Wed, 11/28/2012 - 17:07
I thought it might be helpful to start a separate topic on the Star Analyzer so that people looking for that don't have to page through the other topics in this forum.
My question was mainly for Jeff, but if anyone else can answer I am happy to hear from you too.
The Star Analyzer SA-100 says that it can be used as a normal 1.25-inch filter. Has anyone tried mounting it in a filter wheel? I'm curious because I think that it would improve the ease of use if it was just another slot in the filter wheel that people could have as an option for imaging.
-John
Hi John,
Yes, people have mounted the Star Analyser in filter wheels. This forum was started by someone doing that. The problem was orientation and spacing. As you know the spectrum should be as horizontal as possible (requiring a rotation of hte Star Analyser relative to the CCD) and there are optimum spacings for the Star Analyser and CCD, one to include the zero and first order spectra and one for just the first order.
If the orientation and spacing can be set properly, it doesn't matter where the Star Analyser resides. The big problem is what do you do with the other filter wheel positions? If you are doing astrophotograph and/or photometry, there may be a focusing problem.
Using a filter wheel is certain doable if done properly, however, it may be better to think a speparte system. Even a 60 mm telescope piggyback with a Star Analyser could do a lot of work. As you know it is only when you go to higher resolution that the bigger telescopes are need even on fairly bright stars.
Jeff
This makes sense. The image of the star should be focused at the star analyser for it to work. Therefore the CCD needs to be 3-5 cm behind the analyser so that the spectrum spreads out enough to get good resolution per pixel. And in order to re-focuse to do imaging you need at least the 3-5 cm travel to switch back and forth. Thinking that through, I agree that can be a tall order.
I was hopeful that the analyser software could deal with a spectrum that did not fall exactly along a CCD row. It is disappointing and adds a dimension of difficuly for filter wheel mounting, if the software doesn't handle a mis-aligned grating axis well.
The piggy back idea sounds like a good one but it requires a second CCD camera.
I was asking because I was thinking about getting my hands on one of these myself so I was more directly familiar with it. In my mind I was thinking that this might have some of the advantages of an objective prism, without the fuss of mounting a big piece of glass on the front aperture of your scope, but I'm realizing I had some bad assumptions in my mental comparison there.
Hi John,
The RSpec software has the ability to rotate the spectral image so it can be adjusted in software. It is still best to get it as horizonal as possible without software rotation. Somethings the software does strange things with the rotated spectrum.
As for a separate CCD camera for a piggyback unit, the Orion Star Shoot G3 mono has recently been on sale for $399 with free shipping and no tax. I use one of these with my Lhires III. It has the same chip as the DSI Pro II, but also has TEC cooling. You could also use a DSLR with the piggyback scope. Even web cams have been used. The G3 is what I would recommend. See
http://www.telescope.com/Astrophotography/Astrophotography-Cameras/Orio…
Jeff
[quote=uis01]
This makes sense. The image of the star should be focused at the star analyser for it to work. Therefore the CCD needs to be 3-5 cm behind the analyser so that the spectrum spreads out enough to get good resolution per pixel. And in order to re-focuse to do imaging you need at least the 3-5 cm travel to switch back and forth. Thinking that through, I agree that can be a tall order.
I was hopeful that the analyser software could deal with a spectrum that did not fall exactly along a CCD row. It is disappointing and adds a dimension of difficuly for filter wheel mounting, if the software doesn't handle a mis-aligned grating axis well.
[/quote]
Rather late in replying to this I know but I am still getting hits on my website from the thread so it is probably worthwhile correcting and clarifying this to avoid any confusion, particularly since the new Star Analyser 200 model designed for use in filter wheels has just come out.
The focal plane remains at the CCD. You do not refocus the star image at the Star Analyser plane. The unfocussed beam is dispersed at the Star Analyser and the various wavelengths all come to focus on the CCD (approximately, there are some aberrations, the limitation of the simple converging beam configuration) There will be some small refocussing needed due to different optical thickness of the Star Analyser relative to other filters and the fact that the focal plane of the spectrum is slightly tilted but nothing that a normal focuser would not be able to cope with.
