Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

A Tri-Bahtinov mask for SCT collimation and focusing

astrophotography imaging SCT
  • Please log in to reply
487 replies to this topic

#376 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,046
  • Joined: 24 Nov 2014

Posted 16 July 2020 - 06:02 PM

****. Yea. That's what my flats look like and that's what CCDInspector curves look like.

Should've put money on it :)

 

It's not clear to me that you can fix the problem by yourself.

 

Good luck!

 

cytan



#377 Ben Diss

Ben Diss

    Vostok 1

  • *****
  • Posts: 100
  • Joined: 07 Nov 2019
  • Loc: N40 W83

Posted 16 July 2020 - 06:57 PM

Should've put money on it smile.gif

 

It's not clear to me that you can fix the problem by yourself.

 

Good luck!

 

cytan

 

See, here's the thing (and the reason I latched onto this thread). I bought this OTA 2nd hand and I've never been happy with the corners. I can align the secondary mirror and get the center stars nice and round, but I've never got the corners ones perfect. So, I bought the Hotech and aligned the corrector plate. Thing is, it's all the way over to one side in order to pass the Hotech alignment procedure and my field curvature went way up compared to when it was mechanically centered. The stars are very much improved however.

 

I have also gone down a rabbit hole learning and understanding masks and mask design. I've got Bahtinov's, Carey's, tri-Bahtinov's and hybrids for each focal length. Here's my collection. 

 

masks.jpg


  • Psittacula and meegja like this

#378 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,046
  • Joined: 24 Nov 2014

Posted 16 July 2020 - 07:40 PM

Hi Ben,

 

  Wow, that's a lot of masks! 

 

  I really don't have any advice as to how to fix the problem. I think half of it is identifying the problem and the other half is figuring out whether it is worth fixing. I certainly hope that the problem has been found.

 

cytan


  • Ben Diss likes this

#379 meegja

meegja

    Messenger

  • -----
  • Posts: 486
  • Joined: 02 Jul 2020
  • Loc: The Netherlands

Posted 17 July 2020 - 02:35 AM

I have also gone down a rabbit hole learning and understanding masks and mask design. I've got Bahtinov's, Carey's, tri-Bahtinov's and hybrids for each focal length. Here's my collection. 

That's indeed a collection :)
Do you have a favorite? When it comes to results and/or ease of use?



#380 Ben Diss

Ben Diss

    Vostok 1

  • *****
  • Posts: 100
  • Joined: 07 Nov 2019
  • Loc: N40 W83

Posted 17 July 2020 - 05:24 AM

That's indeed a collection smile.gif
Do you have a favorite? When it comes to results and/or ease of use?

 

Yup. Carey mask. Things I learned on that journey:

 

1. Smaller slots make a huge difference. Smaller the better.

2. Thin is in. I settled on 1mm thick.

3. I printed Duramic PLA+, eSun ABS+ and Inland PETG. I settled on plain old PLA+ for ease of print and durability.

4. Rounded corners do indeed make a difference. The spikes are sharper and more well defined.

 

I rewrote the Bahintov mask SCAD to produce rounded corners and either Bathinov or Carey masks by specifying the angle for both left and right halves. I had to call in help from my math brain son for the trig on that project.


  • meegja likes this

#381 meegja

meegja

    Messenger

  • -----
  • Posts: 486
  • Joined: 02 Jul 2020
  • Loc: The Netherlands

Posted 18 July 2020 - 01:47 PM

Yup. Carey mask. Things I learned on that journey:

 

1. Smaller slots make a huge difference. Smaller the better.

2. Thin is in. I settled on 1mm thick.

3. I printed Duramic PLA+, eSun ABS+ and Inland PETG. I settled on plain old PLA+ for ease of print and durability.

4. Rounded corners do indeed make a difference. The spikes are sharper and more well defined.

 

I rewrote the Bahintov mask SCAD to produce rounded corners and either Bathinov or Carey masks by specifying the angle for both left and right halves. I had to call in help from my math brain son for the trig on that project.

