Llamatron - building my own scanning (helio)spectragraph

[NickK]

The first question is are we allowed to cover DIY solar scopes? (I ask this as a solar scope is 136.7 mW/cm2 at the objective lens, giving 136.7*50.265 = ~6.871W at the slit)

So I have been hatching a plan.. a cunning plan.

It involves three steps:

a) basic spectra

self-scanning 

c) <insert secret here>

I'll be using a thorlabs adjustable slot and reflective grating to start with along with the A80Mf, then using the ye olde 16IC to start with. I have a design and will be looking at the design for the collimation and focusing. Reason for using thorlabs is that it will happily cope with high power lasers (above the wattage here).

Why a spectrograph?

Well it will provide a spectrum for each pixel - effectively giving a 3D image, that shows a image for each wavelength - from CaK to H alpha. The addition of a scanner will mean it will scan a full image on an EQ6 in quick succession ("quick" means 20 seconds * 3000 images in the case of using a 383L! but I'll probably use the titan to start with..) whilst the EQ6 is tracking the sun/DSO.

Still have some more to go.. but the design so far:

The main point here is that the full defraction grating isn't used at the moment so I'll probably add a lens in to spread the slot beam.

[DIYASTRO]

That looks interesting!!!

Not into solar but spectra scope can be used elsewhere!

No little about solar or imaging of, etc, but if your idea of spectra per pixel is sound and potential 3d image then that is interesting.

would be tempted to have both mirrors independent in terms of lateral movement and thus the focusing lens config would need to be independantly moveable along the focus axis. that will make life easer in long run.

Keep us posted... interesting!

Boyd

[Merlin66]

Hmmmm

Interesting but...

Have you checked out the spectrohelioscope designs on the Yahoo group?

http://groups.yahoo.com/neo/groups/spectrohelioscopes/conversations/messages

This will give you some good ideas about moveable slits, their positions and the scanning techniques already in use.

The slits need to be positioned at an image focus......

(The reflected beam from the grating to the imaging lens will still be a collimated beam. )

Onwards and upwards

[NickK]

That's the interesting bit - the FISS scanner uses a collimated beam rather than placing the slit at the convergence crossover.

My thought on the collimated beam to the grating is that a beam expander will make the grating more efficient, then the focusing onto the CCD. I note that a couple of designs use one lens/mirror for both.

[Merlin66]

A grating works at maximum efficiency when in a fully collimated beam, the same size as the grating ie a 30mm diameter beam for a 30mm sq. grating.

You can use the imaging lens you show, but the incoming beam will be a collimated beam. Don't focus this lens on the grating.

[NickK]

I was planning a beam expander to take the collimated input beam and wider to a colimated output beam to increase efficiency. Two concave mirrors could do this - although the image would flip.

The output from the grating is then diverging for each slit, so a focusing lens or mirror then brings them back to focus on the CCD sensor plane.

[NickK]

Hmmmm

Interesting but...

Have you checked out the spectrohelioscope designs on the Yahoo group?

http://groups.yahoo.com/neo/groups/spectrohelioscopes/conversations/messages

I would but yahoo mandates a mobile number.. last thing I'm going to give someone is the option to spam me!

[NickK]

Signed up for the group merlin but it's still pending.. I did note that there's a link to resources that I've been reading on the phone.

[skybadger]

You'll find what you need there. It's funny how designs haven't changed in 50 years.

[NickK]

Hmm slit dispersal - still need to continue reading

[Merlin66]

Nowadays you can electronically scan the image of the sun with a fixed slit spectrohelioscope and then recombine the "slices" to form a complete image at the selected wavelength.

Majic!!

[PeterW]

Ken, did you make progress in the spectohelioscope kit I vaguely remember you were going to make (based on your spectroscope)?

Cheers

Peter

[Merlin66]

Peter,

The components for the spectrohelioscope (and the echelle) are still sitting in the "Project Box".

Busy at the moment supporting the new Spectra-L200 build and finalising a commercial compact FC120 design....

[NickK]

I've been looking at the efficiency profile of the thorlabs, now people say you want 90% on a 1800 rulings grating but naturally 90% may only be at a particular point in the spectrum. Although a 1200 ruling is acceptable

To me, it makes sense to target the highest efficiency in the main band that we're interested in. http://www.thorlabs.de/thorproduct.cfm?partnumber=GR50-1850 returns a decent spectrum around Ha and others.

[Merlin66]

That grating is blazed for 500nm - visual......

If you want to consider a dedicated Ha system, then a 750nm blaze would be better. Look at the GR50-1208.

[NickK]

Not much in the CaK region once you combine the slit efficiency and the camera efficiency at low wavelengths..

Letter Wavelength (nm) Chemical origin Colour range

A 759.37 atmospheric O₂ dark red

B 686.72 atmospheric O₂ red

C 656.28 hydrogen alpha red

D1 589.59 neutral sodium red orange

D2 589.00 neutral sodium yellow

E 526.96 neutral iron green

F 486.13 hydrogen beta cyan

G 431.42 CH molecule blue

H 396.85 ionised calcium dark violet

K 393.37 ionised calcium dark violet

Note that OIII and SII etc all fit within these.. so going above 1000nm for other nebulae wavelengths isn't probably going to help because the scope, the optics and the CCD aren't designed really to pick up these wave lengths. Now you probably see that I'd like to dual task this with DSO nebula too..