Most (All?) spectrum processing software can make geometric corrections to the spectrum image. to correct for tilt, slant, smile etc for example. It is advisable though to rotate the grating to make the spectrum as horizontal (or vertical) as possible to avoid the risk of artifacts, particularly if a small aperture short focal length telescope is used where the spectrum may be very narrow and cover only a couple of rows. This is achieved when using a filter wheel by unscrewing the Star Analyser slightly and fixing in position eg with plumber's PTFE tape on the thread or a spot of hot melt adhesive. The SA200 also has a mounting kit for wheels which take larger filters and this allows the grating to be fixed in the correct orientation.
Note It can be useful to be able to rotate the spectrum relative to the field to avoid contamination of the spectrum from background stars when measuring faint objects in crowded fields but by preference the camera and grating should be rotated together so the spectrum remains horizontal (or vertical) in the field.
Rotating to avoid background stars can be inconvenient and difficult to achieve in robotic setups of course so an alternative in this situation is to use 2 gratings in the wheel at 90 or 180 deg to each other to maximise the opportunity to miss background stars or at least identify where contamination may be occuring.
Cheers
Robin
Yes, As it happens it is the subject of a current thread on the star analyser group.
http://tech.groups.yahoo.com/group/staranalyser/message/2684
This is my standard Star Analyser setup but with the filter wheel spaced at the optimum distance from the camera. (For faint objects you also need to arrange for the camera+wheel to be rotateable to avoid interference from background stars/spectra. )
Cheers
Robin
Thanks! That is
Robin,
Thanks! That is helpful. As I said to Jeff, I was thinking of getting one of these myself to play with and your experience is really useful for helping me uncover the bad assumptions I may be making.
-John
This is what nova Cep 2013 looks like through the Star Analyser. (Taken on 6th Feb, v mag ~12)
The equipment setup used is described here
http://www.threehillsobservatory.co.uk/astro/spectra_42.htm
Dispersion is 17.2A/pixel and the spectrum was focussed at the H alpha line. It was taken as a test after a rebuild and the collimation needs tweaking.
Robin
deleted duplicate post -
note to webmaster - please add delete post option.
It's been a long time since this thread was active, but I wanted to update it with some developments that have occurred in the interim.
The original Star Analyser is a 100 line/mm grating. In early 2014, a 200 line/mm version version of the Star Analyser grating became available.
This higher line/mm version allows it to be mounted more closely to the sensor, something that is often needed when using it in a filter wheel.
The SA-200 is still mounted in 1.25" filter cell, but the cell has a lower profile, allowing it to be used in the more confined clearances that filter wheels sometimes have.
And, there is also an optional, inexpensive mounting plate that can be customized for filter wheels that are designed to take un-mounted drop-in filters.
For more details, see http://www.rspec-astro.com/star-analyser-200. (Also note the calculator on that page whic can help you determine whether the 200 or 100 is optimal for your specific equipment configuration.)
Tom Field
When I first became interested in stellar spectroscopy, too long ago to think about, I purchased an early version of the SA 100. I was able to make some interesting observations of bright stellar objects. I believe that my resolution was about 40 Angstroms per pixel. I was reducing the data using the Line examination tool in Maxim. It worked well enough for casual examination. The grating was good for survey work and I was able to get my feet wet on what it was going to take to perform this type of spectroscopy.
I soon purchased one of the [then new] SBIG Self Guiding Spectrometers [SGS]. I was able to get much deeper than I was able to go with the SA 100 and my resolution increased immensely [~2.3 Angstroms/pixel]. A software package was included with the SGS that allowed for a reasonably rapid reduction of the spectra images. Calibration was not easy. VSpec had just become available online. It was a free package for spectroscopy data reduction. It was a long learning curve. Well written software but not intuitive...at least not for me.
I still have the SGS and have used the it for several projects, but the problem always come back to reduction of the image. Even after the image was reduced and calibrated, it was a chore putting the data into a format that was easy to share. A few years ago, Mike Simonsen messaged me that he had seen a new software advertised at the NEAF meeting. The software was for reducing spectra data. That was RSpec. It fulfilled all of my meager requirements. Calibration was a breeze. It was intuitive. While not freeware, the price was within the range for amateurs.
Being a person that doesn't like to keep unused equipment gathering dust, I sold the SA 100 on Astromart. I will admit that it was not one of my smartest moves. While the SGS is a great instrument for the amateur, it is not something that you just slap on the business end of a scope and start taking spectra. There is a reasonable about of planning involved. I wished on several occasions that I had kept that old SA 100. It would have allowed me to do a quick survey spectrum on a star and see if it was worth the time and energy to get out the SGS to take higher resolution spectra.
