The Lowspec spectrometer

[Merlin66]

Louise,

As I mentioned earlier this , I think this the "-1st" order...

The bright section is the red emission in the CFL, and should appear at the right side of the spectral image. The zero order is well over (out of frame) towards the righthand side of your spectral image as presented above.

Adjust the micrometer to move towards the right of the above .....note the zero order reading - then keep going until you find the spectrum.

Relax, you'll soon be there....

(The tilt of the spectrum across the FOV is the camera tilt.... the slant of the lines within the spectrum is the slit/ grating alignment)

 

 

 

Edited December 16, 2019 by Merlin66
clarification

[saac]

1 hour ago, Thalestris24 said:

Hi Jim

Yeah, you have to be careful what you order... I got the mirrors from Thorlabs - I only fitted them the other day! I got some round and square 1" craft mirrors of Ebay - very cheap! I used them in place of the proper ones until I was ready. The Thorlabs ones are a little expensive for what they are but not that much in absolute terms - plus they fit perfectly. I've got so far but am struggling a bit with the finer points. I don't think mine will be stable enough for astro-spectrometry. Getting and maintaining focus is proving very difficult. I'm sometimes getting some odd results but don't understand why.

Here's one I just did of the CFL bulb:

I think that's the zeroth order on the left and I've flipped it so it should fit Ken's calibration ok. It's all on a slant - not sure why. The spectral lines are slanted too - presumably to do with the seating of the grating. I've tried to adjust it but to no avail. I'm getting old and tired and impatient! I'm losing my grip - keep dropping things, and my cataracts make it hard to see what I'm doing! My back's been killing me too. I think 2020 will be my last year of doing astro things .

Louise

I know things like that can make this sort of work really frustrating - my mum suffers from arthritis in her hands and it really restricts what she is able to do and I can see that it lies ahead of me as well at some stage  Keep at it Louise for as long as you can, the mind needs to be challenged. I think as well you should be really proud of what you have achieved so far; you are going to overcome these problems for sure - I'm relying on you as I build mine behind you  

Jim 

Edited December 16, 2019 by saac

[Thalestris24]

14 hours ago, saac said:

I know things like that can make this sort of work really frustrating - my mum suffers from arthritis in her hands and it really restricts what she is able to do and I can see that it lies ahead of me as well at some stage  Keep at it Louise for as long as you can, the mind needs to be challenged. I think as well you should be really proud of what you have achieved so far; you are going to overcome these problems for sure - I'm relying on you as I build mine behind you  

Jim 

I've calmed down a bit now - just a bit! Had to walk away from it last night.

Louise

[saac]

19 minutes ago, Thalestris24 said:

I've calmed down a bit now - just a bit! Had to walk away from it last night.

Louise

Don't be afraid to set it aside for a day or two. I've seen myself working on a problem on my car and getting frustrated at getting nowhere fast. Not unusual for me to come back after setting it aside for a few days or a week and it's as though the problem was never there .  The gremlins know we are getting frustrated so they bunker down for fun  

Jim 

[Thalestris24]

After struggling with using a lens and guide cam last night I went back to the cfl today. Even though I had a bright source which covered the slit I had trouble acquiring a decent image. Heaven knows how you do it with just a star - a particular star, at that!

Just did another of the CFL. I can't seem to do anything about the sloping spectral lines. It seems to be linked to the way the holder sits and I can't see what I could do about that other than maybe reprinting it all. But maybe that wouldn't make any difference. It could be something to do with the spring affecting it - I'm not sure. I've also developed several horizontal black lines... I don't know what would cause them other than the camera sensor which is a bit worrying. Not sure if the slant and, particularly the black lines, will affect the processing?? If so, it could be end of project If it's still ok to process I'll have to get to grips with BASS.

Louise

[saac]

When I used the Star Analyser grating (about 2 years ago now )  I did the processing with RSpec. I may be wrong but I seem to remember  that RSpec provided  some functionality that allowed you to rotate the spectra until the lines were vertical.  I've never used BAS, I take it that there is nothing similar in the software then that would allow you to rotate the spectra. 

Jim 

[Thalestris24]

8 minutes ago, saac said:

When I used the Star Analyser grating (about 2 years ago now )  I did the processing with RSpec. I may be wrong but I seem to remember  that RSpec provided  some functionality that allowed you to rotate the spectra until the lines were vertical.  I've never used BAS, I take it that there is nothing similar in the software then that would allow you to rotate the spectra. 