Okay, thanks! Did look up the Carey mask and looks interesting. Think I am going to try and make a cardboard version :)



#382 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,046
  • Joined: 24 Nov 2014

Posted 19 July 2020 - 02:30 PM

Hi guys,

   It seems that my Tri-Bahtinov repository will be preserved in the Github Arctic Cold Vault smile.gif

 

track.jpg

 

   Here's more info about the Github Arctic Cold Vault:

 

https://github.blog/...-to-the-arctic/

 

cytan


  • dswtan, Scott Mitchell, calypsob and 3 others like this

#383 meegja

meegja

    Messenger

  • -----
  • Posts: 486
  • Joined: 02 Jul 2020
  • Loc: The Netherlands

Posted 20 July 2020 - 03:58 AM

Hi guys,

   It seems that my Tri-Bahtinov repository will be preserved in the Github Arctic Cold Vault smile.gif

Talking about recognition of great work! waytogo.gif


  • cytan299 likes this

#384 fortranguy

fortranguy

    Lift Off

  • -----
  • Posts: 10
  • Joined: 06 May 2018
  • Loc: Auckland, New Zealand

Posted 21 July 2020 - 06:31 PM

Hi guys,

 

First of all, I would like to thank C.Y. Tan and Satoru Takagi for this amazing tool for focusing and also collimating reflector telescopes.

 

I have developed a web based cover generator based on their codes.

 

Satoru, you have a PR in your github: https://github.com/s...vWebApps/pull/1

 

You can have a look at: http://skylabs.co.nz...inov_cover.html

 

Thanks!

 

Jordi


  • cytan299 likes this

#385 fortranguy

fortranguy

    Lift Off

  • -----
  • Posts: 10
  • Joined: 06 May 2018
  • Loc: Auckland, New Zealand

Posted 23 July 2020 - 07:30 PM

Hi,

 

After trying several materials and cutting technologies, I found that high-density cardboard provides the best results.

 

I removed some holes as they do not contribute in the expected diffraction pattern. I'm currently working on the round corners.

 

This is how it looks now.

 

Cheers,

 

Jordi

 
Tri-Bahtinov Satoru Takagi version

  • calypsob, cytan299 and wcoastsands like this

#386 SteveInNZ

SteveInNZ

    Viking 1

  • *****
  • Posts: 943
  • Joined: 07 Oct 2007
  • Loc: Auckland, New Zealand

Posted 23 July 2020 - 10:12 PM

Wouldn't it get soggy/soft in the dew overnight ?

 

Steve.



#387 Psittacula

Psittacula

    Explorer 1

  • -----
  • Posts: 90
  • Joined: 27 May 2017
  • Loc: Tokyo

Posted 24 July 2020 - 03:12 AM

Hi guys,

 

First of all, I would like to thank C.Y. Tan and Satoru Takagi for this amazing tool for focusing and also collimating reflector telescopes.

 

I have developed a web based cover generator based on their codes.

 

Satoru, you have a PR in your github: https://github.com/s...vWebApps/pull/1

 

You can have a look at: http://skylabs.co.nz...inov_cover.html

 

Thanks!

 

Jordi

Hi Jordi,

 

Thank you very much for contributing to the progress of the open source hardware tri-bahtinov mask.
Please tell me how to use the Tri-Bahtinov Mask Cover you designed.

 

When I generate a cover using your link, I get this output:

mask.png
When it was layered on the sensitivity-enhancing mask, did it look like this or did it look like this?

m1a.png

 

On the other hand, what I imagined as a cover to enable only one set of Bahtinov Mask is the pattern of the red part in the figure below.

m2.png

 

Satoru



#388 Ben Diss

Ben Diss

    Vostok 1

  • *****
  • Posts: 100
  • Joined: 07 Nov 2019
  • Loc: N40 W83

Posted 24 July 2020 - 06:18 AM

Hi,

 

After trying several materials and cutting technologies, I found that high-density cardboard provides the best results.

 

I removed some holes as they do not contribute in the expected diffraction pattern. I'm currently working on the round corners.

 

This is how it looks now.

 

Cheers,

 

Jordi

 

Here's my SCAD file where I implemented rounded corners. Lots of trig here.