So, at the Spring SAS meeting, I parted with some of my retirements funds [hear creaking sound as wallet opens and moths escape from many years of confinement] and purchased one of the SA 200 gratings. The grating is inserted, when needed, in the optical train just ahead of the filter wheel in a filter drawer. I have use it on two occasions since the meeting and it has performed admirably. I am getting spectra with a resolution of 19 Angstroms per pixel (using an SBIG ST-10XME at 1x1 binning), which is plenty good for survey work.
Hi SDB,
It is encouraging to see observers make good use of the SA200. I know you and others do not want to change your system to do spectroscopy and I can certainly understand that. Experimenting with the SA200 is an excelllent way to break into the field of spectroscopy. However, once you are hooked you will want to go to a higher resolution system. It's kind of like observing double stars. When you reach the limit of resolution with your tlelscope, you will need a larger aperture to resolve closer doubles. For advancing in spectroscopy the ALPY 600 is an excellent choice for the next step. But that will require a dedicated tescope for the spectrograph. A jump up from there would be a LISA or Lhires III. If funds are not an issue and you have a large telesope the eShel spectrograph is amazing.
One point I should note is what you are referring ot as resolution is acutaly dispersion (angstroms/pixel). Dispersion is inversally related to resolution. Resolution is defined as the width of a line (usually the FWHM) in angstroms divided by the center wavelength. That means a system's resolution may vary with where the line of interest is in the spectrum.
Keep up the good work and keep us posted on your experiences with the SA200
Jeff Hopkins (187283
HPO
Hi Jeff, here are my results using a SA200 with an SBIG ST-10XME from Mt. Pinos CA on Quasar 3C 273:
http://www.astroimage.info/images/3C-273%28Mt.Pinos%293.jpg
I summed 5 x 5 minute guided exposures through a 33cm apeture classical cassegrain from a moderately dark sky site (90 miles NW of Los Angeles at 8,000 feet elevation). The computed average z-shift in the balmer lines in Ha, Hb, and H-gamma came out to be 0.161, pretty close to its published value of z=0.158. I shot quasar 3C 454.3 (Z=0.86) in eruption near 14th magnitude, but unfortunately, a nearby star's spectrum overlapped it's Red and IR range where all the interesting H-line shifts are seen......maybe I'll try again, but it'll be a long exposure set when it goes back to 16th magnitude!
James
For me, the idea of having the SA-200 in my filter wheel is its biggest attraction. I do mostly PT and the idea of taking the camera and filter wheel off the telecope is a non starter. Having my PM filters and the spectra available at the click of a filter position makes it something that I might do. I have a DSS7 spectragraph, which I bought to take spectra of led's and light box bulbs. I just am never going to take my PT setup apart to get a spectra. Spectra is a fun activity for me. For those serious about it, the separate unit is clearly the way to go.
Gary
I have an ST-10 with a cfw-10 filter wheel. The two are closely mounted and focus compensated in MaxIm DL and ACP. I have an old Rainbow optics grating. What will I get by screwing that into an open slot in the filter wheel? I suspect bad focus. Is there a way to make it work or should I just concentrate on my LHires III on another telescope?
Marlin Costello
Hi Martin,
The Star Anaylser 200 (SA 200) is designed for use in a filter wheel. The SA100 does not work well in a filter wheel. The Rainbow optics is similar to the SA 200, but I have not heard of anyone uing it in a filter wheel. Now if you are just starting with spectroscopy the low-resoultion of the Rainbow Optics or SA spectrographs are great for learning. They can show the zero order and higher order spectra. Used with an eyepiece they are great at star parties. Processing of the low-resolution spectra will provide great practice and learning. RSpec is highly recommented for the spectrum processing.
Once you have master the low-resolution spectroscopy you will be ready to do some serious high-resolution work with the Lhires III. The high-resolution work is very different form the low-resolution work, but will provide much better data.
The spectrum window with a 2400 line/mm grating is only ~100Å to 200 Å wide. That means you need to experiment to make sure you know where you are looking. Unlike the Low-Resolution work, the Lhires III allows you move the spectrum window. It takes a bit of detective work to figure out the wavelengths of the lines you are seeing.
The two most popular lines are the bright red hydrogen alpha line (6562.81Å), which can be easily bracketed by two distinct neon lines (6532.88Å and 6598.95Å) using the built-in neon calibrator of the Lhires III, and the sodium D lines, two closely spaced very distinct yellow lines (5889.950Å and 5895.928Å).
