Long focus cassegrain optimised spectrograph, mixed optics

Design, construction, tuning of spectroscopes
Information and discussion about softwares (telescope remote, autoguiding, acquisition, spectral processing ...)
Post Reply
Hamish Barker
Posts: 226
Joined: Thu Jun 20, 2019 12:11 am

Long focus cassegrain optimised spectrograph, mixed optics

Post by Hamish Barker »

I am thinking of building a spectrograph optimised for long focus, large aperture, high f ratios, for C14 but also possibly to attempt to convince AAVSO to install on the UC Mt John 61cm f16 Optical Craftsman fork mounted cassegrain telescope, optimised for faint transients, as deep as possible.

Just exploring concepts here and throwing it open to the forum.

Basics - required slit size:

First, the front end interface:
C14 f11 = 3740mm focal length, 3" seeing (our club observatory, sea level, adjacent to the sea so assume 3" seeing, giving a typical star FWHM circa 54 micron. Assume circa 17 mag/arcsec^2 light pollution.
OC 61cm f16 = 9760mm, typical 1.5" seeing at Mt John = 71 micron FWHM star. Best seeing is 1" which would be 47 microns. Assume Mt John has perhaps 20 or perhaps even 21 Mag/arcsec^2.

So perhaps the same slit size around 70 microns might be ok.

To get best faint object performance, my objective is to arrange significant focal reduction at the camera end to operate as fast as possible.

Components on hand: SA200 transmission grating, 300mm focal length f4 spherical mirror, f1.9 pentax lens. The big question is, how good is the camera lens over say a full 3900-7000A range? Probably not great. Maybe a more modern 50mm prime might

Layout:
I'm thinking of an off-axis mirror collimator, as I have a 75mm f4 mirror (300mm focal length.) At f11 this would give a collimated beam of 27mm, at f16 would be 18.75mm. Both of these would fit through the SA200. But at f11, depending on the effective grating to camera distance (SA200 mounted directly to the front of the 50mm lens, say 15mm effective?), would be vignetting the telescope down to 310mm effective aperture.

For the c14 and a 460ex camera, S/N 10 limit at 3x1200s exposure would be Mag 14.7 for 5500K star and 0 bolometric correction, assuming 15 micron camera resolution. Limiting factor is the sky glow at 17 Mag/arsec^2.
For the OC610 and the same camera, at 21 mag/arcsec2 and 1.5" seeing, S/N 10 limit would be mag 17.3. Sky background is still a little more than other noise sources.
Both of the above are at 2x2 binning, 5e- read noise, 0.001e-/se/pixel.
resolution is a little low, maybe R=275 at 15 micron camera resolution and 10 micron collimator resolution. I'm not sure if this camera resolution would be achievable.

Questions:
- How hard is it to set up the mirror for off axis collimation?
- Is it a bad idea to use a mixed optic system like this for the intended duty (wideband, faint objects, high f ratio input)?
Robin Leadbeater
Posts: 1926
Joined: Mon Sep 26, 2011 4:41 pm
Contact:

Re: Long focus cassegrain optimised spectrograph, mixed opti

Post by Robin Leadbeater »

Hi Hamish,

When designing a system from scratch particularly something for remote operation there are many practical pitfalls which even designs from experienced spectrograph manufacturers have had problems with (primarily thermal and mechanical stability for example). Why not try an off the shelf solution like the ALPY600? Focal reduction to f4 and a 35um slit would give the same resolution as you are proposing in a highly stable design already proven for remote operation eg by 2SPOT. This would be suitable for a large aperture setup (how large dependent on seeing) Such a system based around a long focal length and reducer would inevitably suffer from some chromatism (for spectroscopy all reflective systems are strongly preferable) but precise flux calibration is not critical for transient identification.

Choosing the right resolution is critical. For example I chose R~150 for my modified ALPY 200 specifically so I could go as deep as possible (giving a limiting magnitude at SNR 10 ~17-18 with a C11) for preliminary transient identification (mainly supernovae) and this resolution was then also used by ZTF in their SED machine. I could have gone for a higher resolution but then the limiting magnitude and therefore the number of potential targets was reduced significantly. This is too low though for example to distinguish with certainty between narrow line supernovae and impostors for example. I would recommend talking to professionals in the field looking for a sweet spot of resolution/limiting magnitude where professionals currently have poor coverage. My strategy was quite successful, picking up a number of supernovae that were not being classified by professionals and the same strategy has been copied by a couple of other amateurs but opportunities have significantly dried up as SED machine and more recently other high productivity professional spectroscopic classification systems have come on line.


Cheers
Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Robin Leadbeater
Posts: 1926
Joined: Mon Sep 26, 2011 4:41 pm
Contact:

Re: Long focus cassegrain optimised spectrograph, mixed opti

Post by Robin Leadbeater »

Robin Leadbeater wrote:
Why not try an off the shelf solution like the ALPY600? Focal reduction to f4 and a 35um slit would give the same resolution as you are proposing in a highly stable design already proven for remote operation eg by 2SPOT. This would be suitable for a large aperture setup (how large dependent on seeing) Such a system based around a long focal length and reducer would inevitably suffer from some chromatism (for spectroscopy all reflective systems are strongly preferable) but precise flux calibration is not critical for transient identification.
With your native 70um spot size you would only need to reduce to ~f8 to give a 35um spot size. This sort of reduction is possible with manageable chromatism (Although not ideal, I run my C11 at f5.5 just using a standard Meade branded reducer and while there is certainly chromatism it is ok for relative flux to typically better than +-10% over 4000-7000A provided the telescope focus is stable between reference and target.)

