Rolling my own HSG thanks to santa :)

Post by **Merlin66** » Fri Jan 24, 2014 9:43 pm

Nick,
The attached SimSpecV4 shows your arrangement:
I've assumed a slit gap of 20micron, a total angle of 38 deg and pixel size of 8.3 (you can change this to suit your camera)
It shows the grating sits at 43.6 deg to the optical axis and for an f10 beam requires a 30mm x 50mm grating size - this is non standard so a 50 x 50 would be needed.
HTH

Nik_SimSpecV4_1c.zip

Nick · Post by **Nick** » Fri Jan 24, 2014 10:25 pm

Thanks Merlin

The grating is a 2"x2" .. when I played with the figures it works out as 58mm but considering the edges of the vixen aren't going to be stunning I figured that even with tilt that this grating would be a good all rounder.

I was playing with the 36degree angle to check it before doing anything with a rivet gun and jigsaw I would later come to regret. The alu bar is marked out with the final locations of the grating, slit and lens so this is looking good.

Once I got beyond the fact that this was lithgow by default things started to make sense (and some coffee).

I took an photo using the DSLR - but at the moment I'm hand holding the grating until I have made the cage hence the image is difficult to see (and I don't think the Mrs would really like me attempting to align the FD onto the EOS body in the dark with the FD aperture lugs pointing dangerously at the EOS!)

The top and bottom bright spectra is because the light can get past the slit (see the last picture). The dimmer spectra is the slit

For those that may have seen the v2 (second time the upgrade has munched it

).. it's the new proposed (green) camera location.

So I'm happy with the calculations.. next up is to get the components bolted onto the optic axis (the box is also bolted on but not there to provide the rigidity.

Here's the test setup

I forgot to say that I will be adding a 2x to the 300mm f5.6 so this will make it 600mm f/11.2 which matches the Vixen (f/11.3) and a reasonable match to the Pentax with the 2x on (1340mm f/12.6). However for now. it's good.

Nick · Post by **Nick** » Fri Jan 24, 2014 10:48 pm

I also managed to get to the bearing of the HD read/write heads:

These are extreme precision and very very light!

Nick · Post by **Nick** » Sat Jan 25, 2014 10:56 pm

Still got a little way to go but here's a zero order image from the test set up using the 383L.

Get a spectrum, however the camera isn't picking up any lines.. but I think this may be down to focus.

Also have the final bits of alu.. next up I'll sort out some of the mountings.. The good thing this has shown the need of both accuracy and the ability to tinker with the distances.

Post by **Merlin66** » Sun Jan 26, 2014 2:38 am

Nick,
That slit image looks about 3metre wide!!!!
Ideally it should only cover 2-4pixel on your camera....
Reset the slit gap and try again.

swisswalter · Post by **swisswalter** » Sun Jan 26, 2014 7:47 am

Hi Nick

congratulations on the first light, even if it is not what you expected. It is still a first light. I can see it :-)

Nick · Post by **Nick** » Mon Jan 27, 2014 8:27 pm

After a hectic weekend.. however the garage is clear.. so that gave me the opportunity to try out my light sources

High Pressure Sodium 400W bulb at full tilt:

Metal Halide 400W bulb at full tilt:

I also had a thought about the phase two - adapting the scanner - it's possible to fit it (mirror rotated 180) to a modified Lithgow design with a single collimating/focusing lens.. hmm interesting options for phase two..

I still to think more but I think this will work.

Attached files

Nick · Post by **Nick** » Tue Jan 28, 2014 9:56 pm

I've been holding up making the next step of securing the components because I've been thinking of the best way todo it. In short I wanted to add some ability to allow adjustment.

I'm going to use a frame of 4 alu rods and clamps, for example:

For some components, the adjustment can be done by rotating a screw with a spring applying opposite force:

The frame will then be clamped to the rails though the side of the spectrometer - thus making a rigid axis on which to mount components.

swisswalter · Post by **swisswalter** » Tue Jan 28, 2014 10:06 pm

Hi Nick

what a light power.

Nick · Post by **Nick** » Fri Jan 31, 2014 10:05 pm

Hi Merlin - I may need to pick your brains..

The attached image is a Thorlabs fixed width 10um slit with a torch on the other side, going through a f5.6 300mm lens, then being reflected off the grating at order 0 before being focused by the f4 300mm and finally imaged by the 383L with bin 1x1 (5.4um pixels). I get the same with the variable slit.. just continuous spectra without hints of lines.

I'm not getting much luck with binning 2x2 or switching cameras with 7.4um pixels :/

I would have expected to see some variation in brightness.. could the collimator be too far from the grating? (in theory this shouldn't matter because the light should be parallel.. unless both lenses are compensating by being focused for non-collimated light).