Jim 

Yeah, I've used RSpec too - trial period long run out now! It did allow rotation of the whole spectrum to compensate for the SA100 not being horizontal. It won't let me fix the slant of the spectral lines, afaik - I don't think anything will. BASS is very similar, I believe, and basically does the same things as RSpec - but for free! I'm currently not very optimistic about getting any spectra from stars. I estimate that exposures will need to be quite long and my guiding from indoors isn't that good... I will give it a try though.

Louise

ps where have you got to now?

 

Edited December 17, 2019 by Thalestris24

[Merlin66]

Louise,

The slant of the lines in your spectrum isn't a show stopper. BASS can easily correct for this  - see the example I posted, it had slant corrected.

What is probably more important at the moment is rotating the grating >>>through zero image >>>> to the 1st order spectrum.

Ken

 

[Thalestris24]

1 hour ago, Merlin66 said:

Louise,

The slant of the lines in your spectrum isn't a show stopper. BASS can easily correct for this  - see the example I posted, it had slant corrected.

What is probably more important at the moment is rotating the grating >>>through zero image >>>> to the 1st order spectrum.

Ken

 

I've had another go this evening. I rotated the grating clockwise as far as it would go. I think I hit the blaze angle as there was no spectrum but just light on the screen. I turned the micrometer anticlockwise from this position until I got to the first spectrum. I got this (flipped) CFL one:

Must be the first order?

I also did one with the 500mm lens on and of a high pressure sodium lamp about 150m away. This is a screen clip of the guide cam screen (EZPlanetary):

Pretty bright, huh? But I had trouble acquiring a spectrum. It's straddling the slit - it's very hard to position targets manually (I have the Lowspec mounted on a Star Adventure / Manfrotto 055. It's my 'bench' for now.). Anyway - this was the best I could get:

I don't understand why the height of the spectrum seems to collapse sometimes? 

There is a HP sodium lamp reference spectrum here Mine seems similar-ish .

I found out what causes horizontal black lines - dust across the slit! Unavoidable here...

Louise

[Merlin66]

Louise,

Firstly the spectrum height...

On the bench you illuminate the full slit height - 6mm hence the height of the spectral image.

When you point to your street lamp the size of the image is much smaller , hence the reduced height.

A star spectrum will only be a few pixel in height, depending on the size of the star disk at focus.

 

Grating rotation:

Not sure what's going on. Assuming the grating is correctly orientated, you should be able to see the zero order image as a bright single line (an image of the slit gap). When you rotate the grating one way you should find the brighter (blazed) 1st order spectrum, the blue region coming into view first ...... and when you rotate it the other way, from the zero order,  it will pick up the "-1st" order spectrum (again the blue region coming into view first) which will appear fainter.  The 1st order blazed is much brighter than the -1st order. This doesn't appear evident in your image.

I think there's no need, at this stage to flip the spectral images, it can become confusing.

The attached image shows a test shot taken with a small spectrograph.....the bright line towards the left is the zero order image of the slit and you can see to the right the beginnings of the 1st order spectrum.

Find that bright zero order line first......

 

[Thalestris24]

5 minutes ago, Merlin66 said:

Louise,

Firstly the spectrum height...

On the bench you illuminate the full slit height - 6mm hence the height of the spectral image.

When you point to your street lamp the size of the image is much smaller , hence the reduced height.

A star spectrum will only be a few pixel in height, depending on the size of the star disk at focus.

 

Grating rotation:

Not sure what's going on. Assuming the grating is correctly orientated, you should be able to see the zero order image as a bright single line (an image of the slit gap). When you rotate the grating one way you should find the brighter (blazed) 1st order spectrum, the blue region coming into view first ...... and when you rotate it the other way, from the zero order,  it will pick up the "-1st" order spectrum (again the blue region coming into view first) which will appear fainter.  The 1st order blazed is much brighter than the -1st order. This doesn't appear evident in your image.

I think there's no need, at this stage to flip the spectral images, it can become confusing.

The attached image shows a test shot taken with a small spectrograph.....the bright line towards the left is the zero order image of the slit and you can see to the right the beginnings of the 1st order spectrum.

Find that bright zero order line first......

 

"A star spectrum will only be a few pixel in height, depending on the size of the star disk at focus. " Of course - silly me! I'm losing my marbles  

I think in mono, it's hard to tell whether a particular line might be the zero order. I'm not sure how far away from the rest of the spectrum it might be? But if it's the first spectrum encountered from the max clockwise position, then that must be the first order spectrum? According to Thorlabs the blaze angle is about 8-9 deg and I'd estimate that's the angle of incidence at the max clockwise position. Could it be that I should have the grating arrow on top and pointing in the opposite direction to the Lowspec build guide? I need someone with a Lowspec and 600 line grating to concur... (Hurry up, Jim!)