Cheers
Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Hamish Barker
Posts: 226
Joined: Thu Jun 20, 2019 12:11 am

Re: Long focus cassegrain optimised spectrograph, mixed opti

Post by Hamish Barker »

Hi Robin, thanks for the constructive input. To paraphrase the old saying about birds and bushes, fair call on the fact that a useable proven instrument in the hand is worth two (or more) prototypes on the workbench. :-)

Indeed, an alpy600 (or modified per your alpy200) is the model of solid and stable spectrograph. I was thinking that lens focal reduction to meet optimum for an alpy would be tough in terms of chromatism, plus that would be say 2 or 3 (I assume a strong focal reducer would need at least 2 elements, plus the alpy's collimator lens) additional lenses upstream of the grating which wouldn't help at the blue end. I was thinking that by making at least one of the elements reflective and using the mirror, overall chromatic performance could be better, number of optical surfaces reduced, and slightly better sampling with typical 3.7-5 micron pixels with or without 2x2 binning for ccd (not for cmos). Avoiding 3d printed parts where possible and possibly using a CFRP tube for the spectrograph body might help to stabilise against mechanical and thermal expansion issues. Keeping the spectrograph with a fixed transmission grating rather than rotating grating could help to keep any wavelength drift easily compensated.

Thanks especially for the tip about the R value which was useful. In that case, my proposed resolution using the SA200 could be ok.

An advantage of prototyping on the club c14, if it performs ok at f11, could be that f16 should be less demanding of the optics.

A bonus of using the native incoming f16 and 70-75 micron slit could be that a pure reflective knife edge slit with air gap might be less problematic than I read may have been the case in the early LHires air gap slit, which was presumably narrower and therefore more sensitive to slit edge un-evenness, causing transversalium.
Robin Leadbeater
Posts: 1926
Joined: Mon Sep 26, 2011 4:41 pm
Contact:

Re: Long focus cassegrain optimised spectrograph, mixed opti

Post by Robin Leadbeater »

The ALPY design is interesting. The compact in line design does make it relatively easy to keep it rigid but I noticed another feature (possibly accidental) that perhaps could be employed in other designs to improve mechanical stability without needing to add extra weight.

If you look at the ALPY design the critical (but light weight) elements between slit and camera in the core module are effectively free floating with the main loads of the imaging and guide cameras being taken by a separate outer tube and box connected to the scope. The only common connection between the outer part and the inner core module is at the camera mounting meaning there is no load on the critical components other than their own (minimal) weight.

I have often wondered if the same approach could be applied generally with an inner and outer box construction with just one common wall where the imaging camera is mounted. The critical components would be mounted in the inner box which would only need sufficient rigidity to carry just the load of the relatively low weight of the spectrograph components without flexing and constructed from materials with good thermal stability. The telescope, guide and imaging camera would be mounted off the outer box, the weight of the cameras being taken by this outer box. (Flexure between telescope and slit is not important as this is taken out by guiding and while flexure between the guide camera and slit is important for maintaining the guide position relative to the slit, it is less critical than the relationship between slit, grating and imaging camera)

Cheers
Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Robin Leadbeater
Posts: 1926
Joined: Mon Sep 26, 2011 4:41 pm
Contact:

Re: Long focus cassegrain optimised spectrograph, mixed opti

Post by Robin Leadbeater »

Hamish Barker wrote: A bonus of using the native incoming f16 and 70-75 micron slit could be that a pure reflective knife edge slit with air gap might be less problematic than I read may have been the case in the early LHires air gap slit, which was presumably narrower and therefore more sensitive to slit edge un-evenness, causing transversalium.
I think most (all?) designs of spectrographs for amateur use now use reflective slits printed on glass rather than air slits. These have excellent uniformity and work well to at least 3800A, though I believe the UVEX may use a special UV transparent glass substrate to extend the range further into the UV. The transmission/reflectance of all the optics comes into the equation though if working in that region.

Cheers
Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Hamish Barker
Posts: 226
Joined: Thu Jun 20, 2019 12:11 am

Re: Long focus cassegrain optimised spectrograph, mixed opti

Post by Hamish Barker »

Yes of course glass reflective is easier to keep clean since the actual slit is in the dust free internal part of the spectrograph.

The idea of air slit is just as a later possible improvement, since a 70-100 micron gap and the internal focal reduction makes it feasible.

I see that the alpy foil plate is actually an air gap slit.

I have an aluminizer so will make my own slit plate, possibly will try using an antireflection coated window as substrate. With the long focal length of the oc61 or c14, I'll make the plate quite large ( circa 30mm dia) so as to give a decent field of view.

Re internal/external structure, yes I have thought about this in the past when trying to build a spectrograph. A further bonus of the alpy structural design is that it is a tube, so torsional loads do not distort it., And it's much stronger than typical spectrograph bodies which are often 3 sided boxes with a lid secured with screws.

My L200 for example is built of 1 or 2mm aluminium plate, and it's simply not stiff enough at the attachment for the guide optic. Really it needs 3 or more mm plate there, as any flex means that the location of the slit image in the guide ccd can tend to move a bit. I improved it from the original construction but still could be improved.
Post Reply