I suggest spending a few hours on the bench witht he Lhires III before puttting it on the telescope. Until you are confident withthe Lhires III on the bench you will be frustrated and waste time at the telescope.
Good luck!
Jeff Hopkins (187283)
Hopkins Phoenix Observatory
Good point Jeff! I was within a hair of buying the Alpy600 medium resolution spectroscope as SAS, but I couldn't justify the cost and the difficulty changing set-ups btw photometry and spectroscopy. I agree that RSpec takes alot of the pain out of learning how to anaylze your spectrum. I will definetly invest in an Alpy600 of Lhires when I get a larger apeture scope away from the strong backround light pollution near downtown Los Angeles. I wonder if RSpec works as well with data on these higher end spectrscopes?
One thing I forgot to mention, I have my SA200 filter placed "backwards" in my filter wheel. Instead of screwing the filter on the side common to the other filter (towards the sensor) I have it screwed on the other side of the filter wheel away from the sensor towards the telescope. This gave me a little better dispersion (9.96 A/Pix) than the normal position.
James
Hi James,
I am in complete agreement about switching between photometry and spectroscopy. During the Epsilon Aurigae Campaign I tried that, but gave up soon as it was way too hard. It takes time to get each setup properly and switching becomes a great chore. I now have my ALPY 600 setup on my 8" LX90 and the Lhires III on my 12" LX200. Once set, I don't change them.
I have 2- SA100's, an ALPY 600 and Lhires III with both 600 and 2400 l/mm gratings. While I really like the ALPY 600, it is easy to set up and use and produces better spectra than the SA and think it is definitely a big step up from the SA, it still does not have the resolution I want. It cannot even resolve the sodium D lines. Also, for my work on the Orion Project, Alnilam's hydrogen alpha variation just cannot be seen with the ALPY 600. The Lhires III, even with the 600 l/mm grating has better resolution than the ALPY 600. But I use the 2400 l/mm grating most of the time. You do not need a big telescope for it. In fact the Lhires III is supposed to be optimized for a 12" scope. I know people use 16" and larger scopes, but if you are approaching meter sized scopes you should consider a different spectrograph and that becomes something that is way beyond the financial reach of most of us.
Since I live in Phoenix the light pollution is probably close to that of LA. As such I tend to only do high-resolution spectroscopy on bright stars with my 12" LX200 GPS with my Lhires III 2400, usually 3rd magnitude or brighter. Because the stars are so bright, the exposure times are reasonable (15 minutes to 2 hours) and the light pollution has a very minor effect on the results. The quest for high-resolution spectroscopy of fainter stars is best left to the professional observatories in dark sky locations and with meter + sized telescopes. There are many bright stars that are very interesting and deserving high-resolution spectroscopy. Most of these are very ignored and need observations. The high-resolution allows precise measurements of radial velocities and changes in the morphology of the lines is easily seen.
Regarding using RSpec for the high-resolution work, I use it all the time and am very happy with it. I find it much easier than VSpec, ISIS and some of the other freeware.
Good luck with your projects. I look forward to the fall observing season. Right now it is way too hot in Phoenix, even at midnight.
Jeff Hopkins (187283)
Hopkins Phoenix Observatory
Star Analyzer SA-200 Enthusiasts,
I have put together two more resources that you might find useful and/or inspiring. The first is a tutorial on how to use SAOimage DS9 (free software) to extract spectra from images taken with your filter wheel grating. The link for that is:
https://edocs.uis.edu/jmart5/www/SpectroscopyWorkshop/gratingtutorial.html
The second page is a growing list of good examples of use of a filter wheel grating that I have collected through my own experience. We all know that the resolution of these grating is too low to do any meaningful Doppler measurements or probe the state of the art of what professional stellar spectroscopists do. However these gratings open up unexplored regimes in time-domain spectroscopy and identifying transients. I have mine on a 0.5 m (20-inch) telescope so I can chase faint blobbies that the rest of you might struggle to get, but I am surprised what I can do and the technology is improving and I am also hopeful that putting these examples out there will inspire others to push their system to do some interesting new projects.