 

http://benodee.com/a...s_mask_4.0.scad

(right-click, save as...)



#389 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,046
  • Joined: 24 Nov 2014

Posted 24 July 2020 - 02:00 PM

Here's my SCAD file where I implemented rounded corners. Lots of trig here.

 

http://benodee.com/a...s_mask_4.0.scad

(right-click, save as...)

Hi Ben,

   I just looked at it, quite a bit of code :). But this is for a Carey mask and not a Tribahtinov.

 

cytan



#390 Ben Diss

Ben Diss

    Vostok 1

  • *****
  • Posts: 100
  • Joined: 07 Nov 2019
  • Loc: N40 W83

Posted 24 July 2020 - 02:56 PM

Hi Ben,

   I just looked at it, quite a bit of code smile.gif. But this is for a Carey mask and not a Tribahtinov.

 

cytan

 

Right. I implemented rounded corners and it was a bit of an undertaking. I thought this might prove helpful as you do this for the tri-B mask. The challenge for me was that with slots you can simply draw slots and cut the edge off by subtracting a cylinder at the inner and out boundary. With a rounded corner that won't work so I drew each slot individually and calculated the appropriate begin and end, accounting for the degrees of slant.

 

Also, with this file you can do either Bahtinov or Carey. You specify the degrees for the left or right. For Carey, 10 and 12 works well. For Bahtinov, 0 and 20 is good.



#391 Jinux

Jinux

    Messenger

  • *****
  • Posts: 480
  • Joined: 25 Jun 2007
  • Loc: Bay Area, CA

Posted 24 July 2020 - 06:33 PM

Hi,

 

After trying several materials and cutting technologies, I found that high-density cardboard provides the best results.

 

I removed some holes as they do not contribute in the expected diffraction pattern. I'm currently working on the round corners.

 

This is how it looks now.

 

Cheers,

 

Jordi

I tried three types of masks made of

 

1. Clear Lexan (clear polycarbonate plastic) 1/8" - CNC cut

2. Opaque Lexan 1/8" - CNC cut

3. 3D printed with black ABS in 5mm thickness

 

All design parameters are the same like Bhatinov factor, outer/inner margin etc. So amount of open/clear area the same.

Among these, 2 and 3 are about the same. There might be but they're in the same ballpark.

On the other hand, #1 has much brighter and longer spikes in totally different league.

 

But I can't use #1 for any use because patterns are not Tri-B patterns anymore. There are many more patterns overlapped on each other.

My guess is, 1) when I ordered this mask to CNC guy, Satoru's pattern generator has slight symmetry issue. Due to bright nature of this material, any imperfection would be apparent. and 2) Lexan is not coated for light transmission. So there are surfaces that light reflects and make offset patterns.

I couldn't confirm which one contributes more though as I dropped the experiment.

Meanwhile, William Optics is selling the clear B-mask with claim 'much brighter', you can see it's not just a marketing phrase. It's indeed a (or several) notch brighter.



#392 Ben Diss

Ben Diss

    Vostok 1

  • *****
  • Posts: 100
  • Joined: 07 Nov 2019
  • Loc: N40 W83

Posted 24 July 2020 - 07:15 PM

I tried three types of masks made of

 

1. Clear Lexan (clear polycarbonate plastic) 1/8" - CNC cut

2. Opaque Lexan 1/8" - CNC cut

3. 3D printed with black ABS in 5mm thickness

 

All design parameters are the same like Bhatinov factor, outer/inner margin etc. So amount of open/clear area the same.

Among these, 2 and 3 are about the same. There might be but they're in the same ballpark.

On the other hand, #1 has much brighter and longer spikes in totally different league.

 

But I can't use #1 for any use because patterns are not Tri-B patterns anymore. There are many more patterns overlapped on each other.

My guess is, 1) when I ordered this mask to CNC guy, Satoru's pattern generator has slight symmetry issue. Due to bright nature of this material, any imperfection would be apparent. and 2) Lexan is not coated for light transmission. So there are surfaces that light reflects and make offset patterns.

I couldn't confirm which one contributes more though as I dropped the experiment.

