The Lowspec spectrometer

[Thalestris24]

This looks like a good video for calibrating the Ender 3 extruder. I'll have a go at following it tomorrow:

Louise

[Merlin66]

Paul,

I assume(d) the current Jeulin reflective slit plate was a copy of the Custom design plate that OVIO manufactured for me. When viewed from the telescope side, the numbers appear correct, the chrome coating and slits are on the rear face - towards the collimator.

 

[Paul Gerlach]

Well, I can only speak for the two discs I bought in December 2017 and they show the exact opposite of what your L200 disc shows.
So I think that the best thing to do for every builder is to check it and act accordantly.
I think I will delete the remark about the orientation of the writing on the disc in the build manual because it could confuse people if Jeulin decides to swap the orientation at will.

Some Polish amateurs are also confused about the orientation:
https://www.forumastronomiczne.pl/index.php?/topic/18311-wydrukuj-sobie-spektroskop/&page=3

My Polish is terrible so I used Google Translate to make sense of it. There is a picture of a LOWSPEC build by an amateur that show the slit disc with readable numbers. They go on telling that this is the reflective surface and that they have oriented it to get rid off the double reflections on the guide image. Later on the realize, after consulting an expert (Merlin66 😉) that the correct orientation should be with the reflective surface facing down.
With that I can only conclude that their disc has the reflective surface in the same orientation as mine.

Edited January 3, 2020 by Paul Gerlach

[ejp1684]

1 hour ago, Merlin66 said:

When viewed from the telescope side, the numbers appear correct, the chrome coating and slits are on the rear face - towards the collimator.

When I view my OVIO disk like this the numbers are mirror images, similar to Paul's comment.

Eric.

[Merlin66]

Paul,

Understood.

I assume your markings on the edge of the slit holder should match the necessary orientation of the slit plate. That would remove ambiguity.

Sorry if I've "muddied the waters" - my only experience is with the Custom designed plates.

Ken

 

[Paul Gerlach]

Ken,
Nothing to be sorry for.

I just thought about it a bit more. Maybe Jeulin  has chosen this orientation because the slits are mostly used in labs in optical setups where laser light hits the slit perpendicular. Then it would make sense to place the numbers correctly on the front side where the laser hits the slit.

[Jagho]

On 03/01/2020 at 09:56, Paul Gerlach said:

Louise,

Download the new Slit holder disk from Thingiverse (16_ovio_holder_slit_disc.stl) and use that. It has larger openings.

Hi Paul (and other astro friends ). I thought that It'll be good to print "16_ovio_guide_platform" also, because It has a bit larger (longer) opening.  Unfortunately, this was not a good idea. The new "16_ovio_guide_platform" can't be fitted to LowSpec2, right?

Edited January 4, 2020 by Jagho

[Paul Gerlach]

Oh my, you're absolutely right! I've been so caught up in this version that I totally forgot that I'd made some dimensional change to the guide platform. I'm so sorry.
Please use your 2.0 version and use the 3.0 version of 17_ovio_holder_slit_disc with that.
Thanks for pointing that out. I've made the corresponding changes to the text on Thingiverse.

By the way, great to see that you and others in Poland are entering the exciting world of spectroscopy. Keep up the good work!

Paul

Edited January 4, 2020 by Paul Gerlach

[Jagho]

Thank you Paul, nothing happened. Maybe I will build Lowspec 3 someday and I already have one part
As far as I know there are 3 people in Poland who are building Lowspec. @drjolo and me are still building or testing, but one of our colleagues (Bajastro) already has reached very good results at high resolution (Almach, Aldebaran, Capella stars):

https://astropolis.pl/topic/69283-widma-gwiazd-ze-spektrografu-low-spec-2/

[Paul Gerlach]

Nice results! I'm a bit surprised though that he uses a f/4 newton. When using the 1800 l/mm grating and use anything faster than f/10 you will lose a lot of light that is just not falling on the grating. Even at f/10 you will lose some light with a 25 x 25 mm grating (width of the light bundle falling on to the grating is (according to SIMSPEC) 26.8 mm. That's why I advice to use a 30 x 30 mm grating in the new assembly manual.