The link for those examples is:
https://edocs.uis.edu/jmart5/www/SpectroscopyWorkshop/gratexamples.html
Note that the last one on that page is an ID of a transient ASASSN 14dz where we actually beat the Asiago group to the classification by at least 20 hours, but we weren't confident enough at that time to report it. Spectroscopy is one domain where ASASSN and other projects become our collaborator and not our competitor.
Hi John,
Just a couple of points on your SA200 writeup.
The mounting kit for wheels which take unmounted filters also includes a threaded ring to clamp the SA200 in the pre-punched mounting plate. This aalows the SA200 to be aligned for example when fitted to filter wheels designed for square filters and should avoid the need for silicon glue except perhaps in rare circumstances where you might want to laminate several plates together to drop the position of the SA200 in the wheel.
Vgnetting of the field can be seen with larger size CCD sensors in some circumstances but since the vignetting occurs before the light is dispersed into the spectrum, no light lis lost provided the light cone from the zero order star image is placed within the unvignetted area. (You can see this effect in some of your images where despite the vignetting at the edes of the field, the spectrum continues into this region undiminished. Indeed a slit spectorgraph is of course an extreme example of this where the slit vignettes the entire field except for the narrow region of interest.) There is normally plenty of room in the CCD field to achieve this, and still include the full spectrum in the field.
Cheers
Robin
A nice reference of the Star Analyser used to classify transients is Gianluca Massi's robotic setup using a Star Analyser in a filter wheel (an SA100 currently at very low dispersion but he is considering an SA200 to increase resolution) His setup has been running for around 18 months. He monitors the unconfirmed transients list and picks up any which are within the range of his equipment. He has a good string of successes under his belt
http://www.virtualtelescope.eu/spectroscopy/
Ultimately sky backgound limits how faint you can go with slitless systems though and background subtracton for supernovae in particular is frought with problems when operating without a slit which is why I have been testing an ALPY spectrograph which I have modified for lower resolution. The results so far look quite promising with spectra of mag 16+ supernovae at snr good enough for identificatrion purposes with a 280mm aperture
http://www.threehillsobservatory.co.uk/astro/spectroscopy_20.htm
Robin
Thanks for the clarification Robin. When I thought about it you are correct that my field is vingetted by the grating but there is no loss in throughput for the spectrum of a target where the zero order image for the target falls inside the area covered by the grating. That is: all the light that the telescope collects for a star in the part of the field occupied by the grating makes it through the grating to the image plane. So there would be no improvement in through-put for a grating that covers more of my field.
This raises the question in my mind for an experiment. I might try masking the area of the grating in order to eliminate overlapp in a crowded field.... effectively making it more like slit spectroscopy. I think that might be an interesting trick for more advanced users to try out.
Hi John,
Yes that has been tried, though it is only partially successful as the vignette from a mask which is not at a focal plane is not sharp like an in focus slit. (you also need to take particular care with background subtraction as the sky background will have a gradient) One possible arrangement (assuming the spectrum is horizontal,) is to place the star at the edge of the field and use a vertical blind which you can move in from the side. That way you can see the full field and position the zero order before moving in the mask so the spectrum as far as possible falls in the fully vignetted area. With your 50mm wide filter aperture you could offset the SA200 in the holder so you can place the zero order at the edge of the field without vignetting.
thne next stage is to arrange a secondary focus to put the mask/slit at, as here for example
http://www.threehillsobservatory.co.uk/astro/spectroscopy_18.htm
and from there you can even go the whole hog and build a fully collimated arrangement aout of junk box bits and pieces
http://www.threehillsobservatory.co.uk/astro/spectroscopy_19.htm
Robin
Hi John,
Your mention of using a mask prompted me to recheck the situtation with the SA200 (which is potentially better than for the SA100 as the optimum mask diameter is half that for the SA100 for a given length of spectrum)
The calculation confirms that for reasonable focal ratios (ie < f10) the full spectrum up to 7500A still lies fully within the penumbra of the out of focus mask, even for the SA200, so the background is only partially eliminated even at the red end. This is with an optimally sized mask at the plane of the SA200 with a width matching the light cone of the zero order image at that location.
Cheers
Robin
Correction - that should read ".... focal ratios (<f6.5) "
At f10 and with an optimally sized mask, the spectrum beyond ~5000A will be in the totally vignetted region, so a potentially useful arrangment with the SA200 used with higher focal ratios. The downside is with higher focal ratios you need higher dispersion (and greater grating to camera distances) to get the same resolution because of the larger image star size
Robin