I may try running with a 2x power mate to make the collimator 600mm, however surely this would simply increase the wise of the slit?

Post by **Merlin66** » Fri Jan 31, 2014 11:33 pm

A couple of things
1. A halogen lamp has no spectral detail - it basically a "flat" Black body continuum....what you need to use is an energy saving fluoro.....
2. The zero image is well over exposed - that's why I think it's looking sooooo wide.
3. The tapered slit gap shows the optics are mis-aligned.
4. Providing the grating is set correctly with the top edge blaze arrow pointing in the direction of the imaging lens, you should obtain a spectrum. As I said the fluoro lamp is the easiest to use and has some good emission lines in the blue yellow and red.

Hope this helps.

Wah · Post by **Wah** » Fri Jan 31, 2014 11:35 pm

Have you tried other light source? I know tungsten/led torch won't give you any spectral lines.

Nick · Post by **Nick** » Sat Feb 01, 2014 9:44 am

Thanks you both! I forgot to say that the image was auto stretched in PI so you could see the focus is good and the slit size easily - but looking at the area of light compared to the slit size itself.. it still looks overexposed on the non-stretched - perhaps order 2 will work because if it!

The 383L is a kodak full frame sensor so has a mechanical shutter which, from experience, needs about 3 seconds to not make any shadow in the corner of the image. At 0.01 etc it's looking like there's still too much light. I was using this camera because it's square and sat on the desk nicely

(optimistic!)

I tried the little titan - it is my usual solar camera as it's sony chip gives fast exposure and frame rate without the need for a mechanical shutter. The slit was large but I found a long exposure bug in my own drivers - nothing like dog fooding! The disadvantage for desk use is it's cylindrical. Next to try is the 4000 - an electronic shutter, but the frame rate is slower - cylindrical again.

Once I have an observable spectra I'll start making the final holders for the components - but for now it's all bench setup so all the off centre optics etc are a given! It was a howling gale so I couldn't setup in the garage with the two large lights.. so I'll get myself a energy saving bulb as the house friendly option!

Nick · Post by **Nick** » Sat Feb 01, 2014 9:27 pm

Ok, I've fixed the issue in the titan (and improved image quality for it).. but here's the zero order without stretching:

I found a CFL bulb but I don't have a lamp that will take it grr..

Sooo close lol.

Post by **Merlin66** » Sat Feb 01, 2014 9:38 pm

That looks better!
You just need to ensure the alignment of the slit to the grating direction and rotate the slit image to the vertical.
(The tapering off of the light at the top may be due to the misalignment of the slit to the light source ( a diffuser over the end - greaseproof paper? would assist at this stage)or vignetting in the optics.....
Onwards and Upwards

Nick · Post by **Nick** » Fri Feb 14, 2014 10:42 am

I see the thread made it across (almost intact - some images are a bit confused lol).

I have just ordered a drill mill .. attempting to drill accurate holes by hand held bosh really wasn't cutting it. The project is getting there - I spent an afternoon evaluating how to create the design - hence the experience leading to a drill mill. I figure I am now into negative brownie points with the mrs.. today of all days..

The idea is to have this complete by end of march for the SGL star party

Nick · Post by **Nick** » Sat Feb 15, 2014 9:58 pm

I've spent some time today getting some of the internal cage progressed.

Now this is just a simple mock up - no solid revetting and just bolts on the side.

: IMG_1774.JPG (3.81 MiB) Viewed 4597 times

The idea is to get a skeleton done - there's some filing todo as the cut has one side slightly too large. The final version will have additional plating to add structural integrity.

: IMG_1775.JPG (3.64 MiB) Viewed 4597 times

Now I did say it needed some final filing.. you can see here it's not quite fitting

: IMG_1776.jpg (2.86 MiB) Viewed 4597 times

But this is the main structure. The final version will be wider.

swisswalter · Post by **swisswalter** » Sun Feb 16, 2014 8:11 am

Hi Nick

that looks great

Nick · Post by **Nick** » Sun Feb 23, 2014 7:42 pm

I've decided to go for a skeleton to start with - it allows easy access to adjust things.. after which I can then mount in the box I've created earlier.

So this weekend saw a rather large amount of progress..

The adjustable grating cage

: IMG_1777.JPG (549.17 KiB) Viewed 4569 times

Fully adjustable lens mount to allow collimation:

: IMG_1778.JPG (445.53 KiB) Viewed 4569 times

Front:

: IMG_1779.JPG (492 KiB) Viewed 4569 times

Current simple screw held slit - this is pending more work. The single piece of aluminium will be doubled up on the top.

I've fabricated two more adjustable lens holders for the focusing lens and I'll make a camera clamp to hold securely.