[Jagho]

You are right. He was doing it at very bright stars and wanted to achieve high resolution (dispersion less than 1A/pix as I know) so lose of light was acceptable. 

I'm building lowspec for a completely different purpose - using grating 150 l/mm, 80mm camera lens and 26um camera pixel (dispersion 20A/pixel, R~130) - I want to make spectra of 15-17mag supernovae in long time exposures (10" SCT).

Edited January 4, 2020 by Jagho

[Paul Gerlach]

Great to see that you're adapting the use of the LOWSPEC to your needs. That's the great thing about a 3D printed design. You can make changes to the design relatively easy.
Please keep me in the loop about your progress and results.

Regards,
Paul

[drjolo]

Hi again!

Yesterday evening I assembled LowSpec device - I still miss focusing bolt, so I need to focus manually moving the camera. And also spectroscope collimation is rough, for testing purpose. But first results are very promising. I have used all standard optics from LowSpec and 600 l/mm EO grating.

I used Altair IMX224 camera.

First test frame with workshop fluorescent lamp:

Mosaic of full visual range:

And plot from Vspec:

Today morning I recorded few Sun spectra. Doublet Na:

and magnesium triplet:

I am pretty happy with this test results. I think after careful collimation and mounting focusing bolt resolution will improve a bit.

Many thanks to Paul for this outstanding project and to Ken for constant support!

 

[Paul Gerlach]

That really looks good! 👍

[Thalestris24]

I bought some Sylvania brand 70W starters from the local Toolstation this morning - their last box of 5 for £1.50 . They have a quoted non-reclosure voltage of >150V. I believe they give off light more in the blue end of the spectrum but I'll have to test one to be sure. Still, here's a grab from the youtube video I saw one on:

Looks pretty blue-violet there! Could be mostly argon and/or mercury. Probably rich in UV so best not to stare at it...

I did find that you can get so-called 'spectrum tubes' of single gases for ~£15 each but they are like 26cm long and need a special, expensive eht (1000-5000V) power supply (though I daresay one could put a diy one together). 

Louise

[Paul Gerlach]

Best to stick with those starter bulbs. Although I think that they will become scarce as more fluorescence lights are being replaced by LED. So if you find some good ones it's best to stockpile them 😉

[Thalestris24]

1 hour ago, Paul Gerlach said:

Best to stick with those starter bulbs. Although I think that they will become scarce as more fluorescence lights are being replaced by LED. So if you find some good ones it's best to stockpile them 😉

Yeah, I bought 10 of the S10 bulbs! Plus 5 of the Sylvania ones. A bulb should last quite long time - as long as a neon bulb under similar conditions. I just tried out the Sylvania bulb. It wasn't so bright, unfortunately and disappointingly, though that was with a 33k resistor. Since the bulbs can run at ~150V without closing, then a smaller resistor should allow a greater current / higher voltage across the bulb. It should then be brighter. I'll have a look to see if I have a suitable resistor I've not received the Chinese 12V input power supply yet. I mistakenly mentioned before that you could run a glow bulb from 110V without a resistor but actually you still need a current limiting resistor in series otherwise the bulb will pass a relatively high current if the source can deliver it.  As that will likely be in the amps range with a 40W-rated supply, the bulb will glow bright but not last very long - at all! It will probably blow up - not good! In general, the greater the current the shorter the bulb life. I'll try the Sylvania bulb via the 240V mains again but with a 18k-22k resistor and see how it behaves.

Here is spectrum at 120s exposure:

And chart with S10:

There are plenty of peaks at the blue end - just mostly a bit small!

Louise

Edited January 5, 2020 by Thalestris24

[Merlin66]

Lucas,

Congratulations on your successful build!

Very promising results...

What slit gap were  you using?

Have you measured a line FWHM and determined your R value?

I used VSpec for many years but now use and recommend BASS Project.

https://groups.io/g/BassSpectro

There's a good solar spectrum resource http://bass2000.obspm.fr/solar_spect.php and a downloadable annotated solar spectrum http://bass2000.obspm.fr/download/solar_spect.pdf

[drjolo]

Thanks Ken! I used 30um slit - this will be the slit at my observatory telescope (Meade ACF 10" f/10). I have not yet measured R, I will do it after final collimation (actually I forgot that R is determined from FWHM ) 

Thank for pointing out BASS project - I will be happy to check another software. Last days Vspec made me little bit nervous.