Just estimating the Ha line ... 500nm 1800 is 50%, 750nm 1200 is 70% but comparing the lower ranges such as CaK 500nm blaze is 50% but the 750nm is 10%.

Just comparing a couple of CCD sensor specs:

Kodak KAF 8300 (ATIK 383L): ~35% for CaK 393nm and again 35% for around Ha 656nm

Sony ICX424 (ATIK 16IC/Titan) : relative to 1.0 at 500nm, 0.6 at 400nm, 0.65 at 656nm

Sony ICX694 (ie ATIK 460EX): relative again.. 0.6 at 400nm, but a far higher 0.9 at 656nm.

Now the sony "relative" makes it a pain to compare against others but given it's generally thought of as being more sensitive than the Kodak with less nose.. let's guess at a 60% max sensitivity.

So as an overall suitability in terms of efficiency, I'm thinking that loosing some efficiency in Ha on the grating could then be made up by a lower noise higher sensitivity CCD (ie the 694) and then we get a broad range of spectrum available, so combining (for first order):

ICX694 1800 500nm blaze:

(60% * 0.9) * 0.50 = 27% in Ha

(60% * 0.6) * 0.50 = 18.5% in CaK 

ICX694 1200 750nm blaze:

(60% * 0.9) * 0.70 = 37.8% in Ha,

(60% * 0.6) * 0.10 = 3.6% in CaK 

Now Veio indicates that the 1200 is the best all around (and allows use of the other orders) but 1800 results first order but only green/blue use in the second, with that I've calculated the same table for the 1200 nm blaze for comparison (70% at 400nm, 50% at 656nm):

ICX694 1200 500nm blaze:

(60% * 0.9) * 0.50 = 27% in Ha

(60% * 0.6) * 0.70 = 25.2% in CaK 

But you're right that the Thorlabs graphs don't give second and third order efficiency but they are usually lower by about 1/2 from what I can tell (unless it's a Eschelle).

[Merlin66]

As you're finding the Cak and Ha are at opposite extremes... different gratings for different wavelengths - a bit like telescopes - there's no ideal scope!

[NickK]

I think given the dual use, I'll target the broad spectrum for now - it's possible to specialise at a later date. I understand this has other complexities (archo vs APO etc for full range). If needed I can increase the total light coming which would lead to higher absolute with a larger scope.

I'd like to mount this on the EQ6, but let's look at the focal lengths required for the resolution next..

[Merlin66]

The good news is all the APO v's Achro does is affect the focusing position for the wavelengths being used. Think about it as "extreme narrowband" - chromatic aberrations don't exist.

[NickK]

I thought I'd add some progress - thanks to santa..

Grating 500nm 1200l/mm reflective:

Adjustable slit:

Also have a 10um slit too.

I only got to check them before they were whisked away until the 25th..

I've been reading up and looking at lenses which work out about $7-10 but this will give me a good basis going forward.

[NickK]

I found this great link for gratings - it shows the order and spectrum overlaps nicely: http://www.eyes-on-the-skies.org/shs/SHS.html

[NickK]

The nice thing about Le Santa.. is you get your presents at midnight..

First up - the 1200l/mm 500nm blaze reflection grating - still sealed in it's protective canopy & air tight bag:

I tried to get a shot where it's breaking spectrum.. but lots of lights, DLSR and late at night..

Next up the variable slit with micrometer:

I'm pretty sure this is slit is getting fully closed and ail give me down to 5um

And finally the 10um fixed slit

[NickK]

Just a quick update - I braved taking the grating out of it's protection and attempted to take a couple of spectra by hand holding the slits and the cannon DLSR:

First - the ceiling light without any slit.

Next up the adjustable slit at about 15-25um (bit of blur):

Then I broke out the 10um fixed and managed to get this hand held:

Not bad

[Merlin66]

Nick,

It's always fun to play with new toys!

When you get some optics involved the results will be much better!

Onwards and Upwards.

[NickK]

So the bits for the Llamatron are taking shape!

It will be done in two phases:

Phase 1 - a standard spectrograph:

Phase 2 - getting the scanning and other bits involved:

Today the two 300mm lenses arrived from cameraworld.co.uk and these are the components laid out in the phase 1 configuration:

Scope: VIxen A80Mf - a 80mm 910mm fl f/11.3 achro
Thorlabs adjustable slit with micrometer (10um fixed slit next it it)
Small black lens: Cannon 300mm f/5.6 prime as collimator (£25!)
Grating 50x50mm 1200lpmm 500nm blaze reflective
Larger black lens: Cannon 300mm f/4 prime as focusing lens (£120 which considering the lens is rather good!)
Cameras
* Red square - ATIK 383L mono
* Red circular - ATIK 4000 mono 
* Blue circular - ATIK Titan

Next the optics side-by-side:

Left to right - 300mm 5.6, Vixen A80mf, 300mm f4

Both cannon lenses have retractable shields and the f/4 has a built in tripod grip and teleconverter slit (there's a lens that pops out). I'm impressed … when not on solar duties.. the F4 may be piggybacking the Pentax 105SDP!

The lenses, as those keen photographers may note, are Cannon FD (not EOS fitting) lenses with manual focus. I'll need to make an adaptor to get the lens iris to retract but there's FD to EOS adaptors.. and EOS to T-thread adaptors..