I think an afternoon at some point and this will be ready for first light..

Nick · Post by **Nick** » Sun Feb 23, 2014 9:12 pm

Ok, attempted a pseudo first light with the cannon Ixus80 just to see what the current form will give.

Firstly - the slight is orientated 90 deg out.. stupidly this occurred to me as I was building it.. no biggie so I rotated the grating 90deg (just for now) and pointed it at the ceiling light.. which just so happens to be a energy saving FL..

Got spectrum! WIth the above noted - hence the non-straight nature:

: IMG_1787_Preview01_firstlight.jpg (878.5 KiB) Viewed 4564 times

Looking at it visually - there's more lines in there.. so the better resolution cameras will be better

swisswalter · Post by **swisswalter** » Sun Feb 23, 2014 9:15 pm

Hi Nick

good Progress. Good luck on the finish

Nick · Post by **Nick** » Sun Feb 23, 2014 9:25 pm

swisswalter wrote:Hi Nick

good Progress. Good luck on the finish

Thanks

Ixus only takes jpegs but.. looking good

: IMG_1790.jpg (2.6 MiB) Viewed 4564 times

Wah · Post by **Wah** » Mon Feb 24, 2014 2:23 am

You may need to rotate your slit to let the spectral lines perpendicular to the spectrum.

Nick · Post by **Nick** » Mon Feb 24, 2014 8:03 am

Wah wrote:You may need to rotate your slit to let the spectral lines perpendicular to the spectrum.

Yup - the slit needs 90 degree rotation and I need to add the final imaging lens+camera support. These shots were just to verify it was working - but it's looking good.

Nick · Post by **Nick** » Tue Feb 25, 2014 10:24 pm

Here's the grating in the holder:

: IMG_1791.JPG (751.7 KiB) Viewed 4536 times

Now tonight, as I worked from home and looked after a sick mrs.. it meant that I could get down to the task at 1830.. I look at my watch and it's 2200!!

I have now made a basic modification for the slit, rotating by 90 deg as required and providing some adjustment. I'll sort out something a bit more secure and adjustable later.. but for now an L piece of wood will suffice:

: IMG_1794.jpg (592.56 KiB) Viewed 4536 times

Now the lower components were assembled, with two holders for the cameras - one for the larger diameter 383L & 4000, the other for the smaller Titan and 16IC:

: IMG_1796.JPG (510.28 KiB) Viewed 4536 times

Not the bottom isn't in it's final position but simply resting against the upper section to give a indication of how it will fit together

I still need to make a couple of holder pieces etc but for tonight that will do!

I've also managed to (a) get a bulb holder for a CF so I can put in on the desk .. but (b) managed to find a electrician's screw driver.. with a neon bulb in it.. I hope to have everything put together and aligned for a proper first light tomorrow.

swisswalter · Post by **swisswalter** » Wed Feb 26, 2014 5:28 am

Hi Nick

great job, looking forward to your first light

Nick · Post by **Nick** » Thu Feb 27, 2014 12:01 am

I still need some tweaking - but this is the first image through the DIY spectroscope

: capture_2014_02_26_23_54_35_0000_4_clone.png (2 MiB) Viewed 3550 times

So that DIY scope and DIY driver and DIY capture app

swisswalter · Post by **swisswalter** » Thu Feb 27, 2014 5:38 am

Hi Nick

congratulations on that first light. Why do we have here horizontal lines

Nick · Post by **Nick** » Thu Feb 27, 2014 7:56 am

They're caused by the dust/uneaven edges on the slit

Nick · Post by **Nick** » Thu Feb 27, 2014 8:26 pm

Holy moly.

I've just taken the two strong yellow Hg lines..

: spectrum_9_07a_Preview01twolines.png (1.76 KiB) Viewed 3526 times

20A apart and it's giving me 182 pixels.. so still unsure how accurate my measurement is - so this is native resolution crop of the two lines:

: capture_2014_02_27_20_07_18_0000_1_Preview01.png (1.44 KiB) Viewed 3526 times

Wider native resolution crop:

: widercrop.png (108.54 KiB) Viewed 3526 times

swisswalter · Post by **swisswalter** » Thu Feb 27, 2014 9:10 pm

Nick wrote:They're caused by the dust/uneaven edges on the slit

O.K. Nick

thank you for the explanation. so that is not really a problem

swisswalter · Post by **swisswalter** » Thu Feb 27, 2014 9:13 pm

Hi Nick

it's getting very interesting. I can't commend on the lines captured, sorry

Nick · Post by **Nick** » Thu Feb 27, 2014 10:40 pm

So.. the jury is in… in it's current 300mm form it's 0.9A, R=6410. I say current form as it's final phase one has a 2x in the design to bring the f5.6 to f11.2 to match the vixen's 910mm f11.4 (final design, in theory is 0.67A, R=8611).