I've been using APT for image capture which lets me horizontally flip the image on screen - SharpCap doesn't have that facility. If I then flip the saved image it then matches the screen and is the 'right way round'?

I'll resume testing on Friday - have my ecg appt on Thursday. Hope I don't find I need a heart transplant!

Thanks

Louise

[Merlin66]

Louise,

Don't flip the images.

Look at your previous CFL images....the fainter lines are towards to blue and the brighter, wider, hazy line is at the red end. This should give you some guidance as to which part of the spectrum you're imaging.

The zero order is a defined bright line.....

When you look at your spectrum of the CFL, the distance from the first line to the red haze is about the same as the distance from the blue line back towards the zero order. ie if your CFL spectrum fills about 2/3 the full CCD frame then the zero order will be found  around 1 to 1.5 (approx) FOV from the blue line.

Don't worry about the blaze angle...(check p 148 > for technical details - not necessary to know at this stage)

Hope your ecg app goes well tomorrow.

Ken

 

 

Edited December 17, 2019 by Merlin66

[Thalestris24]

6 minutes ago, Merlin66 said:

Louise,

Don't flip the images.

Look at your previous CFL images....the fainter lines are towards to blue and the brighter, wider, hazy line is at the red end. This should give you some guidance as to which part of the spectrum you're imaging.

The zero order is a defined bright line.....

When you look at your spectrum of the CFL, the distance from the first line to the red haze is about the same as the distance from the blue line back towards the zero order. ie if your CFL spectrum fills about 2/3 the full CCD frame then the zero order will be found  around 1 to 1.5 (approx) FOV from the blue line.

Hope your ecg app goes well tomorrow.

Ken

 

 

I'm confused - didn't you say previously that the spectrum image only calibrated when flipped? I'll have another quick look at what's visible tomorrow. There may have been an odd line outside of the saved fov. My ecg appt. is Thursday - it's Wednesday tomorrow!

Thanks

Louise

[Merlin66]

Louise,

It's Wednesday down here in Oz.....

The reason initially for flipping the spectrum was due to your sequence of images 1-7 this appeared to show that you were moving towards the blue as you increased the micrometer, rather than towards the red.

(If you had continued this sequence ...... you probably would have found the zero order)

If you can get the micrometer to the zero image then you can confidently determine which order of spectra you're observing  - a faint one on one side (-1st order) and a brighter one on the other side (blazed 1st order)

Generally by winding the micrometer inwards ( lower value on the micrometer dial) from the zero image, you hit the blue side of the spectrum first, then onwards to the red.

I think you had the grating position well away from the "normal" operating range and came into a spectrum from the "other side" ie red end first.

Hope this makes some sense.

Ken

 

Edited December 17, 2019 by Merlin66

[Thalestris24]

2 minutes ago, Merlin66 said:

Louise,

It's Wednesday down here in Oz.....

The reason initially for flipping the spectrum was due to your sequence of images 1-7 this appeared to show that you were moving towards the blue as you increased the micrometer, rather than towards the red.

(If you had continued this sequence ...... you probably would have found the zero order)

If you can get the micrometer to the zero image then you can confidently determine which order of spectra you're observing  - a faint one on one side (-1st order) and a brighter one on the other side (blazed 1st order)

Generally by winding the micrometer inwards ( higher value on the micrometer dial) from the zero image, you hit the blue side of the spectrum first, then onwards to the red.

I think you had the grating position well away from the "normal" operating range and came into a spectrum from the "other side" ie red end first.

Hope this makes some sense.

Ken

 

It's 11\:30pm on Tuesday here I'm still a bit confused. The clockwise rotation only goes so far. I went as far as possible above then turned the micrometer anti-clockwise until I ncountered a spectrum. I've seen some examples which suggest that a second or third order spectrum can be brighter than the first. I'll keep at it and try and find my way.

Thanks

Louise

[Merlin66]

Louise,

OK.

It could be your starting point for the clockwise rotation was already too far from the zero position.

Unwind the micrometer almost completely, then slowly go CW watching what happens - you'll probably find a spectrum....keep going....dark space....then hopefully a zero order line image ....note the micrometer reading....keep going.... dark space.... and you'll find another spectrum. Hopefully the second one you find will be much brighter than the first. This is the one you want.

Once you've got zero image/ spectrum you're in business.

Ken

The 2nd and subsequent spectra are always fainter the the 1st order due to the increase in dispersion.

 

[Thalestris24]

2 minutes ago, Merlin66 said:

Louise,

OK.

