Using a focal reducer with star analyser

[the rev]

Hello,

I am using a 5 inch Mak-cas telescope with my starlight MX9 and a star analyser to take spectra. Recently I added an extra spacer which has improved my data collection. I am now wondering would it be useful or not to use a focal reducer or would this either make the image too small in my scope or move it too far away from the chip? I have attached my recent images of Alhena, Mebsuta and Mekbuda

Kate

[nytecam]

I've never[?] used my SCT @ native 30cm SCT f/10 for spectro via Rainbow grating but f/6.3 FR was fine. The length of the spectrum depends on grating's distance from the CCD sensor. In lowering the spectral resolution eg shorter spectrum - the spectrum is brighter and a shorter exposure is possible or will record a fainter object in a given exposure. The choice is yours.

Currently have a f/3.3 FR running @ f/3.6 and the Rainbow grating shows severe aberrations of coma and astigmatism so there are limits. Obviously the 100 l/mm Staranalyser is less sensitive to these effects than my 200 l/mm Rainbow grating which produces a spectrum twice as long - I think this is right

[Merlin66]

Around f5 seems to be the optimum for the SA100. This gives smaller star images and hence better resolution. Target the distance from the grating to the CCD to give 10A/ pixel dispersion.

[the rev]

Okay. So I've been told the f ratio of my 5 inch is f15, I have a cheap focal reducer which I will try in two different places. Firstly in front so it doesn't extend the distance between the SA and chip and then just behind the SA and will see which gives the better resolution.

Thanks for the help.

[robin_astro]

Okay. So I've been told the f ratio of my 5 inch is f15, I have a cheap focal reducer which I will try in two different places. Firstly in front so it doesn't extend the distance between the SA and chip and then just behind the SA and will see which gives the better resolution.

Thanks for the help.

Hi Kate,

Yep with a focal length of nearly 1900mm and such a small CCD there is the potential to improve resolution by adding a focal reducer (eitehr before or after the grating) to reduce the star image size. I will be interested in the comparison. In theory, provided you maintain the same dispersion, adding it after the grating should give the better result as it keeps the grating in the less converging beam, reducing the aberrations. Calculating the correct distance in this case though is more complex as the reducer also reduces the dispersion so you have to increase the distance between the grating and the CCD. This happens to some extent automatically though due to the extra length introduced by the reducer. The best solution is to try it, adding extra spacers as appropriate to bring the dispersion up to the maximum possible while still allowing the zero order and spectrum to fit on your CCD. Note if you mount the grating after the reducer (ie in the more converging beam) you should not reduce to more to more than f4.5 otherwise the converging beam aberrations start to get too bad.

Cheers

Robin

[robin_astro]

Note if you mount the grating after the reducer (ie in the more converging beam) you should not reduce to more to more than f4.5 otherwise the converging beam aberrations start to get too bad.

Actually that should of course read ..

.....you should not reduce to LESS than f4.5

but you know what I mean :-)

Robin

[robin_astro]

Yep with a focal length of nearly 1900mm and such a small CCD there is the potential to improve resolution by adding a focal reducer

Actually I am not sure exactly what size the CCD is in the MX9. (I was thinking of the MX5/7) I had a quick look and cannot find the info on the net. Do you know what CCD it uses? The other limit to resolution is the star image size relative to the pixel size. There is no advantage in reducing beyond the point where the star image covers less than 2 pixels as you will then be under sampling. Any way, in general reducing the star image size is a good thing provided the grating does not end up in a beam steeper than ~f4.5 and provided you are not under sampling.

Cheers

Robin

[robin_astro]

Actually I am not sure exactly what size the CCD is in the MX9. (I was thinking of the MX5/7) I had a quick look and cannot find the info on the net. Do you know what CCD it uses?

OK found the operating manual

http://www.astro.soton.ac.uk/~mjc/obs/mx916.pdf

According to this the MX9 uses the Sony ICX083 CCD. The pixel count an pixel size given in the manual suggests the camera 2x bins making the effective pixel size very large (23um) With this very large size pixel there is a big risk of undersampling with your scope. Can you confirm if your camera runs 376x290 pixels like the manual suggests or perhaps twice that? (Even 11um pixels are still on the large size though compared with most modern cameras so you may still potentially end up undersampled when using a focal reducer)

Cheers

Robin

[robin_astro]

According to this the MX9 uses the Sony ICX083 CCD. The pixel count an pixel size given in the manual suggests the camera 2x bins making the effective pixel size very large (23um)

or are you using the SXV M9 as described in your signature (a different camera with a different CCD)

http://www.sxccd.com/handbooks/SXV-M9%20handbook.pdf

Starlight Xpress camera notation is very confusing

If you can confirm which camera I can probably advise on the optimum arrangement for your scope

Cheers

Robin

[the rev]

Now I'm confused. I bought it second hand and was old it was the SXVF M9, but reading the manuals I think it is the MX916. There is no label on it, but the arrangement on the back looks more like the MX916.

Thanks for he advice.

Kate

[Merlin66]

Kate,

I used an ol' MX916 for a while....you can (under AstroArt) get it to run unbinned with 11micron effective pixels...

[the rev]

I use astro art to capture and stack so will check what it is running as.

Thank you

Kate

[nytecam]

Hi Robin - just a theoretical ?? Does the recorded spectral resolution depend on the area of the grating illuminated by a star? Kate working @ [native] f/15 with grating say 50mm from sensor = 50/15 =3.3mm diam star disk = 3.3x100 l/mm = 330 lines. At f/6.3 = 790 lines and f/4.5 = 1110 lines - goes this matter?

[robin_astro]

Hi Robin - just a theoretical ?? Does the recorded spectral resolution depend on the area of the grating illuminated by a star? Kate working @ [native] f/15 with grating say 50mm from sensor = 50/15 =3.3mm diam star disk = 3.3x100 l/mm = 330 lines. At f/6.3 = 790 lines and f/4.5 = 1110 lines - goes this matter?

Hi Maurice,

This is indeed an ultimate limit on resolution but from what I have seen this is rarely encountered in practical spectrographs. In the case you gave the theoretical reciprocal resolution R would be 330 but the limitations of star image size and aberrations from the converging beam would mean this would not be reached in practise. (Converging beam Star Analyser spectra are generally around R ~100-200) I do recall one case where the limit was approached though where a Star Analyser was used to make a classical spectrograph, I think with a refracting telescopes as lenses. The long focal lengths meant the full theoretical resolution with the Star Analyser fully illuminated (R ~3000) was almost reached and the Solar spectrum Sodium doublet split. Impressive for a 100 l/mm grating though a rather impractical design

Cheers

Robin

[nytecam]

Many thanks Robin for the explanation eg theory v practise via the Staranalyser/Rainbow gratings in its common application.

It's interesting that sunlight grazing a CD/DVD reveals the solar sodium duo lines to the eye without additional optics in the 2nd/3rd order spectrum