Also I thought I'd take a time out.. and mention some of the lessons I've learn in doing this.

1. When I started out I thought I knew nothing, now I know I know nothing - and the void feels even bigger.

2. Design - the SIM XLS has been a very good tool for the design, both in verifying my understanding of the maths but also point out things I'd not even considered - enough to go away and research that too. When doing the design.. there's expertise out there - pick their brains, they've done it multiple times before.. because it's better than scratching your head.

3. Start simple - block of wood + bolts. I have a nickname for the alu cage in my design.. and it's not repeatable. I would strongly advise not attempting a tube cage design - I have a a 4ft high bench drill mill and I found it virtually impossible. If you're a CNC operator - start with wood.. then get fancy once you have it working..
I would say that the alu components help to keep the weight down, BUT, I would look at that AFTER the wood form is working. The more materials - the more cost, the more time needed to figure out how things will work and the more mistakes.

4. Don't under-estimate it. It may be a couple of lens and a fancy mirror thing but it's hard. Did I say it was hard?

5. Lastly .. pick one size of bolts.. I have euro M6 bolts of differing lengths.. with 4mm bolts for the small components (because they fit my adjustable slit mounting point).

Abbreviated form - Wood is good, start simple, be humble and ask people with experience because you'll learn to admit it.. they really do know more than you.

swisswalter · Post by **swisswalter** » Fri Feb 28, 2014 5:42 am

Hi Nick

moding things is fun and sometimes it is important the set it away and leave it for a couple of days/weeks ;-)

Nick · Post by **Nick** » Fri Feb 28, 2014 10:57 am

I forgot to say - dispersion is 0.1A/pixel.

So for now I think I will get this finalised - adding the 2x in there, fixing into the outter box, flocking/painting the inside. Finding the CoG and adding the mount point.

I think the best approach with the fixed 10um slit will be to make a metal L shape to conduct heat, fixed in with 4x4mm bolts for height adjustment. I'll use a armature arrangement with a bolt thread supplying the fine movement through the back.

Next lighting will be targeting the 400W Metal Halide/High Pressure Sodium bulbs. Then after that the big solar thing.

swisswalter · Post by **swisswalter** » Fri Feb 28, 2014 8:22 pm

Hi Nick

I'm looking forward to the big solar thing. Take your time

Nick · Post by **Nick** » Fri Feb 28, 2014 10:08 pm

Yup, I'll take it in baby steps Walter.

Ouch.. you know I said it was hard.. now I've added the PM 2x to double up the collimating lens and using the 10um fixed slit.. with a f4 imaging lens the difficulty is now obscene. I've had to modify the imaging lens FD bracket to get the back focus required!

This still needs *alot* of adjustment and alignment (reason for the other rubbish in the image) but here's a spectra - look at that line detail appearing!

: capture_2014_02_28_21_50_27_0000_3_Preview01.png (641.61 KiB) Viewed 3499 times

For kicks - attached is the zero order when focusing.

swisswalter · Post by **swisswalter** » Sun Mar 02, 2014 8:09 am

Hi Nick

another step, looks interesting

Nick · Post by **Nick** » Wed Mar 05, 2014 11:23 am

Sorry for the slow progress.. I'm on vacation which means I'm DIYing again.. this time the bathroom floor underfloor heating, towel radiator and tiling.. so it's all locked away from the dust and muck :/

Nick · Post by **Nick** » Sun Mar 09, 2014 8:51 pm

I managed to get the cage mounted into the box, so it's looking almost like it's final form.

Nick · Post by **Nick** » Sun Mar 16, 2014 12:07 pm

First solar test today

Visually you can see the Ha line (slit virtually closed) - I tried looking for the CaK line but it should pop out once the dark work has done.

Things not todo - put plastic paint guard in front of the slit to check for light.. I have today's sun melted into it

Here you can see the sun on the slit and the spectrum off the grating.

: IMG_2580-800x600.jpg (148.59 KiB) Viewed 3356 times

Happy that the image finding is great, the slit can take the heat easily without any problem. Happy to see the Ha line, Sodium doublet etc.

Things needed:
* scope reseated forward about 2 inches for focus changes on the 2x slit.
* Flocking and add additional walls to block light (simply too much light causes the 383L to swamp out - even with the dark pod over the top.
* put on side mount so I can mount it to the EQ6 so the slit is orientated on the drift (in the image below it's 90deg as it's sitting on the ground.
* drill support holes for the camera

swisswalter · Post by **swisswalter** » Mon Mar 17, 2014 5:37 am

Hi Nick

another great step