It could be your starting point for the clockwise rotation was already too far from the zero position.

Unwind the micrometer almost completely, then slowly go CW watching what happens - you'll probably find a spectrum....keep going....dark space....then hopefully a zero order line image ....note the micrometer reading....keep going.... dark space.... and you'll find another spectrum. Hopefully the second one you find will be much brighter than the first. This is the one you want.

Once you've got zero image/ spectrum you're in business.

Ken

The 2nd and subsequent spectra are always fainter the the 1st order due to the increase in dispersion.

 

Ok - will try that!

Thanks

Louise

[Merlin66]

Louise,

Check out the illustration on ASA, p42.

It shows the relationship between the various spectral orders and the zero image.....may help your understanding.

 

Edited December 18, 2019 by Merlin66
poor image

[Merlin66]

This may help....

Edited December 18, 2019 by Merlin66

[Thalestris24]

Hi Ken

I have done the theory (covered it in physics at uni ) but it's just been the practicalities I've had trouble getting my head around. It hasn't helped that everything is right to left on the screen and need to go anticlockwise from a starting position as well as it being a blazed grating. I think I've got there now - tbh, I wasn't expecting the zero order to be quite so far away from the 1st order spectrum. When you mentioned it would be, a light began to glow in my head! I don't think I can actually get the whole of the recordable spectrum on one screen but can think I can get the major features ok.

Anyway, this must definitely be the zero order:

Turning anticlockwise, and shifting right,  I eventually get to the start of the first spectrum - rightmost faint line (spectral lines shift right):

a less bright 0.5s exposure:

Then, shifting further right, I get to the end of the first spectrum (broad faint line on the left):

After that faint line on the left in the last image, there is nothing else close, as far as I could see. The apparent zero order was at 1.76mm on the micrometer. The start of the spectrum was first just visible on the left at 2.08mm - more than a whole screen away. The start shifted to the right (2nd image) at 4.41mm with the third image on 5.11mm. So the spectrum seems to cover 5.11-1.76 = 3.35mm of the 13mm micrometer i.e. just over a quarter of the total micrometer range. There are fainter spectra in an anticlockwise direction. Of course, I've been used to using the SA100 where you can easily get the zero order plus the whole spectrum on one screen - easy processing!

Louise

[saac]

Louise hi, I’m still a good bit behind you. I was hoping to finish printing the remaining internals (lens holders and lid) but last night I found that my laptop hard drive had crashed. It’s completly dead so Ihad to order a replacement which I could do without just before christmas😩 Anyway hopefully it will be here at the weekend and I can get started again. ,I’m using my small iPad to type this and I detest using the touch screen to type. 

I think you are doing amazingly well and Ken’s advice is brilliant. I don’t expect to be in a position to get my optics installed until after the new year, like I said before I suspect I will need to reorder some lenses. I will keep you posted😊

[Thalestris24]

12 minutes ago, saac said:

Louise hi, I’m still a good bit behind you. I was hoping to finish printing the remaining internals (lens holders and lid) but last night I found that my laptop hard drive had crashed. It’s completly dead so Ihad to order a replacement which I could do without just before christmas😩 Anyway hopefully it will be here at the weekend and I can get started again. ,I’m using my small iPad to type this and I detest using the touch screen to type. 

I think you are doing amazingly well and Ken’s advice is brilliant. I don’t expect to be in a position to get my optics installed until after the new year, like I said before I suspect I will need to reorder some lenses. I will keep you posted😊

Hi Jim

Oh dear, sorry about that - things always go wrong when you need them, don't they. Hope you can get up and running again soon!

Yeah, Ken has indeed been brilliant! I really appreciate his help, support - and patience! Eric has been very helpful too. Also all the other guys on here who helped me get to grips with 3d printing. 

Well, at least the Surplus Shed lenses are cheap! Hopefully they will have the type, diameter and focal length you need. Just take your time checking everything matches before you order. I think I used their lens finder. They are quite helpful if you need to ask anything. I didn't quite appreciate how closely the diameters and focal lengths need to match the design. My 121 mm focal length collimating lens seems to be ok but there isn't much room to play with the 125 mm design. Diameters can't be too big either otherwise the two main lenses will crash together! This was a problem I had trying to use the 125 mm finder lens which is 31.2 mm dia - too big! The focal length of the camera lens should match your camera properties. My Atik has a spacing of 20 mm for the sensor - t2 flange distance. So, in the end, a 89 mm focal length proved a comfortable distance to the sensor plane. I've ordered a T2 360 deg rotator. A bit pricey but I think it will be useful. It takes up 5.5mm optical length so fitting it in the camera optical train shouldn't be a problem.