Meanwhile, William Optics is selling the clear B-mask with claim 'much brighter', you can see it's not just a marketing phrase. It's indeed a (or several) notch brighter.

 

Try printing 1mm thick ABS, PLA or PTEG. Thinner was definitely better for me.



#393 Psittacula

Psittacula

    Explorer 1

  • -----
  • Posts: 90
  • Joined: 27 May 2017
  • Loc: Tokyo

Posted 24 July 2020 - 09:28 PM

Meanwhile, William Optics is selling the clear B-mask with claim 'much brighter', you can see it's not just a marketing phrase. It's indeed a (or several) notch brighter.

 

Try printing 1mm thick ABS, PLA or PTEG. Thinner was definitely better for me.

https://www.ontariot...8mm_p_1442.html
 

If you zoom in on the photo of the company's mask on this page, it looks like a grating pattern engraved on a transparent resin surface maybe by laser engraver. This may effectively correspond to a very thin and narrow pitch grating.

 

 

By the way, I own a cutting plotter, so I use masks with very thin plastic sheets cut out of it.

PS: I'm using an opaque plastic file folder as a material, so it's a polypropylene sheet with a thickness of about 0.2mm.

 

Satoru


Edited by Psittacula, 24 July 2020 - 11:32 PM.

  • Ben Diss likes this

#394 Zebul

Zebul

    Lift Off

  • -----
  • Posts: 6
  • Joined: 14 Apr 2019
  • Loc: France

Posted 26 July 2020 - 03:35 AM

Hi all,

My english is far form perfect and I don't well understand when you talk about thick, what is exactly about.

 

Now after reading messages this is clear, and I fully agree, that the thinner your mask is, the better diffraction result you will have.

I don't whant to spend time to explain why it is better but it is opticaly extremly logical. You don't generat parasith bangs that kill contrast.

 

For my personnal usage I use opaque PVC sheet to build my mask and I have well sharpe diffraction spikes.

I confirm that the solution of engraving (not cutting) a cristal polycarbonat will probably do the same well result. I can say too, that some person have done some mask with only nylon fishing line and they have perfect spike too.

If you build a 3D printing mask, I suggest to create bevel on each slits (like that "\  /" or reverse wink.gif ) to optimise your DIY result.

 

Now, I don't understand when you talk about rounded corner, what is about. I imagine that this build is to avoid parasith bangs too...

Is there some one can place an explicative picture ?

 

A last question, is there really a serious difference beetwen large slits and thin ones ? If some one can place comparative pictures...

To my short experience and forums reading, there is no significative improvement.


Edited by Zebul, 26 July 2020 - 03:36 AM.


#395 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,046
  • Joined: 24 Nov 2014

Posted 26 July 2020 - 05:24 PM

Hi all,

My english is far form perfect and I don't well understand when you talk about thick, what is exactly about.

 

Now after reading messages this is clear, and I fully agree, that the thinner your mask is, the better diffraction result you will have.

I don't whant to spend time to explain why it is better but it is opticaly extremly logical. You don't generat parasith bangs that kill contrast.

 

For my personnal usage I use opaque PVC sheet to build my mask and I have well sharpe diffraction spikes.

I confirm that the solution of engraving (not cutting) a cristal polycarbonat will probably do the same well result. I can say too, that some person have done some mask with only nylon fishing line and they have perfect spike too.

If you build a 3D printing mask, I suggest to create bevel on each slits (like that "\  /" or reverse wink.gif ) to optimise your DIY result.

 

Now, I don't understand when you talk about rounded corner, what is about. I imagine that this build is to avoid parasith bangs too...

Is there some one can place an explicative picture ?

 

A last question, is there really a serious difference beetwen large slits and thin ones ? If some one can place comparative pictures...

To my short experience and forums reading, there is no significative improvement.

I think what you say above: I will interpret "bangs" as "spikes".

 

With regards to rounded corners, this basically means that higher order spikes in the spectrum is reduced. So, the Bahtinov spike image is less noisy, i.e. a lot less spurious spikes.

 

The advantage of using thin slits is that you can have more of them. More slits means that more energy is now concentrated on the main spikes. More energy means that the spikes are brighter and thinner.

