H alpha at Star Analyser resolution which is why H alpha seems to be absent in your Deneb spectrum.

 

Cheers

Robin

Edited by robin_astro, 16 July 2017 - 12:47 PM.

[AllanDystrup]

SA100 grating – Mono Camera – K-Stars

 

     It’s mid July, 2 hours past midnight local time, and I’ve set up my classic 4” Vixen FL-102S  refractor in my suburban backyard. The transparancy and seeing are both just above medium, with a thin high haze, a light wind and a NELM ~5^m. Tonight my goal is to have a look at spectra of type K stars, such as the giant Arcturus (α Boo) and the two main sequence dwarfs in the binary 61 Cygni system.

 

     Arcturus is hard for me to catch from my backyard at this time of the year, because it is low (~17°) at the W horizon, and as it is sinking fast, it usually gets entangled in my neighbour’s large larch tree. 61 Cygni however is easy to catch, high up at 70° Alt in the SE wing of the Swan.  Already in my Zeiss C60/250mm finder @ 6.25x magnification, 61 Cyg shows up as a small but clearly separated, beautiful golden double star. At 73x in my ATC-K32mm finder eyepiece on the 4” refractor, the binary is a splendid showpiece of two similar bright, deep orange components with a Sep of 32.5” in 153° PA.

 

     The 61 Cyg A-B binary stars are both main seq. K-type dwarfs showing very weak Balmer H-absorption, but strong absorption in neutral metals such as Na I, Ca I and Fe I.  Cygnus 61 A is type K5V (~4.5 K.Kelvin surface temp) while 61 B is type K7V (~4 K.Kelvin).  61 B shows slightly stronger absorption for Ca I and Mg I. 61 Cyg A-B  are among the coolest and dimmest ordinary hydrogen-fusing dwarfs visible to the naked eye.

 

     Their luminosities are 15 and 9% solar, their masses 60 and 50% solar, and their radii 65 and 60% solar. As dimmer versions of solar type stars, they have magnetic cycles similar to that of the Sun, the brighter 8 years, the fainter 11 years. As starspots move across the stellar faces, the brightnesses vary, yielding rotation periods of 35 days. Magnetically active, each is also capable of popping a mighty "flare" that causes a sudden brightening [quote: http://stars.astro.i...du/sow/hrd.html]

 

 

*click*

 

*click*
 

*click*

 

Allan
 

Edited by AllanDystrup, 20 July 2017 - 03:02 AM.

[DHEB]

Thanks Allan for these most interesting reports. I have some interest in spectroscopy and am reading this thread with great interest.

 

Just a question: why do you observe 2 hours after midnight? Technically, these days the Sun reaches its lowest altitude under the northern horizon at about 00:50 local time (here in Sweden, with summer time, which is the same as for you in Danmark). This is the darkest hour of our bright summer nights. Now, you observe at almost 02:00, which at least here at 60°N is already too bright. I understand you are at about 55°N which is some difference (darker than here) but still seems bright to me. Am I missing something? Not being critical, just curious.

[AllanDystrup]

Hi cincosauces,

 

     I'm at 55°52' N, in a suburb just N of Copenhagen, so that's a rather southern latitude compared to most of Sweden . I am at UT(GMT)+1, and with our 1h daylight saving time on top, we now have nautical twilight in the range of ~22.30 - 04:00 local time. The darkest stretch around solar midnight (astronomical twilight) is at ~00:00 - 02:30 local.

 

     Normally I don't keep awake to observe at this time of the year, and I do not normally set the phone for a wakeup call. I just habitually wake up somewhere around  01:30UT+1 local DST (ie ~solar midnight +/- 1 hour), and then relaxed and in good peace and quiet make a cup o' tea while saddling my small Vixen horse, for riding into the night, -- which often in a couple of hours turns into sunrise in a choir of blackbird song (still a few singing here in late July...)

 

     Anyways, for spectroscopy, astronomical twilight seems to be dark enough, at least as far as I have tried it out this summer. And even nautical dusk/dawn has worked for me, when targeting the very brightest stars, such as Vega.

 

Cheers from your neighbour @ ~56N

Allan

Edited by AllanDystrup, 20 July 2017 - 08:12 AM.