Hope you can have a relaxing Christmas

Louise

[Merlin66]

Louise,

No disrespect intended...I just wanted to reinforce the relative positions of the first order spectra to the zero order image.

OK. Looks like you have managed to get the zero order! 

Your second image above looks like it covers the main Fluoro spectrum from blue to red and should compare favourably with the earlier profile I posted.

From the micrometer settings you quote ie Zero at 176 and the centre of the first order at  411  it seems in the LowSpec design that the grating holder rotates in the opposite direction to the Littrow spectrographs I have. For me the zero order is found at a much higher micro reading (I use a 25mm micro) say around 1500 and depending on the grating used can go down to 1050 to centre the red line of the fluoro lamp.

Obviously if the zero is at only 176 there's no chance of going further to pick up the first order on the other side of the zero order.....

With a slit spectrograph, other than during initial set-up ( as we're doing now), you will probably never look at or use the zero order again. Once you have the micro settings for the various wavelengths ( I centre Fluoro Blue (FB), Fluoro Green (FG), Fluoro Yellow (FY) and Fluoro Red (FR)) to give me a quick calibration guide...ie

Zero 1540

FB 1277

FG 1100

FY 1072

FR 1049

Using these settings allows me to quickly set the grating to the wavelength of interest. When you commission your reference lamp you can refine this calibration.

Now you're up and running..

Rotate the camera 180 deg (if you can) to bring the blue region to the left hand side.

Practise different focus positions to find the best result - at the moment I think your blue focus is tighter than the red. Depending on the quality of your optics and the set-up accuracy of your Collimator/ grating/ imaging lens, you'll find this is a compromise.

At this stage I would be looking at BASS Project to help analyse the spectra you're recording. Another learning curve - but there are a few good tutorials available to help.

Well done!!!

Ken

 

Edited December 18, 2019 by Merlin66

[Thalestris24]

4 minutes ago, Merlin66 said:

Louise,

No disrespect intended...I just wanted to reinforce the relative positions of the first order spectra to the zero order image.

OK. Looks like you have managed to get the zero order! 

From the micrometer settings you quote ie Zero at 176 and the centre of the first order at  411  it seems in the LowSpec design that the grating holder rotates in the opposite direction to the Littrow spectrographs I have. For me the zero order is found at a much higher micro reading (I use a 25mm micro) say around 1500 and depending on the grating used can go down to 1050 to centre the red line of the fluoro lamp.

Obviously if the zero is at only 176 there's no chance of going further to pick up the first order on the other side of the zero order.....

With a slit spectrograph, other than during initial set-up ( as we're doing now), you will probably never look at or use the zero order again. Once you have the micro settings for the various wavelengths ( I centre Fluoro Blue (FB), Fluoro Green (FG), Fluoro Yellow (FY) and Fluoro Red (FR)) to give me a quick calibration guide...ie

Zero 1540

FB 1277

FG 1100

FY 1072

FR 1049

Using these settings allows me to quickly set the grating to the wavelength of interest. When you commission your reference lamp you can refine this calibration.

Now you're up and running..

Rotate the camera 180 deg (if you can) to bring the blue region to the left hand side.

Practise different focus positions to find the best result - at the moment I think your blue focus is tighter than the red. Depending on the quality of your optics and the set-up accuracy of your Collimator/ grating/ imaging lens, you'll find this is a compromise.

At this stage I would be looking at BASS Project to help analyse the spectra you're recording. Another learning curve - but there are a few good tutorials available to help.

Well done!!!

Ken

 

Hi Ken

Hope you're keeping cool down there! The temperatures in some parts are dreadful! 😮 

Yeah, the micrometer only extends out to 13mm. Also, looking down, the collimated incident beam comes in from the right. Diagrams always show it coming in from the left of the normal. The camera isn't rotatable at the moment as it's supported, and sits on, a flat aluminium bar. The Atik 383l+ is quite heavy.... I might have to try and rearrange things or sort out a different way to support it. Yeah, focusing is difficult and varies a lot. Certainly varies with wavelength - maybe also a positional effect? 

Yep, will try out BASS next!

Thanks

Louise

[Merlin66]

I don't have any first hand knowledge of the LowSpec, so I have to assume that Paul designed it correctly.

In my Littrow, with an 1800 l/mm grating which gives a limited wavelength coverage, I have to refocus at every micro setting.....the focus can vary up to 1.8mm across the visible spectrum.

It was 42 deg C yesterday and they're saying tomorrow could be even hotter!!!