Calibrating a LowSpec3.0 Spectrum of Alpha UMa

[SteveBz]

Hi Fellow Astronomers,

I'm trying to get the hang of a LowSpec3.0 spectrometer.  The other night, everything fell into place and guiding on the slit worked (after a bit of star-hopping). The slit is set to 20 um and I took 10 x 20 second exposures, stacking in Bass.

Here is a spectrum of Alpha UMa in the H-alpha region:

I hope the big dark line is the H-alpha line, but I don't know the names of any of the other lines so that I can calibrate it.  Any ideas?  Here is a rough chart from Bass, after removing hot pixels except one at about 540 px:

Thanks for your thoughts.

Kind regards,

Steve.

[drjolo]

Hi Steve,

What is the dispersion of your LowSpec? What grating do you use?

Here is the Dubhe spectrum that I have obtained with 600lpmm grating, so maybe you will be able to compare 

https://astrojolo.com/spectroscopy/big-dipper-cluster-collinder-285-spectra/

Unfortunately Dubhe is a G9 type giant, so there is a plenty of lines in the spectrum  

 

[SteveBz]

1 hour ago, drjolo said:

Hi Steve,

What is the dispersion of your LowSpec? What grating do you use?

It's 1,800. 

Thanks for the Dubhe spectrum. I'm trying to calibrate now

Steve

[SteveBz]

The issue I'm having is matching up lines when the resolutions are different as close together spikes can become joined up and mislead the eye.

I think I need an alternative source for calibration.  I may try to calibrate with the Fraunhofer lines from the Sun, or maybe a luminescent bulb from Wickes.  Stars don't come out until nearly midnight now, so that's a bit painful.

Steve.

[drjolo]

Hm, if you do high res spectra with 1,800 grating, you may try to calibrate on atmospheric telluric lines around Ha band - they should be already present in the spectra, but identification is required. If you have calculated dispersion of your system in A/px or nm/px it would definitely help.

I did it once https://astrojolo.com/spectroscopy/menkalinan-spectroscopic-binary-star/

[robin_astro]

If you are looking specifically around H alpha at high resolution a simple neon lamp will do the job. Point the telescope at a bright A or B star and look for H alpha which will be obvious as there will be no other strong lines near it. Then when you have it centred take a neon lamp as a refence and identify the lines from for example Christian Buil's website here

http://www.astrosurf.com/buil/us/spe2/hresol4.htm

Cheers

Robin

 

[SteveBz]

On 25/05/2023 at 23:52, robin_astro said:

If you are looking specifically around H alpha at high resolution a simple neon lamp will do the job. Point the telescope at a bright A or B star and look for H alpha which will be obvious as there will be no other strong lines near it. Then when you have it centred take a neon lamp as a refence and identify the lines from for example Christian Buil's website here

http://www.astrosurf.com/buil/us/spe2/hresol4.htm

Cheers

Robin

 

Hi Robin,

I just saw that.  It's pretty much what I did.  I found an old Philips low energy fluoscent bulb and got this:

Which resulted in this:

I calculated the R number. The half height width for Ha seems to be about 0.16 so R=656/0.16 or about 4000. Is that right?

Now I need to improve the process with flats, darks and biases etc.

I've also bought soem neon lights, so the calibration will get better.

Tx

Steve

 

 

[robin_astro]

1 hour ago, SteveBz said:

I calculated the R number. The half height width for Ha seems to be about 0.16 so R=656/0.16 or about 4000. Is that right?