 

Here's a link that from Japan that shows the effect of having more slits for the TriBahtinov mask:

 

https://reflexions.j...017/10/04/2280/

 

The disadvantage of having more slits in the same area is that the struts between the slits are thinner and so makes the mask more fragile. It's a tradeoff between strength of the mask and clarity of the spikes.

 

cytan



#396 Jinux

Jinux

    Messenger

  • *****
  • Posts: 480
  • Joined: 25 Jun 2007
  • Loc: Bay Area, CA

Posted 26 July 2020 - 06:54 PM

Hi all,

My english is far form perfect and I don't well understand when you talk about thick, what is exactly about.

 

Now after reading messages this is clear, and I fully agree, that the thinner your mask is, the better diffraction result you will have.

I don't whant to spend time to explain why it is better but it is opticaly extremly logical. You don't generat parasith bangs that kill contrast.

 

For my personnal usage I use opaque PVC sheet to build my mask and I have well sharpe diffraction spikes.

I confirm that the solution of engraving (not cutting) a cristal polycarbonat will probably do the same well result. I can say too, that some person have done some mask with only nylon fishing line and they have perfect spike too.

If you build a 3D printing mask, I suggest to create bevel on each slits (like that "\  /" or reverse wink.gif ) to optimise your DIY result.

 

Now, I don't understand when you talk about rounded corner, what is about. I imagine that this build is to avoid parasith bangs too...

Is there some one can place an explicative picture ?

 

A last question, is there really a serious difference beetwen large slits and thin ones ? If some one can place comparative pictures...

To my short experience and forums reading, there is no significative improvement.

If I understand the 'rounded corners' correctly, it means the end of slit is rounded rather than sharp edged like attached drawings. 

I guess it will reduce higher order diffraction and less spurious spikes but my 5mm thickness edged 3d-printed one works similar order of magnitude as ~3mm thickness rounded CNC'ed one. If we go to 1mm in 3D printing material, it will be flexible and could be touching the corrector plate in SCT case rather than hanging on top of corrector plate retaining ring. I may try but won't go below 2mm. 

One of the challenge of thinner 3D printed mask would be assembly. My (and most of) 3D printer bed is not big enough to print mask of C11. So I had to divide whole section to 6 outer sections and one inner section and assembled using dovetails. If we make it thin, assembly may not fit as flat mask.

 

Yes. Slit width has an effect on spike patterns. You can simulate it using Maskulator program. I actually created multiple masks with different B-factor (typically believed as 150 ~ 200). Indeed, 150 is good number to go with. If you go with 200, it starts to have higher order spurious spikes while not clearly visible in real world and once you go with 300, there are more than 3 spikes in apparent patterns. If you go less of B-factor (wider slit), spikes has less color separation and getting shorter and dimmer.

I'll find those pictures I experimented and post later.

 

-Jinux

Attached Thumbnails

  • rounded.jpg


#397 Jinux

Jinux

    Messenger

  • *****
  • Posts: 480
  • Joined: 25 Jun 2007
  • Loc: Bay Area, CA

Posted 26 July 2020 - 07:06 PM

I confirm that the solution of engraving (not cutting) a cristal polycarbonat will probably do the same well result. I can say too, that some person have done some mask with only nylon fishing line and they have perfect spike too.

 

This inspires me to have a 3D printed mask design that would be much brighter patterns.

What if, we remove all slits but leave stems for section dividers. And then, create hooks to net a fish line through where slit lines are supposed to be? 

Now openings are much wider, so it will be brighter but there are fishing lines where slits are supposed to be like engraving on acrylic. 

There won't be any surfaces to create reflections and hey it's 100% transmission much more than 92% WO is boasting and I'm pretty sure no one can beat this number. lol.gif

 

-Jinux



#398 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,046
  • Joined: 24 Nov 2014

Posted 26 July 2020 - 08:25 PM

Actually, I've never understood why an engraved transparent mask would actually work or is good because light would have to refract through the transparent material. If the upper and lower surfaces are not parallel, you can imagine that the light paths will be different from different locations and thus the focusing (and for collimation) is probably not quite perfect.