[Thalestris24]

17 hours ago, Thalestris24 said:

 

And chart with S10:

There are plenty of peaks at the blue end - just mostly a bit small!

Louise

Um, anyone know why the BASS chart x scale is Angstroms but the label says nm? Is it me or a bug?

Thanks

Louise

[Merlin66]

Louise,

Under chart settings/ X axis "calibrated text" you need to set the units and description to Angstrom = "Wavelength ({0})  

 

[Thalestris24]

10 minutes ago, Merlin66 said:

Louise,

Under chart settings/ X axis "calibrated text" you need to set the units and description to Angstrom = "Wavelength ({0})  

 

If I change units to Angstrom it prints Wavelength (A) correctly but puts an extra zero in the values i.e. 410nm shows as 41000 rather than 4100?

[Thalestris24]

I changed the Ovio slit disk holder to the Lowspec 3 version. I've not changed the Ovio guide platform or slit illuminator though. However I did file them the best I could. I also reposition the Sylvania starter bulb a little - it was a little low down compared to the S10 bulb. Anyway, the result was a brighter image without changing the resistor . But.. the spectrum seems to have shifted to the right (in the red direction). I'm not sure why/how that's happened as I've not touched the micrometer which has been in the same position for a while now. I'm wondering if a slight change of direction of the source illumination could have resulted in spectrum shift? Although the calibration will have changed I've still used the original S10 anyway this time though I'll obviously have to redo it. 

 

 

On the plus side, the blue end peaks are much better now and more like I'd envisioned . If I can identify the wavelengths then it could be a better calibration bulb than the S10. 

I received the wee 12v inverter today. It really is very small. No instructions with it though the +/- 12V inputs are marked and I have the 220/240V AC outputs from online. It only has a total of 8 components, plus 4 diodes for optional DC output, so it should be easy to draw/reverse engineer the circuit and work out 110V AC transformer tappings . Will try and do that in the next day or two.  I've also come across references to it being able to run from voltages much lower than 12V so will investigate that as well.

ps I meant to mention that I'd noticed some shadow top and bottom of the exposure. I'm assuming that will be due to the holes in the slit platform and slit illuminator not being quite big enough? I'll maybe reprint the slit illuminator at some point but I don't really want to take the slit platform apart again if I can get away with it.

Louise

Edited January 6, 2020 by Thalestris24

[drjolo]

16 hours ago, Merlin66 said:

Have you measured a line FWHM and determined your R value?

Today I mounted Relco SC480 starter to my LowSpec and after careful collimation I checked its resolution. I also installed BASS project software - works like a charm.

According to SimSpec excel sheet I calculated resolution of LowSpec with 600 l/mm grating to about 800 at blue and about 1100 at red (I used 25um for lenses resolution). Real measurements with Relco starter indicates higher resolution - about 1400-1600 at red and about 1100-1300 in blue. However this test was performed without any optics - I just put starter in the front of 30um slit.

Edited January 6, 2020 by drjolo

[Paul Gerlach]

4 hours ago, Thalestris24 said:

ps I meant to mention that I'd noticed some shadow top and bottom of the exposure. I'm assuming that will be due to the holes in the slit platform and slit illuminator not being quite big enough? I'll maybe reprint the slit illuminator at some point but I don't really want to take the slit platform apart again if I can get away with it.

Louise

Louise,
Mea culpa, mea culpa... My head is so busy with the LOWSPEC 3.0 that I lost track a bit about what items where redesigned.
As you can see, the opening of the new LOWSPEC 3.0 guide platform is different than that of the 'original' 2.0 version. Which would explain that even after using the new slit disc holder. Also the dimensions of the platform are a bit different.
To make up for it, I've enclosed two STL files of a adapted version of the platform and guide plug as well as the source files of the hole unit.

 

16_ovio_guide_platform (LOWSPEC 2).stl 16_ovio_slit_illuminator (LOWSPEC 2).stl LowSpec 2.0_16_ovio_platform_NEW.f3d