Provided the actual width of the line is significantly narrower than the resolution then yes this works but Ha alpha can be wide particularly in some stars eg like in Vega so you really need a line which is known to be very narrow. Try measuring the width in some of the narrow metal lines in your spectrum and see if this gives a higher resolution or  sky lines (natural or from light pollution) or from narrow lines in the calibration lamp provided it illuminates the spectrograph well (Note a single small lamp at the telescope aperture will not work for this as it like  a very high focal ratio pinhole camera and gives a perfect in focus image even when you are out of focus! If doing this using neons arrange 3 or 4 of them round the aperture edge)

Cheers

Robin

[SteveBz]

9 minutes ago, robin_astro said:

Provided the actual width of the line is significantly narrower than the resolution then yes this works but Ha alpha can be wide particularly in some stars eg like in Vega so you really need a line which is known to be very narrow. Try measuring the width in some of the narrow metal lines in your spectrum and see if this gives a higher resolution or  sky lines (natural or from light pollution) or from narrow lines in the calibration lamp provided it illuminates the spectrograph well (Note a single small lamp at the telescope aperture will not work for this as it like  a very high focal ratio pinhole camera and gives a perfect in focus image even when you are out of focus! If doing this using neons arrange 3 or 4 of them round the aperture edge)

Cheers

Robin

Great advice.  The lines at 639 are much narrower and I get nearer 9k.

I bought 5 neon lights.  I thought I'd 3d print a kind of ring round the rim to embed them in it.

[SteveBz]

Hi People,

It seems to be going in the right direction. Neon calibration:

3D printed stay with embedded bullbs:

Just moving the inner slots out a little.

Steve.

[robin_astro]

It is more common to mount them in a circle to give even illumination independent of  the orientation relative to the direction of the slit.  A nice setup suggested by Christian Buil with his StarEx  is to feed optical fibres to small clips round the edge. The obstruction is small and they can be left mounted without significantly disturbing the telescope throughput. They can even be switched on during the exposure to superimpose lines on the spectra giving very high precision calibration for example where flexure might be an issue during long exposures.

https://groups.io/g/Solex-project/topic/88905429

(In French but the photos tell  the story and Google etc translates well if you need more information)

 

Cheers

Robin

[SteveBz]

On 05/06/2023 at 00:05, robin_astro said:

It is more common to mount them in a circle to give even illumination independent of  the orientation relative to the direction of the slit.  A nice setup suggested by Christian Buil with his StarEx  is to feed optical fibres to small clips round the edge. The obstruction is small and they can be left mounted without significantly disturbing the telescope throughput. They can even be switched on during the exposure to superimpose lines on the spectra giving very high precision calibration for example where flexure might be an issue during long exposures.

https://groups.io/g/Solex-project/topic/88905429

(In French but the photos tell  the story and Google etc translates well if you need more information)

 

Cheers

Robin

Yes, it hasn't been ideal, I might try that.

[athornett]

All above is excellent advice.

As simple method, put a compact Fluorescent bulb on a stand at end of garden and aim your telescope/spectroscope at it and takes it spectra. Easily identifiable lines to put into RSPEC or whatever software you are using to calibrate the spectra at start of an evening's observing.

Andy

[robin_astro]

5 hours ago, athornett said:

All above is excellent advice.

As simple method, put a compact Fluorescent bulb on a stand at end of garden and aim your telescope/spectroscope at it and takes it spectra. Easily identifiable lines to put into RSPEC or whatever software you are using to calibrate the spectra at start of an evening's observing.

Andy

Hi Andy

The major disadvantage of this method for wavelength calibration is that the calibration may (and often does with many spectrographs) shift significantly due to thermal and mechanical distortion so it is important  to be able to calibrate with  the telescope under the actual observing conditions. With the LHIRES for example to reach the kind of precision needed to make these kinds of measurements where 1/5 of a pixel shift is significant

https://britastro.org/observations/observation.php?id=20220907_220530_9acdefe378f91ae9
I took calibration lamp spectra  before and after every observation ie every 10 minutes.

(One of the tests I do on any new spectrograph is to check its stability by taking repeated lamp spectra while the spectrograph cools and then with the telescope pointing in different directions)

Cheers

Robin

Edited October 29, 2023 by robin_astro
typo

[athornett]

Thanks for that explanation Robin