 

cytan


Edited by cytan299, 26 July 2020 - 08:25 PM.


#399 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,046
  • Joined: 24 Nov 2014

Posted 26 July 2020 - 08:27 PM

If I understand the 'rounded corners' correctly, it means the end of slit is rounded rather than sharp edged like attached drawings. 

I guess it will reduce higher order diffraction and less spurious spikes but my 5mm thickness edged 3d-printed one works similar order of magnitude as ~3mm thickness rounded CNC'ed one. If we go to 1mm in 3D printing material, it will be flexible and could be touching the corrector plate in SCT case rather than hanging on top of corrector plate retaining ring. I may try but won't go below 2mm. 

One of the challenge of thinner 3D printed mask would be assembly. My (and most of) 3D printer bed is not big enough to print mask of C11. So I had to divide whole section to 6 outer sections and one inner section and assembled using dovetails. If we make it thin, assembly may not fit as flat mask.

 

Yes. Slit width has an effect on spike patterns. You can simulate it using Maskulator program. I actually created multiple masks with different B-factor (typically believed as 150 ~ 200). Indeed, 150 is good number to go with. If you go with 200, it starts to have higher order spurious spikes while not clearly visible in real world and once you go with 300, there are more than 3 spikes in apparent patterns. If you go less of B-factor (wider slit), spikes has less color separation and getting shorter and dimmer.

I'll find those pictures I experimented and post later.

 

-Jinux

Yes, ends should be round like the second cartoon that you drew to be effective.

 

cytan



#400 fortranguy

fortranguy

    Lift Off

  • -----
  • Posts: 10
  • Joined: 06 May 2018
  • Loc: Auckland, New Zealand

Posted 26 July 2020 - 09:39 PM

Hi Jordi,

 

Thank you very much for contributing to the progress of the open source hardware tri-bahtinov mask.
Please tell me how to use the Tri-Bahtinov Mask Cover you designed.

 

When I generate a cover using your link, I get this output:

attachicon.gifmask.png
When it was layered on the sensitivity-enhancing mask, did it look like this or did it look like this?

attachicon.gifm1a.png

 

On the other hand, what I imagined as a cover to enable only one set of Bahtinov Mask is the pattern of the red part in the figure below.

attachicon.gifm2.png

 

Satoru

Hi Satoru,

 

Yes, you are right. The idea is to generate a classical Bahtinov pattern to evaluate each collimation screw.

 

Please, have a look at the new PR.

 

I will upload the quick guide about the process. Here you have the first draft.

 

Regards,

 

Jordi

 

 

------

 

How to use the Skylabs Improved Sensitivity Tri-Bahtinov Mask for collimation

  1. The initial collimation must be done with another tool, like a laser beam or Cheshire eyepiece. The secondary mirror needs to be correctly adjusted because the tri-Bahtinov mask will only help you to align the primary mirror.
  2. Then introduce an artificial star in the center of the field of view and focus. You can do it with a kitchen aluminium foil wrapping a LED torch and drill a hole in the center with a pin. Move the torch to a reasonable distance.
  3. Set up the tri-Bahtinov mask in front of the telescope, either over the spider or over the correction lens (SCT, MCT, etc.). Try matching the orientation of the three little holes located at the edge of the mask with the adjustment screws of the primary mirror. This will help you to identify which **** should use.
  4. Focus the artificial star following the instructions defined in the previous section until you get the best alignment in the diffraction pattern.
  5. Introduce the mask cover on top of the Tri-Bahtinov mask and align the small holes of the mask and cover. This will generate a classical Bahtinov diffraction pattern.
  6. Move the primary mirror collimation screw aligned to the uncovered mark to adjust the central spike with the other two. → If the image shows the central spike is off to the left, the mirror is too far. → If the central spikes are off to the right, then the mirror is too close.
  7. Rotate the mask cover 120° and repeat step 4.
  8. The changes introduced by the other two screws may introduce minor changes in the first screw. Iterate the steeps 4 and 5 until the three positions of the cover provide a perfect aligned classical Bahtinov diffraction.

  • Psittacula likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: astrophotography, imaging, SCT



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics